diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,147261 @@ +{ + "best_metric": 0.24186615645885468, + "best_model_checkpoint": "./w2v-bert-2.0-luo_19_77h/checkpoint-16000", + "epoch": 17.736486486486488, + "eval_steps": 1000, + "global_step": 21000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0008445945945945946, + "grad_norm": 30.041261672973633, + "learning_rate": 3.0000000000000004e-09, + "loss": 10.2166, + "step": 1 + }, + { + "epoch": 0.0016891891891891893, + "grad_norm": 30.156036376953125, + "learning_rate": 6.000000000000001e-09, + "loss": 10.2294, + "step": 2 + }, + { + "epoch": 0.002533783783783784, + "grad_norm": 26.313936233520508, + "learning_rate": 9e-09, + "loss": 8.9859, + "step": 3 + }, + { + "epoch": 0.0033783783783783786, + "grad_norm": 23.545856475830078, + "learning_rate": 1.2000000000000002e-08, + "loss": 8.2749, + "step": 4 + }, + { + "epoch": 0.004222972972972973, + "grad_norm": 24.423189163208008, + "learning_rate": 1.5000000000000002e-08, + "loss": 8.4263, + "step": 5 + }, + { + "epoch": 0.005067567567567568, + "grad_norm": 22.688091278076172, + "learning_rate": 1.8e-08, + "loss": 7.9682, + "step": 6 + }, + { + "epoch": 0.0059121621621621625, + "grad_norm": 24.062379837036133, + "learning_rate": 2.1e-08, + "loss": 8.1049, + "step": 7 + }, + { + "epoch": 0.006756756756756757, + "grad_norm": 23.942007064819336, + "learning_rate": 2.4000000000000003e-08, + "loss": 8.4051, + "step": 8 + }, + { + "epoch": 0.007601351351351352, + "grad_norm": 22.403472900390625, + "learning_rate": 2.7e-08, + "loss": 8.036, + "step": 9 + }, + { + "epoch": 0.008445945945945946, + "grad_norm": 23.114015579223633, + "learning_rate": 3.0000000000000004e-08, + "loss": 8.0337, + "step": 10 + }, + { + "epoch": 0.009290540540540541, + "grad_norm": 22.005962371826172, + "learning_rate": 3.3000000000000004e-08, + "loss": 7.7341, + "step": 11 + }, + { + "epoch": 0.010135135135135136, + "grad_norm": 22.61867904663086, + "learning_rate": 3.6e-08, + "loss": 8.0461, + "step": 12 + }, + { + "epoch": 0.01097972972972973, + "grad_norm": 21.15814781188965, + "learning_rate": 3.9e-08, + "loss": 7.4812, + "step": 13 + }, + { + "epoch": 0.011824324324324325, + "grad_norm": 23.01306915283203, + "learning_rate": 4.2e-08, + "loss": 8.0035, + "step": 14 + }, + { + "epoch": 0.01266891891891892, + "grad_norm": 23.632301330566406, + "learning_rate": 4.5e-08, + "loss": 8.0151, + "step": 15 + }, + { + "epoch": 0.013513513513513514, + "grad_norm": 21.468507766723633, + "learning_rate": 4.8000000000000006e-08, + "loss": 7.7719, + "step": 16 + }, + { + "epoch": 0.014358108108108109, + "grad_norm": 23.78802490234375, + "learning_rate": 5.1e-08, + "loss": 8.208, + "step": 17 + }, + { + "epoch": 0.015202702702702704, + "grad_norm": 22.624507904052734, + "learning_rate": 5.4e-08, + "loss": 7.7212, + "step": 18 + }, + { + "epoch": 0.016047297297297296, + "grad_norm": 23.669586181640625, + "learning_rate": 5.7e-08, + "loss": 8.0846, + "step": 19 + }, + { + "epoch": 0.016891891891891893, + "grad_norm": 21.6710205078125, + "learning_rate": 6.000000000000001e-08, + "loss": 7.5955, + "step": 20 + }, + { + "epoch": 0.017736486486486486, + "grad_norm": 22.7683162689209, + "learning_rate": 6.3e-08, + "loss": 7.7659, + "step": 21 + }, + { + "epoch": 0.018581081081081082, + "grad_norm": 21.614612579345703, + "learning_rate": 6.600000000000001e-08, + "loss": 7.5101, + "step": 22 + }, + { + "epoch": 0.019425675675675675, + "grad_norm": 22.18198585510254, + "learning_rate": 6.9e-08, + "loss": 7.5836, + "step": 23 + }, + { + "epoch": 0.02027027027027027, + "grad_norm": 21.311038970947266, + "learning_rate": 7.2e-08, + "loss": 7.4794, + "step": 24 + }, + { + "epoch": 0.021114864864864864, + "grad_norm": 23.17034912109375, + "learning_rate": 7.500000000000001e-08, + "loss": 7.8479, + "step": 25 + }, + { + "epoch": 0.02195945945945946, + "grad_norm": 21.3421573638916, + "learning_rate": 7.8e-08, + "loss": 7.5082, + "step": 26 + }, + { + "epoch": 0.022804054054054054, + "grad_norm": 23.82416534423828, + "learning_rate": 8.100000000000001e-08, + "loss": 8.09, + "step": 27 + }, + { + "epoch": 0.02364864864864865, + "grad_norm": 22.870861053466797, + "learning_rate": 8.4e-08, + "loss": 7.7119, + "step": 28 + }, + { + "epoch": 0.024493243243243243, + "grad_norm": 21.287532806396484, + "learning_rate": 8.7e-08, + "loss": 7.4755, + "step": 29 + }, + { + "epoch": 0.02533783783783784, + "grad_norm": 23.006458282470703, + "learning_rate": 9e-08, + "loss": 7.8052, + "step": 30 + }, + { + "epoch": 0.026182432432432432, + "grad_norm": 21.515592575073242, + "learning_rate": 9.3e-08, + "loss": 7.4144, + "step": 31 + }, + { + "epoch": 0.02702702702702703, + "grad_norm": 21.50055694580078, + "learning_rate": 9.600000000000001e-08, + "loss": 7.5667, + "step": 32 + }, + { + "epoch": 0.02787162162162162, + "grad_norm": 22.542566299438477, + "learning_rate": 9.9e-08, + "loss": 7.8069, + "step": 33 + }, + { + "epoch": 0.028716216216216218, + "grad_norm": 22.679594039916992, + "learning_rate": 1.02e-07, + "loss": 7.7379, + "step": 34 + }, + { + "epoch": 0.02956081081081081, + "grad_norm": 21.791154861450195, + "learning_rate": 1.05e-07, + "loss": 7.4066, + "step": 35 + }, + { + "epoch": 0.030405405405405407, + "grad_norm": 21.735776901245117, + "learning_rate": 1.08e-07, + "loss": 7.5155, + "step": 36 + }, + { + "epoch": 0.03125, + "grad_norm": 21.71992301940918, + "learning_rate": 1.11e-07, + "loss": 7.4357, + "step": 37 + }, + { + "epoch": 0.03209459459459459, + "grad_norm": 21.623634338378906, + "learning_rate": 1.14e-07, + "loss": 7.351, + "step": 38 + }, + { + "epoch": 0.032939189189189186, + "grad_norm": 21.16552734375, + "learning_rate": 1.17e-07, + "loss": 7.1919, + "step": 39 + }, + { + "epoch": 0.033783783783783786, + "grad_norm": 22.113243103027344, + "learning_rate": 1.2000000000000002e-07, + "loss": 7.5134, + "step": 40 + }, + { + "epoch": 0.03462837837837838, + "grad_norm": 23.897302627563477, + "learning_rate": 1.23e-07, + "loss": 7.9938, + "step": 41 + }, + { + "epoch": 0.03547297297297297, + "grad_norm": 23.209487915039062, + "learning_rate": 1.26e-07, + "loss": 7.4279, + "step": 42 + }, + { + "epoch": 0.036317567567567564, + "grad_norm": 21.94633674621582, + "learning_rate": 1.29e-07, + "loss": 7.2868, + "step": 43 + }, + { + "epoch": 0.037162162162162164, + "grad_norm": 24.575820922851562, + "learning_rate": 1.3200000000000002e-07, + "loss": 7.9448, + "step": 44 + }, + { + "epoch": 0.03800675675675676, + "grad_norm": 24.243881225585938, + "learning_rate": 1.35e-07, + "loss": 7.9155, + "step": 45 + }, + { + "epoch": 0.03885135135135135, + "grad_norm": 25.585111618041992, + "learning_rate": 1.38e-07, + "loss": 8.0889, + "step": 46 + }, + { + "epoch": 0.03969594594594594, + "grad_norm": 22.785053253173828, + "learning_rate": 1.41e-07, + "loss": 7.3562, + "step": 47 + }, + { + "epoch": 0.04054054054054054, + "grad_norm": 25.740917205810547, + "learning_rate": 1.44e-07, + "loss": 8.0947, + "step": 48 + }, + { + "epoch": 0.041385135135135136, + "grad_norm": 26.560054779052734, + "learning_rate": 1.47e-07, + "loss": 8.1362, + "step": 49 + }, + { + "epoch": 0.04222972972972973, + "grad_norm": 27.582263946533203, + "learning_rate": 1.5000000000000002e-07, + "loss": 8.3005, + "step": 50 + }, + { + "epoch": 0.04307432432432432, + "grad_norm": 29.912200927734375, + "learning_rate": 1.53e-07, + "loss": 9.656, + "step": 51 + }, + { + "epoch": 0.04391891891891892, + "grad_norm": 33.70949172973633, + "learning_rate": 1.56e-07, + "loss": 10.1062, + "step": 52 + }, + { + "epoch": 0.044763513513513514, + "grad_norm": 27.549236297607422, + "learning_rate": 1.59e-07, + "loss": 8.963, + "step": 53 + }, + { + "epoch": 0.04560810810810811, + "grad_norm": 25.627683639526367, + "learning_rate": 1.6200000000000002e-07, + "loss": 8.276, + "step": 54 + }, + { + "epoch": 0.0464527027027027, + "grad_norm": 23.890743255615234, + "learning_rate": 1.6499999999999998e-07, + "loss": 7.8065, + "step": 55 + }, + { + "epoch": 0.0472972972972973, + "grad_norm": 25.216169357299805, + "learning_rate": 1.68e-07, + "loss": 7.9806, + "step": 56 + }, + { + "epoch": 0.04814189189189189, + "grad_norm": 23.96738052368164, + "learning_rate": 1.71e-07, + "loss": 7.6956, + "step": 57 + }, + { + "epoch": 0.048986486486486486, + "grad_norm": 23.76844596862793, + "learning_rate": 1.74e-07, + "loss": 7.5949, + "step": 58 + }, + { + "epoch": 0.04983108108108108, + "grad_norm": 23.767166137695312, + "learning_rate": 1.77e-07, + "loss": 7.517, + "step": 59 + }, + { + "epoch": 0.05067567567567568, + "grad_norm": 24.323551177978516, + "learning_rate": 1.8e-07, + "loss": 7.6777, + "step": 60 + }, + { + "epoch": 0.05152027027027027, + "grad_norm": 26.29371452331543, + "learning_rate": 1.83e-07, + "loss": 8.0524, + "step": 61 + }, + { + "epoch": 0.052364864864864864, + "grad_norm": 23.914844512939453, + "learning_rate": 1.86e-07, + "loss": 7.4471, + "step": 62 + }, + { + "epoch": 0.05320945945945946, + "grad_norm": 24.644559860229492, + "learning_rate": 1.89e-07, + "loss": 7.6847, + "step": 63 + }, + { + "epoch": 0.05405405405405406, + "grad_norm": 24.32817268371582, + "learning_rate": 1.9200000000000003e-07, + "loss": 7.5677, + "step": 64 + }, + { + "epoch": 0.05489864864864865, + "grad_norm": 23.334178924560547, + "learning_rate": 1.9499999999999999e-07, + "loss": 7.4037, + "step": 65 + }, + { + "epoch": 0.05574324324324324, + "grad_norm": 25.466205596923828, + "learning_rate": 1.98e-07, + "loss": 7.679, + "step": 66 + }, + { + "epoch": 0.056587837837837836, + "grad_norm": 26.21953010559082, + "learning_rate": 2.01e-07, + "loss": 7.8208, + "step": 67 + }, + { + "epoch": 0.057432432432432436, + "grad_norm": 24.264188766479492, + "learning_rate": 2.04e-07, + "loss": 7.4462, + "step": 68 + }, + { + "epoch": 0.05827702702702703, + "grad_norm": 25.32801628112793, + "learning_rate": 2.0700000000000001e-07, + "loss": 7.5401, + "step": 69 + }, + { + "epoch": 0.05912162162162162, + "grad_norm": 23.999486923217773, + "learning_rate": 2.1e-07, + "loss": 7.3017, + "step": 70 + }, + { + "epoch": 0.059966216216216214, + "grad_norm": 24.321430206298828, + "learning_rate": 2.1300000000000001e-07, + "loss": 7.3851, + "step": 71 + }, + { + "epoch": 0.060810810810810814, + "grad_norm": 23.702253341674805, + "learning_rate": 2.16e-07, + "loss": 7.1971, + "step": 72 + }, + { + "epoch": 0.06165540540540541, + "grad_norm": 25.426889419555664, + "learning_rate": 2.1900000000000002e-07, + "loss": 7.4711, + "step": 73 + }, + { + "epoch": 0.0625, + "grad_norm": 23.89812660217285, + "learning_rate": 2.22e-07, + "loss": 7.2399, + "step": 74 + }, + { + "epoch": 0.0633445945945946, + "grad_norm": 24.252708435058594, + "learning_rate": 2.25e-07, + "loss": 7.2351, + "step": 75 + }, + { + "epoch": 0.06418918918918919, + "grad_norm": 25.75965690612793, + "learning_rate": 2.28e-07, + "loss": 7.4759, + "step": 76 + }, + { + "epoch": 0.06503378378378379, + "grad_norm": 26.164785385131836, + "learning_rate": 2.3100000000000002e-07, + "loss": 7.3439, + "step": 77 + }, + { + "epoch": 0.06587837837837837, + "grad_norm": 24.45169448852539, + "learning_rate": 2.34e-07, + "loss": 7.1512, + "step": 78 + }, + { + "epoch": 0.06672297297297297, + "grad_norm": 24.157541275024414, + "learning_rate": 2.3700000000000002e-07, + "loss": 6.9968, + "step": 79 + }, + { + "epoch": 0.06756756756756757, + "grad_norm": 27.462305068969727, + "learning_rate": 2.4000000000000003e-07, + "loss": 7.5567, + "step": 80 + }, + { + "epoch": 0.06841216216216216, + "grad_norm": 23.78824234008789, + "learning_rate": 2.43e-07, + "loss": 7.0406, + "step": 81 + }, + { + "epoch": 0.06925675675675676, + "grad_norm": 26.32724380493164, + "learning_rate": 2.46e-07, + "loss": 7.2242, + "step": 82 + }, + { + "epoch": 0.07010135135135136, + "grad_norm": 24.769834518432617, + "learning_rate": 2.49e-07, + "loss": 7.0531, + "step": 83 + }, + { + "epoch": 0.07094594594594594, + "grad_norm": 26.533096313476562, + "learning_rate": 2.52e-07, + "loss": 7.4062, + "step": 84 + }, + { + "epoch": 0.07179054054054054, + "grad_norm": 25.535757064819336, + "learning_rate": 2.5500000000000005e-07, + "loss": 6.9453, + "step": 85 + }, + { + "epoch": 0.07263513513513513, + "grad_norm": 25.1673641204834, + "learning_rate": 2.58e-07, + "loss": 6.9378, + "step": 86 + }, + { + "epoch": 0.07347972972972973, + "grad_norm": 29.48828887939453, + "learning_rate": 2.6099999999999997e-07, + "loss": 7.5344, + "step": 87 + }, + { + "epoch": 0.07432432432432433, + "grad_norm": 24.315221786499023, + "learning_rate": 2.6400000000000003e-07, + "loss": 6.8405, + "step": 88 + }, + { + "epoch": 0.07516891891891891, + "grad_norm": 26.589948654174805, + "learning_rate": 2.67e-07, + "loss": 6.9727, + "step": 89 + }, + { + "epoch": 0.07601351351351351, + "grad_norm": 31.044879913330078, + "learning_rate": 2.7e-07, + "loss": 7.5327, + "step": 90 + }, + { + "epoch": 0.07685810810810811, + "grad_norm": 25.49603271484375, + "learning_rate": 2.73e-07, + "loss": 6.709, + "step": 91 + }, + { + "epoch": 0.0777027027027027, + "grad_norm": 29.160490036010742, + "learning_rate": 2.76e-07, + "loss": 7.3064, + "step": 92 + }, + { + "epoch": 0.0785472972972973, + "grad_norm": 28.57056427001953, + "learning_rate": 2.79e-07, + "loss": 7.242, + "step": 93 + }, + { + "epoch": 0.07939189189189189, + "grad_norm": 25.080602645874023, + "learning_rate": 2.82e-07, + "loss": 6.6483, + "step": 94 + }, + { + "epoch": 0.08023648648648649, + "grad_norm": Infinity, + "learning_rate": 2.82e-07, + "loss": 7.5511, + "step": 95 + }, + { + "epoch": 0.08108108108108109, + "grad_norm": 28.516067504882812, + "learning_rate": 2.85e-07, + "loss": 6.8897, + "step": 96 + }, + { + "epoch": 0.08192567567567567, + "grad_norm": 29.320741653442383, + "learning_rate": 2.88e-07, + "loss": 7.1002, + "step": 97 + }, + { + "epoch": 0.08277027027027027, + "grad_norm": Infinity, + "learning_rate": 2.88e-07, + "loss": 8.7989, + "step": 98 + }, + { + "epoch": 0.08361486486486487, + "grad_norm": 31.507081985473633, + "learning_rate": 2.91e-07, + "loss": 7.3056, + "step": 99 + }, + { + "epoch": 0.08445945945945946, + "grad_norm": 40.39160919189453, + "learning_rate": 2.94e-07, + "loss": 8.3497, + "step": 100 + }, + { + "epoch": 0.08530405405405406, + "grad_norm": 39.08432388305664, + "learning_rate": 2.97e-07, + "loss": 8.7855, + "step": 101 + }, + { + "epoch": 0.08614864864864864, + "grad_norm": 36.78487777709961, + "learning_rate": 3.0000000000000004e-07, + "loss": 8.191, + "step": 102 + }, + { + "epoch": 0.08699324324324324, + "grad_norm": 36.979854583740234, + "learning_rate": 3.03e-07, + "loss": 8.0717, + "step": 103 + }, + { + "epoch": 0.08783783783783784, + "grad_norm": 35.79908752441406, + "learning_rate": 3.06e-07, + "loss": 7.5926, + "step": 104 + }, + { + "epoch": 0.08868243243243243, + "grad_norm": 29.094715118408203, + "learning_rate": 3.0900000000000003e-07, + "loss": 6.8309, + "step": 105 + }, + { + "epoch": 0.08952702702702703, + "grad_norm": 34.87013244628906, + "learning_rate": 3.12e-07, + "loss": 7.3975, + "step": 106 + }, + { + "epoch": 0.09037162162162163, + "grad_norm": 33.40941619873047, + "learning_rate": 3.15e-07, + "loss": 7.3015, + "step": 107 + }, + { + "epoch": 0.09121621621621621, + "grad_norm": 32.447322845458984, + "learning_rate": 3.18e-07, + "loss": 7.0735, + "step": 108 + }, + { + "epoch": 0.09206081081081081, + "grad_norm": 28.247825622558594, + "learning_rate": 3.21e-07, + "loss": 6.5874, + "step": 109 + }, + { + "epoch": 0.0929054054054054, + "grad_norm": 29.923877716064453, + "learning_rate": 3.2400000000000004e-07, + "loss": 6.6682, + "step": 110 + }, + { + "epoch": 0.09375, + "grad_norm": 31.19878387451172, + "learning_rate": 3.27e-07, + "loss": 6.8222, + "step": 111 + }, + { + "epoch": 0.0945945945945946, + "grad_norm": 35.79947280883789, + "learning_rate": 3.2999999999999996e-07, + "loss": 7.2365, + "step": 112 + }, + { + "epoch": 0.09543918918918919, + "grad_norm": 32.8036994934082, + "learning_rate": 3.3300000000000003e-07, + "loss": 6.7647, + "step": 113 + }, + { + "epoch": 0.09628378378378379, + "grad_norm": 32.139404296875, + "learning_rate": 3.36e-07, + "loss": 6.7093, + "step": 114 + }, + { + "epoch": 0.09712837837837837, + "grad_norm": 34.4754524230957, + "learning_rate": 3.39e-07, + "loss": 6.7646, + "step": 115 + }, + { + "epoch": 0.09797297297297297, + "grad_norm": 32.37584686279297, + "learning_rate": 3.42e-07, + "loss": 6.674, + "step": 116 + }, + { + "epoch": 0.09881756756756757, + "grad_norm": 35.467166900634766, + "learning_rate": 3.45e-07, + "loss": 6.971, + "step": 117 + }, + { + "epoch": 0.09966216216216216, + "grad_norm": 35.6771240234375, + "learning_rate": 3.48e-07, + "loss": 6.7502, + "step": 118 + }, + { + "epoch": 0.10050675675675676, + "grad_norm": 33.036224365234375, + "learning_rate": 3.51e-07, + "loss": 6.4815, + "step": 119 + }, + { + "epoch": 0.10135135135135136, + "grad_norm": 35.81736373901367, + "learning_rate": 3.54e-07, + "loss": 6.2469, + "step": 120 + }, + { + "epoch": 0.10219594594594594, + "grad_norm": 35.50452423095703, + "learning_rate": 3.5700000000000003e-07, + "loss": 6.6253, + "step": 121 + }, + { + "epoch": 0.10304054054054054, + "grad_norm": 32.284420013427734, + "learning_rate": 3.6e-07, + "loss": 6.1888, + "step": 122 + }, + { + "epoch": 0.10388513513513513, + "grad_norm": 37.04789733886719, + "learning_rate": 3.63e-07, + "loss": 6.549, + "step": 123 + }, + { + "epoch": 0.10472972972972973, + "grad_norm": 32.927799224853516, + "learning_rate": 3.66e-07, + "loss": 6.2005, + "step": 124 + }, + { + "epoch": 0.10557432432432433, + "grad_norm": 34.512516021728516, + "learning_rate": 3.6900000000000004e-07, + "loss": 6.3103, + "step": 125 + }, + { + "epoch": 0.10641891891891891, + "grad_norm": 37.607975006103516, + "learning_rate": 3.72e-07, + "loss": 6.3475, + "step": 126 + }, + { + "epoch": 0.10726351351351351, + "grad_norm": 38.9622917175293, + "learning_rate": 3.75e-07, + "loss": 6.447, + "step": 127 + }, + { + "epoch": 0.10810810810810811, + "grad_norm": 33.891563415527344, + "learning_rate": 3.78e-07, + "loss": 6.0602, + "step": 128 + }, + { + "epoch": 0.1089527027027027, + "grad_norm": 35.37641525268555, + "learning_rate": 3.81e-07, + "loss": 6.1296, + "step": 129 + }, + { + "epoch": 0.1097972972972973, + "grad_norm": 41.04951858520508, + "learning_rate": 3.8400000000000005e-07, + "loss": 6.4529, + "step": 130 + }, + { + "epoch": 0.11064189189189189, + "grad_norm": 34.485023498535156, + "learning_rate": 3.87e-07, + "loss": 5.9128, + "step": 131 + }, + { + "epoch": 0.11148648648648649, + "grad_norm": 34.580596923828125, + "learning_rate": 3.8999999999999997e-07, + "loss": 5.9016, + "step": 132 + }, + { + "epoch": 0.11233108108108109, + "grad_norm": 32.636627197265625, + "learning_rate": 3.9300000000000004e-07, + "loss": 5.6992, + "step": 133 + }, + { + "epoch": 0.11317567567567567, + "grad_norm": 38.833343505859375, + "learning_rate": 3.96e-07, + "loss": 6.1081, + "step": 134 + }, + { + "epoch": 0.11402027027027027, + "grad_norm": 33.795040130615234, + "learning_rate": 3.99e-07, + "loss": 5.6719, + "step": 135 + }, + { + "epoch": 0.11486486486486487, + "grad_norm": 36.63085174560547, + "learning_rate": 4.02e-07, + "loss": 5.803, + "step": 136 + }, + { + "epoch": 0.11570945945945946, + "grad_norm": 36.71419906616211, + "learning_rate": 4.05e-07, + "loss": 5.7576, + "step": 137 + }, + { + "epoch": 0.11655405405405406, + "grad_norm": 35.51446533203125, + "learning_rate": 4.08e-07, + "loss": 5.6938, + "step": 138 + }, + { + "epoch": 0.11739864864864864, + "grad_norm": 38.03469467163086, + "learning_rate": 4.11e-07, + "loss": 5.8003, + "step": 139 + }, + { + "epoch": 0.11824324324324324, + "grad_norm": 38.660438537597656, + "learning_rate": 4.1400000000000003e-07, + "loss": 5.7635, + "step": 140 + }, + { + "epoch": 0.11908783783783784, + "grad_norm": 40.9460334777832, + "learning_rate": 4.17e-07, + "loss": 5.8708, + "step": 141 + }, + { + "epoch": 0.11993243243243243, + "grad_norm": 36.32587432861328, + "learning_rate": 4.2e-07, + "loss": 5.5664, + "step": 142 + }, + { + "epoch": 0.12077702702702703, + "grad_norm": 33.58983612060547, + "learning_rate": 4.23e-07, + "loss": 5.3982, + "step": 143 + }, + { + "epoch": 0.12162162162162163, + "grad_norm": 35.66792297363281, + "learning_rate": 4.2600000000000003e-07, + "loss": 5.4498, + "step": 144 + }, + { + "epoch": 0.12246621621621621, + "grad_norm": 38.18095016479492, + "learning_rate": 4.2900000000000004e-07, + "loss": 5.496, + "step": 145 + }, + { + "epoch": 0.12331081081081081, + "grad_norm": 37.19709014892578, + "learning_rate": 4.32e-07, + "loss": 5.4009, + "step": 146 + }, + { + "epoch": 0.1241554054054054, + "grad_norm": 39.8029899597168, + "learning_rate": 4.35e-07, + "loss": 5.5473, + "step": 147 + }, + { + "epoch": 0.125, + "grad_norm": 44.81085968017578, + "learning_rate": 4.3800000000000003e-07, + "loss": 5.7568, + "step": 148 + }, + { + "epoch": 0.1258445945945946, + "grad_norm": 42.78243637084961, + "learning_rate": 4.41e-07, + "loss": 5.6571, + "step": 149 + }, + { + "epoch": 0.1266891891891892, + "grad_norm": 51.93869400024414, + "learning_rate": 4.44e-07, + "loss": 6.1946, + "step": 150 + }, + { + "epoch": 0.12753378378378377, + "grad_norm": 60.69337463378906, + "learning_rate": 4.47e-07, + "loss": 6.6398, + "step": 151 + }, + { + "epoch": 0.12837837837837837, + "grad_norm": 54.987709045410156, + "learning_rate": 4.5e-07, + "loss": 6.2874, + "step": 152 + }, + { + "epoch": 0.12922297297297297, + "grad_norm": 45.63478469848633, + "learning_rate": 4.5300000000000005e-07, + "loss": 5.6928, + "step": 153 + }, + { + "epoch": 0.13006756756756757, + "grad_norm": 42.55747985839844, + "learning_rate": 4.56e-07, + "loss": 5.4473, + "step": 154 + }, + { + "epoch": 0.13091216216216217, + "grad_norm": 43.81795883178711, + "learning_rate": 4.5899999999999997e-07, + "loss": 5.4391, + "step": 155 + }, + { + "epoch": 0.13175675675675674, + "grad_norm": 39.85747146606445, + "learning_rate": 4.6200000000000003e-07, + "loss": 5.2575, + "step": 156 + }, + { + "epoch": 0.13260135135135134, + "grad_norm": 35.404029846191406, + "learning_rate": 4.65e-07, + "loss": 5.0423, + "step": 157 + }, + { + "epoch": 0.13344594594594594, + "grad_norm": 35.49181365966797, + "learning_rate": 4.68e-07, + "loss": 5.0365, + "step": 158 + }, + { + "epoch": 0.13429054054054054, + "grad_norm": 39.000736236572266, + "learning_rate": 4.7099999999999997e-07, + "loss": 5.2002, + "step": 159 + }, + { + "epoch": 0.13513513513513514, + "grad_norm": 35.42692947387695, + "learning_rate": 4.7400000000000004e-07, + "loss": 5.0081, + "step": 160 + }, + { + "epoch": 0.13597972972972974, + "grad_norm": 33.154659271240234, + "learning_rate": 4.77e-07, + "loss": 4.9141, + "step": 161 + }, + { + "epoch": 0.13682432432432431, + "grad_norm": 30.762388229370117, + "learning_rate": 4.800000000000001e-07, + "loss": 4.7852, + "step": 162 + }, + { + "epoch": 0.13766891891891891, + "grad_norm": 29.567853927612305, + "learning_rate": 4.83e-07, + "loss": 4.7011, + "step": 163 + }, + { + "epoch": 0.13851351351351351, + "grad_norm": 31.12281036376953, + "learning_rate": 4.86e-07, + "loss": 4.7841, + "step": 164 + }, + { + "epoch": 0.13935810810810811, + "grad_norm": 30.417043685913086, + "learning_rate": 4.89e-07, + "loss": 4.7412, + "step": 165 + }, + { + "epoch": 0.14020270270270271, + "grad_norm": 24.96772003173828, + "learning_rate": 4.92e-07, + "loss": 4.5564, + "step": 166 + }, + { + "epoch": 0.14104729729729729, + "grad_norm": 26.463911056518555, + "learning_rate": 4.95e-07, + "loss": 4.5891, + "step": 167 + }, + { + "epoch": 0.14189189189189189, + "grad_norm": 23.757137298583984, + "learning_rate": 4.98e-07, + "loss": 4.4922, + "step": 168 + }, + { + "epoch": 0.14273648648648649, + "grad_norm": 23.5830135345459, + "learning_rate": 5.01e-07, + "loss": 4.4685, + "step": 169 + }, + { + "epoch": 0.14358108108108109, + "grad_norm": 23.51699447631836, + "learning_rate": 5.04e-07, + "loss": 4.4984, + "step": 170 + }, + { + "epoch": 0.14442567567567569, + "grad_norm": 22.159404754638672, + "learning_rate": 5.07e-07, + "loss": 4.4583, + "step": 171 + }, + { + "epoch": 0.14527027027027026, + "grad_norm": 18.180374145507812, + "learning_rate": 5.100000000000001e-07, + "loss": 4.334, + "step": 172 + }, + { + "epoch": 0.14611486486486486, + "grad_norm": 17.997875213623047, + "learning_rate": 5.13e-07, + "loss": 4.3539, + "step": 173 + }, + { + "epoch": 0.14695945945945946, + "grad_norm": 17.558578491210938, + "learning_rate": 5.16e-07, + "loss": 4.3295, + "step": 174 + }, + { + "epoch": 0.14780405405405406, + "grad_norm": 16.177507400512695, + "learning_rate": 5.19e-07, + "loss": 4.3254, + "step": 175 + }, + { + "epoch": 0.14864864864864866, + "grad_norm": 14.26958179473877, + "learning_rate": 5.219999999999999e-07, + "loss": 4.2765, + "step": 176 + }, + { + "epoch": 0.14949324324324326, + "grad_norm": 14.035073280334473, + "learning_rate": 5.250000000000001e-07, + "loss": 4.3131, + "step": 177 + }, + { + "epoch": 0.15033783783783783, + "grad_norm": 11.331958770751953, + "learning_rate": 5.280000000000001e-07, + "loss": 4.2241, + "step": 178 + }, + { + "epoch": 0.15118243243243243, + "grad_norm": 8.813743591308594, + "learning_rate": 5.31e-07, + "loss": 4.1411, + "step": 179 + }, + { + "epoch": 0.15202702702702703, + "grad_norm": 10.020484924316406, + "learning_rate": 5.34e-07, + "loss": 4.1893, + "step": 180 + }, + { + "epoch": 0.15287162162162163, + "grad_norm": 7.586029052734375, + "learning_rate": 5.37e-07, + "loss": 4.0924, + "step": 181 + }, + { + "epoch": 0.15371621621621623, + "grad_norm": 9.077047348022461, + "learning_rate": 5.4e-07, + "loss": 4.2087, + "step": 182 + }, + { + "epoch": 0.1545608108108108, + "grad_norm": 7.71276330947876, + "learning_rate": 5.43e-07, + "loss": 4.1599, + "step": 183 + }, + { + "epoch": 0.1554054054054054, + "grad_norm": 7.351337432861328, + "learning_rate": 5.46e-07, + "loss": 4.1191, + "step": 184 + }, + { + "epoch": 0.15625, + "grad_norm": 7.672425270080566, + "learning_rate": 5.490000000000001e-07, + "loss": 4.0739, + "step": 185 + }, + { + "epoch": 0.1570945945945946, + "grad_norm": 7.623657703399658, + "learning_rate": 5.52e-07, + "loss": 4.0805, + "step": 186 + }, + { + "epoch": 0.1579391891891892, + "grad_norm": 7.231611251831055, + "learning_rate": 5.55e-07, + "loss": 4.0966, + "step": 187 + }, + { + "epoch": 0.15878378378378377, + "grad_norm": 8.520071029663086, + "learning_rate": 5.58e-07, + "loss": 4.0319, + "step": 188 + }, + { + "epoch": 0.15962837837837837, + "grad_norm": 7.268913269042969, + "learning_rate": 5.61e-07, + "loss": 4.1552, + "step": 189 + }, + { + "epoch": 0.16047297297297297, + "grad_norm": 7.801273822784424, + "learning_rate": 5.64e-07, + "loss": 4.0636, + "step": 190 + }, + { + "epoch": 0.16131756756756757, + "grad_norm": 7.4708251953125, + "learning_rate": 5.67e-07, + "loss": 4.0826, + "step": 191 + }, + { + "epoch": 0.16216216216216217, + "grad_norm": 8.571698188781738, + "learning_rate": 5.7e-07, + "loss": 4.0333, + "step": 192 + }, + { + "epoch": 0.16300675675675674, + "grad_norm": 6.819480895996094, + "learning_rate": 5.73e-07, + "loss": 4.0671, + "step": 193 + }, + { + "epoch": 0.16385135135135134, + "grad_norm": 8.195098876953125, + "learning_rate": 5.76e-07, + "loss": 4.0351, + "step": 194 + }, + { + "epoch": 0.16469594594594594, + "grad_norm": 8.325255393981934, + "learning_rate": 5.790000000000001e-07, + "loss": 4.0069, + "step": 195 + }, + { + "epoch": 0.16554054054054054, + "grad_norm": 7.7631449699401855, + "learning_rate": 5.82e-07, + "loss": 3.9811, + "step": 196 + }, + { + "epoch": 0.16638513513513514, + "grad_norm": 6.879054546356201, + "learning_rate": 5.85e-07, + "loss": 4.0938, + "step": 197 + }, + { + "epoch": 0.16722972972972974, + "grad_norm": 7.688642978668213, + "learning_rate": 5.88e-07, + "loss": 4.1009, + "step": 198 + }, + { + "epoch": 0.16807432432432431, + "grad_norm": 6.741268157958984, + "learning_rate": 5.909999999999999e-07, + "loss": 4.0692, + "step": 199 + }, + { + "epoch": 0.16891891891891891, + "grad_norm": 7.883812427520752, + "learning_rate": 5.94e-07, + "loss": 4.1131, + "step": 200 + }, + { + "epoch": 0.16976351351351351, + "grad_norm": 14.186535835266113, + "learning_rate": 5.970000000000001e-07, + "loss": 4.16, + "step": 201 + }, + { + "epoch": 0.17060810810810811, + "grad_norm": 11.843823432922363, + "learning_rate": 6.000000000000001e-07, + "loss": 4.1283, + "step": 202 + }, + { + "epoch": 0.17145270270270271, + "grad_norm": 7.373616695404053, + "learning_rate": 6.03e-07, + "loss": 3.9636, + "step": 203 + }, + { + "epoch": 0.17229729729729729, + "grad_norm": 6.748037338256836, + "learning_rate": 6.06e-07, + "loss": 3.9425, + "step": 204 + }, + { + "epoch": 0.17314189189189189, + "grad_norm": 6.426863670349121, + "learning_rate": 6.09e-07, + "loss": 3.877, + "step": 205 + }, + { + "epoch": 0.17398648648648649, + "grad_norm": 6.953408241271973, + "learning_rate": 6.12e-07, + "loss": 3.8526, + "step": 206 + }, + { + "epoch": 0.17483108108108109, + "grad_norm": 6.968282222747803, + "learning_rate": 6.15e-07, + "loss": 3.8909, + "step": 207 + }, + { + "epoch": 0.17567567567567569, + "grad_norm": 6.598355293273926, + "learning_rate": 6.180000000000001e-07, + "loss": 3.844, + "step": 208 + }, + { + "epoch": 0.17652027027027026, + "grad_norm": 6.215937614440918, + "learning_rate": 6.21e-07, + "loss": 3.8543, + "step": 209 + }, + { + "epoch": 0.17736486486486486, + "grad_norm": 6.035060405731201, + "learning_rate": 6.24e-07, + "loss": 3.8398, + "step": 210 + }, + { + "epoch": 0.17820945945945946, + "grad_norm": 6.316421985626221, + "learning_rate": 6.27e-07, + "loss": 3.8124, + "step": 211 + }, + { + "epoch": 0.17905405405405406, + "grad_norm": 5.757258415222168, + "learning_rate": 6.3e-07, + "loss": 3.8174, + "step": 212 + }, + { + "epoch": 0.17989864864864866, + "grad_norm": 6.193333625793457, + "learning_rate": 6.33e-07, + "loss": 3.7969, + "step": 213 + }, + { + "epoch": 0.18074324324324326, + "grad_norm": 6.534773826599121, + "learning_rate": 6.36e-07, + "loss": 3.7408, + "step": 214 + }, + { + "epoch": 0.18158783783783783, + "grad_norm": 6.198360443115234, + "learning_rate": 6.39e-07, + "loss": 3.7362, + "step": 215 + }, + { + "epoch": 0.18243243243243243, + "grad_norm": 5.548851013183594, + "learning_rate": 6.42e-07, + "loss": 3.6941, + "step": 216 + }, + { + "epoch": 0.18327702702702703, + "grad_norm": 5.362044811248779, + "learning_rate": 6.45e-07, + "loss": 3.7173, + "step": 217 + }, + { + "epoch": 0.18412162162162163, + "grad_norm": 4.947911262512207, + "learning_rate": 6.480000000000001e-07, + "loss": 3.7536, + "step": 218 + }, + { + "epoch": 0.18496621621621623, + "grad_norm": 5.195919513702393, + "learning_rate": 6.51e-07, + "loss": 3.7011, + "step": 219 + }, + { + "epoch": 0.1858108108108108, + "grad_norm": 5.140451908111572, + "learning_rate": 6.54e-07, + "loss": 3.6381, + "step": 220 + }, + { + "epoch": 0.1866554054054054, + "grad_norm": 6.985661506652832, + "learning_rate": 6.57e-07, + "loss": 3.6498, + "step": 221 + }, + { + "epoch": 0.1875, + "grad_norm": 4.790067672729492, + "learning_rate": 6.599999999999999e-07, + "loss": 3.6777, + "step": 222 + }, + { + "epoch": 0.1883445945945946, + "grad_norm": 6.5684380531311035, + "learning_rate": 6.63e-07, + "loss": 3.6726, + "step": 223 + }, + { + "epoch": 0.1891891891891892, + "grad_norm": 4.633362770080566, + "learning_rate": 6.660000000000001e-07, + "loss": 3.6131, + "step": 224 + }, + { + "epoch": 0.19003378378378377, + "grad_norm": 4.487237930297852, + "learning_rate": 6.690000000000001e-07, + "loss": 3.6082, + "step": 225 + }, + { + "epoch": 0.19087837837837837, + "grad_norm": 4.568017482757568, + "learning_rate": 6.72e-07, + "loss": 3.5599, + "step": 226 + }, + { + "epoch": 0.19172297297297297, + "grad_norm": 5.269558906555176, + "learning_rate": 6.75e-07, + "loss": 3.5888, + "step": 227 + }, + { + "epoch": 0.19256756756756757, + "grad_norm": 4.506144046783447, + "learning_rate": 6.78e-07, + "loss": 3.5736, + "step": 228 + }, + { + "epoch": 0.19341216216216217, + "grad_norm": 6.557576656341553, + "learning_rate": 6.81e-07, + "loss": 3.5978, + "step": 229 + }, + { + "epoch": 0.19425675675675674, + "grad_norm": 4.258693218231201, + "learning_rate": 6.84e-07, + "loss": 3.5083, + "step": 230 + }, + { + "epoch": 0.19510135135135134, + "grad_norm": 4.073233127593994, + "learning_rate": 6.87e-07, + "loss": 3.5602, + "step": 231 + }, + { + "epoch": 0.19594594594594594, + "grad_norm": 4.718168258666992, + "learning_rate": 6.9e-07, + "loss": 3.577, + "step": 232 + }, + { + "epoch": 0.19679054054054054, + "grad_norm": 8.27757740020752, + "learning_rate": 6.93e-07, + "loss": 3.4903, + "step": 233 + }, + { + "epoch": 0.19763513513513514, + "grad_norm": 3.956329345703125, + "learning_rate": 6.96e-07, + "loss": 3.541, + "step": 234 + }, + { + "epoch": 0.19847972972972974, + "grad_norm": 5.696030616760254, + "learning_rate": 6.990000000000001e-07, + "loss": 3.4747, + "step": 235 + }, + { + "epoch": 0.19932432432432431, + "grad_norm": 4.120911598205566, + "learning_rate": 7.02e-07, + "loss": 3.4515, + "step": 236 + }, + { + "epoch": 0.20016891891891891, + "grad_norm": 7.624979019165039, + "learning_rate": 7.05e-07, + "loss": 3.4743, + "step": 237 + }, + { + "epoch": 0.20101351351351351, + "grad_norm": 4.508566856384277, + "learning_rate": 7.08e-07, + "loss": 3.4977, + "step": 238 + }, + { + "epoch": 0.20185810810810811, + "grad_norm": 4.473377227783203, + "learning_rate": 7.11e-07, + "loss": 3.4908, + "step": 239 + }, + { + "epoch": 0.20270270270270271, + "grad_norm": 3.6259267330169678, + "learning_rate": 7.140000000000001e-07, + "loss": 3.4485, + "step": 240 + }, + { + "epoch": 0.20354729729729729, + "grad_norm": 6.081716537475586, + "learning_rate": 7.170000000000001e-07, + "loss": 3.4149, + "step": 241 + }, + { + "epoch": 0.20439189189189189, + "grad_norm": 4.095852375030518, + "learning_rate": 7.2e-07, + "loss": 3.4401, + "step": 242 + }, + { + "epoch": 0.20523648648648649, + "grad_norm": 4.885341644287109, + "learning_rate": 7.23e-07, + "loss": 3.4414, + "step": 243 + }, + { + "epoch": 0.20608108108108109, + "grad_norm": 5.430721759796143, + "learning_rate": 7.26e-07, + "loss": 3.4949, + "step": 244 + }, + { + "epoch": 0.20692567567567569, + "grad_norm": 4.539923191070557, + "learning_rate": 7.29e-07, + "loss": 3.4289, + "step": 245 + }, + { + "epoch": 0.20777027027027026, + "grad_norm": 3.8432259559631348, + "learning_rate": 7.32e-07, + "loss": 3.3665, + "step": 246 + }, + { + "epoch": 0.20861486486486486, + "grad_norm": 3.643447160720825, + "learning_rate": 7.350000000000001e-07, + "loss": 3.483, + "step": 247 + }, + { + "epoch": 0.20945945945945946, + "grad_norm": 3.4876132011413574, + "learning_rate": 7.380000000000001e-07, + "loss": 3.4032, + "step": 248 + }, + { + "epoch": 0.21030405405405406, + "grad_norm": 4.948246002197266, + "learning_rate": 7.41e-07, + "loss": 3.4731, + "step": 249 + }, + { + "epoch": 0.21114864864864866, + "grad_norm": 4.664821147918701, + "learning_rate": 7.44e-07, + "loss": 3.5363, + "step": 250 + }, + { + "epoch": 0.21199324324324326, + "grad_norm": 12.31495189666748, + "learning_rate": 7.47e-07, + "loss": 3.5478, + "step": 251 + }, + { + "epoch": 0.21283783783783783, + "grad_norm": 9.681405067443848, + "learning_rate": 7.5e-07, + "loss": 3.5528, + "step": 252 + }, + { + "epoch": 0.21368243243243243, + "grad_norm": 3.5735626220703125, + "learning_rate": 7.53e-07, + "loss": 3.3519, + "step": 253 + }, + { + "epoch": 0.21452702702702703, + "grad_norm": 3.4453413486480713, + "learning_rate": 7.56e-07, + "loss": 3.3211, + "step": 254 + }, + { + "epoch": 0.21537162162162163, + "grad_norm": 3.443364381790161, + "learning_rate": 7.59e-07, + "loss": 3.3693, + "step": 255 + }, + { + "epoch": 0.21621621621621623, + "grad_norm": 4.0599822998046875, + "learning_rate": 7.62e-07, + "loss": 3.3548, + "step": 256 + }, + { + "epoch": 0.2170608108108108, + "grad_norm": 5.010953903198242, + "learning_rate": 7.65e-07, + "loss": 3.3346, + "step": 257 + }, + { + "epoch": 0.2179054054054054, + "grad_norm": 3.628857374191284, + "learning_rate": 7.680000000000001e-07, + "loss": 3.2759, + "step": 258 + }, + { + "epoch": 0.21875, + "grad_norm": 3.1029486656188965, + "learning_rate": 7.71e-07, + "loss": 3.3162, + "step": 259 + }, + { + "epoch": 0.2195945945945946, + "grad_norm": 3.687361717224121, + "learning_rate": 7.74e-07, + "loss": 3.3186, + "step": 260 + }, + { + "epoch": 0.2204391891891892, + "grad_norm": 4.240677833557129, + "learning_rate": 7.77e-07, + "loss": 3.2374, + "step": 261 + }, + { + "epoch": 0.22128378378378377, + "grad_norm": 3.154611587524414, + "learning_rate": 7.799999999999999e-07, + "loss": 3.2563, + "step": 262 + }, + { + "epoch": 0.22212837837837837, + "grad_norm": 4.333796501159668, + "learning_rate": 7.830000000000001e-07, + "loss": 3.1992, + "step": 263 + }, + { + "epoch": 0.22297297297297297, + "grad_norm": 4.110022068023682, + "learning_rate": 7.860000000000001e-07, + "loss": 3.1996, + "step": 264 + }, + { + "epoch": 0.22381756756756757, + "grad_norm": 2.868145704269409, + "learning_rate": 7.89e-07, + "loss": 3.2386, + "step": 265 + }, + { + "epoch": 0.22466216216216217, + "grad_norm": 2.282886505126953, + "learning_rate": 7.92e-07, + "loss": 3.2283, + "step": 266 + }, + { + "epoch": 0.22550675675675674, + "grad_norm": 2.89989972114563, + "learning_rate": 7.95e-07, + "loss": 3.2682, + "step": 267 + }, + { + "epoch": 0.22635135135135134, + "grad_norm": 4.6485114097595215, + "learning_rate": 7.98e-07, + "loss": 3.199, + "step": 268 + }, + { + "epoch": 0.22719594594594594, + "grad_norm": 2.506139039993286, + "learning_rate": 8.01e-07, + "loss": 3.1895, + "step": 269 + }, + { + "epoch": 0.22804054054054054, + "grad_norm": 2.809164047241211, + "learning_rate": 8.04e-07, + "loss": 3.2107, + "step": 270 + }, + { + "epoch": 0.22888513513513514, + "grad_norm": 3.0435609817504883, + "learning_rate": 8.070000000000001e-07, + "loss": 3.1991, + "step": 271 + }, + { + "epoch": 0.22972972972972974, + "grad_norm": 4.884791851043701, + "learning_rate": 8.1e-07, + "loss": 3.1651, + "step": 272 + }, + { + "epoch": 0.23057432432432431, + "grad_norm": 5.473139762878418, + "learning_rate": 8.13e-07, + "loss": 3.1995, + "step": 273 + }, + { + "epoch": 0.23141891891891891, + "grad_norm": 2.719464063644409, + "learning_rate": 8.16e-07, + "loss": 3.1576, + "step": 274 + }, + { + "epoch": 0.23226351351351351, + "grad_norm": 2.8822715282440186, + "learning_rate": 8.19e-07, + "loss": 3.1123, + "step": 275 + }, + { + "epoch": 0.23310810810810811, + "grad_norm": 4.860684394836426, + "learning_rate": 8.22e-07, + "loss": 3.2042, + "step": 276 + }, + { + "epoch": 0.23395270270270271, + "grad_norm": 3.2784409523010254, + "learning_rate": 8.25e-07, + "loss": 3.1347, + "step": 277 + }, + { + "epoch": 0.23479729729729729, + "grad_norm": 4.444830417633057, + "learning_rate": 8.280000000000001e-07, + "loss": 3.123, + "step": 278 + }, + { + "epoch": 0.23564189189189189, + "grad_norm": 5.739529609680176, + "learning_rate": 8.31e-07, + "loss": 3.2222, + "step": 279 + }, + { + "epoch": 0.23648648648648649, + "grad_norm": 3.4318301677703857, + "learning_rate": 8.34e-07, + "loss": 3.1219, + "step": 280 + }, + { + "epoch": 0.23733108108108109, + "grad_norm": 2.624128580093384, + "learning_rate": 8.370000000000001e-07, + "loss": 3.1389, + "step": 281 + }, + { + "epoch": 0.23817567567567569, + "grad_norm": 6.095109462738037, + "learning_rate": 8.4e-07, + "loss": 3.1004, + "step": 282 + }, + { + "epoch": 0.23902027027027026, + "grad_norm": 12.982423782348633, + "learning_rate": 8.43e-07, + "loss": 3.1456, + "step": 283 + }, + { + "epoch": 0.23986486486486486, + "grad_norm": 2.982940196990967, + "learning_rate": 8.46e-07, + "loss": 3.1224, + "step": 284 + }, + { + "epoch": 0.24070945945945946, + "grad_norm": 2.8266680240631104, + "learning_rate": 8.489999999999999e-07, + "loss": 3.1161, + "step": 285 + }, + { + "epoch": 0.24155405405405406, + "grad_norm": 4.7769622802734375, + "learning_rate": 8.520000000000001e-07, + "loss": 3.1198, + "step": 286 + }, + { + "epoch": 0.24239864864864866, + "grad_norm": 6.072540283203125, + "learning_rate": 8.550000000000001e-07, + "loss": 3.1089, + "step": 287 + }, + { + "epoch": 0.24324324324324326, + "grad_norm": 2.5155792236328125, + "learning_rate": 8.580000000000001e-07, + "loss": 3.142, + "step": 288 + }, + { + "epoch": 0.24408783783783783, + "grad_norm": 2.7020747661590576, + "learning_rate": 8.61e-07, + "loss": 3.0921, + "step": 289 + }, + { + "epoch": 0.24493243243243243, + "grad_norm": 5.426540851593018, + "learning_rate": 8.64e-07, + "loss": 3.1057, + "step": 290 + }, + { + "epoch": 0.24577702702702703, + "grad_norm": 3.0407447814941406, + "learning_rate": 8.67e-07, + "loss": 3.0874, + "step": 291 + }, + { + "epoch": 0.24662162162162163, + "grad_norm": 3.6032521724700928, + "learning_rate": 8.7e-07, + "loss": 3.0925, + "step": 292 + }, + { + "epoch": 0.24746621621621623, + "grad_norm": 5.548582553863525, + "learning_rate": 8.73e-07, + "loss": 3.1212, + "step": 293 + }, + { + "epoch": 0.2483108108108108, + "grad_norm": 3.741297483444214, + "learning_rate": 8.760000000000001e-07, + "loss": 3.1169, + "step": 294 + }, + { + "epoch": 0.2491554054054054, + "grad_norm": 6.661920070648193, + "learning_rate": 8.79e-07, + "loss": 3.0641, + "step": 295 + }, + { + "epoch": 0.25, + "grad_norm": 30.23625373840332, + "learning_rate": 8.82e-07, + "loss": 3.103, + "step": 296 + }, + { + "epoch": 0.25084459459459457, + "grad_norm": 4.753331661224365, + "learning_rate": 8.85e-07, + "loss": 3.0638, + "step": 297 + }, + { + "epoch": 0.2516891891891892, + "grad_norm": 10.742020606994629, + "learning_rate": 8.88e-07, + "loss": 3.0928, + "step": 298 + }, + { + "epoch": 0.25253378378378377, + "grad_norm": 5.0832414627075195, + "learning_rate": 8.91e-07, + "loss": 3.0714, + "step": 299 + }, + { + "epoch": 0.2533783783783784, + "grad_norm": 5.307823657989502, + "learning_rate": 8.94e-07, + "loss": 3.1193, + "step": 300 + }, + { + "epoch": 0.25422297297297297, + "grad_norm": 12.649903297424316, + "learning_rate": 8.97e-07, + "loss": 3.2067, + "step": 301 + }, + { + "epoch": 0.25506756756756754, + "grad_norm": 10.827836036682129, + "learning_rate": 9e-07, + "loss": 3.1884, + "step": 302 + }, + { + "epoch": 0.25591216216216217, + "grad_norm": 7.382022380828857, + "learning_rate": 9.03e-07, + "loss": 3.0506, + "step": 303 + }, + { + "epoch": 0.25675675675675674, + "grad_norm": 7.186032295227051, + "learning_rate": 9.060000000000001e-07, + "loss": 3.0408, + "step": 304 + }, + { + "epoch": 0.25760135135135137, + "grad_norm": 3.7775020599365234, + "learning_rate": 9.09e-07, + "loss": 3.0022, + "step": 305 + }, + { + "epoch": 0.25844594594594594, + "grad_norm": 3.4667515754699707, + "learning_rate": 9.12e-07, + "loss": 3.0028, + "step": 306 + }, + { + "epoch": 0.2592905405405405, + "grad_norm": 3.7518577575683594, + "learning_rate": 9.15e-07, + "loss": 3.0664, + "step": 307 + }, + { + "epoch": 0.26013513513513514, + "grad_norm": 3.438864231109619, + "learning_rate": 9.179999999999999e-07, + "loss": 3.0063, + "step": 308 + }, + { + "epoch": 0.2609797297297297, + "grad_norm": 5.039534568786621, + "learning_rate": 9.210000000000001e-07, + "loss": 3.0349, + "step": 309 + }, + { + "epoch": 0.26182432432432434, + "grad_norm": 3.365210771560669, + "learning_rate": 9.240000000000001e-07, + "loss": 3.0094, + "step": 310 + }, + { + "epoch": 0.2626689189189189, + "grad_norm": 12.429679870605469, + "learning_rate": 9.270000000000001e-07, + "loss": 3.0232, + "step": 311 + }, + { + "epoch": 0.2635135135135135, + "grad_norm": 8.666618347167969, + "learning_rate": 9.3e-07, + "loss": 3.0373, + "step": 312 + }, + { + "epoch": 0.2643581081081081, + "grad_norm": 3.9830923080444336, + "learning_rate": 9.33e-07, + "loss": 2.9989, + "step": 313 + }, + { + "epoch": 0.2652027027027027, + "grad_norm": 5.235404968261719, + "learning_rate": 9.36e-07, + "loss": 2.9844, + "step": 314 + }, + { + "epoch": 0.2660472972972973, + "grad_norm": 4.292595863342285, + "learning_rate": 9.39e-07, + "loss": 3.0039, + "step": 315 + }, + { + "epoch": 0.2668918918918919, + "grad_norm": 3.370432138442993, + "learning_rate": 9.419999999999999e-07, + "loss": 2.9844, + "step": 316 + }, + { + "epoch": 0.2677364864864865, + "grad_norm": 4.029133319854736, + "learning_rate": 9.450000000000001e-07, + "loss": 2.9931, + "step": 317 + }, + { + "epoch": 0.2685810810810811, + "grad_norm": 4.292975902557373, + "learning_rate": 9.480000000000001e-07, + "loss": 2.9878, + "step": 318 + }, + { + "epoch": 0.26942567567567566, + "grad_norm": 22.067821502685547, + "learning_rate": 9.51e-07, + "loss": 3.0038, + "step": 319 + }, + { + "epoch": 0.2702702702702703, + "grad_norm": 4.961012840270996, + "learning_rate": 9.54e-07, + "loss": 3.0059, + "step": 320 + }, + { + "epoch": 0.27111486486486486, + "grad_norm": 3.0115411281585693, + "learning_rate": 9.57e-07, + "loss": 3.0179, + "step": 321 + }, + { + "epoch": 0.2719594594594595, + "grad_norm": 4.749654293060303, + "learning_rate": 9.600000000000001e-07, + "loss": 2.9885, + "step": 322 + }, + { + "epoch": 0.27280405405405406, + "grad_norm": 3.6144912242889404, + "learning_rate": 9.63e-07, + "loss": 3.0281, + "step": 323 + }, + { + "epoch": 0.27364864864864863, + "grad_norm": 2.7031500339508057, + "learning_rate": 9.66e-07, + "loss": 2.9912, + "step": 324 + }, + { + "epoch": 0.27449324324324326, + "grad_norm": 3.159846067428589, + "learning_rate": 9.690000000000002e-07, + "loss": 3.0011, + "step": 325 + }, + { + "epoch": 0.27533783783783783, + "grad_norm": 5.864487171173096, + "learning_rate": 9.72e-07, + "loss": 2.9705, + "step": 326 + }, + { + "epoch": 0.27618243243243246, + "grad_norm": 2.543203592300415, + "learning_rate": 9.75e-07, + "loss": 2.9966, + "step": 327 + }, + { + "epoch": 0.27702702702702703, + "grad_norm": 4.886862277984619, + "learning_rate": 9.78e-07, + "loss": 2.9799, + "step": 328 + }, + { + "epoch": 0.2778716216216216, + "grad_norm": 3.160799264907837, + "learning_rate": 9.81e-07, + "loss": 2.979, + "step": 329 + }, + { + "epoch": 0.27871621621621623, + "grad_norm": 14.417391777038574, + "learning_rate": 9.84e-07, + "loss": 2.9562, + "step": 330 + }, + { + "epoch": 0.2795608108108108, + "grad_norm": 2.8676137924194336, + "learning_rate": 9.87e-07, + "loss": 2.982, + "step": 331 + }, + { + "epoch": 0.28040540540540543, + "grad_norm": 3.9169321060180664, + "learning_rate": 9.9e-07, + "loss": 2.9725, + "step": 332 + }, + { + "epoch": 0.28125, + "grad_norm": 5.00766134262085, + "learning_rate": 9.929999999999999e-07, + "loss": 3.0085, + "step": 333 + }, + { + "epoch": 0.28209459459459457, + "grad_norm": 5.7545318603515625, + "learning_rate": 9.96e-07, + "loss": 2.9841, + "step": 334 + }, + { + "epoch": 0.2829391891891892, + "grad_norm": 4.9501261711120605, + "learning_rate": 9.99e-07, + "loss": 2.9724, + "step": 335 + }, + { + "epoch": 0.28378378378378377, + "grad_norm": 4.723694801330566, + "learning_rate": 1.002e-06, + "loss": 2.9731, + "step": 336 + }, + { + "epoch": 0.2846283783783784, + "grad_norm": 4.284634113311768, + "learning_rate": 1.0050000000000001e-06, + "loss": 2.9762, + "step": 337 + }, + { + "epoch": 0.28547297297297297, + "grad_norm": 23.19192123413086, + "learning_rate": 1.008e-06, + "loss": 2.9554, + "step": 338 + }, + { + "epoch": 0.28631756756756754, + "grad_norm": 6.5551958084106445, + "learning_rate": 1.0110000000000001e-06, + "loss": 2.9514, + "step": 339 + }, + { + "epoch": 0.28716216216216217, + "grad_norm": 18.30768585205078, + "learning_rate": 1.014e-06, + "loss": 2.9849, + "step": 340 + }, + { + "epoch": 0.28800675675675674, + "grad_norm": 6.836459636688232, + "learning_rate": 1.017e-06, + "loss": 2.9895, + "step": 341 + }, + { + "epoch": 0.28885135135135137, + "grad_norm": 22.666933059692383, + "learning_rate": 1.0200000000000002e-06, + "loss": 2.9828, + "step": 342 + }, + { + "epoch": 0.28969594594594594, + "grad_norm": 3.5748236179351807, + "learning_rate": 1.023e-06, + "loss": 2.9632, + "step": 343 + }, + { + "epoch": 0.2905405405405405, + "grad_norm": 12.486822128295898, + "learning_rate": 1.026e-06, + "loss": 2.9741, + "step": 344 + }, + { + "epoch": 0.29138513513513514, + "grad_norm": 7.287432670593262, + "learning_rate": 1.029e-06, + "loss": 2.987, + "step": 345 + }, + { + "epoch": 0.2922297297297297, + "grad_norm": 10.717141151428223, + "learning_rate": 1.032e-06, + "loss": 2.9929, + "step": 346 + }, + { + "epoch": 0.29307432432432434, + "grad_norm": 8.657833099365234, + "learning_rate": 1.035e-06, + "loss": 2.9835, + "step": 347 + }, + { + "epoch": 0.2939189189189189, + "grad_norm": 6.957840919494629, + "learning_rate": 1.038e-06, + "loss": 2.9719, + "step": 348 + }, + { + "epoch": 0.2947635135135135, + "grad_norm": 30.359800338745117, + "learning_rate": 1.041e-06, + "loss": 3.0106, + "step": 349 + }, + { + "epoch": 0.2956081081081081, + "grad_norm": 30.25848388671875, + "learning_rate": 1.0439999999999999e-06, + "loss": 2.9319, + "step": 350 + }, + { + "epoch": 0.2964527027027027, + "grad_norm": 10.868090629577637, + "learning_rate": 1.047e-06, + "loss": 3.0504, + "step": 351 + }, + { + "epoch": 0.2972972972972973, + "grad_norm": 12.999271392822266, + "learning_rate": 1.0500000000000001e-06, + "loss": 2.9922, + "step": 352 + }, + { + "epoch": 0.2981418918918919, + "grad_norm": 8.730743408203125, + "learning_rate": 1.053e-06, + "loss": 2.9739, + "step": 353 + }, + { + "epoch": 0.2989864864864865, + "grad_norm": 10.077520370483398, + "learning_rate": 1.0560000000000001e-06, + "loss": 2.9777, + "step": 354 + }, + { + "epoch": 0.2998310810810811, + "grad_norm": 7.2232441902160645, + "learning_rate": 1.059e-06, + "loss": 2.9517, + "step": 355 + }, + { + "epoch": 0.30067567567567566, + "grad_norm": 5.6169304847717285, + "learning_rate": 1.062e-06, + "loss": 2.9595, + "step": 356 + }, + { + "epoch": 0.3015202702702703, + "grad_norm": 3.754972457885742, + "learning_rate": 1.065e-06, + "loss": 2.9394, + "step": 357 + }, + { + "epoch": 0.30236486486486486, + "grad_norm": 3.4870903491973877, + "learning_rate": 1.068e-06, + "loss": 2.9372, + "step": 358 + }, + { + "epoch": 0.3032094594594595, + "grad_norm": 5.308501720428467, + "learning_rate": 1.0710000000000002e-06, + "loss": 2.9514, + "step": 359 + }, + { + "epoch": 0.30405405405405406, + "grad_norm": 7.2072343826293945, + "learning_rate": 1.074e-06, + "loss": 2.9341, + "step": 360 + }, + { + "epoch": 0.30489864864864863, + "grad_norm": 4.556038856506348, + "learning_rate": 1.077e-06, + "loss": 2.9452, + "step": 361 + }, + { + "epoch": 0.30574324324324326, + "grad_norm": 5.287181854248047, + "learning_rate": 1.08e-06, + "loss": 2.9503, + "step": 362 + }, + { + "epoch": 0.30658783783783783, + "grad_norm": 3.2110817432403564, + "learning_rate": 1.083e-06, + "loss": 2.8918, + "step": 363 + }, + { + "epoch": 0.30743243243243246, + "grad_norm": 5.052977085113525, + "learning_rate": 1.086e-06, + "loss": 2.9495, + "step": 364 + }, + { + "epoch": 0.30827702702702703, + "grad_norm": 6.079784870147705, + "learning_rate": 1.089e-06, + "loss": 2.9078, + "step": 365 + }, + { + "epoch": 0.3091216216216216, + "grad_norm": 6.086483955383301, + "learning_rate": 1.092e-06, + "loss": 2.912, + "step": 366 + }, + { + "epoch": 0.30996621621621623, + "grad_norm": 2.289714813232422, + "learning_rate": 1.0949999999999999e-06, + "loss": 2.9249, + "step": 367 + }, + { + "epoch": 0.3108108108108108, + "grad_norm": 24.835132598876953, + "learning_rate": 1.0980000000000001e-06, + "loss": 2.9094, + "step": 368 + }, + { + "epoch": 0.31165540540540543, + "grad_norm": 5.982048511505127, + "learning_rate": 1.1010000000000001e-06, + "loss": 2.9061, + "step": 369 + }, + { + "epoch": 0.3125, + "grad_norm": 4.019280433654785, + "learning_rate": 1.104e-06, + "loss": 2.9385, + "step": 370 + }, + { + "epoch": 0.31334459459459457, + "grad_norm": 5.862982273101807, + "learning_rate": 1.1070000000000002e-06, + "loss": 2.9068, + "step": 371 + }, + { + "epoch": 0.3141891891891892, + "grad_norm": 3.3039450645446777, + "learning_rate": 1.11e-06, + "loss": 2.9054, + "step": 372 + }, + { + "epoch": 0.31503378378378377, + "grad_norm": 7.7915754318237305, + "learning_rate": 1.113e-06, + "loss": 2.9183, + "step": 373 + }, + { + "epoch": 0.3158783783783784, + "grad_norm": 4.3896307945251465, + "learning_rate": 1.116e-06, + "loss": 2.8848, + "step": 374 + }, + { + "epoch": 0.31672297297297297, + "grad_norm": 7.42189359664917, + "learning_rate": 1.119e-06, + "loss": 2.9385, + "step": 375 + }, + { + "epoch": 0.31756756756756754, + "grad_norm": 8.95450496673584, + "learning_rate": 1.122e-06, + "loss": 2.8614, + "step": 376 + }, + { + "epoch": 0.31841216216216217, + "grad_norm": 9.680588722229004, + "learning_rate": 1.125e-06, + "loss": 2.9289, + "step": 377 + }, + { + "epoch": 0.31925675675675674, + "grad_norm": 6.833977222442627, + "learning_rate": 1.128e-06, + "loss": 2.9146, + "step": 378 + }, + { + "epoch": 0.32010135135135137, + "grad_norm": 4.731185436248779, + "learning_rate": 1.131e-06, + "loss": 2.8875, + "step": 379 + }, + { + "epoch": 0.32094594594594594, + "grad_norm": 2.581483840942383, + "learning_rate": 1.134e-06, + "loss": 2.8992, + "step": 380 + }, + { + "epoch": 0.3217905405405405, + "grad_norm": 4.493574619293213, + "learning_rate": 1.137e-06, + "loss": 2.9215, + "step": 381 + }, + { + "epoch": 0.32263513513513514, + "grad_norm": 6.431309223175049, + "learning_rate": 1.14e-06, + "loss": 2.9385, + "step": 382 + }, + { + "epoch": 0.3234797297297297, + "grad_norm": 7.152055263519287, + "learning_rate": 1.1430000000000001e-06, + "loss": 2.9184, + "step": 383 + }, + { + "epoch": 0.32432432432432434, + "grad_norm": 16.45577049255371, + "learning_rate": 1.146e-06, + "loss": 2.9203, + "step": 384 + }, + { + "epoch": 0.3251689189189189, + "grad_norm": 8.75451374053955, + "learning_rate": 1.1490000000000001e-06, + "loss": 2.9314, + "step": 385 + }, + { + "epoch": 0.3260135135135135, + "grad_norm": 6.141480922698975, + "learning_rate": 1.152e-06, + "loss": 2.8966, + "step": 386 + }, + { + "epoch": 0.3268581081081081, + "grad_norm": 8.34850788116455, + "learning_rate": 1.155e-06, + "loss": 2.8784, + "step": 387 + }, + { + "epoch": 0.3277027027027027, + "grad_norm": 8.640499114990234, + "learning_rate": 1.1580000000000002e-06, + "loss": 2.9112, + "step": 388 + }, + { + "epoch": 0.3285472972972973, + "grad_norm": 4.203456401824951, + "learning_rate": 1.161e-06, + "loss": 2.9119, + "step": 389 + }, + { + "epoch": 0.3293918918918919, + "grad_norm": 5.857361316680908, + "learning_rate": 1.164e-06, + "loss": 2.885, + "step": 390 + }, + { + "epoch": 0.3302364864864865, + "grad_norm": 7.505223274230957, + "learning_rate": 1.167e-06, + "loss": 2.8814, + "step": 391 + }, + { + "epoch": 0.3310810810810811, + "grad_norm": 5.490025043487549, + "learning_rate": 1.17e-06, + "loss": 2.9219, + "step": 392 + }, + { + "epoch": 0.33192567567567566, + "grad_norm": 9.197518348693848, + "learning_rate": 1.173e-06, + "loss": 2.9052, + "step": 393 + }, + { + "epoch": 0.3327702702702703, + "grad_norm": 16.83067512512207, + "learning_rate": 1.176e-06, + "loss": 2.9016, + "step": 394 + }, + { + "epoch": 0.33361486486486486, + "grad_norm": 9.640503883361816, + "learning_rate": 1.179e-06, + "loss": 2.9276, + "step": 395 + }, + { + "epoch": 0.3344594594594595, + "grad_norm": 10.660700798034668, + "learning_rate": 1.1819999999999999e-06, + "loss": 2.9061, + "step": 396 + }, + { + "epoch": 0.33530405405405406, + "grad_norm": 28.515207290649414, + "learning_rate": 1.185e-06, + "loss": 2.8886, + "step": 397 + }, + { + "epoch": 0.33614864864864863, + "grad_norm": 3.2154574394226074, + "learning_rate": 1.188e-06, + "loss": 2.9224, + "step": 398 + }, + { + "epoch": 0.33699324324324326, + "grad_norm": 6.231912612915039, + "learning_rate": 1.191e-06, + "loss": 2.9111, + "step": 399 + }, + { + "epoch": 0.33783783783783783, + "grad_norm": 8.636836051940918, + "learning_rate": 1.1940000000000001e-06, + "loss": 3.0012, + "step": 400 + }, + { + "epoch": 0.33868243243243246, + "grad_norm": 13.399665832519531, + "learning_rate": 1.197e-06, + "loss": 3.0295, + "step": 401 + }, + { + "epoch": 0.33952702702702703, + "grad_norm": 5.762270927429199, + "learning_rate": 1.2000000000000002e-06, + "loss": 2.9282, + "step": 402 + }, + { + "epoch": 0.3403716216216216, + "grad_norm": 4.649792194366455, + "learning_rate": 1.203e-06, + "loss": 2.8744, + "step": 403 + }, + { + "epoch": 0.34121621621621623, + "grad_norm": 7.4348273277282715, + "learning_rate": 1.206e-06, + "loss": 2.8617, + "step": 404 + }, + { + "epoch": 0.3420608108108108, + "grad_norm": 8.782305717468262, + "learning_rate": 1.2090000000000002e-06, + "loss": 2.8736, + "step": 405 + }, + { + "epoch": 0.34290540540540543, + "grad_norm": 13.771907806396484, + "learning_rate": 1.212e-06, + "loss": 2.863, + "step": 406 + }, + { + "epoch": 0.34375, + "grad_norm": 5.662527561187744, + "learning_rate": 1.215e-06, + "loss": 2.8971, + "step": 407 + }, + { + "epoch": 0.34459459459459457, + "grad_norm": 8.083502769470215, + "learning_rate": 1.218e-06, + "loss": 2.8749, + "step": 408 + }, + { + "epoch": 0.3454391891891892, + "grad_norm": 6.402782440185547, + "learning_rate": 1.221e-06, + "loss": 2.8999, + "step": 409 + }, + { + "epoch": 0.34628378378378377, + "grad_norm": 5.754297733306885, + "learning_rate": 1.224e-06, + "loss": 2.8733, + "step": 410 + }, + { + "epoch": 0.3471283783783784, + "grad_norm": 3.4721643924713135, + "learning_rate": 1.227e-06, + "loss": 2.8552, + "step": 411 + }, + { + "epoch": 0.34797297297297297, + "grad_norm": 2.891150951385498, + "learning_rate": 1.23e-06, + "loss": 2.8883, + "step": 412 + }, + { + "epoch": 0.34881756756756754, + "grad_norm": 8.8919677734375, + "learning_rate": 1.2329999999999999e-06, + "loss": 2.8533, + "step": 413 + }, + { + "epoch": 0.34966216216216217, + "grad_norm": 3.6528217792510986, + "learning_rate": 1.2360000000000001e-06, + "loss": 2.9004, + "step": 414 + }, + { + "epoch": 0.35050675675675674, + "grad_norm": 9.373733520507812, + "learning_rate": 1.2390000000000001e-06, + "loss": 2.8243, + "step": 415 + }, + { + "epoch": 0.35135135135135137, + "grad_norm": 7.086609363555908, + "learning_rate": 1.242e-06, + "loss": 2.8749, + "step": 416 + }, + { + "epoch": 0.35219594594594594, + "grad_norm": 2.963568687438965, + "learning_rate": 1.2450000000000002e-06, + "loss": 2.8538, + "step": 417 + }, + { + "epoch": 0.3530405405405405, + "grad_norm": 5.242905616760254, + "learning_rate": 1.248e-06, + "loss": 2.828, + "step": 418 + }, + { + "epoch": 0.35388513513513514, + "grad_norm": 2.6869542598724365, + "learning_rate": 1.251e-06, + "loss": 2.8717, + "step": 419 + }, + { + "epoch": 0.3547297297297297, + "grad_norm": 2.9123752117156982, + "learning_rate": 1.254e-06, + "loss": 2.8349, + "step": 420 + }, + { + "epoch": 0.35557432432432434, + "grad_norm": 7.346799373626709, + "learning_rate": 1.257e-06, + "loss": 2.8665, + "step": 421 + }, + { + "epoch": 0.3564189189189189, + "grad_norm": 1.8919286727905273, + "learning_rate": 1.26e-06, + "loss": 2.8165, + "step": 422 + }, + { + "epoch": 0.3572635135135135, + "grad_norm": 5.813717365264893, + "learning_rate": 1.263e-06, + "loss": 2.8708, + "step": 423 + }, + { + "epoch": 0.3581081081081081, + "grad_norm": 5.506077766418457, + "learning_rate": 1.266e-06, + "loss": 2.8632, + "step": 424 + }, + { + "epoch": 0.3589527027027027, + "grad_norm": 2.443960428237915, + "learning_rate": 1.269e-06, + "loss": 2.8693, + "step": 425 + }, + { + "epoch": 0.3597972972972973, + "grad_norm": 4.45931339263916, + "learning_rate": 1.272e-06, + "loss": 2.8597, + "step": 426 + }, + { + "epoch": 0.3606418918918919, + "grad_norm": 7.797896385192871, + "learning_rate": 1.275e-06, + "loss": 2.8341, + "step": 427 + }, + { + "epoch": 0.3614864864864865, + "grad_norm": 7.734846115112305, + "learning_rate": 1.278e-06, + "loss": 2.8438, + "step": 428 + }, + { + "epoch": 0.3623310810810811, + "grad_norm": 5.587929725646973, + "learning_rate": 1.281e-06, + "loss": 2.8498, + "step": 429 + }, + { + "epoch": 0.36317567567567566, + "grad_norm": 7.9393181800842285, + "learning_rate": 1.284e-06, + "loss": 2.8605, + "step": 430 + }, + { + "epoch": 0.3640202702702703, + "grad_norm": 4.924855709075928, + "learning_rate": 1.2870000000000001e-06, + "loss": 2.869, + "step": 431 + }, + { + "epoch": 0.36486486486486486, + "grad_norm": 6.086109161376953, + "learning_rate": 1.29e-06, + "loss": 2.8801, + "step": 432 + }, + { + "epoch": 0.3657094594594595, + "grad_norm": 8.032732009887695, + "learning_rate": 1.293e-06, + "loss": 2.8212, + "step": 433 + }, + { + "epoch": 0.36655405405405406, + "grad_norm": 3.7947981357574463, + "learning_rate": 1.2960000000000002e-06, + "loss": 2.8992, + "step": 434 + }, + { + "epoch": 0.36739864864864863, + "grad_norm": 14.060548782348633, + "learning_rate": 1.299e-06, + "loss": 2.8623, + "step": 435 + }, + { + "epoch": 0.36824324324324326, + "grad_norm": 7.624170780181885, + "learning_rate": 1.302e-06, + "loss": 2.8734, + "step": 436 + }, + { + "epoch": 0.36908783783783783, + "grad_norm": 4.458164691925049, + "learning_rate": 1.305e-06, + "loss": 2.8931, + "step": 437 + }, + { + "epoch": 0.36993243243243246, + "grad_norm": 6.2936110496521, + "learning_rate": 1.308e-06, + "loss": 2.8584, + "step": 438 + }, + { + "epoch": 0.37077702702702703, + "grad_norm": 5.623869895935059, + "learning_rate": 1.311e-06, + "loss": 2.8644, + "step": 439 + }, + { + "epoch": 0.3716216216216216, + "grad_norm": 7.622312545776367, + "learning_rate": 1.314e-06, + "loss": 2.8425, + "step": 440 + }, + { + "epoch": 0.37246621621621623, + "grad_norm": 6.497400283813477, + "learning_rate": 1.317e-06, + "loss": 2.8748, + "step": 441 + }, + { + "epoch": 0.3733108108108108, + "grad_norm": 14.701950073242188, + "learning_rate": 1.3199999999999999e-06, + "loss": 2.8568, + "step": 442 + }, + { + "epoch": 0.37415540540540543, + "grad_norm": 2.3270835876464844, + "learning_rate": 1.323e-06, + "loss": 2.8185, + "step": 443 + }, + { + "epoch": 0.375, + "grad_norm": 13.485559463500977, + "learning_rate": 1.326e-06, + "loss": 2.8701, + "step": 444 + }, + { + "epoch": 0.37584459459459457, + "grad_norm": 4.986533164978027, + "learning_rate": 1.3290000000000001e-06, + "loss": 2.819, + "step": 445 + }, + { + "epoch": 0.3766891891891892, + "grad_norm": 3.6960654258728027, + "learning_rate": 1.3320000000000001e-06, + "loss": 2.8694, + "step": 446 + }, + { + "epoch": 0.37753378378378377, + "grad_norm": 6.399331569671631, + "learning_rate": 1.335e-06, + "loss": 2.8661, + "step": 447 + }, + { + "epoch": 0.3783783783783784, + "grad_norm": 10.952079772949219, + "learning_rate": 1.3380000000000001e-06, + "loss": 2.8564, + "step": 448 + }, + { + "epoch": 0.37922297297297297, + "grad_norm": 7.544609069824219, + "learning_rate": 1.341e-06, + "loss": 2.9203, + "step": 449 + }, + { + "epoch": 0.38006756756756754, + "grad_norm": 20.06273078918457, + "learning_rate": 1.344e-06, + "loss": 2.9192, + "step": 450 + }, + { + "epoch": 0.38091216216216217, + "grad_norm": 14.131715774536133, + "learning_rate": 1.3470000000000002e-06, + "loss": 3.0126, + "step": 451 + }, + { + "epoch": 0.38175675675675674, + "grad_norm": 6.238643646240234, + "learning_rate": 1.35e-06, + "loss": 2.9395, + "step": 452 + }, + { + "epoch": 0.38260135135135137, + "grad_norm": 2.8054912090301514, + "learning_rate": 1.353e-06, + "loss": 2.8316, + "step": 453 + }, + { + "epoch": 0.38344594594594594, + "grad_norm": 9.227554321289062, + "learning_rate": 1.356e-06, + "loss": 2.8444, + "step": 454 + }, + { + "epoch": 0.3842905405405405, + "grad_norm": 9.365824699401855, + "learning_rate": 1.359e-06, + "loss": 2.8031, + "step": 455 + }, + { + "epoch": 0.38513513513513514, + "grad_norm": 10.169881820678711, + "learning_rate": 1.362e-06, + "loss": 2.8811, + "step": 456 + }, + { + "epoch": 0.3859797297297297, + "grad_norm": 9.027901649475098, + "learning_rate": 1.365e-06, + "loss": 2.8171, + "step": 457 + }, + { + "epoch": 0.38682432432432434, + "grad_norm": 6.066862106323242, + "learning_rate": 1.368e-06, + "loss": 2.8259, + "step": 458 + }, + { + "epoch": 0.3876689189189189, + "grad_norm": 5.39225959777832, + "learning_rate": 1.3709999999999999e-06, + "loss": 2.8071, + "step": 459 + }, + { + "epoch": 0.3885135135135135, + "grad_norm": 10.827184677124023, + "learning_rate": 1.374e-06, + "loss": 2.8278, + "step": 460 + }, + { + "epoch": 0.3893581081081081, + "grad_norm": 5.567850112915039, + "learning_rate": 1.3770000000000001e-06, + "loss": 2.8398, + "step": 461 + }, + { + "epoch": 0.3902027027027027, + "grad_norm": 6.060507297515869, + "learning_rate": 1.38e-06, + "loss": 2.9011, + "step": 462 + }, + { + "epoch": 0.3910472972972973, + "grad_norm": 7.492789268493652, + "learning_rate": 1.3830000000000001e-06, + "loss": 2.8197, + "step": 463 + }, + { + "epoch": 0.3918918918918919, + "grad_norm": 7.2701096534729, + "learning_rate": 1.386e-06, + "loss": 2.8147, + "step": 464 + }, + { + "epoch": 0.3927364864864865, + "grad_norm": 4.302719593048096, + "learning_rate": 1.389e-06, + "loss": 2.799, + "step": 465 + }, + { + "epoch": 0.3935810810810811, + "grad_norm": 12.36325454711914, + "learning_rate": 1.392e-06, + "loss": 2.8459, + "step": 466 + }, + { + "epoch": 0.39442567567567566, + "grad_norm": 7.409857273101807, + "learning_rate": 1.395e-06, + "loss": 2.8005, + "step": 467 + }, + { + "epoch": 0.3952702702702703, + "grad_norm": 6.1235032081604, + "learning_rate": 1.3980000000000002e-06, + "loss": 2.7905, + "step": 468 + }, + { + "epoch": 0.39611486486486486, + "grad_norm": 4.628472328186035, + "learning_rate": 1.401e-06, + "loss": 2.7922, + "step": 469 + }, + { + "epoch": 0.3969594594594595, + "grad_norm": 6.1194891929626465, + "learning_rate": 1.404e-06, + "loss": 2.8151, + "step": 470 + }, + { + "epoch": 0.39780405405405406, + "grad_norm": 4.1689252853393555, + "learning_rate": 1.407e-06, + "loss": 2.8252, + "step": 471 + }, + { + "epoch": 0.39864864864864863, + "grad_norm": 4.425589084625244, + "learning_rate": 1.41e-06, + "loss": 2.7897, + "step": 472 + }, + { + "epoch": 0.39949324324324326, + "grad_norm": 8.902186393737793, + "learning_rate": 1.413e-06, + "loss": 2.7778, + "step": 473 + }, + { + "epoch": 0.40033783783783783, + "grad_norm": 11.088008880615234, + "learning_rate": 1.416e-06, + "loss": 2.8133, + "step": 474 + }, + { + "epoch": 0.40118243243243246, + "grad_norm": 3.505232095718384, + "learning_rate": 1.419e-06, + "loss": 2.7921, + "step": 475 + }, + { + "epoch": 0.40202702702702703, + "grad_norm": 13.966998100280762, + "learning_rate": 1.422e-06, + "loss": 2.794, + "step": 476 + }, + { + "epoch": 0.4028716216216216, + "grad_norm": 4.248779773712158, + "learning_rate": 1.4250000000000001e-06, + "loss": 2.7931, + "step": 477 + }, + { + "epoch": 0.40371621621621623, + "grad_norm": 5.585564136505127, + "learning_rate": 1.4280000000000001e-06, + "loss": 2.7804, + "step": 478 + }, + { + "epoch": 0.4045608108108108, + "grad_norm": 17.161651611328125, + "learning_rate": 1.431e-06, + "loss": 2.8173, + "step": 479 + }, + { + "epoch": 0.40540540540540543, + "grad_norm": 6.102417945861816, + "learning_rate": 1.4340000000000002e-06, + "loss": 2.7674, + "step": 480 + }, + { + "epoch": 0.40625, + "grad_norm": 7.277127742767334, + "learning_rate": 1.437e-06, + "loss": 2.7926, + "step": 481 + }, + { + "epoch": 0.40709459459459457, + "grad_norm": 6.0430474281311035, + "learning_rate": 1.44e-06, + "loss": 2.7882, + "step": 482 + }, + { + "epoch": 0.4079391891891892, + "grad_norm": 5.013068199157715, + "learning_rate": 1.443e-06, + "loss": 2.7841, + "step": 483 + }, + { + "epoch": 0.40878378378378377, + "grad_norm": 4.746347427368164, + "learning_rate": 1.446e-06, + "loss": 2.7917, + "step": 484 + }, + { + "epoch": 0.4096283783783784, + "grad_norm": 4.270693778991699, + "learning_rate": 1.449e-06, + "loss": 2.774, + "step": 485 + }, + { + "epoch": 0.41047297297297297, + "grad_norm": 12.575785636901855, + "learning_rate": 1.452e-06, + "loss": 2.7806, + "step": 486 + }, + { + "epoch": 0.41131756756756754, + "grad_norm": 8.850872039794922, + "learning_rate": 1.455e-06, + "loss": 2.7766, + "step": 487 + }, + { + "epoch": 0.41216216216216217, + "grad_norm": 21.560516357421875, + "learning_rate": 1.458e-06, + "loss": 2.8002, + "step": 488 + }, + { + "epoch": 0.41300675675675674, + "grad_norm": 10.071349143981934, + "learning_rate": 1.461e-06, + "loss": 2.7758, + "step": 489 + }, + { + "epoch": 0.41385135135135137, + "grad_norm": 4.770452499389648, + "learning_rate": 1.464e-06, + "loss": 2.7561, + "step": 490 + }, + { + "epoch": 0.41469594594594594, + "grad_norm": 14.489822387695312, + "learning_rate": 1.467e-06, + "loss": 2.7975, + "step": 491 + }, + { + "epoch": 0.4155405405405405, + "grad_norm": 2.9929802417755127, + "learning_rate": 1.4700000000000001e-06, + "loss": 2.7678, + "step": 492 + }, + { + "epoch": 0.41638513513513514, + "grad_norm": 6.704476833343506, + "learning_rate": 1.473e-06, + "loss": 2.757, + "step": 493 + }, + { + "epoch": 0.4172297297297297, + "grad_norm": 5.000030517578125, + "learning_rate": 1.4760000000000001e-06, + "loss": 2.7551, + "step": 494 + }, + { + "epoch": 0.41807432432432434, + "grad_norm": 11.731222152709961, + "learning_rate": 1.479e-06, + "loss": 2.7979, + "step": 495 + }, + { + "epoch": 0.4189189189189189, + "grad_norm": 4.338691711425781, + "learning_rate": 1.482e-06, + "loss": 2.7756, + "step": 496 + }, + { + "epoch": 0.4197635135135135, + "grad_norm": 3.959395170211792, + "learning_rate": 1.4850000000000002e-06, + "loss": 2.7905, + "step": 497 + }, + { + "epoch": 0.4206081081081081, + "grad_norm": 19.91620635986328, + "learning_rate": 1.488e-06, + "loss": 2.74, + "step": 498 + }, + { + "epoch": 0.4214527027027027, + "grad_norm": 9.639913558959961, + "learning_rate": 1.491e-06, + "loss": 2.7846, + "step": 499 + }, + { + "epoch": 0.4222972972972973, + "grad_norm": 5.794993877410889, + "learning_rate": 1.494e-06, + "loss": 2.7784, + "step": 500 + }, + { + "epoch": 0.4231418918918919, + "grad_norm": 11.16996955871582, + "learning_rate": 1.497e-06, + "loss": 2.8778, + "step": 501 + }, + { + "epoch": 0.4239864864864865, + "grad_norm": 11.319851875305176, + "learning_rate": 1.5e-06, + "loss": 2.8329, + "step": 502 + }, + { + "epoch": 0.4248310810810811, + "grad_norm": 8.348740577697754, + "learning_rate": 1.503e-06, + "loss": 2.7652, + "step": 503 + }, + { + "epoch": 0.42567567567567566, + "grad_norm": 13.276633262634277, + "learning_rate": 1.506e-06, + "loss": 2.7577, + "step": 504 + }, + { + "epoch": 0.4265202702702703, + "grad_norm": NaN, + "learning_rate": 1.506e-06, + "loss": 2.7484, + "step": 505 + }, + { + "epoch": 0.42736486486486486, + "grad_norm": 6.370997428894043, + "learning_rate": 1.5089999999999999e-06, + "loss": 2.7474, + "step": 506 + }, + { + "epoch": 0.4282094594594595, + "grad_norm": 12.133169174194336, + "learning_rate": 1.512e-06, + "loss": 2.7436, + "step": 507 + }, + { + "epoch": 0.42905405405405406, + "grad_norm": 4.18903112411499, + "learning_rate": 1.5150000000000001e-06, + "loss": 2.754, + "step": 508 + }, + { + "epoch": 0.42989864864864863, + "grad_norm": 3.702763080596924, + "learning_rate": 1.518e-06, + "loss": 2.7186, + "step": 509 + }, + { + "epoch": 0.43074324324324326, + "grad_norm": 9.818225860595703, + "learning_rate": 1.5210000000000001e-06, + "loss": 2.7341, + "step": 510 + }, + { + "epoch": 0.43158783783783783, + "grad_norm": 4.300854206085205, + "learning_rate": 1.524e-06, + "loss": 2.7232, + "step": 511 + }, + { + "epoch": 0.43243243243243246, + "grad_norm": 5.108453750610352, + "learning_rate": 1.5270000000000002e-06, + "loss": 2.734, + "step": 512 + }, + { + "epoch": 0.43327702702702703, + "grad_norm": 4.305500507354736, + "learning_rate": 1.53e-06, + "loss": 2.7551, + "step": 513 + }, + { + "epoch": 0.4341216216216216, + "grad_norm": 7.147480487823486, + "learning_rate": 1.533e-06, + "loss": 2.7129, + "step": 514 + }, + { + "epoch": 0.43496621621621623, + "grad_norm": 5.4375691413879395, + "learning_rate": 1.5360000000000002e-06, + "loss": 2.736, + "step": 515 + }, + { + "epoch": 0.4358108108108108, + "grad_norm": 6.5706892013549805, + "learning_rate": 1.539e-06, + "loss": 2.7333, + "step": 516 + }, + { + "epoch": 0.43665540540540543, + "grad_norm": 5.715762615203857, + "learning_rate": 1.542e-06, + "loss": 2.7583, + "step": 517 + }, + { + "epoch": 0.4375, + "grad_norm": 4.601909160614014, + "learning_rate": 1.545e-06, + "loss": 2.7209, + "step": 518 + }, + { + "epoch": 0.43834459459459457, + "grad_norm": 3.6297607421875, + "learning_rate": 1.548e-06, + "loss": 2.727, + "step": 519 + }, + { + "epoch": 0.4391891891891892, + "grad_norm": 5.696209907531738, + "learning_rate": 1.551e-06, + "loss": 2.7713, + "step": 520 + }, + { + "epoch": 0.44003378378378377, + "grad_norm": 3.746169328689575, + "learning_rate": 1.554e-06, + "loss": 2.7114, + "step": 521 + }, + { + "epoch": 0.4408783783783784, + "grad_norm": 6.602119445800781, + "learning_rate": 1.557e-06, + "loss": 2.7411, + "step": 522 + }, + { + "epoch": 0.44172297297297297, + "grad_norm": 28.328222274780273, + "learning_rate": 1.5599999999999999e-06, + "loss": 2.7355, + "step": 523 + }, + { + "epoch": 0.44256756756756754, + "grad_norm": 2.223724842071533, + "learning_rate": 1.5630000000000001e-06, + "loss": 2.6732, + "step": 524 + }, + { + "epoch": 0.44341216216216217, + "grad_norm": 5.391133785247803, + "learning_rate": 1.5660000000000001e-06, + "loss": 2.7333, + "step": 525 + }, + { + "epoch": 0.44425675675675674, + "grad_norm": 23.767484664916992, + "learning_rate": 1.569e-06, + "loss": 2.7181, + "step": 526 + }, + { + "epoch": 0.44510135135135137, + "grad_norm": 7.038394451141357, + "learning_rate": 1.5720000000000002e-06, + "loss": 2.7198, + "step": 527 + }, + { + "epoch": 0.44594594594594594, + "grad_norm": 3.705350399017334, + "learning_rate": 1.575e-06, + "loss": 2.7354, + "step": 528 + }, + { + "epoch": 0.4467905405405405, + "grad_norm": 14.262215614318848, + "learning_rate": 1.578e-06, + "loss": 2.717, + "step": 529 + }, + { + "epoch": 0.44763513513513514, + "grad_norm": 4.495677947998047, + "learning_rate": 1.581e-06, + "loss": 2.7266, + "step": 530 + }, + { + "epoch": 0.4484797297297297, + "grad_norm": 5.568889617919922, + "learning_rate": 1.584e-06, + "loss": 2.6944, + "step": 531 + }, + { + "epoch": 0.44932432432432434, + "grad_norm": 5.374840259552002, + "learning_rate": 1.5870000000000002e-06, + "loss": 2.7166, + "step": 532 + }, + { + "epoch": 0.4501689189189189, + "grad_norm": 15.227386474609375, + "learning_rate": 1.59e-06, + "loss": 2.6913, + "step": 533 + }, + { + "epoch": 0.4510135135135135, + "grad_norm": 11.313265800476074, + "learning_rate": 1.593e-06, + "loss": 2.684, + "step": 534 + }, + { + "epoch": 0.4518581081081081, + "grad_norm": 4.823352336883545, + "learning_rate": 1.596e-06, + "loss": 2.694, + "step": 535 + }, + { + "epoch": 0.4527027027027027, + "grad_norm": 5.779737949371338, + "learning_rate": 1.599e-06, + "loss": 2.7141, + "step": 536 + }, + { + "epoch": 0.4535472972972973, + "grad_norm": 9.257954597473145, + "learning_rate": 1.602e-06, + "loss": 2.7097, + "step": 537 + }, + { + "epoch": 0.4543918918918919, + "grad_norm": 10.581798553466797, + "learning_rate": 1.605e-06, + "loss": 2.6852, + "step": 538 + }, + { + "epoch": 0.4552364864864865, + "grad_norm": 7.242800235748291, + "learning_rate": 1.608e-06, + "loss": 2.7128, + "step": 539 + }, + { + "epoch": 0.4560810810810811, + "grad_norm": 32.23623275756836, + "learning_rate": 1.611e-06, + "loss": 2.7027, + "step": 540 + }, + { + "epoch": 0.45692567567567566, + "grad_norm": 4.901346683502197, + "learning_rate": 1.6140000000000001e-06, + "loss": 2.6808, + "step": 541 + }, + { + "epoch": 0.4577702702702703, + "grad_norm": 5.322842597961426, + "learning_rate": 1.6170000000000001e-06, + "loss": 2.6606, + "step": 542 + }, + { + "epoch": 0.45861486486486486, + "grad_norm": 14.407244682312012, + "learning_rate": 1.62e-06, + "loss": 2.688, + "step": 543 + }, + { + "epoch": 0.4594594594594595, + "grad_norm": 4.879325866699219, + "learning_rate": 1.6230000000000002e-06, + "loss": 2.6793, + "step": 544 + }, + { + "epoch": 0.46030405405405406, + "grad_norm": 13.707354545593262, + "learning_rate": 1.626e-06, + "loss": 2.7003, + "step": 545 + }, + { + "epoch": 0.46114864864864863, + "grad_norm": 10.787659645080566, + "learning_rate": 1.629e-06, + "loss": 2.6821, + "step": 546 + }, + { + "epoch": 0.46199324324324326, + "grad_norm": 7.668027400970459, + "learning_rate": 1.632e-06, + "loss": 2.688, + "step": 547 + }, + { + "epoch": 0.46283783783783783, + "grad_norm": 6.573879718780518, + "learning_rate": 1.635e-06, + "loss": 2.7137, + "step": 548 + }, + { + "epoch": 0.46368243243243246, + "grad_norm": 5.135451793670654, + "learning_rate": 1.638e-06, + "loss": 2.6829, + "step": 549 + }, + { + "epoch": 0.46452702702702703, + "grad_norm": 11.152881622314453, + "learning_rate": 1.641e-06, + "loss": 2.7032, + "step": 550 + }, + { + "epoch": 0.4653716216216216, + "grad_norm": 5.305014610290527, + "learning_rate": 1.644e-06, + "loss": 2.7735, + "step": 551 + }, + { + "epoch": 0.46621621621621623, + "grad_norm": 5.128194808959961, + "learning_rate": 1.6469999999999999e-06, + "loss": 2.7325, + "step": 552 + }, + { + "epoch": 0.4670608108108108, + "grad_norm": 6.190968036651611, + "learning_rate": 1.65e-06, + "loss": 2.7021, + "step": 553 + }, + { + "epoch": 0.46790540540540543, + "grad_norm": 4.584140777587891, + "learning_rate": 1.653e-06, + "loss": 2.6331, + "step": 554 + }, + { + "epoch": 0.46875, + "grad_norm": 4.168137550354004, + "learning_rate": 1.6560000000000001e-06, + "loss": 2.6392, + "step": 555 + }, + { + "epoch": 0.46959459459459457, + "grad_norm": 6.725541114807129, + "learning_rate": 1.6590000000000001e-06, + "loss": 2.6329, + "step": 556 + }, + { + "epoch": 0.4704391891891892, + "grad_norm": 4.766119480133057, + "learning_rate": 1.662e-06, + "loss": 2.6429, + "step": 557 + }, + { + "epoch": 0.47128378378378377, + "grad_norm": 5.0897016525268555, + "learning_rate": 1.6650000000000002e-06, + "loss": 2.6186, + "step": 558 + }, + { + "epoch": 0.4721283783783784, + "grad_norm": 8.18778133392334, + "learning_rate": 1.668e-06, + "loss": 2.6889, + "step": 559 + }, + { + "epoch": 0.47297297297297297, + "grad_norm": 13.40765380859375, + "learning_rate": 1.671e-06, + "loss": 2.6383, + "step": 560 + }, + { + "epoch": 0.47381756756756754, + "grad_norm": 3.5309979915618896, + "learning_rate": 1.6740000000000002e-06, + "loss": 2.6259, + "step": 561 + }, + { + "epoch": 0.47466216216216217, + "grad_norm": 3.991955518722534, + "learning_rate": 1.677e-06, + "loss": 2.6402, + "step": 562 + }, + { + "epoch": 0.47550675675675674, + "grad_norm": 4.3028740882873535, + "learning_rate": 1.68e-06, + "loss": 2.6023, + "step": 563 + }, + { + "epoch": 0.47635135135135137, + "grad_norm": 7.393435478210449, + "learning_rate": 1.683e-06, + "loss": 2.6, + "step": 564 + }, + { + "epoch": 0.47719594594594594, + "grad_norm": 11.725375175476074, + "learning_rate": 1.686e-06, + "loss": 2.616, + "step": 565 + }, + { + "epoch": 0.4780405405405405, + "grad_norm": 8.35815715789795, + "learning_rate": 1.689e-06, + "loss": 2.6018, + "step": 566 + }, + { + "epoch": 0.47888513513513514, + "grad_norm": 6.578551292419434, + "learning_rate": 1.692e-06, + "loss": 2.5819, + "step": 567 + }, + { + "epoch": 0.4797297297297297, + "grad_norm": 3.6284830570220947, + "learning_rate": 1.695e-06, + "loss": 2.6111, + "step": 568 + }, + { + "epoch": 0.48057432432432434, + "grad_norm": 3.441498279571533, + "learning_rate": 1.6979999999999999e-06, + "loss": 2.5704, + "step": 569 + }, + { + "epoch": 0.4814189189189189, + "grad_norm": 9.052006721496582, + "learning_rate": 1.701e-06, + "loss": 2.5899, + "step": 570 + }, + { + "epoch": 0.4822635135135135, + "grad_norm": 9.55909252166748, + "learning_rate": 1.7040000000000001e-06, + "loss": 2.5915, + "step": 571 + }, + { + "epoch": 0.4831081081081081, + "grad_norm": 3.3446545600891113, + "learning_rate": 1.707e-06, + "loss": 2.5705, + "step": 572 + }, + { + "epoch": 0.4839527027027027, + "grad_norm": 7.323623180389404, + "learning_rate": 1.7100000000000001e-06, + "loss": 2.6058, + "step": 573 + }, + { + "epoch": 0.4847972972972973, + "grad_norm": 4.487628936767578, + "learning_rate": 1.713e-06, + "loss": 2.5596, + "step": 574 + }, + { + "epoch": 0.4856418918918919, + "grad_norm": 3.388967752456665, + "learning_rate": 1.7160000000000002e-06, + "loss": 2.5425, + "step": 575 + }, + { + "epoch": 0.4864864864864865, + "grad_norm": 2.965804100036621, + "learning_rate": 1.719e-06, + "loss": 2.5455, + "step": 576 + }, + { + "epoch": 0.4873310810810811, + "grad_norm": 3.8988540172576904, + "learning_rate": 1.722e-06, + "loss": 2.5647, + "step": 577 + }, + { + "epoch": 0.48817567567567566, + "grad_norm": 4.884891510009766, + "learning_rate": 1.7250000000000002e-06, + "loss": 2.527, + "step": 578 + }, + { + "epoch": 0.4890202702702703, + "grad_norm": 9.025464057922363, + "learning_rate": 1.728e-06, + "loss": 2.5066, + "step": 579 + }, + { + "epoch": 0.48986486486486486, + "grad_norm": 6.73118782043457, + "learning_rate": 1.731e-06, + "loss": 2.4935, + "step": 580 + }, + { + "epoch": 0.4907094594594595, + "grad_norm": 4.213400363922119, + "learning_rate": 1.734e-06, + "loss": 2.5075, + "step": 581 + }, + { + "epoch": 0.49155405405405406, + "grad_norm": 5.480779647827148, + "learning_rate": 1.737e-06, + "loss": 2.4982, + "step": 582 + }, + { + "epoch": 0.49239864864864863, + "grad_norm": 4.28178596496582, + "learning_rate": 1.74e-06, + "loss": 2.4976, + "step": 583 + }, + { + "epoch": 0.49324324324324326, + "grad_norm": 8.385747909545898, + "learning_rate": 1.743e-06, + "loss": 2.4663, + "step": 584 + }, + { + "epoch": 0.49408783783783783, + "grad_norm": 4.014645099639893, + "learning_rate": 1.746e-06, + "loss": 2.4793, + "step": 585 + }, + { + "epoch": 0.49493243243243246, + "grad_norm": 4.956210613250732, + "learning_rate": 1.749e-06, + "loss": 2.5019, + "step": 586 + }, + { + "epoch": 0.49577702702702703, + "grad_norm": 4.933856964111328, + "learning_rate": 1.7520000000000001e-06, + "loss": 2.4774, + "step": 587 + }, + { + "epoch": 0.4966216216216216, + "grad_norm": 12.505708694458008, + "learning_rate": 1.7550000000000001e-06, + "loss": 2.4841, + "step": 588 + }, + { + "epoch": 0.49746621621621623, + "grad_norm": 3.4025816917419434, + "learning_rate": 1.758e-06, + "loss": 2.424, + "step": 589 + }, + { + "epoch": 0.4983108108108108, + "grad_norm": 7.530827522277832, + "learning_rate": 1.7610000000000002e-06, + "loss": 2.4414, + "step": 590 + }, + { + "epoch": 0.49915540540540543, + "grad_norm": 5.113575458526611, + "learning_rate": 1.764e-06, + "loss": 2.4362, + "step": 591 + }, + { + "epoch": 0.5, + "grad_norm": 6.3427605628967285, + "learning_rate": 1.767e-06, + "loss": 2.4483, + "step": 592 + }, + { + "epoch": 0.5008445945945946, + "grad_norm": 3.3286240100860596, + "learning_rate": 1.77e-06, + "loss": 2.4217, + "step": 593 + }, + { + "epoch": 0.5016891891891891, + "grad_norm": 6.46979284286499, + "learning_rate": 1.773e-06, + "loss": 2.3942, + "step": 594 + }, + { + "epoch": 0.5025337837837838, + "grad_norm": 9.662845611572266, + "learning_rate": 1.776e-06, + "loss": 2.4226, + "step": 595 + }, + { + "epoch": 0.5033783783783784, + "grad_norm": 4.734137535095215, + "learning_rate": 1.779e-06, + "loss": 2.3829, + "step": 596 + }, + { + "epoch": 0.504222972972973, + "grad_norm": 7.9803056716918945, + "learning_rate": 1.782e-06, + "loss": 2.3982, + "step": 597 + }, + { + "epoch": 0.5050675675675675, + "grad_norm": 14.68302059173584, + "learning_rate": 1.785e-06, + "loss": 2.442, + "step": 598 + }, + { + "epoch": 0.5059121621621622, + "grad_norm": 6.628355979919434, + "learning_rate": 1.788e-06, + "loss": 2.3981, + "step": 599 + }, + { + "epoch": 0.5067567567567568, + "grad_norm": 13.832152366638184, + "learning_rate": 1.791e-06, + "loss": 2.4588, + "step": 600 + }, + { + "epoch": 0.5076013513513513, + "grad_norm": 8.758715629577637, + "learning_rate": 1.794e-06, + "loss": 2.589, + "step": 601 + }, + { + "epoch": 0.5084459459459459, + "grad_norm": 6.305700302124023, + "learning_rate": 1.7970000000000001e-06, + "loss": 2.4732, + "step": 602 + }, + { + "epoch": 0.5092905405405406, + "grad_norm": 7.424586296081543, + "learning_rate": 1.8e-06, + "loss": 2.3962, + "step": 603 + }, + { + "epoch": 0.5101351351351351, + "grad_norm": 3.992903709411621, + "learning_rate": 1.8030000000000001e-06, + "loss": 2.3319, + "step": 604 + }, + { + "epoch": 0.5109797297297297, + "grad_norm": 5.5471320152282715, + "learning_rate": 1.806e-06, + "loss": 2.3303, + "step": 605 + }, + { + "epoch": 0.5118243243243243, + "grad_norm": 6.069870471954346, + "learning_rate": 1.809e-06, + "loss": 2.2844, + "step": 606 + }, + { + "epoch": 0.512668918918919, + "grad_norm": 28.409570693969727, + "learning_rate": 1.8120000000000002e-06, + "loss": 2.3364, + "step": 607 + }, + { + "epoch": 0.5135135135135135, + "grad_norm": 5.478114604949951, + "learning_rate": 1.815e-06, + "loss": 2.2474, + "step": 608 + }, + { + "epoch": 0.5143581081081081, + "grad_norm": 7.618147850036621, + "learning_rate": 1.818e-06, + "loss": 2.318, + "step": 609 + }, + { + "epoch": 0.5152027027027027, + "grad_norm": 7.373167514801025, + "learning_rate": 1.821e-06, + "loss": 2.2168, + "step": 610 + }, + { + "epoch": 0.5160472972972973, + "grad_norm": 7.242875576019287, + "learning_rate": 1.824e-06, + "loss": 2.2805, + "step": 611 + }, + { + "epoch": 0.5168918918918919, + "grad_norm": 9.23494815826416, + "learning_rate": 1.827e-06, + "loss": 2.2525, + "step": 612 + }, + { + "epoch": 0.5177364864864865, + "grad_norm": 3.5349230766296387, + "learning_rate": 1.83e-06, + "loss": 2.2399, + "step": 613 + }, + { + "epoch": 0.518581081081081, + "grad_norm": 13.601563453674316, + "learning_rate": 1.833e-06, + "loss": 2.1735, + "step": 614 + }, + { + "epoch": 0.5194256756756757, + "grad_norm": 6.553210258483887, + "learning_rate": 1.8359999999999999e-06, + "loss": 2.1992, + "step": 615 + }, + { + "epoch": 0.5202702702702703, + "grad_norm": 4.217235088348389, + "learning_rate": 1.839e-06, + "loss": 2.2044, + "step": 616 + }, + { + "epoch": 0.5211148648648649, + "grad_norm": 3.680985927581787, + "learning_rate": 1.8420000000000001e-06, + "loss": 2.1828, + "step": 617 + }, + { + "epoch": 0.5219594594594594, + "grad_norm": 5.101829528808594, + "learning_rate": 1.8450000000000001e-06, + "loss": 2.1751, + "step": 618 + }, + { + "epoch": 0.5228040540540541, + "grad_norm": 8.42625904083252, + "learning_rate": 1.8480000000000001e-06, + "loss": 2.1744, + "step": 619 + }, + { + "epoch": 0.5236486486486487, + "grad_norm": 11.554248809814453, + "learning_rate": 1.851e-06, + "loss": 2.1865, + "step": 620 + }, + { + "epoch": 0.5244932432432432, + "grad_norm": 9.341460227966309, + "learning_rate": 1.8540000000000002e-06, + "loss": 2.1511, + "step": 621 + }, + { + "epoch": 0.5253378378378378, + "grad_norm": 3.499393939971924, + "learning_rate": 1.857e-06, + "loss": 2.107, + "step": 622 + }, + { + "epoch": 0.5261824324324325, + "grad_norm": 12.367324829101562, + "learning_rate": 1.86e-06, + "loss": 2.1709, + "step": 623 + }, + { + "epoch": 0.527027027027027, + "grad_norm": 6.36176061630249, + "learning_rate": 1.8630000000000002e-06, + "loss": 2.1091, + "step": 624 + }, + { + "epoch": 0.5278716216216216, + "grad_norm": 9.075783729553223, + "learning_rate": 1.866e-06, + "loss": 2.0895, + "step": 625 + }, + { + "epoch": 0.5287162162162162, + "grad_norm": 6.63992977142334, + "learning_rate": 1.869e-06, + "loss": 2.1737, + "step": 626 + }, + { + "epoch": 0.5295608108108109, + "grad_norm": 4.7956390380859375, + "learning_rate": 1.872e-06, + "loss": 2.0908, + "step": 627 + }, + { + "epoch": 0.5304054054054054, + "grad_norm": 17.47795295715332, + "learning_rate": 1.875e-06, + "loss": 2.1276, + "step": 628 + }, + { + "epoch": 0.53125, + "grad_norm": 6.4665141105651855, + "learning_rate": 1.878e-06, + "loss": 2.0754, + "step": 629 + }, + { + "epoch": 0.5320945945945946, + "grad_norm": 6.9309163093566895, + "learning_rate": 1.8810000000000003e-06, + "loss": 2.1277, + "step": 630 + }, + { + "epoch": 0.5329391891891891, + "grad_norm": 3.979571580886841, + "learning_rate": 1.8839999999999999e-06, + "loss": 2.0158, + "step": 631 + }, + { + "epoch": 0.5337837837837838, + "grad_norm": 5.625901222229004, + "learning_rate": 1.8869999999999999e-06, + "loss": 2.0783, + "step": 632 + }, + { + "epoch": 0.5346283783783784, + "grad_norm": 6.448584079742432, + "learning_rate": 1.8900000000000001e-06, + "loss": 2.0265, + "step": 633 + }, + { + "epoch": 0.535472972972973, + "grad_norm": 9.897486686706543, + "learning_rate": 1.8930000000000001e-06, + "loss": 2.007, + "step": 634 + }, + { + "epoch": 0.5363175675675675, + "grad_norm": 3.9134521484375, + "learning_rate": 1.8960000000000001e-06, + "loss": 2.0081, + "step": 635 + }, + { + "epoch": 0.5371621621621622, + "grad_norm": 6.58530330657959, + "learning_rate": 1.899e-06, + "loss": 2.0812, + "step": 636 + }, + { + "epoch": 0.5380067567567568, + "grad_norm": 8.48988151550293, + "learning_rate": 1.902e-06, + "loss": 1.991, + "step": 637 + }, + { + "epoch": 0.5388513513513513, + "grad_norm": 5.9926276206970215, + "learning_rate": 1.905e-06, + "loss": 2.0069, + "step": 638 + }, + { + "epoch": 0.5396959459459459, + "grad_norm": 7.315001964569092, + "learning_rate": 1.908e-06, + "loss": 1.9973, + "step": 639 + }, + { + "epoch": 0.5405405405405406, + "grad_norm": 5.82346248626709, + "learning_rate": 1.9110000000000004e-06, + "loss": 1.9494, + "step": 640 + }, + { + "epoch": 0.5413851351351351, + "grad_norm": 4.0179033279418945, + "learning_rate": 1.914e-06, + "loss": 2.0001, + "step": 641 + }, + { + "epoch": 0.5422297297297297, + "grad_norm": 10.116684913635254, + "learning_rate": 1.917e-06, + "loss": 1.9872, + "step": 642 + }, + { + "epoch": 0.5430743243243243, + "grad_norm": 5.311193466186523, + "learning_rate": 1.9200000000000003e-06, + "loss": 2.0128, + "step": 643 + }, + { + "epoch": 0.543918918918919, + "grad_norm": 8.84736156463623, + "learning_rate": 1.923e-06, + "loss": 1.9945, + "step": 644 + }, + { + "epoch": 0.5447635135135135, + "grad_norm": 7.712672233581543, + "learning_rate": 1.926e-06, + "loss": 1.9044, + "step": 645 + }, + { + "epoch": 0.5456081081081081, + "grad_norm": 9.507221221923828, + "learning_rate": 1.929e-06, + "loss": 1.9764, + "step": 646 + }, + { + "epoch": 0.5464527027027027, + "grad_norm": 5.90064811706543, + "learning_rate": 1.932e-06, + "loss": 1.8883, + "step": 647 + }, + { + "epoch": 0.5472972972972973, + "grad_norm": 4.975958824157715, + "learning_rate": 1.935e-06, + "loss": 1.9746, + "step": 648 + }, + { + "epoch": 0.5481418918918919, + "grad_norm": 5.491756439208984, + "learning_rate": 1.9380000000000003e-06, + "loss": 2.009, + "step": 649 + }, + { + "epoch": 0.5489864864864865, + "grad_norm": 6.358887195587158, + "learning_rate": 1.9409999999999997e-06, + "loss": 2.0644, + "step": 650 + }, + { + "epoch": 0.549831081081081, + "grad_norm": 6.818787574768066, + "learning_rate": 1.944e-06, + "loss": 2.1534, + "step": 651 + }, + { + "epoch": 0.5506756756756757, + "grad_norm": 7.018422603607178, + "learning_rate": 1.947e-06, + "loss": 2.0453, + "step": 652 + }, + { + "epoch": 0.5515202702702703, + "grad_norm": 7.552890777587891, + "learning_rate": 1.95e-06, + "loss": 1.9989, + "step": 653 + }, + { + "epoch": 0.5523648648648649, + "grad_norm": 12.287879943847656, + "learning_rate": 1.953e-06, + "loss": 1.8743, + "step": 654 + }, + { + "epoch": 0.5532094594594594, + "grad_norm": 36.03631591796875, + "learning_rate": 1.956e-06, + "loss": 1.8029, + "step": 655 + }, + { + "epoch": 0.5540540540540541, + "grad_norm": 8.1564359664917, + "learning_rate": 1.959e-06, + "loss": 1.7513, + "step": 656 + }, + { + "epoch": 0.5548986486486487, + "grad_norm": 4.034939289093018, + "learning_rate": 1.962e-06, + "loss": 1.7373, + "step": 657 + }, + { + "epoch": 0.5557432432432432, + "grad_norm": 37.007938385009766, + "learning_rate": 1.9650000000000002e-06, + "loss": 1.7349, + "step": 658 + }, + { + "epoch": 0.5565878378378378, + "grad_norm": 7.123076438903809, + "learning_rate": 1.968e-06, + "loss": 1.7514, + "step": 659 + }, + { + "epoch": 0.5574324324324325, + "grad_norm": 5.414472579956055, + "learning_rate": 1.971e-06, + "loss": 1.6962, + "step": 660 + }, + { + "epoch": 0.558277027027027, + "grad_norm": 10.786808967590332, + "learning_rate": 1.974e-06, + "loss": 1.7026, + "step": 661 + }, + { + "epoch": 0.5591216216216216, + "grad_norm": 13.799015045166016, + "learning_rate": 1.977e-06, + "loss": 1.6839, + "step": 662 + }, + { + "epoch": 0.5599662162162162, + "grad_norm": 5.038327217102051, + "learning_rate": 1.98e-06, + "loss": 1.6443, + "step": 663 + }, + { + "epoch": 0.5608108108108109, + "grad_norm": 14.914339065551758, + "learning_rate": 1.9830000000000003e-06, + "loss": 1.6671, + "step": 664 + }, + { + "epoch": 0.5616554054054054, + "grad_norm": 3.8969640731811523, + "learning_rate": 1.9859999999999997e-06, + "loss": 1.5841, + "step": 665 + }, + { + "epoch": 0.5625, + "grad_norm": 6.912862300872803, + "learning_rate": 1.989e-06, + "loss": 1.6328, + "step": 666 + }, + { + "epoch": 0.5633445945945946, + "grad_norm": 3.7361624240875244, + "learning_rate": 1.992e-06, + "loss": 1.6138, + "step": 667 + }, + { + "epoch": 0.5641891891891891, + "grad_norm": 7.261900901794434, + "learning_rate": 1.995e-06, + "loss": 1.6025, + "step": 668 + }, + { + "epoch": 0.5650337837837838, + "grad_norm": 4.051952362060547, + "learning_rate": 1.998e-06, + "loss": 1.5825, + "step": 669 + }, + { + "epoch": 0.5658783783783784, + "grad_norm": 2.8993775844573975, + "learning_rate": 2.001e-06, + "loss": 1.5195, + "step": 670 + }, + { + "epoch": 0.566722972972973, + "grad_norm": 12.0349760055542, + "learning_rate": 2.004e-06, + "loss": 1.5763, + "step": 671 + }, + { + "epoch": 0.5675675675675675, + "grad_norm": 15.875702857971191, + "learning_rate": 2.007e-06, + "loss": 1.5701, + "step": 672 + }, + { + "epoch": 0.5684121621621622, + "grad_norm": 3.6093907356262207, + "learning_rate": 2.0100000000000002e-06, + "loss": 1.5239, + "step": 673 + }, + { + "epoch": 0.5692567567567568, + "grad_norm": 9.14776611328125, + "learning_rate": 2.0130000000000005e-06, + "loss": 1.5024, + "step": 674 + }, + { + "epoch": 0.5701013513513513, + "grad_norm": 7.615556240081787, + "learning_rate": 2.016e-06, + "loss": 1.5518, + "step": 675 + }, + { + "epoch": 0.5709459459459459, + "grad_norm": 4.983196258544922, + "learning_rate": 2.019e-06, + "loss": 1.5019, + "step": 676 + }, + { + "epoch": 0.5717905405405406, + "grad_norm": 6.137028694152832, + "learning_rate": 2.0220000000000003e-06, + "loss": 1.5546, + "step": 677 + }, + { + "epoch": 0.5726351351351351, + "grad_norm": 4.681580543518066, + "learning_rate": 2.025e-06, + "loss": 1.4529, + "step": 678 + }, + { + "epoch": 0.5734797297297297, + "grad_norm": 4.116276264190674, + "learning_rate": 2.028e-06, + "loss": 1.4477, + "step": 679 + }, + { + "epoch": 0.5743243243243243, + "grad_norm": 4.409270286560059, + "learning_rate": 2.031e-06, + "loss": 1.3931, + "step": 680 + }, + { + "epoch": 0.575168918918919, + "grad_norm": 4.703739166259766, + "learning_rate": 2.034e-06, + "loss": 1.5013, + "step": 681 + }, + { + "epoch": 0.5760135135135135, + "grad_norm": 4.834676265716553, + "learning_rate": 2.037e-06, + "loss": 1.3889, + "step": 682 + }, + { + "epoch": 0.5768581081081081, + "grad_norm": 4.548800945281982, + "learning_rate": 2.0400000000000004e-06, + "loss": 1.3984, + "step": 683 + }, + { + "epoch": 0.5777027027027027, + "grad_norm": 16.031496047973633, + "learning_rate": 2.0429999999999998e-06, + "loss": 1.4136, + "step": 684 + }, + { + "epoch": 0.5785472972972973, + "grad_norm": 3.9161672592163086, + "learning_rate": 2.046e-06, + "loss": 1.3806, + "step": 685 + }, + { + "epoch": 0.5793918918918919, + "grad_norm": 7.458212375640869, + "learning_rate": 2.049e-06, + "loss": 1.3921, + "step": 686 + }, + { + "epoch": 0.5802364864864865, + "grad_norm": 5.378131866455078, + "learning_rate": 2.052e-06, + "loss": 1.3414, + "step": 687 + }, + { + "epoch": 0.581081081081081, + "grad_norm": 11.60522747039795, + "learning_rate": 2.0550000000000002e-06, + "loss": 1.4013, + "step": 688 + }, + { + "epoch": 0.5819256756756757, + "grad_norm": 13.728323936462402, + "learning_rate": 2.058e-06, + "loss": 1.3272, + "step": 689 + }, + { + "epoch": 0.5827702702702703, + "grad_norm": 6.221124649047852, + "learning_rate": 2.061e-06, + "loss": 1.3304, + "step": 690 + }, + { + "epoch": 0.5836148648648649, + "grad_norm": 5.482115745544434, + "learning_rate": 2.064e-06, + "loss": 1.3385, + "step": 691 + }, + { + "epoch": 0.5844594594594594, + "grad_norm": 4.831545352935791, + "learning_rate": 2.0670000000000003e-06, + "loss": 1.3353, + "step": 692 + }, + { + "epoch": 0.5853040540540541, + "grad_norm": 3.467912197113037, + "learning_rate": 2.07e-06, + "loss": 1.3361, + "step": 693 + }, + { + "epoch": 0.5861486486486487, + "grad_norm": 4.4003987312316895, + "learning_rate": 2.073e-06, + "loss": 1.3032, + "step": 694 + }, + { + "epoch": 0.5869932432432432, + "grad_norm": 3.3206112384796143, + "learning_rate": 2.076e-06, + "loss": 1.3154, + "step": 695 + }, + { + "epoch": 0.5878378378378378, + "grad_norm": 5.548069477081299, + "learning_rate": 2.079e-06, + "loss": 1.2913, + "step": 696 + }, + { + "epoch": 0.5886824324324325, + "grad_norm": 7.607484817504883, + "learning_rate": 2.082e-06, + "loss": 1.5821, + "step": 697 + }, + { + "epoch": 0.589527027027027, + "grad_norm": 5.077162742614746, + "learning_rate": 2.0850000000000004e-06, + "loss": 1.2395, + "step": 698 + }, + { + "epoch": 0.5903716216216216, + "grad_norm": 4.804690837860107, + "learning_rate": 2.0879999999999997e-06, + "loss": 1.2665, + "step": 699 + }, + { + "epoch": 0.5912162162162162, + "grad_norm": 8.59417724609375, + "learning_rate": 2.091e-06, + "loss": 1.3427, + "step": 700 + }, + { + "epoch": 0.5920608108108109, + "grad_norm": 9.494065284729004, + "learning_rate": 2.094e-06, + "loss": 1.829, + "step": 701 + }, + { + "epoch": 0.5929054054054054, + "grad_norm": 6.9612507820129395, + "learning_rate": 2.097e-06, + "loss": 1.4861, + "step": 702 + }, + { + "epoch": 0.59375, + "grad_norm": 7.309410572052002, + "learning_rate": 2.1000000000000002e-06, + "loss": 1.2007, + "step": 703 + }, + { + "epoch": 0.5945945945945946, + "grad_norm": 4.303187847137451, + "learning_rate": 2.103e-06, + "loss": 1.1032, + "step": 704 + }, + { + "epoch": 0.5954391891891891, + "grad_norm": 5.203845500946045, + "learning_rate": 2.106e-06, + "loss": 1.1395, + "step": 705 + }, + { + "epoch": 0.5962837837837838, + "grad_norm": 5.856383323669434, + "learning_rate": 2.109e-06, + "loss": 1.0745, + "step": 706 + }, + { + "epoch": 0.5971283783783784, + "grad_norm": 5.252254486083984, + "learning_rate": 2.1120000000000003e-06, + "loss": 1.0932, + "step": 707 + }, + { + "epoch": 0.597972972972973, + "grad_norm": 4.119598388671875, + "learning_rate": 2.1149999999999997e-06, + "loss": 1.0887, + "step": 708 + }, + { + "epoch": 0.5988175675675675, + "grad_norm": 6.04875373840332, + "learning_rate": 2.118e-06, + "loss": 1.0954, + "step": 709 + }, + { + "epoch": 0.5996621621621622, + "grad_norm": 5.257498741149902, + "learning_rate": 2.121e-06, + "loss": 1.0539, + "step": 710 + }, + { + "epoch": 0.6005067567567568, + "grad_norm": 30.43881607055664, + "learning_rate": 2.124e-06, + "loss": 1.0441, + "step": 711 + }, + { + "epoch": 0.6013513513513513, + "grad_norm": 5.048116207122803, + "learning_rate": 2.127e-06, + "loss": 1.0336, + "step": 712 + }, + { + "epoch": 0.6021959459459459, + "grad_norm": 4.458845615386963, + "learning_rate": 2.13e-06, + "loss": 1.0199, + "step": 713 + }, + { + "epoch": 0.6030405405405406, + "grad_norm": 3.537868022918701, + "learning_rate": 2.133e-06, + "loss": 1.0219, + "step": 714 + }, + { + "epoch": 0.6038851351351351, + "grad_norm": 3.635286569595337, + "learning_rate": 2.136e-06, + "loss": 0.9711, + "step": 715 + }, + { + "epoch": 0.6047297297297297, + "grad_norm": 17.949939727783203, + "learning_rate": 2.139e-06, + "loss": 0.9686, + "step": 716 + }, + { + "epoch": 0.6055743243243243, + "grad_norm": 4.264177322387695, + "learning_rate": 2.1420000000000004e-06, + "loss": 1.0077, + "step": 717 + }, + { + "epoch": 0.606418918918919, + "grad_norm": 4.288735389709473, + "learning_rate": 2.145e-06, + "loss": 0.9703, + "step": 718 + }, + { + "epoch": 0.6072635135135135, + "grad_norm": 3.5944790840148926, + "learning_rate": 2.148e-06, + "loss": 0.9677, + "step": 719 + }, + { + "epoch": 0.6081081081081081, + "grad_norm": 4.007157802581787, + "learning_rate": 2.1510000000000002e-06, + "loss": 0.9899, + "step": 720 + }, + { + "epoch": 0.6089527027027027, + "grad_norm": 3.863156318664551, + "learning_rate": 2.154e-06, + "loss": 0.9376, + "step": 721 + }, + { + "epoch": 0.6097972972972973, + "grad_norm": 7.851414680480957, + "learning_rate": 2.1570000000000003e-06, + "loss": 0.9421, + "step": 722 + }, + { + "epoch": 0.6106418918918919, + "grad_norm": 7.9047980308532715, + "learning_rate": 2.16e-06, + "loss": 0.9165, + "step": 723 + }, + { + "epoch": 0.6114864864864865, + "grad_norm": 3.0743162631988525, + "learning_rate": 2.163e-06, + "loss": 0.9432, + "step": 724 + }, + { + "epoch": 0.612331081081081, + "grad_norm": 6.349869728088379, + "learning_rate": 2.166e-06, + "loss": 0.9418, + "step": 725 + }, + { + "epoch": 0.6131756756756757, + "grad_norm": 6.487222194671631, + "learning_rate": 2.1690000000000003e-06, + "loss": 0.8973, + "step": 726 + }, + { + "epoch": 0.6140202702702703, + "grad_norm": 2.9594638347625732, + "learning_rate": 2.172e-06, + "loss": 0.9131, + "step": 727 + }, + { + "epoch": 0.6148648648648649, + "grad_norm": 27.964290618896484, + "learning_rate": 2.175e-06, + "loss": 0.9453, + "step": 728 + }, + { + "epoch": 0.6157094594594594, + "grad_norm": 3.062879800796509, + "learning_rate": 2.178e-06, + "loss": 0.9122, + "step": 729 + }, + { + "epoch": 0.6165540540540541, + "grad_norm": 5.117274761199951, + "learning_rate": 2.181e-06, + "loss": 0.9328, + "step": 730 + }, + { + "epoch": 0.6173986486486487, + "grad_norm": 2.9588890075683594, + "learning_rate": 2.184e-06, + "loss": 0.8994, + "step": 731 + }, + { + "epoch": 0.6182432432432432, + "grad_norm": 2.8696770668029785, + "learning_rate": 2.1870000000000004e-06, + "loss": 0.86, + "step": 732 + }, + { + "epoch": 0.6190878378378378, + "grad_norm": 4.352910041809082, + "learning_rate": 2.1899999999999998e-06, + "loss": 0.9206, + "step": 733 + }, + { + "epoch": 0.6199324324324325, + "grad_norm": 11.425676345825195, + "learning_rate": 2.193e-06, + "loss": 0.9058, + "step": 734 + }, + { + "epoch": 0.620777027027027, + "grad_norm": 6.7926530838012695, + "learning_rate": 2.1960000000000002e-06, + "loss": 0.9008, + "step": 735 + }, + { + "epoch": 0.6216216216216216, + "grad_norm": 8.50149917602539, + "learning_rate": 2.199e-06, + "loss": 0.9372, + "step": 736 + }, + { + "epoch": 0.6224662162162162, + "grad_norm": 23.517080307006836, + "learning_rate": 2.2020000000000003e-06, + "loss": 0.9013, + "step": 737 + }, + { + "epoch": 0.6233108108108109, + "grad_norm": 6.931684494018555, + "learning_rate": 2.205e-06, + "loss": 0.9377, + "step": 738 + }, + { + "epoch": 0.6241554054054054, + "grad_norm": 2.940016031265259, + "learning_rate": 2.208e-06, + "loss": 0.8435, + "step": 739 + }, + { + "epoch": 0.625, + "grad_norm": 4.810202598571777, + "learning_rate": 2.211e-06, + "loss": 0.8691, + "step": 740 + }, + { + "epoch": 0.6258445945945946, + "grad_norm": 3.5142815113067627, + "learning_rate": 2.2140000000000003e-06, + "loss": 0.7826, + "step": 741 + }, + { + "epoch": 0.6266891891891891, + "grad_norm": 2.723093032836914, + "learning_rate": 2.2169999999999997e-06, + "loss": 0.814, + "step": 742 + }, + { + "epoch": 0.6275337837837838, + "grad_norm": 6.035440444946289, + "learning_rate": 2.22e-06, + "loss": 0.8012, + "step": 743 + }, + { + "epoch": 0.6283783783783784, + "grad_norm": 3.9914798736572266, + "learning_rate": 2.223e-06, + "loss": 0.8743, + "step": 744 + }, + { + "epoch": 0.629222972972973, + "grad_norm": 3.9104278087615967, + "learning_rate": 2.226e-06, + "loss": 0.9357, + "step": 745 + }, + { + "epoch": 0.6300675675675675, + "grad_norm": 4.174567222595215, + "learning_rate": 2.229e-06, + "loss": 0.8268, + "step": 746 + }, + { + "epoch": 0.6309121621621622, + "grad_norm": 6.695793151855469, + "learning_rate": 2.232e-06, + "loss": 0.8372, + "step": 747 + }, + { + "epoch": 0.6317567567567568, + "grad_norm": 4.269417762756348, + "learning_rate": 2.2349999999999998e-06, + "loss": 0.8713, + "step": 748 + }, + { + "epoch": 0.6326013513513513, + "grad_norm": 4.508630275726318, + "learning_rate": 2.238e-06, + "loss": 0.889, + "step": 749 + }, + { + "epoch": 0.6334459459459459, + "grad_norm": 5.624368667602539, + "learning_rate": 2.2410000000000002e-06, + "loss": 0.8896, + "step": 750 + }, + { + "epoch": 0.6342905405405406, + "grad_norm": 7.898432731628418, + "learning_rate": 2.244e-06, + "loss": 1.3851, + "step": 751 + }, + { + "epoch": 0.6351351351351351, + "grad_norm": 4.771108627319336, + "learning_rate": 2.247e-06, + "loss": 0.9797, + "step": 752 + }, + { + "epoch": 0.6359797297297297, + "grad_norm": 3.6501803398132324, + "learning_rate": 2.25e-06, + "loss": 0.7582, + "step": 753 + }, + { + "epoch": 0.6368243243243243, + "grad_norm": 3.1603565216064453, + "learning_rate": 2.253e-06, + "loss": 0.7282, + "step": 754 + }, + { + "epoch": 0.637668918918919, + "grad_norm": 3.265428066253662, + "learning_rate": 2.256e-06, + "loss": 0.7367, + "step": 755 + }, + { + "epoch": 0.6385135135135135, + "grad_norm": 7.252074718475342, + "learning_rate": 2.2590000000000003e-06, + "loss": 0.7344, + "step": 756 + }, + { + "epoch": 0.6393581081081081, + "grad_norm": 6.129350185394287, + "learning_rate": 2.262e-06, + "loss": 0.7376, + "step": 757 + }, + { + "epoch": 0.6402027027027027, + "grad_norm": 3.0135514736175537, + "learning_rate": 2.265e-06, + "loss": 0.6633, + "step": 758 + }, + { + "epoch": 0.6410472972972973, + "grad_norm": 2.9709243774414062, + "learning_rate": 2.268e-06, + "loss": 0.6777, + "step": 759 + }, + { + "epoch": 0.6418918918918919, + "grad_norm": 4.98696756362915, + "learning_rate": 2.2710000000000004e-06, + "loss": 0.7074, + "step": 760 + }, + { + "epoch": 0.6427364864864865, + "grad_norm": 5.997633934020996, + "learning_rate": 2.274e-06, + "loss": 0.6684, + "step": 761 + }, + { + "epoch": 0.643581081081081, + "grad_norm": 6.782534122467041, + "learning_rate": 2.277e-06, + "loss": 0.6895, + "step": 762 + }, + { + "epoch": 0.6444256756756757, + "grad_norm": 3.005436658859253, + "learning_rate": 2.28e-06, + "loss": 0.6918, + "step": 763 + }, + { + "epoch": 0.6452702702702703, + "grad_norm": 2.521484851837158, + "learning_rate": 2.283e-06, + "loss": 0.6333, + "step": 764 + }, + { + "epoch": 0.6461148648648649, + "grad_norm": 3.5694329738616943, + "learning_rate": 2.2860000000000002e-06, + "loss": 0.6496, + "step": 765 + }, + { + "epoch": 0.6469594594594594, + "grad_norm": 5.528240203857422, + "learning_rate": 2.2890000000000004e-06, + "loss": 0.7226, + "step": 766 + }, + { + "epoch": 0.6478040540540541, + "grad_norm": 5.213047504425049, + "learning_rate": 2.292e-06, + "loss": 0.7084, + "step": 767 + }, + { + "epoch": 0.6486486486486487, + "grad_norm": 3.5703866481781006, + "learning_rate": 2.295e-06, + "loss": 0.7179, + "step": 768 + }, + { + "epoch": 0.6494932432432432, + "grad_norm": 4.396407604217529, + "learning_rate": 2.2980000000000003e-06, + "loss": 0.6651, + "step": 769 + }, + { + "epoch": 0.6503378378378378, + "grad_norm": 3.6432571411132812, + "learning_rate": 2.301e-06, + "loss": 0.6744, + "step": 770 + }, + { + "epoch": 0.6511824324324325, + "grad_norm": 3.0347988605499268, + "learning_rate": 2.304e-06, + "loss": 0.6405, + "step": 771 + }, + { + "epoch": 0.652027027027027, + "grad_norm": 3.418261766433716, + "learning_rate": 2.307e-06, + "loss": 0.6403, + "step": 772 + }, + { + "epoch": 0.6528716216216216, + "grad_norm": 5.252307891845703, + "learning_rate": 2.31e-06, + "loss": 0.662, + "step": 773 + }, + { + "epoch": 0.6537162162162162, + "grad_norm": 3.67775821685791, + "learning_rate": 2.313e-06, + "loss": 0.639, + "step": 774 + }, + { + "epoch": 0.6545608108108109, + "grad_norm": 4.3481292724609375, + "learning_rate": 2.3160000000000004e-06, + "loss": 0.6281, + "step": 775 + }, + { + "epoch": 0.6554054054054054, + "grad_norm": 2.1848931312561035, + "learning_rate": 2.3189999999999997e-06, + "loss": 0.6137, + "step": 776 + }, + { + "epoch": 0.65625, + "grad_norm": 7.199524402618408, + "learning_rate": 2.322e-06, + "loss": 0.7031, + "step": 777 + }, + { + "epoch": 0.6570945945945946, + "grad_norm": 4.664922714233398, + "learning_rate": 2.325e-06, + "loss": 0.6198, + "step": 778 + }, + { + "epoch": 0.6579391891891891, + "grad_norm": 2.5665695667266846, + "learning_rate": 2.328e-06, + "loss": 0.6418, + "step": 779 + }, + { + "epoch": 0.6587837837837838, + "grad_norm": 4.647937297821045, + "learning_rate": 2.3310000000000002e-06, + "loss": 0.6688, + "step": 780 + }, + { + "epoch": 0.6596283783783784, + "grad_norm": 3.7760863304138184, + "learning_rate": 2.334e-06, + "loss": 0.6563, + "step": 781 + }, + { + "epoch": 0.660472972972973, + "grad_norm": 2.998284101486206, + "learning_rate": 2.337e-06, + "loss": 0.5742, + "step": 782 + }, + { + "epoch": 0.6613175675675675, + "grad_norm": 5.240099906921387, + "learning_rate": 2.34e-06, + "loss": 0.6773, + "step": 783 + }, + { + "epoch": 0.6621621621621622, + "grad_norm": 2.7156858444213867, + "learning_rate": 2.3430000000000003e-06, + "loss": 0.6221, + "step": 784 + }, + { + "epoch": 0.6630067567567568, + "grad_norm": 3.467273473739624, + "learning_rate": 2.346e-06, + "loss": 0.6603, + "step": 785 + }, + { + "epoch": 0.6638513513513513, + "grad_norm": 8.995895385742188, + "learning_rate": 2.349e-06, + "loss": 0.636, + "step": 786 + }, + { + "epoch": 0.6646959459459459, + "grad_norm": 3.149250030517578, + "learning_rate": 2.352e-06, + "loss": 0.6689, + "step": 787 + }, + { + "epoch": 0.6655405405405406, + "grad_norm": 3.017575263977051, + "learning_rate": 2.355e-06, + "loss": 0.7126, + "step": 788 + }, + { + "epoch": 0.6663851351351351, + "grad_norm": 3.933441400527954, + "learning_rate": 2.358e-06, + "loss": 0.6207, + "step": 789 + }, + { + "epoch": 0.6672297297297297, + "grad_norm": 3.0327939987182617, + "learning_rate": 2.3610000000000003e-06, + "loss": 0.6274, + "step": 790 + }, + { + "epoch": 0.6680743243243243, + "grad_norm": 5.854973793029785, + "learning_rate": 2.3639999999999997e-06, + "loss": 0.667, + "step": 791 + }, + { + "epoch": 0.668918918918919, + "grad_norm": 8.013816833496094, + "learning_rate": 2.367e-06, + "loss": 0.604, + "step": 792 + }, + { + "epoch": 0.6697635135135135, + "grad_norm": 7.039916515350342, + "learning_rate": 2.37e-06, + "loss": 0.6194, + "step": 793 + }, + { + "epoch": 0.6706081081081081, + "grad_norm": 8.158195495605469, + "learning_rate": 2.373e-06, + "loss": 0.6654, + "step": 794 + }, + { + "epoch": 0.6714527027027027, + "grad_norm": 3.802830457687378, + "learning_rate": 2.376e-06, + "loss": 0.6637, + "step": 795 + }, + { + "epoch": 0.6722972972972973, + "grad_norm": 7.659937858581543, + "learning_rate": 2.379e-06, + "loss": 0.703, + "step": 796 + }, + { + "epoch": 0.6731418918918919, + "grad_norm": 7.328330993652344, + "learning_rate": 2.382e-06, + "loss": 0.699, + "step": 797 + }, + { + "epoch": 0.6739864864864865, + "grad_norm": 3.4454867839813232, + "learning_rate": 2.385e-06, + "loss": 0.6489, + "step": 798 + }, + { + "epoch": 0.674831081081081, + "grad_norm": 23.224084854125977, + "learning_rate": 2.3880000000000003e-06, + "loss": 0.7197, + "step": 799 + }, + { + "epoch": 0.6756756756756757, + "grad_norm": 13.91353702545166, + "learning_rate": 2.391e-06, + "loss": 0.7697, + "step": 800 + }, + { + "epoch": 0.6765202702702703, + "grad_norm": 8.176276206970215, + "learning_rate": 2.394e-06, + "loss": 1.3262, + "step": 801 + }, + { + "epoch": 0.6773648648648649, + "grad_norm": 5.565192699432373, + "learning_rate": 2.397e-06, + "loss": 0.8093, + "step": 802 + }, + { + "epoch": 0.6782094594594594, + "grad_norm": 3.3387985229492188, + "learning_rate": 2.4000000000000003e-06, + "loss": 0.725, + "step": 803 + }, + { + "epoch": 0.6790540540540541, + "grad_norm": 4.488727569580078, + "learning_rate": 2.403e-06, + "loss": 0.5761, + "step": 804 + }, + { + "epoch": 0.6798986486486487, + "grad_norm": 3.47489857673645, + "learning_rate": 2.406e-06, + "loss": 0.585, + "step": 805 + }, + { + "epoch": 0.6807432432432432, + "grad_norm": 2.583421468734741, + "learning_rate": 2.409e-06, + "loss": 0.5768, + "step": 806 + }, + { + "epoch": 0.6815878378378378, + "grad_norm": 3.4671075344085693, + "learning_rate": 2.412e-06, + "loss": 0.5079, + "step": 807 + }, + { + "epoch": 0.6824324324324325, + "grad_norm": 4.5468878746032715, + "learning_rate": 2.415e-06, + "loss": 0.6022, + "step": 808 + }, + { + "epoch": 0.683277027027027, + "grad_norm": 4.267704486846924, + "learning_rate": 2.4180000000000004e-06, + "loss": 0.5725, + "step": 809 + }, + { + "epoch": 0.6841216216216216, + "grad_norm": 2.755401134490967, + "learning_rate": 2.4209999999999998e-06, + "loss": 0.5441, + "step": 810 + }, + { + "epoch": 0.6849662162162162, + "grad_norm": 3.214634418487549, + "learning_rate": 2.424e-06, + "loss": 0.5524, + "step": 811 + }, + { + "epoch": 0.6858108108108109, + "grad_norm": 3.806959867477417, + "learning_rate": 2.4270000000000002e-06, + "loss": 0.5165, + "step": 812 + }, + { + "epoch": 0.6866554054054054, + "grad_norm": 10.201401710510254, + "learning_rate": 2.43e-06, + "loss": 0.5893, + "step": 813 + }, + { + "epoch": 0.6875, + "grad_norm": 3.0985934734344482, + "learning_rate": 2.4330000000000003e-06, + "loss": 0.6354, + "step": 814 + }, + { + "epoch": 0.6883445945945946, + "grad_norm": 3.772660970687866, + "learning_rate": 2.436e-06, + "loss": 0.542, + "step": 815 + }, + { + "epoch": 0.6891891891891891, + "grad_norm": 3.5363986492156982, + "learning_rate": 2.439e-06, + "loss": 0.5033, + "step": 816 + }, + { + "epoch": 0.6900337837837838, + "grad_norm": 3.652510166168213, + "learning_rate": 2.442e-06, + "loss": 0.5428, + "step": 817 + }, + { + "epoch": 0.6908783783783784, + "grad_norm": 3.403696298599243, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.5864, + "step": 818 + }, + { + "epoch": 0.691722972972973, + "grad_norm": 3.365229606628418, + "learning_rate": 2.448e-06, + "loss": 0.5592, + "step": 819 + }, + { + "epoch": 0.6925675675675675, + "grad_norm": 2.2869858741760254, + "learning_rate": 2.451e-06, + "loss": 0.5384, + "step": 820 + }, + { + "epoch": 0.6934121621621622, + "grad_norm": 3.409757137298584, + "learning_rate": 2.454e-06, + "loss": 0.4989, + "step": 821 + }, + { + "epoch": 0.6942567567567568, + "grad_norm": 5.010795593261719, + "learning_rate": 2.457e-06, + "loss": 0.6218, + "step": 822 + }, + { + "epoch": 0.6951013513513513, + "grad_norm": 3.225161075592041, + "learning_rate": 2.46e-06, + "loss": 0.5242, + "step": 823 + }, + { + "epoch": 0.6959459459459459, + "grad_norm": 5.928736686706543, + "learning_rate": 2.4630000000000004e-06, + "loss": 0.5691, + "step": 824 + }, + { + "epoch": 0.6967905405405406, + "grad_norm": 3.5420708656311035, + "learning_rate": 2.4659999999999998e-06, + "loss": 0.4903, + "step": 825 + }, + { + "epoch": 0.6976351351351351, + "grad_norm": 7.1074724197387695, + "learning_rate": 2.469e-06, + "loss": 0.5824, + "step": 826 + }, + { + "epoch": 0.6984797297297297, + "grad_norm": 6.070399761199951, + "learning_rate": 2.4720000000000002e-06, + "loss": 0.4987, + "step": 827 + }, + { + "epoch": 0.6993243243243243, + "grad_norm": 6.523897647857666, + "learning_rate": 2.475e-06, + "loss": 0.5081, + "step": 828 + }, + { + "epoch": 0.700168918918919, + "grad_norm": 2.541332960128784, + "learning_rate": 2.4780000000000002e-06, + "loss": 0.5366, + "step": 829 + }, + { + "epoch": 0.7010135135135135, + "grad_norm": 3.727370262145996, + "learning_rate": 2.481e-06, + "loss": 0.531, + "step": 830 + }, + { + "epoch": 0.7018581081081081, + "grad_norm": 4.177402019500732, + "learning_rate": 2.484e-06, + "loss": 0.4882, + "step": 831 + }, + { + "epoch": 0.7027027027027027, + "grad_norm": 3.2874255180358887, + "learning_rate": 2.487e-06, + "loss": 0.5532, + "step": 832 + }, + { + "epoch": 0.7035472972972973, + "grad_norm": 3.497114419937134, + "learning_rate": 2.4900000000000003e-06, + "loss": 0.514, + "step": 833 + }, + { + "epoch": 0.7043918918918919, + "grad_norm": 4.828718185424805, + "learning_rate": 2.4929999999999997e-06, + "loss": 0.4967, + "step": 834 + }, + { + "epoch": 0.7052364864864865, + "grad_norm": 2.184201717376709, + "learning_rate": 2.496e-06, + "loss": 0.4994, + "step": 835 + }, + { + "epoch": 0.706081081081081, + "grad_norm": 3.149679660797119, + "learning_rate": 2.499e-06, + "loss": 0.5264, + "step": 836 + }, + { + "epoch": 0.7069256756756757, + "grad_norm": 6.153497219085693, + "learning_rate": 2.502e-06, + "loss": 0.5363, + "step": 837 + }, + { + "epoch": 0.7077702702702703, + "grad_norm": 3.1986303329467773, + "learning_rate": 2.505e-06, + "loss": 0.5391, + "step": 838 + }, + { + "epoch": 0.7086148648648649, + "grad_norm": 4.187166690826416, + "learning_rate": 2.508e-06, + "loss": 0.5603, + "step": 839 + }, + { + "epoch": 0.7094594594594594, + "grad_norm": 4.0729289054870605, + "learning_rate": 2.5109999999999998e-06, + "loss": 0.5159, + "step": 840 + }, + { + "epoch": 0.7103040540540541, + "grad_norm": 4.785458087921143, + "learning_rate": 2.514e-06, + "loss": 0.5537, + "step": 841 + }, + { + "epoch": 0.7111486486486487, + "grad_norm": 4.091228485107422, + "learning_rate": 2.517e-06, + "loss": 0.56, + "step": 842 + }, + { + "epoch": 0.7119932432432432, + "grad_norm": 3.181784152984619, + "learning_rate": 2.52e-06, + "loss": 0.524, + "step": 843 + }, + { + "epoch": 0.7128378378378378, + "grad_norm": 6.5293121337890625, + "learning_rate": 2.523e-06, + "loss": 0.552, + "step": 844 + }, + { + "epoch": 0.7136824324324325, + "grad_norm": 2.4628098011016846, + "learning_rate": 2.526e-06, + "loss": 0.5142, + "step": 845 + }, + { + "epoch": 0.714527027027027, + "grad_norm": 3.7066078186035156, + "learning_rate": 2.5290000000000003e-06, + "loss": 0.5491, + "step": 846 + }, + { + "epoch": 0.7153716216216216, + "grad_norm": 7.073708534240723, + "learning_rate": 2.532e-06, + "loss": 0.5306, + "step": 847 + }, + { + "epoch": 0.7162162162162162, + "grad_norm": 2.844613790512085, + "learning_rate": 2.5350000000000003e-06, + "loss": 0.5307, + "step": 848 + }, + { + "epoch": 0.7170608108108109, + "grad_norm": 3.562270402908325, + "learning_rate": 2.538e-06, + "loss": 0.6552, + "step": 849 + }, + { + "epoch": 0.7179054054054054, + "grad_norm": 19.02691650390625, + "learning_rate": 2.541e-06, + "loss": 0.7825, + "step": 850 + }, + { + "epoch": 0.71875, + "grad_norm": 55.15711975097656, + "learning_rate": 2.544e-06, + "loss": 1.1458, + "step": 851 + }, + { + "epoch": 0.7195945945945946, + "grad_norm": 13.81608772277832, + "learning_rate": 2.5470000000000003e-06, + "loss": 0.8639, + "step": 852 + }, + { + "epoch": 0.7204391891891891, + "grad_norm": 2.7492177486419678, + "learning_rate": 2.55e-06, + "loss": 0.5985, + "step": 853 + }, + { + "epoch": 0.7212837837837838, + "grad_norm": 3.1377856731414795, + "learning_rate": 2.553e-06, + "loss": 0.4998, + "step": 854 + }, + { + "epoch": 0.7221283783783784, + "grad_norm": 17.51803207397461, + "learning_rate": 2.556e-06, + "loss": 0.571, + "step": 855 + }, + { + "epoch": 0.722972972972973, + "grad_norm": 3.809901475906372, + "learning_rate": 2.559e-06, + "loss": 0.5083, + "step": 856 + }, + { + "epoch": 0.7238175675675675, + "grad_norm": 5.294857025146484, + "learning_rate": 2.562e-06, + "loss": 0.4588, + "step": 857 + }, + { + "epoch": 0.7246621621621622, + "grad_norm": 3.7559895515441895, + "learning_rate": 2.5650000000000004e-06, + "loss": 0.4687, + "step": 858 + }, + { + "epoch": 0.7255067567567568, + "grad_norm": 2.357083559036255, + "learning_rate": 2.568e-06, + "loss": 0.4594, + "step": 859 + }, + { + "epoch": 0.7263513513513513, + "grad_norm": 3.084779739379883, + "learning_rate": 2.571e-06, + "loss": 0.457, + "step": 860 + }, + { + "epoch": 0.7271959459459459, + "grad_norm": 4.916243553161621, + "learning_rate": 2.5740000000000003e-06, + "loss": 0.4287, + "step": 861 + }, + { + "epoch": 0.7280405405405406, + "grad_norm": 5.821186542510986, + "learning_rate": 2.577e-06, + "loss": 0.5208, + "step": 862 + }, + { + "epoch": 0.7288851351351351, + "grad_norm": 5.166162490844727, + "learning_rate": 2.58e-06, + "loss": 0.4793, + "step": 863 + }, + { + "epoch": 0.7297297297297297, + "grad_norm": 2.974898338317871, + "learning_rate": 2.583e-06, + "loss": 0.4659, + "step": 864 + }, + { + "epoch": 0.7305743243243243, + "grad_norm": 2.615957021713257, + "learning_rate": 2.586e-06, + "loss": 0.4299, + "step": 865 + }, + { + "epoch": 0.731418918918919, + "grad_norm": 2.0424411296844482, + "learning_rate": 2.589e-06, + "loss": 0.4189, + "step": 866 + }, + { + "epoch": 0.7322635135135135, + "grad_norm": 15.728718757629395, + "learning_rate": 2.5920000000000003e-06, + "loss": 0.4744, + "step": 867 + }, + { + "epoch": 0.7331081081081081, + "grad_norm": 1.9623162746429443, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.44, + "step": 868 + }, + { + "epoch": 0.7339527027027027, + "grad_norm": 2.5169553756713867, + "learning_rate": 2.598e-06, + "loss": 0.4864, + "step": 869 + }, + { + "epoch": 0.7347972972972973, + "grad_norm": 3.419424057006836, + "learning_rate": 2.601e-06, + "loss": 0.4578, + "step": 870 + }, + { + "epoch": 0.7356418918918919, + "grad_norm": 3.695387840270996, + "learning_rate": 2.604e-06, + "loss": 0.4721, + "step": 871 + }, + { + "epoch": 0.7364864864864865, + "grad_norm": 4.312934875488281, + "learning_rate": 2.607e-06, + "loss": 0.4684, + "step": 872 + }, + { + "epoch": 0.737331081081081, + "grad_norm": 5.8416547775268555, + "learning_rate": 2.61e-06, + "loss": 0.4629, + "step": 873 + }, + { + "epoch": 0.7381756756756757, + "grad_norm": 4.334590911865234, + "learning_rate": 2.613e-06, + "loss": 0.4653, + "step": 874 + }, + { + "epoch": 0.7390202702702703, + "grad_norm": 2.7522895336151123, + "learning_rate": 2.616e-06, + "loss": 0.4801, + "step": 875 + }, + { + "epoch": 0.7398648648648649, + "grad_norm": 3.2054061889648438, + "learning_rate": 2.6190000000000003e-06, + "loss": 0.4669, + "step": 876 + }, + { + "epoch": 0.7407094594594594, + "grad_norm": 2.645944833755493, + "learning_rate": 2.622e-06, + "loss": 0.4411, + "step": 877 + }, + { + "epoch": 0.7415540540540541, + "grad_norm": 2.3458149433135986, + "learning_rate": 2.625e-06, + "loss": 0.4733, + "step": 878 + }, + { + "epoch": 0.7423986486486487, + "grad_norm": 3.465505599975586, + "learning_rate": 2.628e-06, + "loss": 0.4048, + "step": 879 + }, + { + "epoch": 0.7432432432432432, + "grad_norm": 8.365090370178223, + "learning_rate": 2.631e-06, + "loss": 0.4819, + "step": 880 + }, + { + "epoch": 0.7440878378378378, + "grad_norm": 8.341907501220703, + "learning_rate": 2.634e-06, + "loss": 0.4512, + "step": 881 + }, + { + "epoch": 0.7449324324324325, + "grad_norm": 5.6314191818237305, + "learning_rate": 2.6370000000000003e-06, + "loss": 0.4383, + "step": 882 + }, + { + "epoch": 0.745777027027027, + "grad_norm": 4.011555194854736, + "learning_rate": 2.6399999999999997e-06, + "loss": 0.5621, + "step": 883 + }, + { + "epoch": 0.7466216216216216, + "grad_norm": 2.473520517349243, + "learning_rate": 2.643e-06, + "loss": 0.4577, + "step": 884 + }, + { + "epoch": 0.7474662162162162, + "grad_norm": 2.6149260997772217, + "learning_rate": 2.646e-06, + "loss": 0.4402, + "step": 885 + }, + { + "epoch": 0.7483108108108109, + "grad_norm": 3.0843722820281982, + "learning_rate": 2.649e-06, + "loss": 0.4673, + "step": 886 + }, + { + "epoch": 0.7491554054054054, + "grad_norm": 2.9637701511383057, + "learning_rate": 2.652e-06, + "loss": 0.5049, + "step": 887 + }, + { + "epoch": 0.75, + "grad_norm": 3.208170175552368, + "learning_rate": 2.655e-06, + "loss": 0.4827, + "step": 888 + }, + { + "epoch": 0.7508445945945946, + "grad_norm": 35.38051986694336, + "learning_rate": 2.6580000000000002e-06, + "loss": 0.4879, + "step": 889 + }, + { + "epoch": 0.7516891891891891, + "grad_norm": 3.3243038654327393, + "learning_rate": 2.661e-06, + "loss": 0.4501, + "step": 890 + }, + { + "epoch": 0.7525337837837838, + "grad_norm": 6.760805606842041, + "learning_rate": 2.6640000000000002e-06, + "loss": 0.4601, + "step": 891 + }, + { + "epoch": 0.7533783783783784, + "grad_norm": 4.376120567321777, + "learning_rate": 2.6670000000000005e-06, + "loss": 0.4854, + "step": 892 + }, + { + "epoch": 0.754222972972973, + "grad_norm": 4.403983116149902, + "learning_rate": 2.67e-06, + "loss": 0.4481, + "step": 893 + }, + { + "epoch": 0.7550675675675675, + "grad_norm": 3.245659828186035, + "learning_rate": 2.673e-06, + "loss": 0.4717, + "step": 894 + }, + { + "epoch": 0.7559121621621622, + "grad_norm": 3.2111103534698486, + "learning_rate": 2.6760000000000003e-06, + "loss": 0.4098, + "step": 895 + }, + { + "epoch": 0.7567567567567568, + "grad_norm": 6.298848628997803, + "learning_rate": 2.679e-06, + "loss": 0.504, + "step": 896 + }, + { + "epoch": 0.7576013513513513, + "grad_norm": 3.531400680541992, + "learning_rate": 2.682e-06, + "loss": 0.4698, + "step": 897 + }, + { + "epoch": 0.7584459459459459, + "grad_norm": 9.439323425292969, + "learning_rate": 2.685e-06, + "loss": 0.5042, + "step": 898 + }, + { + "epoch": 0.7592905405405406, + "grad_norm": 21.46000099182129, + "learning_rate": 2.688e-06, + "loss": 0.561, + "step": 899 + }, + { + "epoch": 0.7601351351351351, + "grad_norm": 5.715875625610352, + "learning_rate": 2.691e-06, + "loss": 0.7504, + "step": 900 + }, + { + "epoch": 0.7609797297297297, + "grad_norm": 11.788183212280273, + "learning_rate": 2.6940000000000004e-06, + "loss": 1.4098, + "step": 901 + }, + { + "epoch": 0.7618243243243243, + "grad_norm": 20.271183013916016, + "learning_rate": 2.6969999999999998e-06, + "loss": 0.9004, + "step": 902 + }, + { + "epoch": 0.762668918918919, + "grad_norm": 4.690286159515381, + "learning_rate": 2.7e-06, + "loss": 0.572, + "step": 903 + }, + { + "epoch": 0.7635135135135135, + "grad_norm": 2.946943521499634, + "learning_rate": 2.703e-06, + "loss": 0.5096, + "step": 904 + }, + { + "epoch": 0.7643581081081081, + "grad_norm": 3.802534341812134, + "learning_rate": 2.706e-06, + "loss": 0.4806, + "step": 905 + }, + { + "epoch": 0.7652027027027027, + "grad_norm": 3.126021385192871, + "learning_rate": 2.7090000000000002e-06, + "loss": 0.4075, + "step": 906 + }, + { + "epoch": 0.7660472972972973, + "grad_norm": 2.7308380603790283, + "learning_rate": 2.712e-06, + "loss": 0.416, + "step": 907 + }, + { + "epoch": 0.7668918918918919, + "grad_norm": 9.571621894836426, + "learning_rate": 2.715e-06, + "loss": 0.4878, + "step": 908 + }, + { + "epoch": 0.7677364864864865, + "grad_norm": 3.5891973972320557, + "learning_rate": 2.718e-06, + "loss": 0.4935, + "step": 909 + }, + { + "epoch": 0.768581081081081, + "grad_norm": 2.413743257522583, + "learning_rate": 2.7210000000000003e-06, + "loss": 0.3803, + "step": 910 + }, + { + "epoch": 0.7694256756756757, + "grad_norm": 2.606523036956787, + "learning_rate": 2.724e-06, + "loss": 0.4451, + "step": 911 + }, + { + "epoch": 0.7702702702702703, + "grad_norm": 2.9152355194091797, + "learning_rate": 2.727e-06, + "loss": 0.4436, + "step": 912 + }, + { + "epoch": 0.7711148648648649, + "grad_norm": 5.7266459465026855, + "learning_rate": 2.73e-06, + "loss": 0.4157, + "step": 913 + }, + { + "epoch": 0.7719594594594594, + "grad_norm": 4.879587173461914, + "learning_rate": 2.733e-06, + "loss": 0.4138, + "step": 914 + }, + { + "epoch": 0.7728040540540541, + "grad_norm": 2.3446836471557617, + "learning_rate": 2.736e-06, + "loss": 0.3991, + "step": 915 + }, + { + "epoch": 0.7736486486486487, + "grad_norm": 2.380108594894409, + "learning_rate": 2.7390000000000004e-06, + "loss": 0.4046, + "step": 916 + }, + { + "epoch": 0.7744932432432432, + "grad_norm": 2.473540782928467, + "learning_rate": 2.7419999999999998e-06, + "loss": 0.3955, + "step": 917 + }, + { + "epoch": 0.7753378378378378, + "grad_norm": 2.7485885620117188, + "learning_rate": 2.745e-06, + "loss": 0.4536, + "step": 918 + }, + { + "epoch": 0.7761824324324325, + "grad_norm": 2.585313081741333, + "learning_rate": 2.748e-06, + "loss": 0.4062, + "step": 919 + }, + { + "epoch": 0.777027027027027, + "grad_norm": 2.742379903793335, + "learning_rate": 2.751e-06, + "loss": 0.3962, + "step": 920 + }, + { + "epoch": 0.7778716216216216, + "grad_norm": 2.213240623474121, + "learning_rate": 2.7540000000000002e-06, + "loss": 0.3703, + "step": 921 + }, + { + "epoch": 0.7787162162162162, + "grad_norm": 2.545355796813965, + "learning_rate": 2.757e-06, + "loss": 0.4037, + "step": 922 + }, + { + "epoch": 0.7795608108108109, + "grad_norm": 7.319743633270264, + "learning_rate": 2.76e-06, + "loss": 0.4035, + "step": 923 + }, + { + "epoch": 0.7804054054054054, + "grad_norm": 5.70536994934082, + "learning_rate": 2.763e-06, + "loss": 0.4367, + "step": 924 + }, + { + "epoch": 0.78125, + "grad_norm": 2.9482364654541016, + "learning_rate": 2.7660000000000003e-06, + "loss": 0.4243, + "step": 925 + }, + { + "epoch": 0.7820945945945946, + "grad_norm": 2.9586501121520996, + "learning_rate": 2.7689999999999997e-06, + "loss": 0.4452, + "step": 926 + }, + { + "epoch": 0.7829391891891891, + "grad_norm": 3.626885414123535, + "learning_rate": 2.772e-06, + "loss": 0.3793, + "step": 927 + }, + { + "epoch": 0.7837837837837838, + "grad_norm": 6.151919364929199, + "learning_rate": 2.775e-06, + "loss": 0.386, + "step": 928 + }, + { + "epoch": 0.7846283783783784, + "grad_norm": 2.452312707901001, + "learning_rate": 2.778e-06, + "loss": 0.436, + "step": 929 + }, + { + "epoch": 0.785472972972973, + "grad_norm": 5.526254177093506, + "learning_rate": 2.781e-06, + "loss": 0.4472, + "step": 930 + }, + { + "epoch": 0.7863175675675675, + "grad_norm": 2.7901716232299805, + "learning_rate": 2.784e-06, + "loss": 0.4535, + "step": 931 + }, + { + "epoch": 0.7871621621621622, + "grad_norm": 2.913891077041626, + "learning_rate": 2.787e-06, + "loss": 0.4623, + "step": 932 + }, + { + "epoch": 0.7880067567567568, + "grad_norm": 5.771517276763916, + "learning_rate": 2.79e-06, + "loss": 0.4111, + "step": 933 + }, + { + "epoch": 0.7888513513513513, + "grad_norm": 5.966020584106445, + "learning_rate": 2.793e-06, + "loss": 0.4772, + "step": 934 + }, + { + "epoch": 0.7896959459459459, + "grad_norm": 3.626677989959717, + "learning_rate": 2.7960000000000004e-06, + "loss": 0.4354, + "step": 935 + }, + { + "epoch": 0.7905405405405406, + "grad_norm": 4.778061866760254, + "learning_rate": 2.799e-06, + "loss": 0.3943, + "step": 936 + }, + { + "epoch": 0.7913851351351351, + "grad_norm": 7.907177925109863, + "learning_rate": 2.802e-06, + "loss": 0.4193, + "step": 937 + }, + { + "epoch": 0.7922297297297297, + "grad_norm": 4.2070465087890625, + "learning_rate": 2.8050000000000002e-06, + "loss": 0.4057, + "step": 938 + }, + { + "epoch": 0.7930743243243243, + "grad_norm": 3.1167190074920654, + "learning_rate": 2.808e-06, + "loss": 0.4168, + "step": 939 + }, + { + "epoch": 0.793918918918919, + "grad_norm": 3.933281660079956, + "learning_rate": 2.8110000000000003e-06, + "loss": 0.4427, + "step": 940 + }, + { + "epoch": 0.7947635135135135, + "grad_norm": 2.4652369022369385, + "learning_rate": 2.814e-06, + "loss": 0.4855, + "step": 941 + }, + { + "epoch": 0.7956081081081081, + "grad_norm": 3.2439510822296143, + "learning_rate": 2.817e-06, + "loss": 0.4605, + "step": 942 + }, + { + "epoch": 0.7964527027027027, + "grad_norm": 3.180105686187744, + "learning_rate": 2.82e-06, + "loss": 0.406, + "step": 943 + }, + { + "epoch": 0.7972972972972973, + "grad_norm": 3.4609155654907227, + "learning_rate": 2.8230000000000003e-06, + "loss": 0.4376, + "step": 944 + }, + { + "epoch": 0.7981418918918919, + "grad_norm": 5.133181095123291, + "learning_rate": 2.826e-06, + "loss": 0.4436, + "step": 945 + }, + { + "epoch": 0.7989864864864865, + "grad_norm": 3.329972505569458, + "learning_rate": 2.829e-06, + "loss": 0.4757, + "step": 946 + }, + { + "epoch": 0.799831081081081, + "grad_norm": 4.459960460662842, + "learning_rate": 2.832e-06, + "loss": 0.4294, + "step": 947 + }, + { + "epoch": 0.8006756756756757, + "grad_norm": 6.41441011428833, + "learning_rate": 2.835e-06, + "loss": 0.4338, + "step": 948 + }, + { + "epoch": 0.8015202702702703, + "grad_norm": 6.200233459472656, + "learning_rate": 2.838e-06, + "loss": 0.4665, + "step": 949 + }, + { + "epoch": 0.8023648648648649, + "grad_norm": 6.074668884277344, + "learning_rate": 2.8410000000000004e-06, + "loss": 0.6139, + "step": 950 + }, + { + "epoch": 0.8032094594594594, + "grad_norm": 9.062395095825195, + "learning_rate": 2.844e-06, + "loss": 1.1244, + "step": 951 + }, + { + "epoch": 0.8040540540540541, + "grad_norm": 7.925955772399902, + "learning_rate": 2.847e-06, + "loss": 0.7087, + "step": 952 + }, + { + "epoch": 0.8048986486486487, + "grad_norm": 13.363890647888184, + "learning_rate": 2.8500000000000002e-06, + "loss": 0.4924, + "step": 953 + }, + { + "epoch": 0.8057432432432432, + "grad_norm": 2.6526026725769043, + "learning_rate": 2.853e-06, + "loss": 0.4522, + "step": 954 + }, + { + "epoch": 0.8065878378378378, + "grad_norm": 1.9380518198013306, + "learning_rate": 2.8560000000000003e-06, + "loss": 0.4073, + "step": 955 + }, + { + "epoch": 0.8074324324324325, + "grad_norm": 3.3135671615600586, + "learning_rate": 2.859e-06, + "loss": 0.39, + "step": 956 + }, + { + "epoch": 0.808277027027027, + "grad_norm": 2.910343647003174, + "learning_rate": 2.862e-06, + "loss": 0.4741, + "step": 957 + }, + { + "epoch": 0.8091216216216216, + "grad_norm": 2.250469923019409, + "learning_rate": 2.865e-06, + "loss": 0.4164, + "step": 958 + }, + { + "epoch": 0.8099662162162162, + "grad_norm": 2.1899232864379883, + "learning_rate": 2.8680000000000003e-06, + "loss": 0.3975, + "step": 959 + }, + { + "epoch": 0.8108108108108109, + "grad_norm": 2.4764249324798584, + "learning_rate": 2.8709999999999997e-06, + "loss": 0.4268, + "step": 960 + }, + { + "epoch": 0.8116554054054054, + "grad_norm": 2.420175552368164, + "learning_rate": 2.874e-06, + "loss": 0.4103, + "step": 961 + }, + { + "epoch": 0.8125, + "grad_norm": 3.1002845764160156, + "learning_rate": 2.877e-06, + "loss": 0.4134, + "step": 962 + }, + { + "epoch": 0.8133445945945946, + "grad_norm": 2.7466914653778076, + "learning_rate": 2.88e-06, + "loss": 0.3802, + "step": 963 + }, + { + "epoch": 0.8141891891891891, + "grad_norm": 2.2749295234680176, + "learning_rate": 2.883e-06, + "loss": 0.3667, + "step": 964 + }, + { + "epoch": 0.8150337837837838, + "grad_norm": 2.0592689514160156, + "learning_rate": 2.886e-06, + "loss": 0.3884, + "step": 965 + }, + { + "epoch": 0.8158783783783784, + "grad_norm": 2.203829288482666, + "learning_rate": 2.8889999999999998e-06, + "loss": 0.4112, + "step": 966 + }, + { + "epoch": 0.816722972972973, + "grad_norm": 3.822779893875122, + "learning_rate": 2.892e-06, + "loss": 0.3669, + "step": 967 + }, + { + "epoch": 0.8175675675675675, + "grad_norm": 2.5669620037078857, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.3848, + "step": 968 + }, + { + "epoch": 0.8184121621621622, + "grad_norm": 2.08593487739563, + "learning_rate": 2.898e-06, + "loss": 0.4098, + "step": 969 + }, + { + "epoch": 0.8192567567567568, + "grad_norm": 2.4915738105773926, + "learning_rate": 2.901e-06, + "loss": 0.3805, + "step": 970 + }, + { + "epoch": 0.8201013513513513, + "grad_norm": 2.885053873062134, + "learning_rate": 2.904e-06, + "loss": 0.3806, + "step": 971 + }, + { + "epoch": 0.8209459459459459, + "grad_norm": 9.08874225616455, + "learning_rate": 2.907e-06, + "loss": 0.3613, + "step": 972 + }, + { + "epoch": 0.8217905405405406, + "grad_norm": 2.276305913925171, + "learning_rate": 2.91e-06, + "loss": 0.3979, + "step": 973 + }, + { + "epoch": 0.8226351351351351, + "grad_norm": 4.339950084686279, + "learning_rate": 2.9130000000000003e-06, + "loss": 0.3836, + "step": 974 + }, + { + "epoch": 0.8234797297297297, + "grad_norm": 2.1599605083465576, + "learning_rate": 2.916e-06, + "loss": 0.3679, + "step": 975 + }, + { + "epoch": 0.8243243243243243, + "grad_norm": 3.6263608932495117, + "learning_rate": 2.919e-06, + "loss": 0.4208, + "step": 976 + }, + { + "epoch": 0.825168918918919, + "grad_norm": 2.404494047164917, + "learning_rate": 2.922e-06, + "loss": 0.3716, + "step": 977 + }, + { + "epoch": 0.8260135135135135, + "grad_norm": 3.409390926361084, + "learning_rate": 2.9250000000000004e-06, + "loss": 0.3893, + "step": 978 + }, + { + "epoch": 0.8268581081081081, + "grad_norm": 4.52224588394165, + "learning_rate": 2.928e-06, + "loss": 0.4133, + "step": 979 + }, + { + "epoch": 0.8277027027027027, + "grad_norm": 2.1830379962921143, + "learning_rate": 2.931e-06, + "loss": 0.4162, + "step": 980 + }, + { + "epoch": 0.8285472972972973, + "grad_norm": 2.695852041244507, + "learning_rate": 2.934e-06, + "loss": 0.3693, + "step": 981 + }, + { + "epoch": 0.8293918918918919, + "grad_norm": 2.0979788303375244, + "learning_rate": 2.937e-06, + "loss": 0.4169, + "step": 982 + }, + { + "epoch": 0.8302364864864865, + "grad_norm": 2.9397168159484863, + "learning_rate": 2.9400000000000002e-06, + "loss": 0.3691, + "step": 983 + }, + { + "epoch": 0.831081081081081, + "grad_norm": 4.680749893188477, + "learning_rate": 2.9430000000000005e-06, + "loss": 0.3694, + "step": 984 + }, + { + "epoch": 0.8319256756756757, + "grad_norm": 6.66609001159668, + "learning_rate": 2.946e-06, + "loss": 0.3808, + "step": 985 + }, + { + "epoch": 0.8327702702702703, + "grad_norm": 2.0889697074890137, + "learning_rate": 2.949e-06, + "loss": 0.3686, + "step": 986 + }, + { + "epoch": 0.8336148648648649, + "grad_norm": 7.404335975646973, + "learning_rate": 2.9520000000000003e-06, + "loss": 0.415, + "step": 987 + }, + { + "epoch": 0.8344594594594594, + "grad_norm": 2.410938262939453, + "learning_rate": 2.955e-06, + "loss": 0.3965, + "step": 988 + }, + { + "epoch": 0.8353040540540541, + "grad_norm": 2.214912176132202, + "learning_rate": 2.958e-06, + "loss": 0.4163, + "step": 989 + }, + { + "epoch": 0.8361486486486487, + "grad_norm": 2.8889362812042236, + "learning_rate": 2.961e-06, + "loss": 0.4452, + "step": 990 + }, + { + "epoch": 0.8369932432432432, + "grad_norm": 2.082115411758423, + "learning_rate": 2.964e-06, + "loss": 0.3986, + "step": 991 + }, + { + "epoch": 0.8378378378378378, + "grad_norm": 2.8310582637786865, + "learning_rate": 2.967e-06, + "loss": 0.4541, + "step": 992 + }, + { + "epoch": 0.8386824324324325, + "grad_norm": 6.203836441040039, + "learning_rate": 2.9700000000000004e-06, + "loss": 0.4673, + "step": 993 + }, + { + "epoch": 0.839527027027027, + "grad_norm": 2.9049460887908936, + "learning_rate": 2.9729999999999997e-06, + "loss": 0.3973, + "step": 994 + }, + { + "epoch": 0.8403716216216216, + "grad_norm": 3.585684061050415, + "learning_rate": 2.976e-06, + "loss": 0.4524, + "step": 995 + }, + { + "epoch": 0.8412162162162162, + "grad_norm": 4.523475170135498, + "learning_rate": 2.979e-06, + "loss": 0.4871, + "step": 996 + }, + { + "epoch": 0.8420608108108109, + "grad_norm": 3.365731716156006, + "learning_rate": 2.982e-06, + "loss": 0.4479, + "step": 997 + }, + { + "epoch": 0.8429054054054054, + "grad_norm": 3.8719980716705322, + "learning_rate": 2.9850000000000002e-06, + "loss": 0.4533, + "step": 998 + }, + { + "epoch": 0.84375, + "grad_norm": 3.979008197784424, + "learning_rate": 2.988e-06, + "loss": 0.513, + "step": 999 + }, + { + "epoch": 0.8445945945945946, + "grad_norm": 5.626384258270264, + "learning_rate": 2.991e-06, + "loss": 0.469, + "step": 1000 + }, + { + "epoch": 0.8445945945945946, + "eval_cer": 0.19926110352582121, + "eval_loss": 0.8816050291061401, + "eval_runtime": 7.5169, + "eval_samples_per_second": 13.436, + "eval_steps_per_second": 0.532, + "eval_wer": 0.6084656084656085, + "step": 1000 + }, + { + "epoch": 0.8454391891891891, + "grad_norm": 8.031661987304688, + "learning_rate": 2.994e-06, + "loss": 1.0552, + "step": 1001 + }, + { + "epoch": 0.8462837837837838, + "grad_norm": 5.172723293304443, + "learning_rate": 2.9970000000000003e-06, + "loss": 0.8773, + "step": 1002 + }, + { + "epoch": 0.8471283783783784, + "grad_norm": 2.9077281951904297, + "learning_rate": 3e-06, + "loss": 0.4483, + "step": 1003 + }, + { + "epoch": 0.847972972972973, + "grad_norm": 2.970555543899536, + "learning_rate": 3.003e-06, + "loss": 0.3897, + "step": 1004 + }, + { + "epoch": 0.8488175675675675, + "grad_norm": 2.8808038234710693, + "learning_rate": 3.006e-06, + "loss": 0.4078, + "step": 1005 + }, + { + "epoch": 0.8496621621621622, + "grad_norm": 2.5956809520721436, + "learning_rate": 3.009e-06, + "loss": 0.36, + "step": 1006 + }, + { + "epoch": 0.8505067567567568, + "grad_norm": 2.1589038372039795, + "learning_rate": 3.012e-06, + "loss": 0.3709, + "step": 1007 + }, + { + "epoch": 0.8513513513513513, + "grad_norm": 2.2923901081085205, + "learning_rate": 3.0150000000000004e-06, + "loss": 0.3425, + "step": 1008 + }, + { + "epoch": 0.8521959459459459, + "grad_norm": 1.8016109466552734, + "learning_rate": 3.0179999999999997e-06, + "loss": 0.3638, + "step": 1009 + }, + { + "epoch": 0.8530405405405406, + "grad_norm": 2.5301201343536377, + "learning_rate": 3.021e-06, + "loss": 0.3411, + "step": 1010 + }, + { + "epoch": 0.8538851351351351, + "grad_norm": 2.2637414932250977, + "learning_rate": 3.024e-06, + "loss": 0.3894, + "step": 1011 + }, + { + "epoch": 0.8547297297297297, + "grad_norm": 2.3878705501556396, + "learning_rate": 3.027e-06, + "loss": 0.3449, + "step": 1012 + }, + { + "epoch": 0.8555743243243243, + "grad_norm": 2.5614755153656006, + "learning_rate": 3.0300000000000002e-06, + "loss": 0.4109, + "step": 1013 + }, + { + "epoch": 0.856418918918919, + "grad_norm": 2.701650381088257, + "learning_rate": 3.033e-06, + "loss": 0.3093, + "step": 1014 + }, + { + "epoch": 0.8572635135135135, + "grad_norm": 1.7050480842590332, + "learning_rate": 3.036e-06, + "loss": 0.3264, + "step": 1015 + }, + { + "epoch": 0.8581081081081081, + "grad_norm": 1.7211273908615112, + "learning_rate": 3.039e-06, + "loss": 0.3564, + "step": 1016 + }, + { + "epoch": 0.8589527027027027, + "grad_norm": 2.2679202556610107, + "learning_rate": 3.0420000000000003e-06, + "loss": 0.3552, + "step": 1017 + }, + { + "epoch": 0.8597972972972973, + "grad_norm": 2.469440460205078, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.3598, + "step": 1018 + }, + { + "epoch": 0.8606418918918919, + "grad_norm": 3.3401472568511963, + "learning_rate": 3.048e-06, + "loss": 0.5454, + "step": 1019 + }, + { + "epoch": 0.8614864864864865, + "grad_norm": 2.2062795162200928, + "learning_rate": 3.051e-06, + "loss": 0.3832, + "step": 1020 + }, + { + "epoch": 0.862331081081081, + "grad_norm": 3.9953551292419434, + "learning_rate": 3.0540000000000003e-06, + "loss": 0.3574, + "step": 1021 + }, + { + "epoch": 0.8631756756756757, + "grad_norm": 2.563718795776367, + "learning_rate": 3.057e-06, + "loss": 0.3954, + "step": 1022 + }, + { + "epoch": 0.8640202702702703, + "grad_norm": 3.5903234481811523, + "learning_rate": 3.06e-06, + "loss": 0.4461, + "step": 1023 + }, + { + "epoch": 0.8648648648648649, + "grad_norm": 3.104625940322876, + "learning_rate": 3.063e-06, + "loss": 0.3285, + "step": 1024 + }, + { + "epoch": 0.8657094594594594, + "grad_norm": 2.4805166721343994, + "learning_rate": 3.066e-06, + "loss": 0.3564, + "step": 1025 + }, + { + "epoch": 0.8665540540540541, + "grad_norm": 17.13140296936035, + "learning_rate": 3.069e-06, + "loss": 0.3591, + "step": 1026 + }, + { + "epoch": 0.8673986486486487, + "grad_norm": 5.436604976654053, + "learning_rate": 3.0720000000000004e-06, + "loss": 0.3282, + "step": 1027 + }, + { + "epoch": 0.8682432432432432, + "grad_norm": 2.3757832050323486, + "learning_rate": 3.0749999999999998e-06, + "loss": 0.3283, + "step": 1028 + }, + { + "epoch": 0.8690878378378378, + "grad_norm": 2.6499125957489014, + "learning_rate": 3.078e-06, + "loss": 0.3592, + "step": 1029 + }, + { + "epoch": 0.8699324324324325, + "grad_norm": 2.9911584854125977, + "learning_rate": 3.0810000000000002e-06, + "loss": 0.346, + "step": 1030 + }, + { + "epoch": 0.870777027027027, + "grad_norm": 2.558669090270996, + "learning_rate": 3.084e-06, + "loss": 0.3355, + "step": 1031 + }, + { + "epoch": 0.8716216216216216, + "grad_norm": 2.800593376159668, + "learning_rate": 3.0870000000000003e-06, + "loss": 0.3848, + "step": 1032 + }, + { + "epoch": 0.8724662162162162, + "grad_norm": 3.0010972023010254, + "learning_rate": 3.09e-06, + "loss": 0.3862, + "step": 1033 + }, + { + "epoch": 0.8733108108108109, + "grad_norm": 2.488438367843628, + "learning_rate": 3.093e-06, + "loss": 0.3387, + "step": 1034 + }, + { + "epoch": 0.8741554054054054, + "grad_norm": 2.431739330291748, + "learning_rate": 3.096e-06, + "loss": 0.4201, + "step": 1035 + }, + { + "epoch": 0.875, + "grad_norm": 3.150390386581421, + "learning_rate": 3.0990000000000003e-06, + "loss": 0.3549, + "step": 1036 + }, + { + "epoch": 0.8758445945945946, + "grad_norm": 2.6257100105285645, + "learning_rate": 3.102e-06, + "loss": 0.3329, + "step": 1037 + }, + { + "epoch": 0.8766891891891891, + "grad_norm": 2.246281623840332, + "learning_rate": 3.105e-06, + "loss": 0.3655, + "step": 1038 + }, + { + "epoch": 0.8775337837837838, + "grad_norm": 2.325094699859619, + "learning_rate": 3.108e-06, + "loss": 0.3731, + "step": 1039 + }, + { + "epoch": 0.8783783783783784, + "grad_norm": 2.557020425796509, + "learning_rate": 3.111e-06, + "loss": 0.3813, + "step": 1040 + }, + { + "epoch": 0.879222972972973, + "grad_norm": 3.9005987644195557, + "learning_rate": 3.114e-06, + "loss": 0.3472, + "step": 1041 + }, + { + "epoch": 0.8800675675675675, + "grad_norm": 2.7207813262939453, + "learning_rate": 3.1170000000000004e-06, + "loss": 0.3631, + "step": 1042 + }, + { + "epoch": 0.8809121621621622, + "grad_norm": 6.04010009765625, + "learning_rate": 3.1199999999999998e-06, + "loss": 0.3439, + "step": 1043 + }, + { + "epoch": 0.8817567567567568, + "grad_norm": 5.995325088500977, + "learning_rate": 3.123e-06, + "loss": 0.4142, + "step": 1044 + }, + { + "epoch": 0.8826013513513513, + "grad_norm": 3.977583408355713, + "learning_rate": 3.1260000000000002e-06, + "loss": 0.4057, + "step": 1045 + }, + { + "epoch": 0.8834459459459459, + "grad_norm": 3.1264164447784424, + "learning_rate": 3.129e-06, + "loss": 0.4775, + "step": 1046 + }, + { + "epoch": 0.8842905405405406, + "grad_norm": 2.883408308029175, + "learning_rate": 3.1320000000000003e-06, + "loss": 0.4243, + "step": 1047 + }, + { + "epoch": 0.8851351351351351, + "grad_norm": 5.313899517059326, + "learning_rate": 3.135e-06, + "loss": 0.4506, + "step": 1048 + }, + { + "epoch": 0.8859797297297297, + "grad_norm": 4.622241020202637, + "learning_rate": 3.138e-06, + "loss": 0.4473, + "step": 1049 + }, + { + "epoch": 0.8868243243243243, + "grad_norm": 5.382021427154541, + "learning_rate": 3.141e-06, + "loss": 0.5296, + "step": 1050 + }, + { + "epoch": 0.887668918918919, + "grad_norm": 10.024415016174316, + "learning_rate": 3.1440000000000003e-06, + "loss": 1.2326, + "step": 1051 + }, + { + "epoch": 0.8885135135135135, + "grad_norm": 6.670192718505859, + "learning_rate": 3.1469999999999997e-06, + "loss": 0.8613, + "step": 1052 + }, + { + "epoch": 0.8893581081081081, + "grad_norm": 2.724771022796631, + "learning_rate": 3.15e-06, + "loss": 0.5212, + "step": 1053 + }, + { + "epoch": 0.8902027027027027, + "grad_norm": 8.164164543151855, + "learning_rate": 3.153e-06, + "loss": 0.4427, + "step": 1054 + }, + { + "epoch": 0.8910472972972973, + "grad_norm": 6.540574073791504, + "learning_rate": 3.156e-06, + "loss": 0.3927, + "step": 1055 + }, + { + "epoch": 0.8918918918918919, + "grad_norm": 8.625256538391113, + "learning_rate": 3.159e-06, + "loss": 0.3816, + "step": 1056 + }, + { + "epoch": 0.8927364864864865, + "grad_norm": 2.943035364151001, + "learning_rate": 3.162e-06, + "loss": 0.3326, + "step": 1057 + }, + { + "epoch": 0.893581081081081, + "grad_norm": 1.8329907655715942, + "learning_rate": 3.1649999999999998e-06, + "loss": 0.3583, + "step": 1058 + }, + { + "epoch": 0.8944256756756757, + "grad_norm": 2.98535418510437, + "learning_rate": 3.168e-06, + "loss": 0.3703, + "step": 1059 + }, + { + "epoch": 0.8952702702702703, + "grad_norm": 2.8117966651916504, + "learning_rate": 3.1710000000000002e-06, + "loss": 0.3402, + "step": 1060 + }, + { + "epoch": 0.8961148648648649, + "grad_norm": 2.1557085514068604, + "learning_rate": 3.1740000000000004e-06, + "loss": 0.3696, + "step": 1061 + }, + { + "epoch": 0.8969594594594594, + "grad_norm": 2.1235058307647705, + "learning_rate": 3.177e-06, + "loss": 0.3559, + "step": 1062 + }, + { + "epoch": 0.8978040540540541, + "grad_norm": 2.2516281604766846, + "learning_rate": 3.18e-06, + "loss": 0.3635, + "step": 1063 + }, + { + "epoch": 0.8986486486486487, + "grad_norm": 2.8615381717681885, + "learning_rate": 3.1830000000000003e-06, + "loss": 0.3135, + "step": 1064 + }, + { + "epoch": 0.8994932432432432, + "grad_norm": 2.9006049633026123, + "learning_rate": 3.186e-06, + "loss": 0.339, + "step": 1065 + }, + { + "epoch": 0.9003378378378378, + "grad_norm": 3.5949385166168213, + "learning_rate": 3.1890000000000003e-06, + "loss": 0.3303, + "step": 1066 + }, + { + "epoch": 0.9011824324324325, + "grad_norm": 2.3325986862182617, + "learning_rate": 3.192e-06, + "loss": 0.3254, + "step": 1067 + }, + { + "epoch": 0.902027027027027, + "grad_norm": 2.3219642639160156, + "learning_rate": 3.195e-06, + "loss": 0.3455, + "step": 1068 + }, + { + "epoch": 0.9028716216216216, + "grad_norm": 1.955398440361023, + "learning_rate": 3.198e-06, + "loss": 0.2683, + "step": 1069 + }, + { + "epoch": 0.9037162162162162, + "grad_norm": 1.8171265125274658, + "learning_rate": 3.2010000000000004e-06, + "loss": 0.3201, + "step": 1070 + }, + { + "epoch": 0.9045608108108109, + "grad_norm": 3.268045663833618, + "learning_rate": 3.204e-06, + "loss": 0.3466, + "step": 1071 + }, + { + "epoch": 0.9054054054054054, + "grad_norm": 3.1142117977142334, + "learning_rate": 3.207e-06, + "loss": 0.3891, + "step": 1072 + }, + { + "epoch": 0.90625, + "grad_norm": 2.3110246658325195, + "learning_rate": 3.21e-06, + "loss": 0.3205, + "step": 1073 + }, + { + "epoch": 0.9070945945945946, + "grad_norm": 3.2976553440093994, + "learning_rate": 3.213e-06, + "loss": 0.318, + "step": 1074 + }, + { + "epoch": 0.9079391891891891, + "grad_norm": 2.028456449508667, + "learning_rate": 3.216e-06, + "loss": 0.2989, + "step": 1075 + }, + { + "epoch": 0.9087837837837838, + "grad_norm": 2.307882785797119, + "learning_rate": 3.2190000000000004e-06, + "loss": 0.3152, + "step": 1076 + }, + { + "epoch": 0.9096283783783784, + "grad_norm": 5.687402248382568, + "learning_rate": 3.222e-06, + "loss": 0.3458, + "step": 1077 + }, + { + "epoch": 0.910472972972973, + "grad_norm": 2.471031904220581, + "learning_rate": 3.225e-06, + "loss": 0.2918, + "step": 1078 + }, + { + "epoch": 0.9113175675675675, + "grad_norm": 2.4872219562530518, + "learning_rate": 3.2280000000000003e-06, + "loss": 0.3464, + "step": 1079 + }, + { + "epoch": 0.9121621621621622, + "grad_norm": 4.8079633712768555, + "learning_rate": 3.231e-06, + "loss": 0.3479, + "step": 1080 + }, + { + "epoch": 0.9130067567567568, + "grad_norm": 2.401413679122925, + "learning_rate": 3.2340000000000003e-06, + "loss": 0.3505, + "step": 1081 + }, + { + "epoch": 0.9138513513513513, + "grad_norm": 2.1417362689971924, + "learning_rate": 3.237e-06, + "loss": 0.2858, + "step": 1082 + }, + { + "epoch": 0.9146959459459459, + "grad_norm": 2.2866523265838623, + "learning_rate": 3.24e-06, + "loss": 0.3413, + "step": 1083 + }, + { + "epoch": 0.9155405405405406, + "grad_norm": 2.159181833267212, + "learning_rate": 3.243e-06, + "loss": 0.3287, + "step": 1084 + }, + { + "epoch": 0.9163851351351351, + "grad_norm": 2.332334518432617, + "learning_rate": 3.2460000000000003e-06, + "loss": 0.3181, + "step": 1085 + }, + { + "epoch": 0.9172297297297297, + "grad_norm": 3.474949598312378, + "learning_rate": 3.2489999999999997e-06, + "loss": 0.407, + "step": 1086 + }, + { + "epoch": 0.9180743243243243, + "grad_norm": 4.522956371307373, + "learning_rate": 3.252e-06, + "loss": 0.3338, + "step": 1087 + }, + { + "epoch": 0.918918918918919, + "grad_norm": 3.545987367630005, + "learning_rate": 3.255e-06, + "loss": 0.3024, + "step": 1088 + }, + { + "epoch": 0.9197635135135135, + "grad_norm": 2.8072292804718018, + "learning_rate": 3.258e-06, + "loss": 0.2974, + "step": 1089 + }, + { + "epoch": 0.9206081081081081, + "grad_norm": 2.5033440589904785, + "learning_rate": 3.261e-06, + "loss": 0.2922, + "step": 1090 + }, + { + "epoch": 0.9214527027027027, + "grad_norm": 2.447547674179077, + "learning_rate": 3.264e-06, + "loss": 0.3536, + "step": 1091 + }, + { + "epoch": 0.9222972972972973, + "grad_norm": 2.343759298324585, + "learning_rate": 3.267e-06, + "loss": 0.3758, + "step": 1092 + }, + { + "epoch": 0.9231418918918919, + "grad_norm": 2.415005683898926, + "learning_rate": 3.27e-06, + "loss": 0.3147, + "step": 1093 + }, + { + "epoch": 0.9239864864864865, + "grad_norm": 2.6333112716674805, + "learning_rate": 3.2730000000000003e-06, + "loss": 0.3478, + "step": 1094 + }, + { + "epoch": 0.924831081081081, + "grad_norm": 2.338019371032715, + "learning_rate": 3.276e-06, + "loss": 0.3075, + "step": 1095 + }, + { + "epoch": 0.9256756756756757, + "grad_norm": 2.7696714401245117, + "learning_rate": 3.279e-06, + "loss": 0.3286, + "step": 1096 + }, + { + "epoch": 0.9265202702702703, + "grad_norm": 2.252162218093872, + "learning_rate": 3.282e-06, + "loss": 0.3689, + "step": 1097 + }, + { + "epoch": 0.9273648648648649, + "grad_norm": 2.839733123779297, + "learning_rate": 3.285e-06, + "loss": 0.4387, + "step": 1098 + }, + { + "epoch": 0.9282094594594594, + "grad_norm": 4.375608444213867, + "learning_rate": 3.288e-06, + "loss": 0.4367, + "step": 1099 + }, + { + "epoch": 0.9290540540540541, + "grad_norm": 4.026871681213379, + "learning_rate": 3.2910000000000003e-06, + "loss": 0.4676, + "step": 1100 + }, + { + "epoch": 0.9298986486486487, + "grad_norm": 9.165740013122559, + "learning_rate": 3.2939999999999997e-06, + "loss": 1.1296, + "step": 1101 + }, + { + "epoch": 0.9307432432432432, + "grad_norm": 4.4579548835754395, + "learning_rate": 3.297e-06, + "loss": 0.5463, + "step": 1102 + }, + { + "epoch": 0.9315878378378378, + "grad_norm": 2.172818183898926, + "learning_rate": 3.3e-06, + "loss": 0.3235, + "step": 1103 + }, + { + "epoch": 0.9324324324324325, + "grad_norm": 4.012520790100098, + "learning_rate": 3.3030000000000004e-06, + "loss": 0.5489, + "step": 1104 + }, + { + "epoch": 0.933277027027027, + "grad_norm": 2.4861109256744385, + "learning_rate": 3.306e-06, + "loss": 0.3274, + "step": 1105 + }, + { + "epoch": 0.9341216216216216, + "grad_norm": 2.6970138549804688, + "learning_rate": 3.309e-06, + "loss": 0.3287, + "step": 1106 + }, + { + "epoch": 0.9349662162162162, + "grad_norm": 2.5089333057403564, + "learning_rate": 3.3120000000000002e-06, + "loss": 0.2736, + "step": 1107 + }, + { + "epoch": 0.9358108108108109, + "grad_norm": 2.1055116653442383, + "learning_rate": 3.315e-06, + "loss": 0.3379, + "step": 1108 + }, + { + "epoch": 0.9366554054054054, + "grad_norm": 2.0926036834716797, + "learning_rate": 3.3180000000000003e-06, + "loss": 0.2725, + "step": 1109 + }, + { + "epoch": 0.9375, + "grad_norm": 1.7224787473678589, + "learning_rate": 3.3210000000000005e-06, + "loss": 0.2383, + "step": 1110 + }, + { + "epoch": 0.9383445945945946, + "grad_norm": 1.8799762725830078, + "learning_rate": 3.324e-06, + "loss": 0.3079, + "step": 1111 + }, + { + "epoch": 0.9391891891891891, + "grad_norm": 2.154421091079712, + "learning_rate": 3.327e-06, + "loss": 0.309, + "step": 1112 + }, + { + "epoch": 0.9400337837837838, + "grad_norm": 2.697490692138672, + "learning_rate": 3.3300000000000003e-06, + "loss": 0.3026, + "step": 1113 + }, + { + "epoch": 0.9408783783783784, + "grad_norm": 1.7967572212219238, + "learning_rate": 3.333e-06, + "loss": 0.3093, + "step": 1114 + }, + { + "epoch": 0.941722972972973, + "grad_norm": 2.3189332485198975, + "learning_rate": 3.336e-06, + "loss": 0.2814, + "step": 1115 + }, + { + "epoch": 0.9425675675675675, + "grad_norm": 2.6212010383605957, + "learning_rate": 3.339e-06, + "loss": 0.3472, + "step": 1116 + }, + { + "epoch": 0.9434121621621622, + "grad_norm": 4.523200035095215, + "learning_rate": 3.342e-06, + "loss": 0.3106, + "step": 1117 + }, + { + "epoch": 0.9442567567567568, + "grad_norm": 2.134326696395874, + "learning_rate": 3.345e-06, + "loss": 0.2774, + "step": 1118 + }, + { + "epoch": 0.9451013513513513, + "grad_norm": 2.310716390609741, + "learning_rate": 3.3480000000000004e-06, + "loss": 0.2952, + "step": 1119 + }, + { + "epoch": 0.9459459459459459, + "grad_norm": 1.911856770515442, + "learning_rate": 3.3509999999999998e-06, + "loss": 0.3163, + "step": 1120 + }, + { + "epoch": 0.9467905405405406, + "grad_norm": 3.4822354316711426, + "learning_rate": 3.354e-06, + "loss": 0.3545, + "step": 1121 + }, + { + "epoch": 0.9476351351351351, + "grad_norm": 1.7684383392333984, + "learning_rate": 3.3570000000000002e-06, + "loss": 0.2951, + "step": 1122 + }, + { + "epoch": 0.9484797297297297, + "grad_norm": 3.828181505203247, + "learning_rate": 3.36e-06, + "loss": 0.3547, + "step": 1123 + }, + { + "epoch": 0.9493243243243243, + "grad_norm": 2.129465341567993, + "learning_rate": 3.3630000000000002e-06, + "loss": 0.2815, + "step": 1124 + }, + { + "epoch": 0.950168918918919, + "grad_norm": 2.4353227615356445, + "learning_rate": 3.366e-06, + "loss": 0.3258, + "step": 1125 + }, + { + "epoch": 0.9510135135135135, + "grad_norm": 3.0216715335845947, + "learning_rate": 3.369e-06, + "loss": 0.3257, + "step": 1126 + }, + { + "epoch": 0.9518581081081081, + "grad_norm": 2.1579179763793945, + "learning_rate": 3.372e-06, + "loss": 0.2944, + "step": 1127 + }, + { + "epoch": 0.9527027027027027, + "grad_norm": 2.217369556427002, + "learning_rate": 3.3750000000000003e-06, + "loss": 0.2961, + "step": 1128 + }, + { + "epoch": 0.9535472972972973, + "grad_norm": 1.879008412361145, + "learning_rate": 3.378e-06, + "loss": 0.2575, + "step": 1129 + }, + { + "epoch": 0.9543918918918919, + "grad_norm": 2.927401542663574, + "learning_rate": 3.381e-06, + "loss": 0.331, + "step": 1130 + }, + { + "epoch": 0.9552364864864865, + "grad_norm": 2.4738593101501465, + "learning_rate": 3.384e-06, + "loss": 0.3373, + "step": 1131 + }, + { + "epoch": 0.956081081081081, + "grad_norm": 2.176619291305542, + "learning_rate": 3.387e-06, + "loss": 0.29, + "step": 1132 + }, + { + "epoch": 0.9569256756756757, + "grad_norm": 3.7879507541656494, + "learning_rate": 3.39e-06, + "loss": 0.3479, + "step": 1133 + }, + { + "epoch": 0.9577702702702703, + "grad_norm": 2.5116519927978516, + "learning_rate": 3.3930000000000004e-06, + "loss": 0.3457, + "step": 1134 + }, + { + "epoch": 0.9586148648648649, + "grad_norm": 2.246290683746338, + "learning_rate": 3.3959999999999998e-06, + "loss": 0.3013, + "step": 1135 + }, + { + "epoch": 0.9594594594594594, + "grad_norm": 2.660771369934082, + "learning_rate": 3.399e-06, + "loss": 0.2999, + "step": 1136 + }, + { + "epoch": 0.9603040540540541, + "grad_norm": 3.447068929672241, + "learning_rate": 3.402e-06, + "loss": 0.3536, + "step": 1137 + }, + { + "epoch": 0.9611486486486487, + "grad_norm": 3.3965096473693848, + "learning_rate": 3.405e-06, + "loss": 0.2923, + "step": 1138 + }, + { + "epoch": 0.9619932432432432, + "grad_norm": 3.7911880016326904, + "learning_rate": 3.4080000000000002e-06, + "loss": 0.3707, + "step": 1139 + }, + { + "epoch": 0.9628378378378378, + "grad_norm": 2.842957019805908, + "learning_rate": 3.411e-06, + "loss": 0.333, + "step": 1140 + }, + { + "epoch": 0.9636824324324325, + "grad_norm": 2.8302502632141113, + "learning_rate": 3.414e-06, + "loss": 0.3283, + "step": 1141 + }, + { + "epoch": 0.964527027027027, + "grad_norm": 13.840734481811523, + "learning_rate": 3.417e-06, + "loss": 0.3587, + "step": 1142 + }, + { + "epoch": 0.9653716216216216, + "grad_norm": 2.6750354766845703, + "learning_rate": 3.4200000000000003e-06, + "loss": 0.2872, + "step": 1143 + }, + { + "epoch": 0.9662162162162162, + "grad_norm": 10.78979778289795, + "learning_rate": 3.4229999999999997e-06, + "loss": 0.3373, + "step": 1144 + }, + { + "epoch": 0.9670608108108109, + "grad_norm": 2.378499746322632, + "learning_rate": 3.426e-06, + "loss": 0.3159, + "step": 1145 + }, + { + "epoch": 0.9679054054054054, + "grad_norm": 3.6243317127227783, + "learning_rate": 3.429e-06, + "loss": 0.3377, + "step": 1146 + }, + { + "epoch": 0.96875, + "grad_norm": 3.327846050262451, + "learning_rate": 3.4320000000000003e-06, + "loss": 0.4435, + "step": 1147 + }, + { + "epoch": 0.9695945945945946, + "grad_norm": 6.80631685256958, + "learning_rate": 3.435e-06, + "loss": 0.3339, + "step": 1148 + }, + { + "epoch": 0.9704391891891891, + "grad_norm": 2.4513871669769287, + "learning_rate": 3.438e-06, + "loss": 0.3598, + "step": 1149 + }, + { + "epoch": 0.9712837837837838, + "grad_norm": 4.110013484954834, + "learning_rate": 3.441e-06, + "loss": 0.4935, + "step": 1150 + }, + { + "epoch": 0.9721283783783784, + "grad_norm": 5.7769646644592285, + "learning_rate": 3.444e-06, + "loss": 0.7298, + "step": 1151 + }, + { + "epoch": 0.972972972972973, + "grad_norm": 6.268688678741455, + "learning_rate": 3.447e-06, + "loss": 0.6741, + "step": 1152 + }, + { + "epoch": 0.9738175675675675, + "grad_norm": 2.2457644939422607, + "learning_rate": 3.4500000000000004e-06, + "loss": 0.3422, + "step": 1153 + }, + { + "epoch": 0.9746621621621622, + "grad_norm": 1.894405722618103, + "learning_rate": 3.453e-06, + "loss": 0.278, + "step": 1154 + }, + { + "epoch": 0.9755067567567568, + "grad_norm": 4.170511722564697, + "learning_rate": 3.456e-06, + "loss": 0.3353, + "step": 1155 + }, + { + "epoch": 0.9763513513513513, + "grad_norm": 2.1298274993896484, + "learning_rate": 3.4590000000000003e-06, + "loss": 0.2613, + "step": 1156 + }, + { + "epoch": 0.9771959459459459, + "grad_norm": 2.1958158016204834, + "learning_rate": 3.462e-06, + "loss": 0.2963, + "step": 1157 + }, + { + "epoch": 0.9780405405405406, + "grad_norm": 1.9845540523529053, + "learning_rate": 3.4650000000000003e-06, + "loss": 0.2384, + "step": 1158 + }, + { + "epoch": 0.9788851351351351, + "grad_norm": 2.6482510566711426, + "learning_rate": 3.468e-06, + "loss": 0.2707, + "step": 1159 + }, + { + "epoch": 0.9797297297297297, + "grad_norm": 7.958125114440918, + "learning_rate": 3.471e-06, + "loss": 0.2988, + "step": 1160 + }, + { + "epoch": 0.9805743243243243, + "grad_norm": 8.353843688964844, + "learning_rate": 3.474e-06, + "loss": 0.2968, + "step": 1161 + }, + { + "epoch": 0.981418918918919, + "grad_norm": 2.24999737739563, + "learning_rate": 3.4770000000000003e-06, + "loss": 0.2879, + "step": 1162 + }, + { + "epoch": 0.9822635135135135, + "grad_norm": 1.9914922714233398, + "learning_rate": 3.48e-06, + "loss": 0.2865, + "step": 1163 + }, + { + "epoch": 0.9831081081081081, + "grad_norm": 4.56576681137085, + "learning_rate": 3.483e-06, + "loss": 0.2976, + "step": 1164 + }, + { + "epoch": 0.9839527027027027, + "grad_norm": 2.3143727779388428, + "learning_rate": 3.486e-06, + "loss": 0.2885, + "step": 1165 + }, + { + "epoch": 0.9847972972972973, + "grad_norm": 2.6489555835723877, + "learning_rate": 3.489e-06, + "loss": 0.3157, + "step": 1166 + }, + { + "epoch": 0.9856418918918919, + "grad_norm": 2.6455962657928467, + "learning_rate": 3.492e-06, + "loss": 0.2966, + "step": 1167 + }, + { + "epoch": 0.9864864864864865, + "grad_norm": 2.463806390762329, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.3165, + "step": 1168 + }, + { + "epoch": 0.987331081081081, + "grad_norm": 3.5306084156036377, + "learning_rate": 3.498e-06, + "loss": 0.2957, + "step": 1169 + }, + { + "epoch": 0.9881756756756757, + "grad_norm": 2.368108034133911, + "learning_rate": 3.501e-06, + "loss": 0.298, + "step": 1170 + }, + { + "epoch": 0.9890202702702703, + "grad_norm": 2.1226396560668945, + "learning_rate": 3.5040000000000002e-06, + "loss": 0.2654, + "step": 1171 + }, + { + "epoch": 0.9898648648648649, + "grad_norm": 2.296680450439453, + "learning_rate": 3.507e-06, + "loss": 0.3061, + "step": 1172 + }, + { + "epoch": 0.9907094594594594, + "grad_norm": 2.0552752017974854, + "learning_rate": 3.5100000000000003e-06, + "loss": 0.3235, + "step": 1173 + }, + { + "epoch": 0.9915540540540541, + "grad_norm": 2.1768038272857666, + "learning_rate": 3.513e-06, + "loss": 0.2989, + "step": 1174 + }, + { + "epoch": 0.9923986486486487, + "grad_norm": 2.387533187866211, + "learning_rate": 3.516e-06, + "loss": 0.301, + "step": 1175 + }, + { + "epoch": 0.9932432432432432, + "grad_norm": 2.138110637664795, + "learning_rate": 3.519e-06, + "loss": 0.3365, + "step": 1176 + }, + { + "epoch": 0.9940878378378378, + "grad_norm": 2.159053325653076, + "learning_rate": 3.5220000000000003e-06, + "loss": 0.3304, + "step": 1177 + }, + { + "epoch": 0.9949324324324325, + "grad_norm": 4.178711891174316, + "learning_rate": 3.5249999999999997e-06, + "loss": 0.2916, + "step": 1178 + }, + { + "epoch": 0.995777027027027, + "grad_norm": 2.7175209522247314, + "learning_rate": 3.528e-06, + "loss": 0.3489, + "step": 1179 + }, + { + "epoch": 0.9966216216216216, + "grad_norm": 3.5765061378479004, + "learning_rate": 3.531e-06, + "loss": 0.4027, + "step": 1180 + }, + { + "epoch": 0.9974662162162162, + "grad_norm": 5.281097888946533, + "learning_rate": 3.534e-06, + "loss": 0.2944, + "step": 1181 + }, + { + "epoch": 0.9983108108108109, + "grad_norm": 2.939997434616089, + "learning_rate": 3.537e-06, + "loss": 0.3109, + "step": 1182 + }, + { + "epoch": 0.9991554054054054, + "grad_norm": 4.888102054595947, + "learning_rate": 3.54e-06, + "loss": 0.4662, + "step": 1183 + }, + { + "epoch": 1.0, + "grad_norm": 5.007531642913818, + "learning_rate": 3.543e-06, + "loss": 0.5094, + "step": 1184 + }, + { + "epoch": 1.0008445945945945, + "grad_norm": 10.51378059387207, + "learning_rate": 3.546e-06, + "loss": 1.2024, + "step": 1185 + }, + { + "epoch": 1.0016891891891893, + "grad_norm": 4.923516273498535, + "learning_rate": 3.5490000000000002e-06, + "loss": 0.6256, + "step": 1186 + }, + { + "epoch": 1.0025337837837838, + "grad_norm": 2.535069227218628, + "learning_rate": 3.552e-06, + "loss": 0.3868, + "step": 1187 + }, + { + "epoch": 1.0033783783783783, + "grad_norm": 4.070481777191162, + "learning_rate": 3.555e-06, + "loss": 0.3164, + "step": 1188 + }, + { + "epoch": 1.004222972972973, + "grad_norm": 3.140296220779419, + "learning_rate": 3.558e-06, + "loss": 0.2661, + "step": 1189 + }, + { + "epoch": 1.0050675675675675, + "grad_norm": 2.4715402126312256, + "learning_rate": 3.5610000000000003e-06, + "loss": 0.3882, + "step": 1190 + }, + { + "epoch": 1.005912162162162, + "grad_norm": 2.344244956970215, + "learning_rate": 3.564e-06, + "loss": 0.2538, + "step": 1191 + }, + { + "epoch": 1.0067567567567568, + "grad_norm": 5.392148494720459, + "learning_rate": 3.5670000000000003e-06, + "loss": 0.295, + "step": 1192 + }, + { + "epoch": 1.0076013513513513, + "grad_norm": 2.7868683338165283, + "learning_rate": 3.57e-06, + "loss": 0.2733, + "step": 1193 + }, + { + "epoch": 1.008445945945946, + "grad_norm": 1.696308970451355, + "learning_rate": 3.573e-06, + "loss": 0.2896, + "step": 1194 + }, + { + "epoch": 1.0092905405405406, + "grad_norm": 1.580560564994812, + "learning_rate": 3.576e-06, + "loss": 0.2518, + "step": 1195 + }, + { + "epoch": 1.010135135135135, + "grad_norm": 2.008948802947998, + "learning_rate": 3.5790000000000004e-06, + "loss": 0.2357, + "step": 1196 + }, + { + "epoch": 1.0109797297297298, + "grad_norm": 2.1189844608306885, + "learning_rate": 3.582e-06, + "loss": 0.3831, + "step": 1197 + }, + { + "epoch": 1.0118243243243243, + "grad_norm": 2.5990095138549805, + "learning_rate": 3.585e-06, + "loss": 0.3274, + "step": 1198 + }, + { + "epoch": 1.0126689189189189, + "grad_norm": 1.4718010425567627, + "learning_rate": 3.588e-06, + "loss": 0.2424, + "step": 1199 + }, + { + "epoch": 1.0135135135135136, + "grad_norm": 1.7520567178726196, + "learning_rate": 3.591e-06, + "loss": 0.2479, + "step": 1200 + }, + { + "epoch": 1.0143581081081081, + "grad_norm": 1.9113444089889526, + "learning_rate": 3.5940000000000002e-06, + "loss": 0.2453, + "step": 1201 + }, + { + "epoch": 1.0152027027027026, + "grad_norm": 1.9601818323135376, + "learning_rate": 3.5970000000000005e-06, + "loss": 0.2576, + "step": 1202 + }, + { + "epoch": 1.0160472972972974, + "grad_norm": 2.192950963973999, + "learning_rate": 3.6e-06, + "loss": 0.2919, + "step": 1203 + }, + { + "epoch": 1.0168918918918919, + "grad_norm": 1.9301856756210327, + "learning_rate": 3.603e-06, + "loss": 0.292, + "step": 1204 + }, + { + "epoch": 1.0177364864864864, + "grad_norm": 6.557028770446777, + "learning_rate": 3.6060000000000003e-06, + "loss": 0.249, + "step": 1205 + }, + { + "epoch": 1.0185810810810811, + "grad_norm": 2.2136268615722656, + "learning_rate": 3.609e-06, + "loss": 0.2431, + "step": 1206 + }, + { + "epoch": 1.0194256756756757, + "grad_norm": 3.1655874252319336, + "learning_rate": 3.612e-06, + "loss": 0.2856, + "step": 1207 + }, + { + "epoch": 1.0202702702702702, + "grad_norm": 1.9824138879776, + "learning_rate": 3.615e-06, + "loss": 0.2831, + "step": 1208 + }, + { + "epoch": 1.021114864864865, + "grad_norm": 2.0816800594329834, + "learning_rate": 3.618e-06, + "loss": 0.2767, + "step": 1209 + }, + { + "epoch": 1.0219594594594594, + "grad_norm": 2.6379828453063965, + "learning_rate": 3.621e-06, + "loss": 0.3044, + "step": 1210 + }, + { + "epoch": 1.022804054054054, + "grad_norm": 1.7938792705535889, + "learning_rate": 3.6240000000000004e-06, + "loss": 0.2228, + "step": 1211 + }, + { + "epoch": 1.0236486486486487, + "grad_norm": 7.77863073348999, + "learning_rate": 3.6269999999999997e-06, + "loss": 0.2648, + "step": 1212 + }, + { + "epoch": 1.0244932432432432, + "grad_norm": 2.66886043548584, + "learning_rate": 3.63e-06, + "loss": 0.3087, + "step": 1213 + }, + { + "epoch": 1.025337837837838, + "grad_norm": 2.5360047817230225, + "learning_rate": 3.633e-06, + "loss": 0.3186, + "step": 1214 + }, + { + "epoch": 1.0261824324324325, + "grad_norm": 3.965250253677368, + "learning_rate": 3.636e-06, + "loss": 0.2849, + "step": 1215 + }, + { + "epoch": 1.027027027027027, + "grad_norm": 1.6053746938705444, + "learning_rate": 3.6390000000000002e-06, + "loss": 0.236, + "step": 1216 + }, + { + "epoch": 1.0278716216216217, + "grad_norm": 2.0597264766693115, + "learning_rate": 3.642e-06, + "loss": 0.2344, + "step": 1217 + }, + { + "epoch": 1.0287162162162162, + "grad_norm": 2.2259819507598877, + "learning_rate": 3.645e-06, + "loss": 0.2471, + "step": 1218 + }, + { + "epoch": 1.0295608108108107, + "grad_norm": 2.3842592239379883, + "learning_rate": 3.648e-06, + "loss": 0.2872, + "step": 1219 + }, + { + "epoch": 1.0304054054054055, + "grad_norm": 1.8917419910430908, + "learning_rate": 3.6510000000000003e-06, + "loss": 0.2884, + "step": 1220 + }, + { + "epoch": 1.03125, + "grad_norm": 1.7990641593933105, + "learning_rate": 3.654e-06, + "loss": 0.276, + "step": 1221 + }, + { + "epoch": 1.0320945945945945, + "grad_norm": 2.5712361335754395, + "learning_rate": 3.657e-06, + "loss": 0.2875, + "step": 1222 + }, + { + "epoch": 1.0329391891891893, + "grad_norm": 2.3495888710021973, + "learning_rate": 3.66e-06, + "loss": 0.2558, + "step": 1223 + }, + { + "epoch": 1.0337837837837838, + "grad_norm": 2.257875442504883, + "learning_rate": 3.663e-06, + "loss": 0.2808, + "step": 1224 + }, + { + "epoch": 1.0346283783783783, + "grad_norm": 2.606130361557007, + "learning_rate": 3.666e-06, + "loss": 0.2921, + "step": 1225 + }, + { + "epoch": 1.035472972972973, + "grad_norm": 2.141749620437622, + "learning_rate": 3.6690000000000004e-06, + "loss": 0.3236, + "step": 1226 + }, + { + "epoch": 1.0363175675675675, + "grad_norm": 3.074852705001831, + "learning_rate": 3.6719999999999997e-06, + "loss": 0.3297, + "step": 1227 + }, + { + "epoch": 1.037162162162162, + "grad_norm": 2.8122124671936035, + "learning_rate": 3.675e-06, + "loss": 0.3458, + "step": 1228 + }, + { + "epoch": 1.0380067567567568, + "grad_norm": 2.1983814239501953, + "learning_rate": 3.678e-06, + "loss": 0.3218, + "step": 1229 + }, + { + "epoch": 1.0388513513513513, + "grad_norm": 2.399354934692383, + "learning_rate": 3.681e-06, + "loss": 0.3356, + "step": 1230 + }, + { + "epoch": 1.039695945945946, + "grad_norm": 2.1986024379730225, + "learning_rate": 3.6840000000000002e-06, + "loss": 0.3454, + "step": 1231 + }, + { + "epoch": 1.0405405405405406, + "grad_norm": 6.288662910461426, + "learning_rate": 3.687e-06, + "loss": 0.3282, + "step": 1232 + }, + { + "epoch": 1.041385135135135, + "grad_norm": 3.4516570568084717, + "learning_rate": 3.6900000000000002e-06, + "loss": 0.3378, + "step": 1233 + }, + { + "epoch": 1.0422297297297298, + "grad_norm": 5.194690704345703, + "learning_rate": 3.693e-06, + "loss": 0.4083, + "step": 1234 + }, + { + "epoch": 1.0430743243243243, + "grad_norm": 4.264784336090088, + "learning_rate": 3.6960000000000003e-06, + "loss": 0.923, + "step": 1235 + }, + { + "epoch": 1.0439189189189189, + "grad_norm": 3.5655083656311035, + "learning_rate": 3.6990000000000005e-06, + "loss": 0.5449, + "step": 1236 + }, + { + "epoch": 1.0447635135135136, + "grad_norm": 2.5541441440582275, + "learning_rate": 3.702e-06, + "loss": 0.4266, + "step": 1237 + }, + { + "epoch": 1.0456081081081081, + "grad_norm": 1.6034356355667114, + "learning_rate": 3.705e-06, + "loss": 0.2961, + "step": 1238 + }, + { + "epoch": 1.0464527027027026, + "grad_norm": 2.9066433906555176, + "learning_rate": 3.7080000000000003e-06, + "loss": 0.2798, + "step": 1239 + }, + { + "epoch": 1.0472972972972974, + "grad_norm": 1.6528135538101196, + "learning_rate": 3.711e-06, + "loss": 0.274, + "step": 1240 + }, + { + "epoch": 1.0481418918918919, + "grad_norm": 1.7048476934432983, + "learning_rate": 3.714e-06, + "loss": 0.2418, + "step": 1241 + }, + { + "epoch": 1.0489864864864864, + "grad_norm": 1.6418216228485107, + "learning_rate": 3.717e-06, + "loss": 0.2468, + "step": 1242 + }, + { + "epoch": 1.0498310810810811, + "grad_norm": 2.0188064575195312, + "learning_rate": 3.72e-06, + "loss": 0.2878, + "step": 1243 + }, + { + "epoch": 1.0506756756756757, + "grad_norm": 1.617560863494873, + "learning_rate": 3.723e-06, + "loss": 0.3219, + "step": 1244 + }, + { + "epoch": 1.0515202702702702, + "grad_norm": 1.623550534248352, + "learning_rate": 3.7260000000000004e-06, + "loss": 0.2758, + "step": 1245 + }, + { + "epoch": 1.052364864864865, + "grad_norm": 1.8247880935668945, + "learning_rate": 3.7289999999999998e-06, + "loss": 0.2205, + "step": 1246 + }, + { + "epoch": 1.0532094594594594, + "grad_norm": 1.8466205596923828, + "learning_rate": 3.732e-06, + "loss": 0.267, + "step": 1247 + }, + { + "epoch": 1.054054054054054, + "grad_norm": 1.923510193824768, + "learning_rate": 3.7350000000000002e-06, + "loss": 0.257, + "step": 1248 + }, + { + "epoch": 1.0548986486486487, + "grad_norm": 2.230746269226074, + "learning_rate": 3.738e-06, + "loss": 0.2358, + "step": 1249 + }, + { + "epoch": 1.0557432432432432, + "grad_norm": 1.9809764623641968, + "learning_rate": 3.7410000000000003e-06, + "loss": 0.2262, + "step": 1250 + }, + { + "epoch": 1.056587837837838, + "grad_norm": 1.6441256999969482, + "learning_rate": 3.744e-06, + "loss": 0.2744, + "step": 1251 + }, + { + "epoch": 1.0574324324324325, + "grad_norm": 1.7983030080795288, + "learning_rate": 3.747e-06, + "loss": 0.3002, + "step": 1252 + }, + { + "epoch": 1.058277027027027, + "grad_norm": 1.583234190940857, + "learning_rate": 3.75e-06, + "loss": 0.2363, + "step": 1253 + }, + { + "epoch": 1.0591216216216217, + "grad_norm": 1.6494765281677246, + "learning_rate": 3.753e-06, + "loss": 0.2259, + "step": 1254 + }, + { + "epoch": 1.0599662162162162, + "grad_norm": 2.5835461616516113, + "learning_rate": 3.756e-06, + "loss": 0.2497, + "step": 1255 + }, + { + "epoch": 1.0608108108108107, + "grad_norm": 2.3659286499023438, + "learning_rate": 3.759e-06, + "loss": 0.2408, + "step": 1256 + }, + { + "epoch": 1.0616554054054055, + "grad_norm": 1.8525545597076416, + "learning_rate": 3.7620000000000006e-06, + "loss": 0.2514, + "step": 1257 + }, + { + "epoch": 1.0625, + "grad_norm": 3.298552989959717, + "learning_rate": 3.765e-06, + "loss": 0.2435, + "step": 1258 + }, + { + "epoch": 1.0633445945945945, + "grad_norm": 1.7289321422576904, + "learning_rate": 3.7679999999999998e-06, + "loss": 0.2473, + "step": 1259 + }, + { + "epoch": 1.0641891891891893, + "grad_norm": 5.088571548461914, + "learning_rate": 3.7710000000000004e-06, + "loss": 0.2326, + "step": 1260 + }, + { + "epoch": 1.0650337837837838, + "grad_norm": 2.4512243270874023, + "learning_rate": 3.7739999999999998e-06, + "loss": 0.2929, + "step": 1261 + }, + { + "epoch": 1.0658783783783783, + "grad_norm": 2.7771553993225098, + "learning_rate": 3.7770000000000004e-06, + "loss": 0.2452, + "step": 1262 + }, + { + "epoch": 1.066722972972973, + "grad_norm": 2.7037246227264404, + "learning_rate": 3.7800000000000002e-06, + "loss": 0.2756, + "step": 1263 + }, + { + "epoch": 1.0675675675675675, + "grad_norm": 2.2165489196777344, + "learning_rate": 3.7829999999999996e-06, + "loss": 0.2642, + "step": 1264 + }, + { + "epoch": 1.068412162162162, + "grad_norm": 1.7473406791687012, + "learning_rate": 3.7860000000000003e-06, + "loss": 0.235, + "step": 1265 + }, + { + "epoch": 1.0692567567567568, + "grad_norm": 2.550541400909424, + "learning_rate": 3.789e-06, + "loss": 0.2538, + "step": 1266 + }, + { + "epoch": 1.0701013513513513, + "grad_norm": 1.5600236654281616, + "learning_rate": 3.7920000000000003e-06, + "loss": 0.2208, + "step": 1267 + }, + { + "epoch": 1.070945945945946, + "grad_norm": 3.788280725479126, + "learning_rate": 3.795e-06, + "loss": 0.2701, + "step": 1268 + }, + { + "epoch": 1.0717905405405406, + "grad_norm": 6.107298851013184, + "learning_rate": 3.798e-06, + "loss": 0.2941, + "step": 1269 + }, + { + "epoch": 1.072635135135135, + "grad_norm": 2.8429603576660156, + "learning_rate": 3.801e-06, + "loss": 0.2419, + "step": 1270 + }, + { + "epoch": 1.0734797297297298, + "grad_norm": 2.0310142040252686, + "learning_rate": 3.804e-06, + "loss": 0.2588, + "step": 1271 + }, + { + "epoch": 1.0743243243243243, + "grad_norm": 2.074951171875, + "learning_rate": 3.8070000000000006e-06, + "loss": 0.2793, + "step": 1272 + }, + { + "epoch": 1.0751689189189189, + "grad_norm": 2.4625377655029297, + "learning_rate": 3.81e-06, + "loss": 0.2628, + "step": 1273 + }, + { + "epoch": 1.0760135135135136, + "grad_norm": 2.5619561672210693, + "learning_rate": 3.8129999999999997e-06, + "loss": 0.2723, + "step": 1274 + }, + { + "epoch": 1.0768581081081081, + "grad_norm": 2.116504192352295, + "learning_rate": 3.816e-06, + "loss": 0.2823, + "step": 1275 + }, + { + "epoch": 1.0777027027027026, + "grad_norm": 2.622720956802368, + "learning_rate": 3.819e-06, + "loss": 0.2553, + "step": 1276 + }, + { + "epoch": 1.0785472972972974, + "grad_norm": 2.3566834926605225, + "learning_rate": 3.822000000000001e-06, + "loss": 0.3018, + "step": 1277 + }, + { + "epoch": 1.0793918918918919, + "grad_norm": 2.6402292251586914, + "learning_rate": 3.825e-06, + "loss": 0.287, + "step": 1278 + }, + { + "epoch": 1.0802364864864864, + "grad_norm": 2.7909083366394043, + "learning_rate": 3.828e-06, + "loss": 0.2502, + "step": 1279 + }, + { + "epoch": 1.0810810810810811, + "grad_norm": 4.550517559051514, + "learning_rate": 3.831e-06, + "loss": 0.247, + "step": 1280 + }, + { + "epoch": 1.0819256756756757, + "grad_norm": 2.4082095623016357, + "learning_rate": 3.834e-06, + "loss": 0.2923, + "step": 1281 + }, + { + "epoch": 1.0827702702702702, + "grad_norm": 3.5004544258117676, + "learning_rate": 3.837000000000001e-06, + "loss": 0.3168, + "step": 1282 + }, + { + "epoch": 1.083614864864865, + "grad_norm": 6.045251369476318, + "learning_rate": 3.8400000000000005e-06, + "loss": 0.338, + "step": 1283 + }, + { + "epoch": 1.0844594594594594, + "grad_norm": 5.250992774963379, + "learning_rate": 3.8429999999999995e-06, + "loss": 0.517, + "step": 1284 + }, + { + "epoch": 1.085304054054054, + "grad_norm": 9.759419441223145, + "learning_rate": 3.846e-06, + "loss": 0.9682, + "step": 1285 + }, + { + "epoch": 1.0861486486486487, + "grad_norm": 11.277482986450195, + "learning_rate": 3.849e-06, + "loss": 0.7388, + "step": 1286 + }, + { + "epoch": 1.0869932432432432, + "grad_norm": 1.9551360607147217, + "learning_rate": 3.852e-06, + "loss": 0.2956, + "step": 1287 + }, + { + "epoch": 1.087837837837838, + "grad_norm": 2.2454543113708496, + "learning_rate": 3.855e-06, + "loss": 0.2918, + "step": 1288 + }, + { + "epoch": 1.0886824324324325, + "grad_norm": 2.29059100151062, + "learning_rate": 3.858e-06, + "loss": 0.3051, + "step": 1289 + }, + { + "epoch": 1.089527027027027, + "grad_norm": 2.880754232406616, + "learning_rate": 3.861e-06, + "loss": 0.2621, + "step": 1290 + }, + { + "epoch": 1.0903716216216217, + "grad_norm": 1.8715039491653442, + "learning_rate": 3.864e-06, + "loss": 0.2558, + "step": 1291 + }, + { + "epoch": 1.0912162162162162, + "grad_norm": 2.3966877460479736, + "learning_rate": 3.8669999999999996e-06, + "loss": 0.3247, + "step": 1292 + }, + { + "epoch": 1.0920608108108107, + "grad_norm": 1.582405924797058, + "learning_rate": 3.87e-06, + "loss": 0.2094, + "step": 1293 + }, + { + "epoch": 1.0929054054054055, + "grad_norm": 9.963890075683594, + "learning_rate": 3.873e-06, + "loss": 0.2627, + "step": 1294 + }, + { + "epoch": 1.09375, + "grad_norm": 1.5399059057235718, + "learning_rate": 3.876000000000001e-06, + "loss": 0.2023, + "step": 1295 + }, + { + "epoch": 1.0945945945945945, + "grad_norm": 1.5043679475784302, + "learning_rate": 3.8790000000000005e-06, + "loss": 0.1904, + "step": 1296 + }, + { + "epoch": 1.0954391891891893, + "grad_norm": 1.433156967163086, + "learning_rate": 3.8819999999999994e-06, + "loss": 0.2366, + "step": 1297 + }, + { + "epoch": 1.0962837837837838, + "grad_norm": 1.4972285032272339, + "learning_rate": 3.885e-06, + "loss": 0.2456, + "step": 1298 + }, + { + "epoch": 1.0971283783783783, + "grad_norm": 1.9010940790176392, + "learning_rate": 3.888e-06, + "loss": 0.2373, + "step": 1299 + }, + { + "epoch": 1.097972972972973, + "grad_norm": 1.632267713546753, + "learning_rate": 3.8910000000000005e-06, + "loss": 0.2754, + "step": 1300 + }, + { + "epoch": 1.0988175675675675, + "grad_norm": 1.6194746494293213, + "learning_rate": 3.894e-06, + "loss": 0.2597, + "step": 1301 + }, + { + "epoch": 1.099662162162162, + "grad_norm": 3.7517242431640625, + "learning_rate": 3.897e-06, + "loss": 0.2661, + "step": 1302 + }, + { + "epoch": 1.1005067567567568, + "grad_norm": 2.0656731128692627, + "learning_rate": 3.9e-06, + "loss": 0.2083, + "step": 1303 + }, + { + "epoch": 1.1013513513513513, + "grad_norm": 1.7041270732879639, + "learning_rate": 3.903e-06, + "loss": 0.2073, + "step": 1304 + }, + { + "epoch": 1.102195945945946, + "grad_norm": 2.0554022789001465, + "learning_rate": 3.906e-06, + "loss": 0.2237, + "step": 1305 + }, + { + "epoch": 1.1030405405405406, + "grad_norm": 1.9084242582321167, + "learning_rate": 3.909e-06, + "loss": 0.2445, + "step": 1306 + }, + { + "epoch": 1.103885135135135, + "grad_norm": 1.8788431882858276, + "learning_rate": 3.912e-06, + "loss": 0.2788, + "step": 1307 + }, + { + "epoch": 1.1047297297297298, + "grad_norm": 1.610705852508545, + "learning_rate": 3.915000000000001e-06, + "loss": 0.2202, + "step": 1308 + }, + { + "epoch": 1.1055743243243243, + "grad_norm": 1.6026777029037476, + "learning_rate": 3.918e-06, + "loss": 0.2038, + "step": 1309 + }, + { + "epoch": 1.1064189189189189, + "grad_norm": 1.8522703647613525, + "learning_rate": 3.921e-06, + "loss": 0.2828, + "step": 1310 + }, + { + "epoch": 1.1072635135135136, + "grad_norm": 1.9751962423324585, + "learning_rate": 3.924e-06, + "loss": 0.2654, + "step": 1311 + }, + { + "epoch": 1.1081081081081081, + "grad_norm": 1.7325501441955566, + "learning_rate": 3.927e-06, + "loss": 0.2461, + "step": 1312 + }, + { + "epoch": 1.1089527027027026, + "grad_norm": 1.8267872333526611, + "learning_rate": 3.9300000000000005e-06, + "loss": 0.2169, + "step": 1313 + }, + { + "epoch": 1.1097972972972974, + "grad_norm": 2.417104721069336, + "learning_rate": 3.933e-06, + "loss": 0.2731, + "step": 1314 + }, + { + "epoch": 1.1106418918918919, + "grad_norm": 2.700547933578491, + "learning_rate": 3.936e-06, + "loss": 0.2525, + "step": 1315 + }, + { + "epoch": 1.1114864864864864, + "grad_norm": 1.8484296798706055, + "learning_rate": 3.939e-06, + "loss": 0.2817, + "step": 1316 + }, + { + "epoch": 1.1123310810810811, + "grad_norm": 2.798255443572998, + "learning_rate": 3.942e-06, + "loss": 0.2052, + "step": 1317 + }, + { + "epoch": 1.1131756756756757, + "grad_norm": 2.786862373352051, + "learning_rate": 3.945e-06, + "loss": 0.2394, + "step": 1318 + }, + { + "epoch": 1.1140202702702702, + "grad_norm": 3.179854154586792, + "learning_rate": 3.948e-06, + "loss": 0.2536, + "step": 1319 + }, + { + "epoch": 1.114864864864865, + "grad_norm": 2.0600314140319824, + "learning_rate": 3.951000000000001e-06, + "loss": 0.2682, + "step": 1320 + }, + { + "epoch": 1.1157094594594594, + "grad_norm": 5.162700176239014, + "learning_rate": 3.954e-06, + "loss": 0.3045, + "step": 1321 + }, + { + "epoch": 1.116554054054054, + "grad_norm": 2.1168763637542725, + "learning_rate": 3.9569999999999996e-06, + "loss": 0.2525, + "step": 1322 + }, + { + "epoch": 1.1173986486486487, + "grad_norm": 2.3407132625579834, + "learning_rate": 3.96e-06, + "loss": 0.2997, + "step": 1323 + }, + { + "epoch": 1.1182432432432432, + "grad_norm": 2.332578420639038, + "learning_rate": 3.963e-06, + "loss": 0.2931, + "step": 1324 + }, + { + "epoch": 1.119087837837838, + "grad_norm": 1.9731706380844116, + "learning_rate": 3.966000000000001e-06, + "loss": 0.2579, + "step": 1325 + }, + { + "epoch": 1.1199324324324325, + "grad_norm": 2.2594234943389893, + "learning_rate": 3.9690000000000005e-06, + "loss": 0.2244, + "step": 1326 + }, + { + "epoch": 1.120777027027027, + "grad_norm": 2.0798585414886475, + "learning_rate": 3.971999999999999e-06, + "loss": 0.285, + "step": 1327 + }, + { + "epoch": 1.1216216216216217, + "grad_norm": 2.3343048095703125, + "learning_rate": 3.975e-06, + "loss": 0.2858, + "step": 1328 + }, + { + "epoch": 1.1224662162162162, + "grad_norm": 3.444812297821045, + "learning_rate": 3.978e-06, + "loss": 0.3169, + "step": 1329 + }, + { + "epoch": 1.1233108108108107, + "grad_norm": 3.613131046295166, + "learning_rate": 3.9810000000000005e-06, + "loss": 0.3178, + "step": 1330 + }, + { + "epoch": 1.1241554054054055, + "grad_norm": 2.588315010070801, + "learning_rate": 3.984e-06, + "loss": 0.266, + "step": 1331 + }, + { + "epoch": 1.125, + "grad_norm": 3.8306024074554443, + "learning_rate": 3.987e-06, + "loss": 0.3046, + "step": 1332 + }, + { + "epoch": 1.1258445945945945, + "grad_norm": 3.9125113487243652, + "learning_rate": 3.99e-06, + "loss": 0.3564, + "step": 1333 + }, + { + "epoch": 1.1266891891891893, + "grad_norm": 3.8790194988250732, + "learning_rate": 3.993e-06, + "loss": 0.3804, + "step": 1334 + }, + { + "epoch": 1.1275337837837838, + "grad_norm": 8.303483009338379, + "learning_rate": 3.996e-06, + "loss": 0.9447, + "step": 1335 + }, + { + "epoch": 1.1283783783783783, + "grad_norm": 2.5319740772247314, + "learning_rate": 3.999e-06, + "loss": 0.599, + "step": 1336 + }, + { + "epoch": 1.129222972972973, + "grad_norm": 2.4417896270751953, + "learning_rate": 4.002e-06, + "loss": 0.4585, + "step": 1337 + }, + { + "epoch": 1.1300675675675675, + "grad_norm": 2.6679375171661377, + "learning_rate": 4.005000000000001e-06, + "loss": 0.2291, + "step": 1338 + }, + { + "epoch": 1.1309121621621623, + "grad_norm": 1.8498201370239258, + "learning_rate": 4.008e-06, + "loss": 0.244, + "step": 1339 + }, + { + "epoch": 1.1317567567567568, + "grad_norm": 1.4788143634796143, + "learning_rate": 4.011e-06, + "loss": 0.2293, + "step": 1340 + }, + { + "epoch": 1.1326013513513513, + "grad_norm": 1.6511422395706177, + "learning_rate": 4.014e-06, + "loss": 0.2667, + "step": 1341 + }, + { + "epoch": 1.133445945945946, + "grad_norm": 1.6509705781936646, + "learning_rate": 4.017e-06, + "loss": 0.2016, + "step": 1342 + }, + { + "epoch": 1.1342905405405406, + "grad_norm": 1.5674350261688232, + "learning_rate": 4.0200000000000005e-06, + "loss": 0.2121, + "step": 1343 + }, + { + "epoch": 1.135135135135135, + "grad_norm": 1.6363579034805298, + "learning_rate": 4.023e-06, + "loss": 0.2438, + "step": 1344 + }, + { + "epoch": 1.1359797297297298, + "grad_norm": 1.722085952758789, + "learning_rate": 4.026000000000001e-06, + "loss": 0.2361, + "step": 1345 + }, + { + "epoch": 1.1368243243243243, + "grad_norm": 1.6746989488601685, + "learning_rate": 4.029e-06, + "loss": 0.2345, + "step": 1346 + }, + { + "epoch": 1.1376689189189189, + "grad_norm": 1.313625693321228, + "learning_rate": 4.032e-06, + "loss": 0.2249, + "step": 1347 + }, + { + "epoch": 1.1385135135135136, + "grad_norm": 1.8895809650421143, + "learning_rate": 4.035e-06, + "loss": 0.2412, + "step": 1348 + }, + { + "epoch": 1.1393581081081081, + "grad_norm": 1.3034863471984863, + "learning_rate": 4.038e-06, + "loss": 0.1954, + "step": 1349 + }, + { + "epoch": 1.1402027027027026, + "grad_norm": 2.4074389934539795, + "learning_rate": 4.041e-06, + "loss": 0.2659, + "step": 1350 + }, + { + "epoch": 1.1410472972972974, + "grad_norm": 1.5558754205703735, + "learning_rate": 4.044000000000001e-06, + "loss": 0.1964, + "step": 1351 + }, + { + "epoch": 1.1418918918918919, + "grad_norm": 1.6169443130493164, + "learning_rate": 4.0469999999999995e-06, + "loss": 0.2567, + "step": 1352 + }, + { + "epoch": 1.1427364864864864, + "grad_norm": 1.6679073572158813, + "learning_rate": 4.05e-06, + "loss": 0.2539, + "step": 1353 + }, + { + "epoch": 1.1435810810810811, + "grad_norm": 3.9571475982666016, + "learning_rate": 4.053e-06, + "loss": 0.2506, + "step": 1354 + }, + { + "epoch": 1.1444256756756757, + "grad_norm": 2.436182975769043, + "learning_rate": 4.056e-06, + "loss": 0.2612, + "step": 1355 + }, + { + "epoch": 1.1452702702702702, + "grad_norm": 1.825347661972046, + "learning_rate": 4.0590000000000004e-06, + "loss": 0.2151, + "step": 1356 + }, + { + "epoch": 1.146114864864865, + "grad_norm": 1.5894114971160889, + "learning_rate": 4.062e-06, + "loss": 0.2247, + "step": 1357 + }, + { + "epoch": 1.1469594594594594, + "grad_norm": 1.5909581184387207, + "learning_rate": 4.065e-06, + "loss": 0.2138, + "step": 1358 + }, + { + "epoch": 1.147804054054054, + "grad_norm": 1.850847601890564, + "learning_rate": 4.068e-06, + "loss": 0.2199, + "step": 1359 + }, + { + "epoch": 1.1486486486486487, + "grad_norm": 1.836927890777588, + "learning_rate": 4.071e-06, + "loss": 0.2356, + "step": 1360 + }, + { + "epoch": 1.1494932432432432, + "grad_norm": 2.411816120147705, + "learning_rate": 4.074e-06, + "loss": 0.2472, + "step": 1361 + }, + { + "epoch": 1.1503378378378377, + "grad_norm": 1.587610125541687, + "learning_rate": 4.077e-06, + "loss": 0.1807, + "step": 1362 + }, + { + "epoch": 1.1511824324324325, + "grad_norm": 2.125458240509033, + "learning_rate": 4.080000000000001e-06, + "loss": 0.2133, + "step": 1363 + }, + { + "epoch": 1.152027027027027, + "grad_norm": 3.080408811569214, + "learning_rate": 4.083e-06, + "loss": 0.2222, + "step": 1364 + }, + { + "epoch": 1.1528716216216217, + "grad_norm": 1.9769071340560913, + "learning_rate": 4.0859999999999995e-06, + "loss": 0.2349, + "step": 1365 + }, + { + "epoch": 1.1537162162162162, + "grad_norm": 2.831348180770874, + "learning_rate": 4.089e-06, + "loss": 0.241, + "step": 1366 + }, + { + "epoch": 1.1545608108108107, + "grad_norm": 2.1594409942626953, + "learning_rate": 4.092e-06, + "loss": 0.2223, + "step": 1367 + }, + { + "epoch": 1.1554054054054055, + "grad_norm": 3.466806173324585, + "learning_rate": 4.095000000000001e-06, + "loss": 0.3114, + "step": 1368 + }, + { + "epoch": 1.15625, + "grad_norm": 1.628710150718689, + "learning_rate": 4.098e-06, + "loss": 0.2331, + "step": 1369 + }, + { + "epoch": 1.1570945945945945, + "grad_norm": 2.109632968902588, + "learning_rate": 4.100999999999999e-06, + "loss": 0.2816, + "step": 1370 + }, + { + "epoch": 1.1579391891891893, + "grad_norm": 2.114964246749878, + "learning_rate": 4.104e-06, + "loss": 0.2482, + "step": 1371 + }, + { + "epoch": 1.1587837837837838, + "grad_norm": 2.1899337768554688, + "learning_rate": 4.107e-06, + "loss": 0.2558, + "step": 1372 + }, + { + "epoch": 1.1596283783783783, + "grad_norm": 2.079357147216797, + "learning_rate": 4.1100000000000005e-06, + "loss": 0.2597, + "step": 1373 + }, + { + "epoch": 1.160472972972973, + "grad_norm": 2.4732472896575928, + "learning_rate": 4.113e-06, + "loss": 0.2805, + "step": 1374 + }, + { + "epoch": 1.1613175675675675, + "grad_norm": 3.101559638977051, + "learning_rate": 4.116e-06, + "loss": 0.3246, + "step": 1375 + }, + { + "epoch": 1.1621621621621623, + "grad_norm": 3.6572983264923096, + "learning_rate": 4.119e-06, + "loss": 0.3211, + "step": 1376 + }, + { + "epoch": 1.1630067567567568, + "grad_norm": 2.55678129196167, + "learning_rate": 4.122e-06, + "loss": 0.2674, + "step": 1377 + }, + { + "epoch": 1.1638513513513513, + "grad_norm": 2.6462063789367676, + "learning_rate": 4.125e-06, + "loss": 0.2257, + "step": 1378 + }, + { + "epoch": 1.164695945945946, + "grad_norm": 4.701154708862305, + "learning_rate": 4.128e-06, + "loss": 0.3679, + "step": 1379 + }, + { + "epoch": 1.1655405405405406, + "grad_norm": 2.7302701473236084, + "learning_rate": 4.131e-06, + "loss": 0.3065, + "step": 1380 + }, + { + "epoch": 1.166385135135135, + "grad_norm": 2.281104803085327, + "learning_rate": 4.1340000000000006e-06, + "loss": 0.2554, + "step": 1381 + }, + { + "epoch": 1.1672297297297298, + "grad_norm": 4.682715892791748, + "learning_rate": 4.137e-06, + "loss": 0.2871, + "step": 1382 + }, + { + "epoch": 1.1680743243243243, + "grad_norm": 2.981750965118408, + "learning_rate": 4.14e-06, + "loss": 0.3339, + "step": 1383 + }, + { + "epoch": 1.1689189189189189, + "grad_norm": 3.8852734565734863, + "learning_rate": 4.143e-06, + "loss": 0.4233, + "step": 1384 + }, + { + "epoch": 1.1697635135135136, + "grad_norm": 7.588109493255615, + "learning_rate": 4.146e-06, + "loss": 0.8751, + "step": 1385 + }, + { + "epoch": 1.1706081081081081, + "grad_norm": 9.11522388458252, + "learning_rate": 4.1490000000000004e-06, + "loss": 0.7324, + "step": 1386 + }, + { + "epoch": 1.1714527027027026, + "grad_norm": 2.9392201900482178, + "learning_rate": 4.152e-06, + "loss": 0.3144, + "step": 1387 + }, + { + "epoch": 1.1722972972972974, + "grad_norm": 11.069239616394043, + "learning_rate": 4.155000000000001e-06, + "loss": 0.2924, + "step": 1388 + }, + { + "epoch": 1.1731418918918919, + "grad_norm": 2.8486416339874268, + "learning_rate": 4.158e-06, + "loss": 0.2878, + "step": 1389 + }, + { + "epoch": 1.1739864864864864, + "grad_norm": 3.573345422744751, + "learning_rate": 4.161e-06, + "loss": 0.2719, + "step": 1390 + }, + { + "epoch": 1.1748310810810811, + "grad_norm": 3.01855206489563, + "learning_rate": 4.164e-06, + "loss": 0.2393, + "step": 1391 + }, + { + "epoch": 1.1756756756756757, + "grad_norm": 1.7733923196792603, + "learning_rate": 4.167e-06, + "loss": 0.235, + "step": 1392 + }, + { + "epoch": 1.1765202702702702, + "grad_norm": 1.84441339969635, + "learning_rate": 4.170000000000001e-06, + "loss": 0.2219, + "step": 1393 + }, + { + "epoch": 1.177364864864865, + "grad_norm": 2.3848133087158203, + "learning_rate": 4.1730000000000005e-06, + "loss": 0.2816, + "step": 1394 + }, + { + "epoch": 1.1782094594594594, + "grad_norm": 3.1700191497802734, + "learning_rate": 4.1759999999999995e-06, + "loss": 0.2736, + "step": 1395 + }, + { + "epoch": 1.179054054054054, + "grad_norm": 1.6309913396835327, + "learning_rate": 4.179e-06, + "loss": 0.212, + "step": 1396 + }, + { + "epoch": 1.1798986486486487, + "grad_norm": 2.264561414718628, + "learning_rate": 4.182e-06, + "loss": 0.2085, + "step": 1397 + }, + { + "epoch": 1.1807432432432432, + "grad_norm": 1.3613708019256592, + "learning_rate": 4.185000000000001e-06, + "loss": 0.1672, + "step": 1398 + }, + { + "epoch": 1.1815878378378377, + "grad_norm": 1.3251581192016602, + "learning_rate": 4.188e-06, + "loss": 0.182, + "step": 1399 + }, + { + "epoch": 1.1824324324324325, + "grad_norm": 1.3410048484802246, + "learning_rate": 4.191e-06, + "loss": 0.2065, + "step": 1400 + }, + { + "epoch": 1.183277027027027, + "grad_norm": 1.4416950941085815, + "learning_rate": 4.194e-06, + "loss": 0.2148, + "step": 1401 + }, + { + "epoch": 1.1841216216216217, + "grad_norm": 1.6065226793289185, + "learning_rate": 4.197e-06, + "loss": 0.1773, + "step": 1402 + }, + { + "epoch": 1.1849662162162162, + "grad_norm": 1.9729933738708496, + "learning_rate": 4.2000000000000004e-06, + "loss": 0.2635, + "step": 1403 + }, + { + "epoch": 1.1858108108108107, + "grad_norm": 1.6674429178237915, + "learning_rate": 4.203e-06, + "loss": 0.2259, + "step": 1404 + }, + { + "epoch": 1.1866554054054055, + "grad_norm": 1.763635277748108, + "learning_rate": 4.206e-06, + "loss": 0.2253, + "step": 1405 + }, + { + "epoch": 1.1875, + "grad_norm": 3.9895071983337402, + "learning_rate": 4.209000000000001e-06, + "loss": 0.2177, + "step": 1406 + }, + { + "epoch": 1.1883445945945945, + "grad_norm": 2.987583637237549, + "learning_rate": 4.212e-06, + "loss": 0.2337, + "step": 1407 + }, + { + "epoch": 1.1891891891891893, + "grad_norm": 1.5025608539581299, + "learning_rate": 4.215e-06, + "loss": 0.2015, + "step": 1408 + }, + { + "epoch": 1.1900337837837838, + "grad_norm": 2.472787618637085, + "learning_rate": 4.218e-06, + "loss": 0.2017, + "step": 1409 + }, + { + "epoch": 1.1908783783783783, + "grad_norm": 1.7498120069503784, + "learning_rate": 4.221e-06, + "loss": 0.2062, + "step": 1410 + }, + { + "epoch": 1.191722972972973, + "grad_norm": 2.2020459175109863, + "learning_rate": 4.2240000000000006e-06, + "loss": 0.236, + "step": 1411 + }, + { + "epoch": 1.1925675675675675, + "grad_norm": 1.841989278793335, + "learning_rate": 4.227e-06, + "loss": 0.2209, + "step": 1412 + }, + { + "epoch": 1.1934121621621623, + "grad_norm": 2.057370185852051, + "learning_rate": 4.229999999999999e-06, + "loss": 0.1998, + "step": 1413 + }, + { + "epoch": 1.1942567567567568, + "grad_norm": 1.9946751594543457, + "learning_rate": 4.233e-06, + "loss": 0.2045, + "step": 1414 + }, + { + "epoch": 1.1951013513513513, + "grad_norm": 1.9583179950714111, + "learning_rate": 4.236e-06, + "loss": 0.21, + "step": 1415 + }, + { + "epoch": 1.195945945945946, + "grad_norm": 1.550641417503357, + "learning_rate": 4.239e-06, + "loss": 0.1975, + "step": 1416 + }, + { + "epoch": 1.1967905405405406, + "grad_norm": 1.9993994235992432, + "learning_rate": 4.242e-06, + "loss": 0.2106, + "step": 1417 + }, + { + "epoch": 1.197635135135135, + "grad_norm": 7.023256301879883, + "learning_rate": 4.245e-06, + "loss": 0.2784, + "step": 1418 + }, + { + "epoch": 1.1984797297297298, + "grad_norm": 2.1227986812591553, + "learning_rate": 4.248e-06, + "loss": 0.2097, + "step": 1419 + }, + { + "epoch": 1.1993243243243243, + "grad_norm": 1.9550102949142456, + "learning_rate": 4.251e-06, + "loss": 0.2348, + "step": 1420 + }, + { + "epoch": 1.2001689189189189, + "grad_norm": 1.6375267505645752, + "learning_rate": 4.254e-06, + "loss": 0.2208, + "step": 1421 + }, + { + "epoch": 1.2010135135135136, + "grad_norm": 1.8926252126693726, + "learning_rate": 4.257e-06, + "loss": 0.2143, + "step": 1422 + }, + { + "epoch": 1.2018581081081081, + "grad_norm": 2.716078758239746, + "learning_rate": 4.26e-06, + "loss": 0.2263, + "step": 1423 + }, + { + "epoch": 1.2027027027027026, + "grad_norm": 2.497647285461426, + "learning_rate": 4.2630000000000005e-06, + "loss": 0.225, + "step": 1424 + }, + { + "epoch": 1.2035472972972974, + "grad_norm": 2.3699045181274414, + "learning_rate": 4.266e-06, + "loss": 0.2338, + "step": 1425 + }, + { + "epoch": 1.2043918918918919, + "grad_norm": 2.140399217605591, + "learning_rate": 4.269e-06, + "loss": 0.2524, + "step": 1426 + }, + { + "epoch": 1.2052364864864864, + "grad_norm": 2.162302017211914, + "learning_rate": 4.272e-06, + "loss": 0.2649, + "step": 1427 + }, + { + "epoch": 1.2060810810810811, + "grad_norm": 4.294188022613525, + "learning_rate": 4.275e-06, + "loss": 0.2553, + "step": 1428 + }, + { + "epoch": 1.2069256756756757, + "grad_norm": 2.0555288791656494, + "learning_rate": 4.278e-06, + "loss": 0.1959, + "step": 1429 + }, + { + "epoch": 1.2077702702702702, + "grad_norm": 3.157799005508423, + "learning_rate": 4.281e-06, + "loss": 0.2764, + "step": 1430 + }, + { + "epoch": 1.208614864864865, + "grad_norm": 3.340256929397583, + "learning_rate": 4.284000000000001e-06, + "loss": 0.208, + "step": 1431 + }, + { + "epoch": 1.2094594594594594, + "grad_norm": 2.5497279167175293, + "learning_rate": 4.287e-06, + "loss": 0.3105, + "step": 1432 + }, + { + "epoch": 1.210304054054054, + "grad_norm": 3.427278518676758, + "learning_rate": 4.29e-06, + "loss": 0.4057, + "step": 1433 + }, + { + "epoch": 1.2111486486486487, + "grad_norm": 3.688011646270752, + "learning_rate": 4.293e-06, + "loss": 0.4025, + "step": 1434 + }, + { + "epoch": 1.2119932432432432, + "grad_norm": 6.17764949798584, + "learning_rate": 4.296e-06, + "loss": 0.9413, + "step": 1435 + }, + { + "epoch": 1.2128378378378377, + "grad_norm": 3.470679998397827, + "learning_rate": 4.299000000000001e-06, + "loss": 0.6293, + "step": 1436 + }, + { + "epoch": 1.2136824324324325, + "grad_norm": 3.907270908355713, + "learning_rate": 4.3020000000000005e-06, + "loss": 0.3664, + "step": 1437 + }, + { + "epoch": 1.214527027027027, + "grad_norm": 3.142171859741211, + "learning_rate": 4.3049999999999994e-06, + "loss": 0.2447, + "step": 1438 + }, + { + "epoch": 1.2153716216216217, + "grad_norm": 3.209247350692749, + "learning_rate": 4.308e-06, + "loss": 0.2454, + "step": 1439 + }, + { + "epoch": 1.2162162162162162, + "grad_norm": 2.158942699432373, + "learning_rate": 4.311e-06, + "loss": 0.2345, + "step": 1440 + }, + { + "epoch": 1.2170608108108107, + "grad_norm": 1.6647306680679321, + "learning_rate": 4.3140000000000005e-06, + "loss": 0.1889, + "step": 1441 + }, + { + "epoch": 1.2179054054054055, + "grad_norm": 1.8848260641098022, + "learning_rate": 4.317e-06, + "loss": 0.2368, + "step": 1442 + }, + { + "epoch": 1.21875, + "grad_norm": 2.0487043857574463, + "learning_rate": 4.32e-06, + "loss": 0.2228, + "step": 1443 + }, + { + "epoch": 1.2195945945945945, + "grad_norm": 1.7472063302993774, + "learning_rate": 4.323e-06, + "loss": 0.2159, + "step": 1444 + }, + { + "epoch": 1.2204391891891893, + "grad_norm": 1.922025442123413, + "learning_rate": 4.326e-06, + "loss": 0.2217, + "step": 1445 + }, + { + "epoch": 1.2212837837837838, + "grad_norm": 2.1723275184631348, + "learning_rate": 4.329e-06, + "loss": 0.2154, + "step": 1446 + }, + { + "epoch": 1.2221283783783783, + "grad_norm": 1.9861878156661987, + "learning_rate": 4.332e-06, + "loss": 0.2513, + "step": 1447 + }, + { + "epoch": 1.222972972972973, + "grad_norm": 2.1793644428253174, + "learning_rate": 4.335e-06, + "loss": 0.2391, + "step": 1448 + }, + { + "epoch": 1.2238175675675675, + "grad_norm": 2.056459665298462, + "learning_rate": 4.338000000000001e-06, + "loss": 0.2332, + "step": 1449 + }, + { + "epoch": 1.2246621621621623, + "grad_norm": 1.9594537019729614, + "learning_rate": 4.341e-06, + "loss": 0.2208, + "step": 1450 + }, + { + "epoch": 1.2255067567567568, + "grad_norm": 2.0486419200897217, + "learning_rate": 4.344e-06, + "loss": 0.2347, + "step": 1451 + }, + { + "epoch": 1.2263513513513513, + "grad_norm": 1.6279255151748657, + "learning_rate": 4.347e-06, + "loss": 0.1892, + "step": 1452 + }, + { + "epoch": 1.227195945945946, + "grad_norm": 1.3484725952148438, + "learning_rate": 4.35e-06, + "loss": 0.1964, + "step": 1453 + }, + { + "epoch": 1.2280405405405406, + "grad_norm": 2.069291830062866, + "learning_rate": 4.3530000000000005e-06, + "loss": 0.2241, + "step": 1454 + }, + { + "epoch": 1.228885135135135, + "grad_norm": 1.9499818086624146, + "learning_rate": 4.356e-06, + "loss": 0.2312, + "step": 1455 + }, + { + "epoch": 1.2297297297297298, + "grad_norm": 1.7753318548202515, + "learning_rate": 4.359e-06, + "loss": 0.1977, + "step": 1456 + }, + { + "epoch": 1.2305743243243243, + "grad_norm": 1.3446979522705078, + "learning_rate": 4.362e-06, + "loss": 0.1761, + "step": 1457 + }, + { + "epoch": 1.2314189189189189, + "grad_norm": 1.6324083805084229, + "learning_rate": 4.365e-06, + "loss": 0.2034, + "step": 1458 + }, + { + "epoch": 1.2322635135135136, + "grad_norm": 1.6396920680999756, + "learning_rate": 4.368e-06, + "loss": 0.2408, + "step": 1459 + }, + { + "epoch": 1.2331081081081081, + "grad_norm": 1.888637900352478, + "learning_rate": 4.371e-06, + "loss": 0.1935, + "step": 1460 + }, + { + "epoch": 1.2339527027027026, + "grad_norm": 2.4468350410461426, + "learning_rate": 4.374000000000001e-06, + "loss": 0.2249, + "step": 1461 + }, + { + "epoch": 1.2347972972972974, + "grad_norm": 2.7995104789733887, + "learning_rate": 4.377e-06, + "loss": 0.2117, + "step": 1462 + }, + { + "epoch": 1.2356418918918919, + "grad_norm": 1.3100922107696533, + "learning_rate": 4.3799999999999996e-06, + "loss": 0.1873, + "step": 1463 + }, + { + "epoch": 1.2364864864864864, + "grad_norm": 1.918799638748169, + "learning_rate": 4.383e-06, + "loss": 0.2265, + "step": 1464 + }, + { + "epoch": 1.2373310810810811, + "grad_norm": 1.4016367197036743, + "learning_rate": 4.386e-06, + "loss": 0.1869, + "step": 1465 + }, + { + "epoch": 1.2381756756756757, + "grad_norm": 1.9194039106369019, + "learning_rate": 4.389000000000001e-06, + "loss": 0.3022, + "step": 1466 + }, + { + "epoch": 1.2390202702702702, + "grad_norm": 2.415968656539917, + "learning_rate": 4.3920000000000005e-06, + "loss": 0.2264, + "step": 1467 + }, + { + "epoch": 1.239864864864865, + "grad_norm": 1.7271966934204102, + "learning_rate": 4.395e-06, + "loss": 0.2392, + "step": 1468 + }, + { + "epoch": 1.2407094594594594, + "grad_norm": 1.6319506168365479, + "learning_rate": 4.398e-06, + "loss": 0.2194, + "step": 1469 + }, + { + "epoch": 1.241554054054054, + "grad_norm": 1.5037660598754883, + "learning_rate": 4.401e-06, + "loss": 0.2044, + "step": 1470 + }, + { + "epoch": 1.2423986486486487, + "grad_norm": 2.361487627029419, + "learning_rate": 4.4040000000000005e-06, + "loss": 0.2594, + "step": 1471 + }, + { + "epoch": 1.2432432432432432, + "grad_norm": 2.3227102756500244, + "learning_rate": 4.407e-06, + "loss": 0.2048, + "step": 1472 + }, + { + "epoch": 1.2440878378378377, + "grad_norm": 1.87447988986969, + "learning_rate": 4.41e-06, + "loss": 0.192, + "step": 1473 + }, + { + "epoch": 1.2449324324324325, + "grad_norm": 1.8433982133865356, + "learning_rate": 4.413000000000001e-06, + "loss": 0.2691, + "step": 1474 + }, + { + "epoch": 1.245777027027027, + "grad_norm": 1.6613513231277466, + "learning_rate": 4.416e-06, + "loss": 0.1756, + "step": 1475 + }, + { + "epoch": 1.2466216216216217, + "grad_norm": 3.158424139022827, + "learning_rate": 4.4189999999999995e-06, + "loss": 0.2267, + "step": 1476 + }, + { + "epoch": 1.2474662162162162, + "grad_norm": 3.133007764816284, + "learning_rate": 4.422e-06, + "loss": 0.2291, + "step": 1477 + }, + { + "epoch": 1.2483108108108107, + "grad_norm": 2.327798843383789, + "learning_rate": 4.425e-06, + "loss": 0.3096, + "step": 1478 + }, + { + "epoch": 1.2491554054054055, + "grad_norm": 1.9810945987701416, + "learning_rate": 4.428000000000001e-06, + "loss": 0.2268, + "step": 1479 + }, + { + "epoch": 1.25, + "grad_norm": 3.6880016326904297, + "learning_rate": 4.4310000000000004e-06, + "loss": 0.2172, + "step": 1480 + }, + { + "epoch": 1.2508445945945945, + "grad_norm": 2.2748100757598877, + "learning_rate": 4.433999999999999e-06, + "loss": 0.2275, + "step": 1481 + }, + { + "epoch": 1.2516891891891893, + "grad_norm": 1.9816619157791138, + "learning_rate": 4.437e-06, + "loss": 0.2177, + "step": 1482 + }, + { + "epoch": 1.2525337837837838, + "grad_norm": 2.296166181564331, + "learning_rate": 4.44e-06, + "loss": 0.2676, + "step": 1483 + }, + { + "epoch": 1.2533783783783785, + "grad_norm": 3.9171066284179688, + "learning_rate": 4.4430000000000005e-06, + "loss": 0.3543, + "step": 1484 + }, + { + "epoch": 1.254222972972973, + "grad_norm": 7.0013933181762695, + "learning_rate": 4.446e-06, + "loss": 0.9201, + "step": 1485 + }, + { + "epoch": 1.2550675675675675, + "grad_norm": 3.2997255325317383, + "learning_rate": 4.449e-06, + "loss": 0.5152, + "step": 1486 + }, + { + "epoch": 1.2559121621621623, + "grad_norm": 1.7087286710739136, + "learning_rate": 4.452e-06, + "loss": 0.2526, + "step": 1487 + }, + { + "epoch": 1.2567567567567568, + "grad_norm": 1.6652605533599854, + "learning_rate": 4.455e-06, + "loss": 0.2609, + "step": 1488 + }, + { + "epoch": 1.2576013513513513, + "grad_norm": 1.242306113243103, + "learning_rate": 4.458e-06, + "loss": 0.2047, + "step": 1489 + }, + { + "epoch": 1.258445945945946, + "grad_norm": 1.1821153163909912, + "learning_rate": 4.461e-06, + "loss": 0.1615, + "step": 1490 + }, + { + "epoch": 1.2592905405405406, + "grad_norm": 1.2177079916000366, + "learning_rate": 4.464e-06, + "loss": 0.1962, + "step": 1491 + }, + { + "epoch": 1.260135135135135, + "grad_norm": 1.463070273399353, + "learning_rate": 4.467000000000001e-06, + "loss": 0.2336, + "step": 1492 + }, + { + "epoch": 1.2609797297297298, + "grad_norm": 4.184276580810547, + "learning_rate": 4.4699999999999996e-06, + "loss": 0.184, + "step": 1493 + }, + { + "epoch": 1.2618243243243243, + "grad_norm": 2.9107327461242676, + "learning_rate": 4.473e-06, + "loss": 0.2165, + "step": 1494 + }, + { + "epoch": 1.2626689189189189, + "grad_norm": 1.1868648529052734, + "learning_rate": 4.476e-06, + "loss": 0.173, + "step": 1495 + }, + { + "epoch": 1.2635135135135136, + "grad_norm": 1.710089087486267, + "learning_rate": 4.479e-06, + "loss": 0.2168, + "step": 1496 + }, + { + "epoch": 1.2643581081081081, + "grad_norm": 1.7420116662979126, + "learning_rate": 4.4820000000000005e-06, + "loss": 0.1928, + "step": 1497 + }, + { + "epoch": 1.2652027027027026, + "grad_norm": 3.5109593868255615, + "learning_rate": 4.485e-06, + "loss": 0.163, + "step": 1498 + }, + { + "epoch": 1.2660472972972974, + "grad_norm": 1.2711050510406494, + "learning_rate": 4.488e-06, + "loss": 0.1545, + "step": 1499 + }, + { + "epoch": 1.2668918918918919, + "grad_norm": 1.4757506847381592, + "learning_rate": 4.491e-06, + "loss": 0.225, + "step": 1500 + }, + { + "epoch": 1.2677364864864864, + "grad_norm": 1.6691631078720093, + "learning_rate": 4.494e-06, + "loss": 0.2657, + "step": 1501 + }, + { + "epoch": 1.2685810810810811, + "grad_norm": 3.478837728500366, + "learning_rate": 4.497e-06, + "loss": 0.1945, + "step": 1502 + }, + { + "epoch": 1.2694256756756757, + "grad_norm": 1.4692988395690918, + "learning_rate": 4.5e-06, + "loss": 0.1673, + "step": 1503 + }, + { + "epoch": 1.2702702702702702, + "grad_norm": 1.7333409786224365, + "learning_rate": 4.503000000000001e-06, + "loss": 0.1774, + "step": 1504 + }, + { + "epoch": 1.271114864864865, + "grad_norm": 1.3540641069412231, + "learning_rate": 4.506e-06, + "loss": 0.1843, + "step": 1505 + }, + { + "epoch": 1.2719594594594594, + "grad_norm": 1.4434082508087158, + "learning_rate": 4.5089999999999995e-06, + "loss": 0.1927, + "step": 1506 + }, + { + "epoch": 1.272804054054054, + "grad_norm": 1.3037199974060059, + "learning_rate": 4.512e-06, + "loss": 0.1823, + "step": 1507 + }, + { + "epoch": 1.2736486486486487, + "grad_norm": 1.5562844276428223, + "learning_rate": 4.515e-06, + "loss": 0.2174, + "step": 1508 + }, + { + "epoch": 1.2744932432432432, + "grad_norm": 1.3409132957458496, + "learning_rate": 4.518000000000001e-06, + "loss": 0.1829, + "step": 1509 + }, + { + "epoch": 1.2753378378378377, + "grad_norm": 1.4616189002990723, + "learning_rate": 4.521e-06, + "loss": 0.186, + "step": 1510 + }, + { + "epoch": 1.2761824324324325, + "grad_norm": 2.0001001358032227, + "learning_rate": 4.524e-06, + "loss": 0.2005, + "step": 1511 + }, + { + "epoch": 1.277027027027027, + "grad_norm": 2.1958422660827637, + "learning_rate": 4.527e-06, + "loss": 0.1981, + "step": 1512 + }, + { + "epoch": 1.2778716216216215, + "grad_norm": 1.8548825979232788, + "learning_rate": 4.53e-06, + "loss": 0.1773, + "step": 1513 + }, + { + "epoch": 1.2787162162162162, + "grad_norm": 1.6317858695983887, + "learning_rate": 4.5330000000000005e-06, + "loss": 0.1956, + "step": 1514 + }, + { + "epoch": 1.2795608108108107, + "grad_norm": 1.7208548784255981, + "learning_rate": 4.536e-06, + "loss": 0.2054, + "step": 1515 + }, + { + "epoch": 1.2804054054054055, + "grad_norm": 1.5220632553100586, + "learning_rate": 4.539e-06, + "loss": 0.1711, + "step": 1516 + }, + { + "epoch": 1.28125, + "grad_norm": 1.6743417978286743, + "learning_rate": 4.542000000000001e-06, + "loss": 0.2327, + "step": 1517 + }, + { + "epoch": 1.2820945945945945, + "grad_norm": 1.815883994102478, + "learning_rate": 4.545e-06, + "loss": 0.1983, + "step": 1518 + }, + { + "epoch": 1.2829391891891893, + "grad_norm": 1.9231069087982178, + "learning_rate": 4.548e-06, + "loss": 0.2258, + "step": 1519 + }, + { + "epoch": 1.2837837837837838, + "grad_norm": 1.5869537591934204, + "learning_rate": 4.551e-06, + "loss": 0.1818, + "step": 1520 + }, + { + "epoch": 1.2846283783783785, + "grad_norm": 1.9035803079605103, + "learning_rate": 4.554e-06, + "loss": 0.2445, + "step": 1521 + }, + { + "epoch": 1.285472972972973, + "grad_norm": 2.129357099533081, + "learning_rate": 4.557000000000001e-06, + "loss": 0.2164, + "step": 1522 + }, + { + "epoch": 1.2863175675675675, + "grad_norm": 1.7226934432983398, + "learning_rate": 4.56e-06, + "loss": 0.1676, + "step": 1523 + }, + { + "epoch": 1.2871621621621623, + "grad_norm": 1.9836223125457764, + "learning_rate": 4.563e-06, + "loss": 0.2592, + "step": 1524 + }, + { + "epoch": 1.2880067567567568, + "grad_norm": 1.8918324708938599, + "learning_rate": 4.566e-06, + "loss": 0.2121, + "step": 1525 + }, + { + "epoch": 1.2888513513513513, + "grad_norm": 2.102487087249756, + "learning_rate": 4.569e-06, + "loss": 0.2468, + "step": 1526 + }, + { + "epoch": 1.289695945945946, + "grad_norm": 2.3074216842651367, + "learning_rate": 4.5720000000000004e-06, + "loss": 0.2341, + "step": 1527 + }, + { + "epoch": 1.2905405405405406, + "grad_norm": 2.7858667373657227, + "learning_rate": 4.575e-06, + "loss": 0.2135, + "step": 1528 + }, + { + "epoch": 1.291385135135135, + "grad_norm": 1.9250261783599854, + "learning_rate": 4.578000000000001e-06, + "loss": 0.2468, + "step": 1529 + }, + { + "epoch": 1.2922297297297298, + "grad_norm": 2.26497745513916, + "learning_rate": 4.581e-06, + "loss": 0.198, + "step": 1530 + }, + { + "epoch": 1.2930743243243243, + "grad_norm": 2.367528200149536, + "learning_rate": 4.584e-06, + "loss": 0.2688, + "step": 1531 + }, + { + "epoch": 1.2939189189189189, + "grad_norm": 2.1268258094787598, + "learning_rate": 4.587e-06, + "loss": 0.2167, + "step": 1532 + }, + { + "epoch": 1.2947635135135136, + "grad_norm": 5.003695487976074, + "learning_rate": 4.59e-06, + "loss": 0.2894, + "step": 1533 + }, + { + "epoch": 1.2956081081081081, + "grad_norm": 4.472856044769287, + "learning_rate": 4.593000000000001e-06, + "loss": 0.4532, + "step": 1534 + }, + { + "epoch": 1.2964527027027026, + "grad_norm": 4.997481822967529, + "learning_rate": 4.5960000000000006e-06, + "loss": 0.7379, + "step": 1535 + }, + { + "epoch": 1.2972972972972974, + "grad_norm": 2.632214069366455, + "learning_rate": 4.5989999999999995e-06, + "loss": 0.4796, + "step": 1536 + }, + { + "epoch": 1.2981418918918919, + "grad_norm": 1.6930913925170898, + "learning_rate": 4.602e-06, + "loss": 0.2775, + "step": 1537 + }, + { + "epoch": 1.2989864864864864, + "grad_norm": 2.116298198699951, + "learning_rate": 4.605e-06, + "loss": 0.1662, + "step": 1538 + }, + { + "epoch": 1.2998310810810811, + "grad_norm": 1.9470033645629883, + "learning_rate": 4.608e-06, + "loss": 0.198, + "step": 1539 + }, + { + "epoch": 1.3006756756756757, + "grad_norm": 1.6916438341140747, + "learning_rate": 4.611e-06, + "loss": 0.2538, + "step": 1540 + }, + { + "epoch": 1.3015202702702702, + "grad_norm": 1.4593232870101929, + "learning_rate": 4.614e-06, + "loss": 0.21, + "step": 1541 + }, + { + "epoch": 1.302364864864865, + "grad_norm": 1.7889823913574219, + "learning_rate": 4.617e-06, + "loss": 0.1602, + "step": 1542 + }, + { + "epoch": 1.3032094594594594, + "grad_norm": 1.2461577653884888, + "learning_rate": 4.62e-06, + "loss": 0.1872, + "step": 1543 + }, + { + "epoch": 1.304054054054054, + "grad_norm": 1.6295520067214966, + "learning_rate": 4.623e-06, + "loss": 0.2291, + "step": 1544 + }, + { + "epoch": 1.3048986486486487, + "grad_norm": 1.2872191667556763, + "learning_rate": 4.626e-06, + "loss": 0.1638, + "step": 1545 + }, + { + "epoch": 1.3057432432432432, + "grad_norm": 1.2668664455413818, + "learning_rate": 4.629e-06, + "loss": 0.1868, + "step": 1546 + }, + { + "epoch": 1.3065878378378377, + "grad_norm": 1.6938451528549194, + "learning_rate": 4.632000000000001e-06, + "loss": 0.2166, + "step": 1547 + }, + { + "epoch": 1.3074324324324325, + "grad_norm": 2.157235860824585, + "learning_rate": 4.635e-06, + "loss": 0.2945, + "step": 1548 + }, + { + "epoch": 1.308277027027027, + "grad_norm": 1.5058062076568604, + "learning_rate": 4.6379999999999995e-06, + "loss": 0.1973, + "step": 1549 + }, + { + "epoch": 1.3091216216216215, + "grad_norm": 1.3128142356872559, + "learning_rate": 4.641e-06, + "loss": 0.2023, + "step": 1550 + }, + { + "epoch": 1.3099662162162162, + "grad_norm": 1.4869714975357056, + "learning_rate": 4.644e-06, + "loss": 0.1944, + "step": 1551 + }, + { + "epoch": 1.3108108108108107, + "grad_norm": 1.8422240018844604, + "learning_rate": 4.6470000000000006e-06, + "loss": 0.253, + "step": 1552 + }, + { + "epoch": 1.3116554054054055, + "grad_norm": 4.726424217224121, + "learning_rate": 4.65e-06, + "loss": 0.2078, + "step": 1553 + }, + { + "epoch": 1.3125, + "grad_norm": 1.9179027080535889, + "learning_rate": 4.653e-06, + "loss": 0.2245, + "step": 1554 + }, + { + "epoch": 1.3133445945945945, + "grad_norm": 1.274351716041565, + "learning_rate": 4.656e-06, + "loss": 0.1521, + "step": 1555 + }, + { + "epoch": 1.3141891891891893, + "grad_norm": 1.4804633855819702, + "learning_rate": 4.659e-06, + "loss": 0.1547, + "step": 1556 + }, + { + "epoch": 1.3150337837837838, + "grad_norm": 1.5648022890090942, + "learning_rate": 4.6620000000000004e-06, + "loss": 0.2659, + "step": 1557 + }, + { + "epoch": 1.3158783783783785, + "grad_norm": 1.6082358360290527, + "learning_rate": 4.665e-06, + "loss": 0.1932, + "step": 1558 + }, + { + "epoch": 1.316722972972973, + "grad_norm": 1.3976863622665405, + "learning_rate": 4.668e-06, + "loss": 0.1588, + "step": 1559 + }, + { + "epoch": 1.3175675675675675, + "grad_norm": 1.7617270946502686, + "learning_rate": 4.671000000000001e-06, + "loss": 0.1734, + "step": 1560 + }, + { + "epoch": 1.3184121621621623, + "grad_norm": 1.773714303970337, + "learning_rate": 4.674e-06, + "loss": 0.2241, + "step": 1561 + }, + { + "epoch": 1.3192567567567568, + "grad_norm": 1.7340244054794312, + "learning_rate": 4.677e-06, + "loss": 0.1805, + "step": 1562 + }, + { + "epoch": 1.3201013513513513, + "grad_norm": 1.3160573244094849, + "learning_rate": 4.68e-06, + "loss": 0.195, + "step": 1563 + }, + { + "epoch": 1.320945945945946, + "grad_norm": 1.4585729837417603, + "learning_rate": 4.683e-06, + "loss": 0.1895, + "step": 1564 + }, + { + "epoch": 1.3217905405405406, + "grad_norm": 1.562736988067627, + "learning_rate": 4.6860000000000005e-06, + "loss": 0.2026, + "step": 1565 + }, + { + "epoch": 1.322635135135135, + "grad_norm": 1.7149602174758911, + "learning_rate": 4.689e-06, + "loss": 0.1859, + "step": 1566 + }, + { + "epoch": 1.3234797297297298, + "grad_norm": 2.140077590942383, + "learning_rate": 4.692e-06, + "loss": 0.187, + "step": 1567 + }, + { + "epoch": 1.3243243243243243, + "grad_norm": 1.7028133869171143, + "learning_rate": 4.695e-06, + "loss": 0.2105, + "step": 1568 + }, + { + "epoch": 1.3251689189189189, + "grad_norm": 1.5647525787353516, + "learning_rate": 4.698e-06, + "loss": 0.213, + "step": 1569 + }, + { + "epoch": 1.3260135135135136, + "grad_norm": 1.3427051305770874, + "learning_rate": 4.701e-06, + "loss": 0.1808, + "step": 1570 + }, + { + "epoch": 1.3268581081081081, + "grad_norm": 2.3056492805480957, + "learning_rate": 4.704e-06, + "loss": 0.2348, + "step": 1571 + }, + { + "epoch": 1.3277027027027026, + "grad_norm": 1.6560503244400024, + "learning_rate": 4.707000000000001e-06, + "loss": 0.1971, + "step": 1572 + }, + { + "epoch": 1.3285472972972974, + "grad_norm": 1.5411895513534546, + "learning_rate": 4.71e-06, + "loss": 0.1836, + "step": 1573 + }, + { + "epoch": 1.3293918918918919, + "grad_norm": 1.5494648218154907, + "learning_rate": 4.713e-06, + "loss": 0.2185, + "step": 1574 + }, + { + "epoch": 1.3302364864864864, + "grad_norm": 1.7651010751724243, + "learning_rate": 4.716e-06, + "loss": 0.2312, + "step": 1575 + }, + { + "epoch": 1.3310810810810811, + "grad_norm": 1.6296474933624268, + "learning_rate": 4.719e-06, + "loss": 0.1866, + "step": 1576 + }, + { + "epoch": 1.3319256756756757, + "grad_norm": 1.9299606084823608, + "learning_rate": 4.722000000000001e-06, + "loss": 0.2505, + "step": 1577 + }, + { + "epoch": 1.3327702702702702, + "grad_norm": 1.7887043952941895, + "learning_rate": 4.7250000000000005e-06, + "loss": 0.2106, + "step": 1578 + }, + { + "epoch": 1.333614864864865, + "grad_norm": 2.7638516426086426, + "learning_rate": 4.7279999999999995e-06, + "loss": 0.2071, + "step": 1579 + }, + { + "epoch": 1.3344594594594594, + "grad_norm": 2.217580795288086, + "learning_rate": 4.731e-06, + "loss": 0.2481, + "step": 1580 + }, + { + "epoch": 1.335304054054054, + "grad_norm": 1.7119313478469849, + "learning_rate": 4.734e-06, + "loss": 0.1998, + "step": 1581 + }, + { + "epoch": 1.3361486486486487, + "grad_norm": 2.1830079555511475, + "learning_rate": 4.7370000000000006e-06, + "loss": 0.2847, + "step": 1582 + }, + { + "epoch": 1.3369932432432432, + "grad_norm": 2.138601303100586, + "learning_rate": 4.74e-06, + "loss": 0.2126, + "step": 1583 + }, + { + "epoch": 1.3378378378378377, + "grad_norm": 3.9024746417999268, + "learning_rate": 4.743e-06, + "loss": 0.3791, + "step": 1584 + }, + { + "epoch": 1.3386824324324325, + "grad_norm": 3.373231887817383, + "learning_rate": 4.746e-06, + "loss": 0.9149, + "step": 1585 + }, + { + "epoch": 1.339527027027027, + "grad_norm": 3.3031623363494873, + "learning_rate": 4.749e-06, + "loss": 0.4816, + "step": 1586 + }, + { + "epoch": 1.3403716216216215, + "grad_norm": 2.192108392715454, + "learning_rate": 4.752e-06, + "loss": 0.3661, + "step": 1587 + }, + { + "epoch": 1.3412162162162162, + "grad_norm": 1.1590098142623901, + "learning_rate": 4.755e-06, + "loss": 0.2365, + "step": 1588 + }, + { + "epoch": 1.3420608108108107, + "grad_norm": 1.6874486207962036, + "learning_rate": 4.758e-06, + "loss": 0.2075, + "step": 1589 + }, + { + "epoch": 1.3429054054054055, + "grad_norm": 1.4894559383392334, + "learning_rate": 4.761000000000001e-06, + "loss": 0.2172, + "step": 1590 + }, + { + "epoch": 1.34375, + "grad_norm": 1.435017704963684, + "learning_rate": 4.764e-06, + "loss": 0.1907, + "step": 1591 + }, + { + "epoch": 1.3445945945945945, + "grad_norm": 1.477776050567627, + "learning_rate": 4.767e-06, + "loss": 0.1674, + "step": 1592 + }, + { + "epoch": 1.3454391891891893, + "grad_norm": 1.3346420526504517, + "learning_rate": 4.77e-06, + "loss": 0.1601, + "step": 1593 + }, + { + "epoch": 1.3462837837837838, + "grad_norm": 1.3952150344848633, + "learning_rate": 4.773e-06, + "loss": 0.1816, + "step": 1594 + }, + { + "epoch": 1.3471283783783785, + "grad_norm": 1.7796143293380737, + "learning_rate": 4.7760000000000005e-06, + "loss": 0.1848, + "step": 1595 + }, + { + "epoch": 1.347972972972973, + "grad_norm": 1.9879733324050903, + "learning_rate": 4.779e-06, + "loss": 0.1935, + "step": 1596 + }, + { + "epoch": 1.3488175675675675, + "grad_norm": 1.4623630046844482, + "learning_rate": 4.782e-06, + "loss": 0.1774, + "step": 1597 + }, + { + "epoch": 1.3496621621621623, + "grad_norm": 1.515995740890503, + "learning_rate": 4.785e-06, + "loss": 0.1919, + "step": 1598 + }, + { + "epoch": 1.3505067567567568, + "grad_norm": 1.1635233163833618, + "learning_rate": 4.788e-06, + "loss": 0.1448, + "step": 1599 + }, + { + "epoch": 1.3513513513513513, + "grad_norm": 1.5840973854064941, + "learning_rate": 4.791e-06, + "loss": 0.1743, + "step": 1600 + }, + { + "epoch": 1.352195945945946, + "grad_norm": 1.4772590398788452, + "learning_rate": 4.794e-06, + "loss": 0.2074, + "step": 1601 + }, + { + "epoch": 1.3530405405405406, + "grad_norm": 1.3433573246002197, + "learning_rate": 4.797e-06, + "loss": 0.1446, + "step": 1602 + }, + { + "epoch": 1.353885135135135, + "grad_norm": 1.4935389757156372, + "learning_rate": 4.800000000000001e-06, + "loss": 0.2154, + "step": 1603 + }, + { + "epoch": 1.3547297297297298, + "grad_norm": 1.320406198501587, + "learning_rate": 4.803e-06, + "loss": 0.1878, + "step": 1604 + }, + { + "epoch": 1.3555743243243243, + "grad_norm": 1.8941783905029297, + "learning_rate": 4.806e-06, + "loss": 0.2034, + "step": 1605 + }, + { + "epoch": 1.3564189189189189, + "grad_norm": 1.4314018487930298, + "learning_rate": 4.809e-06, + "loss": 0.1938, + "step": 1606 + }, + { + "epoch": 1.3572635135135136, + "grad_norm": 3.1836860179901123, + "learning_rate": 4.812e-06, + "loss": 0.1674, + "step": 1607 + }, + { + "epoch": 1.3581081081081081, + "grad_norm": 2.0748777389526367, + "learning_rate": 4.8150000000000005e-06, + "loss": 0.2129, + "step": 1608 + }, + { + "epoch": 1.3589527027027026, + "grad_norm": 1.2451967000961304, + "learning_rate": 4.818e-06, + "loss": 0.1636, + "step": 1609 + }, + { + "epoch": 1.3597972972972974, + "grad_norm": 1.3352161645889282, + "learning_rate": 4.821e-06, + "loss": 0.1608, + "step": 1610 + }, + { + "epoch": 1.3606418918918919, + "grad_norm": 1.716330885887146, + "learning_rate": 4.824e-06, + "loss": 0.2041, + "step": 1611 + }, + { + "epoch": 1.3614864864864864, + "grad_norm": 1.3873698711395264, + "learning_rate": 4.827e-06, + "loss": 0.1751, + "step": 1612 + }, + { + "epoch": 1.3623310810810811, + "grad_norm": 1.6107710599899292, + "learning_rate": 4.83e-06, + "loss": 0.1821, + "step": 1613 + }, + { + "epoch": 1.3631756756756757, + "grad_norm": 2.232847213745117, + "learning_rate": 4.833e-06, + "loss": 0.2059, + "step": 1614 + }, + { + "epoch": 1.3640202702702702, + "grad_norm": 1.7780243158340454, + "learning_rate": 4.836000000000001e-06, + "loss": 0.1877, + "step": 1615 + }, + { + "epoch": 1.364864864864865, + "grad_norm": 2.0209171772003174, + "learning_rate": 4.839e-06, + "loss": 0.1906, + "step": 1616 + }, + { + "epoch": 1.3657094594594594, + "grad_norm": 1.5084786415100098, + "learning_rate": 4.8419999999999996e-06, + "loss": 0.1863, + "step": 1617 + }, + { + "epoch": 1.366554054054054, + "grad_norm": 2.389759063720703, + "learning_rate": 4.845e-06, + "loss": 0.3469, + "step": 1618 + }, + { + "epoch": 1.3673986486486487, + "grad_norm": 1.5516207218170166, + "learning_rate": 4.848e-06, + "loss": 0.2245, + "step": 1619 + }, + { + "epoch": 1.3682432432432432, + "grad_norm": 1.413655400276184, + "learning_rate": 4.851000000000001e-06, + "loss": 0.1686, + "step": 1620 + }, + { + "epoch": 1.3690878378378377, + "grad_norm": 1.4165836572647095, + "learning_rate": 4.8540000000000005e-06, + "loss": 0.1788, + "step": 1621 + }, + { + "epoch": 1.3699324324324325, + "grad_norm": 2.6838691234588623, + "learning_rate": 4.856999999999999e-06, + "loss": 0.1755, + "step": 1622 + }, + { + "epoch": 1.370777027027027, + "grad_norm": 2.4537580013275146, + "learning_rate": 4.86e-06, + "loss": 0.199, + "step": 1623 + }, + { + "epoch": 1.3716216216216215, + "grad_norm": 1.5651308298110962, + "learning_rate": 4.863e-06, + "loss": 0.1926, + "step": 1624 + }, + { + "epoch": 1.3724662162162162, + "grad_norm": 1.8156590461730957, + "learning_rate": 4.8660000000000005e-06, + "loss": 0.2161, + "step": 1625 + }, + { + "epoch": 1.3733108108108107, + "grad_norm": 2.429020404815674, + "learning_rate": 4.869e-06, + "loss": 0.2478, + "step": 1626 + }, + { + "epoch": 1.3741554054054055, + "grad_norm": 1.7156118154525757, + "learning_rate": 4.872e-06, + "loss": 0.2377, + "step": 1627 + }, + { + "epoch": 1.375, + "grad_norm": 1.4375683069229126, + "learning_rate": 4.875e-06, + "loss": 0.1776, + "step": 1628 + }, + { + "epoch": 1.3758445945945945, + "grad_norm": 8.677502632141113, + "learning_rate": 4.878e-06, + "loss": 0.2548, + "step": 1629 + }, + { + "epoch": 1.3766891891891893, + "grad_norm": 2.5117623805999756, + "learning_rate": 4.881e-06, + "loss": 0.2708, + "step": 1630 + }, + { + "epoch": 1.3775337837837838, + "grad_norm": 2.020538568496704, + "learning_rate": 4.884e-06, + "loss": 0.1953, + "step": 1631 + }, + { + "epoch": 1.3783783783783785, + "grad_norm": 2.5296080112457275, + "learning_rate": 4.887e-06, + "loss": 0.2369, + "step": 1632 + }, + { + "epoch": 1.379222972972973, + "grad_norm": 1.9464046955108643, + "learning_rate": 4.890000000000001e-06, + "loss": 0.2115, + "step": 1633 + }, + { + "epoch": 1.3800675675675675, + "grad_norm": 4.4374871253967285, + "learning_rate": 4.8929999999999996e-06, + "loss": 0.4128, + "step": 1634 + }, + { + "epoch": 1.3809121621621623, + "grad_norm": 6.606476783752441, + "learning_rate": 4.896e-06, + "loss": 0.7502, + "step": 1635 + }, + { + "epoch": 1.3817567567567568, + "grad_norm": 5.777895927429199, + "learning_rate": 4.899e-06, + "loss": 0.5848, + "step": 1636 + }, + { + "epoch": 1.3826013513513513, + "grad_norm": 2.248185873031616, + "learning_rate": 4.902e-06, + "loss": 0.2453, + "step": 1637 + }, + { + "epoch": 1.383445945945946, + "grad_norm": 1.5856684446334839, + "learning_rate": 4.9050000000000005e-06, + "loss": 0.196, + "step": 1638 + }, + { + "epoch": 1.3842905405405406, + "grad_norm": 5.930941104888916, + "learning_rate": 4.908e-06, + "loss": 0.2968, + "step": 1639 + }, + { + "epoch": 1.385135135135135, + "grad_norm": 1.4815367460250854, + "learning_rate": 4.911e-06, + "loss": 0.2777, + "step": 1640 + }, + { + "epoch": 1.3859797297297298, + "grad_norm": 2.646932601928711, + "learning_rate": 4.914e-06, + "loss": 0.1896, + "step": 1641 + }, + { + "epoch": 1.3868243243243243, + "grad_norm": 1.9308559894561768, + "learning_rate": 4.917e-06, + "loss": 0.194, + "step": 1642 + }, + { + "epoch": 1.3876689189189189, + "grad_norm": 1.1267740726470947, + "learning_rate": 4.92e-06, + "loss": 0.1706, + "step": 1643 + }, + { + "epoch": 1.3885135135135136, + "grad_norm": 1.4090063571929932, + "learning_rate": 4.923e-06, + "loss": 0.1729, + "step": 1644 + }, + { + "epoch": 1.3893581081081081, + "grad_norm": 1.2829595804214478, + "learning_rate": 4.926000000000001e-06, + "loss": 0.1768, + "step": 1645 + }, + { + "epoch": 1.3902027027027026, + "grad_norm": 1.3032886981964111, + "learning_rate": 4.929000000000001e-06, + "loss": 0.1758, + "step": 1646 + }, + { + "epoch": 1.3910472972972974, + "grad_norm": 1.2857180833816528, + "learning_rate": 4.9319999999999995e-06, + "loss": 0.1758, + "step": 1647 + }, + { + "epoch": 1.3918918918918919, + "grad_norm": 1.9099396467208862, + "learning_rate": 4.935e-06, + "loss": 0.1748, + "step": 1648 + }, + { + "epoch": 1.3927364864864864, + "grad_norm": 1.3197104930877686, + "learning_rate": 4.938e-06, + "loss": 0.1648, + "step": 1649 + }, + { + "epoch": 1.3935810810810811, + "grad_norm": 1.404900312423706, + "learning_rate": 4.941000000000001e-06, + "loss": 0.2217, + "step": 1650 + }, + { + "epoch": 1.3944256756756757, + "grad_norm": 1.300954818725586, + "learning_rate": 4.9440000000000004e-06, + "loss": 0.139, + "step": 1651 + }, + { + "epoch": 1.3952702702702702, + "grad_norm": 1.3419387340545654, + "learning_rate": 4.947e-06, + "loss": 0.1877, + "step": 1652 + }, + { + "epoch": 1.396114864864865, + "grad_norm": 1.2363172769546509, + "learning_rate": 4.95e-06, + "loss": 0.1666, + "step": 1653 + }, + { + "epoch": 1.3969594594594594, + "grad_norm": 1.2362052202224731, + "learning_rate": 4.953e-06, + "loss": 0.1555, + "step": 1654 + }, + { + "epoch": 1.397804054054054, + "grad_norm": 1.3820769786834717, + "learning_rate": 4.9560000000000005e-06, + "loss": 0.1617, + "step": 1655 + }, + { + "epoch": 1.3986486486486487, + "grad_norm": 9.01955509185791, + "learning_rate": 4.959e-06, + "loss": 0.2069, + "step": 1656 + }, + { + "epoch": 1.3994932432432432, + "grad_norm": 1.477208137512207, + "learning_rate": 4.962e-06, + "loss": 0.1889, + "step": 1657 + }, + { + "epoch": 1.4003378378378377, + "grad_norm": 1.495621919631958, + "learning_rate": 4.965000000000001e-06, + "loss": 0.1741, + "step": 1658 + }, + { + "epoch": 1.4011824324324325, + "grad_norm": 1.2375730276107788, + "learning_rate": 4.968e-06, + "loss": 0.1604, + "step": 1659 + }, + { + "epoch": 1.402027027027027, + "grad_norm": 1.6407475471496582, + "learning_rate": 4.9709999999999995e-06, + "loss": 0.2278, + "step": 1660 + }, + { + "epoch": 1.4028716216216215, + "grad_norm": 1.3395596742630005, + "learning_rate": 4.974e-06, + "loss": 0.1516, + "step": 1661 + }, + { + "epoch": 1.4037162162162162, + "grad_norm": 1.6312614679336548, + "learning_rate": 4.977e-06, + "loss": 0.187, + "step": 1662 + }, + { + "epoch": 1.4045608108108107, + "grad_norm": 1.7105951309204102, + "learning_rate": 4.980000000000001e-06, + "loss": 0.1982, + "step": 1663 + }, + { + "epoch": 1.4054054054054055, + "grad_norm": 3.870178461074829, + "learning_rate": 4.983e-06, + "loss": 0.1699, + "step": 1664 + }, + { + "epoch": 1.40625, + "grad_norm": 1.3674215078353882, + "learning_rate": 4.985999999999999e-06, + "loss": 0.1462, + "step": 1665 + }, + { + "epoch": 1.4070945945945945, + "grad_norm": 1.3360575437545776, + "learning_rate": 4.989e-06, + "loss": 0.14, + "step": 1666 + }, + { + "epoch": 1.4079391891891893, + "grad_norm": 1.6585969924926758, + "learning_rate": 4.992e-06, + "loss": 0.1794, + "step": 1667 + }, + { + "epoch": 1.4087837837837838, + "grad_norm": 2.0767927169799805, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.232, + "step": 1668 + }, + { + "epoch": 1.4096283783783785, + "grad_norm": 1.5925108194351196, + "learning_rate": 4.998e-06, + "loss": 0.1782, + "step": 1669 + }, + { + "epoch": 1.410472972972973, + "grad_norm": 1.3230664730072021, + "learning_rate": 5.001e-06, + "loss": 0.1683, + "step": 1670 + }, + { + "epoch": 1.4113175675675675, + "grad_norm": 1.4573302268981934, + "learning_rate": 5.004e-06, + "loss": 0.1588, + "step": 1671 + }, + { + "epoch": 1.4121621621621623, + "grad_norm": 1.7314451932907104, + "learning_rate": 5.007e-06, + "loss": 0.1745, + "step": 1672 + }, + { + "epoch": 1.4130067567567568, + "grad_norm": 1.734103798866272, + "learning_rate": 5.01e-06, + "loss": 0.1881, + "step": 1673 + }, + { + "epoch": 1.4138513513513513, + "grad_norm": 1.5550816059112549, + "learning_rate": 5.013e-06, + "loss": 0.2133, + "step": 1674 + }, + { + "epoch": 1.414695945945946, + "grad_norm": 1.3851512670516968, + "learning_rate": 5.016e-06, + "loss": 0.1603, + "step": 1675 + }, + { + "epoch": 1.4155405405405406, + "grad_norm": 2.397512197494507, + "learning_rate": 5.0190000000000006e-06, + "loss": 0.2087, + "step": 1676 + }, + { + "epoch": 1.416385135135135, + "grad_norm": 2.37619948387146, + "learning_rate": 5.0219999999999995e-06, + "loss": 0.1783, + "step": 1677 + }, + { + "epoch": 1.4172297297297298, + "grad_norm": 2.3001227378845215, + "learning_rate": 5.025e-06, + "loss": 0.2192, + "step": 1678 + }, + { + "epoch": 1.4180743243243243, + "grad_norm": 1.772675633430481, + "learning_rate": 5.028e-06, + "loss": 0.2023, + "step": 1679 + }, + { + "epoch": 1.4189189189189189, + "grad_norm": 1.5861210823059082, + "learning_rate": 5.031e-06, + "loss": 0.2222, + "step": 1680 + }, + { + "epoch": 1.4197635135135136, + "grad_norm": 2.119107961654663, + "learning_rate": 5.034e-06, + "loss": 0.1555, + "step": 1681 + }, + { + "epoch": 1.4206081081081081, + "grad_norm": 2.1693403720855713, + "learning_rate": 5.037e-06, + "loss": 0.2269, + "step": 1682 + }, + { + "epoch": 1.4214527027027026, + "grad_norm": 1.6941003799438477, + "learning_rate": 5.04e-06, + "loss": 0.2028, + "step": 1683 + }, + { + "epoch": 1.4222972972972974, + "grad_norm": 2.8855621814727783, + "learning_rate": 5.043e-06, + "loss": 0.4087, + "step": 1684 + }, + { + "epoch": 1.4231418918918919, + "grad_norm": 7.005020618438721, + "learning_rate": 5.046e-06, + "loss": 1.0026, + "step": 1685 + }, + { + "epoch": 1.4239864864864864, + "grad_norm": 2.6257386207580566, + "learning_rate": 5.049e-06, + "loss": 0.5439, + "step": 1686 + }, + { + "epoch": 1.4248310810810811, + "grad_norm": 2.8622007369995117, + "learning_rate": 5.052e-06, + "loss": 0.2863, + "step": 1687 + }, + { + "epoch": 1.4256756756756757, + "grad_norm": 2.7709267139434814, + "learning_rate": 5.055000000000001e-06, + "loss": 0.2292, + "step": 1688 + }, + { + "epoch": 1.4265202702702702, + "grad_norm": 3.245650291442871, + "learning_rate": 5.0580000000000005e-06, + "loss": 0.2917, + "step": 1689 + }, + { + "epoch": 1.427364864864865, + "grad_norm": 1.7779567241668701, + "learning_rate": 5.0609999999999995e-06, + "loss": 0.1843, + "step": 1690 + }, + { + "epoch": 1.4282094594594594, + "grad_norm": 1.4708261489868164, + "learning_rate": 5.064e-06, + "loss": 0.1616, + "step": 1691 + }, + { + "epoch": 1.429054054054054, + "grad_norm": 1.2617918252944946, + "learning_rate": 5.067e-06, + "loss": 0.1907, + "step": 1692 + }, + { + "epoch": 1.4298986486486487, + "grad_norm": 1.2609525918960571, + "learning_rate": 5.070000000000001e-06, + "loss": 0.1845, + "step": 1693 + }, + { + "epoch": 1.4307432432432432, + "grad_norm": 1.0361642837524414, + "learning_rate": 5.073e-06, + "loss": 0.1593, + "step": 1694 + }, + { + "epoch": 1.4315878378378377, + "grad_norm": 1.5848252773284912, + "learning_rate": 5.076e-06, + "loss": 0.1402, + "step": 1695 + }, + { + "epoch": 1.4324324324324325, + "grad_norm": 1.6258968114852905, + "learning_rate": 5.079e-06, + "loss": 0.153, + "step": 1696 + }, + { + "epoch": 1.433277027027027, + "grad_norm": 1.5922714471817017, + "learning_rate": 5.082e-06, + "loss": 0.1734, + "step": 1697 + }, + { + "epoch": 1.4341216216216215, + "grad_norm": 1.1719039678573608, + "learning_rate": 5.0850000000000004e-06, + "loss": 0.1621, + "step": 1698 + }, + { + "epoch": 1.4349662162162162, + "grad_norm": 1.2151070833206177, + "learning_rate": 5.088e-06, + "loss": 0.1688, + "step": 1699 + }, + { + "epoch": 1.4358108108108107, + "grad_norm": 1.3244596719741821, + "learning_rate": 5.091e-06, + "loss": 0.1587, + "step": 1700 + }, + { + "epoch": 1.4366554054054055, + "grad_norm": 1.9959625005722046, + "learning_rate": 5.094000000000001e-06, + "loss": 0.1892, + "step": 1701 + }, + { + "epoch": 1.4375, + "grad_norm": 2.093501091003418, + "learning_rate": 5.097e-06, + "loss": 0.2046, + "step": 1702 + }, + { + "epoch": 1.4383445945945945, + "grad_norm": 1.257461428642273, + "learning_rate": 5.1e-06, + "loss": 0.144, + "step": 1703 + }, + { + "epoch": 1.4391891891891893, + "grad_norm": 2.132206916809082, + "learning_rate": 5.103e-06, + "loss": 0.2118, + "step": 1704 + }, + { + "epoch": 1.4400337837837838, + "grad_norm": 1.624934434890747, + "learning_rate": 5.106e-06, + "loss": 0.1779, + "step": 1705 + }, + { + "epoch": 1.4408783783783785, + "grad_norm": 2.017587184906006, + "learning_rate": 5.1090000000000006e-06, + "loss": 0.1675, + "step": 1706 + }, + { + "epoch": 1.441722972972973, + "grad_norm": 1.8997743129730225, + "learning_rate": 5.112e-06, + "loss": 0.1724, + "step": 1707 + }, + { + "epoch": 1.4425675675675675, + "grad_norm": 1.300408124923706, + "learning_rate": 5.115e-06, + "loss": 0.1688, + "step": 1708 + }, + { + "epoch": 1.4434121621621623, + "grad_norm": 1.5031449794769287, + "learning_rate": 5.118e-06, + "loss": 0.1682, + "step": 1709 + }, + { + "epoch": 1.4442567567567568, + "grad_norm": 1.310365080833435, + "learning_rate": 5.121e-06, + "loss": 0.1426, + "step": 1710 + }, + { + "epoch": 1.4451013513513513, + "grad_norm": 2.407468318939209, + "learning_rate": 5.124e-06, + "loss": 0.2139, + "step": 1711 + }, + { + "epoch": 1.445945945945946, + "grad_norm": 1.6843092441558838, + "learning_rate": 5.127e-06, + "loss": 0.1571, + "step": 1712 + }, + { + "epoch": 1.4467905405405406, + "grad_norm": 1.750770092010498, + "learning_rate": 5.130000000000001e-06, + "loss": 0.1624, + "step": 1713 + }, + { + "epoch": 1.447635135135135, + "grad_norm": 1.4899258613586426, + "learning_rate": 5.133e-06, + "loss": 0.1877, + "step": 1714 + }, + { + "epoch": 1.4484797297297298, + "grad_norm": 1.5062834024429321, + "learning_rate": 5.136e-06, + "loss": 0.1873, + "step": 1715 + }, + { + "epoch": 1.4493243243243243, + "grad_norm": 1.348258137702942, + "learning_rate": 5.139e-06, + "loss": 0.2055, + "step": 1716 + }, + { + "epoch": 1.4501689189189189, + "grad_norm": 1.4572800397872925, + "learning_rate": 5.142e-06, + "loss": 0.1827, + "step": 1717 + }, + { + "epoch": 1.4510135135135136, + "grad_norm": 1.652618408203125, + "learning_rate": 5.145000000000001e-06, + "loss": 0.1619, + "step": 1718 + }, + { + "epoch": 1.4518581081081081, + "grad_norm": 2.8526625633239746, + "learning_rate": 5.1480000000000005e-06, + "loss": 0.2813, + "step": 1719 + }, + { + "epoch": 1.4527027027027026, + "grad_norm": 1.1546136140823364, + "learning_rate": 5.1509999999999995e-06, + "loss": 0.139, + "step": 1720 + }, + { + "epoch": 1.4535472972972974, + "grad_norm": 1.4774423837661743, + "learning_rate": 5.154e-06, + "loss": 0.1783, + "step": 1721 + }, + { + "epoch": 1.4543918918918919, + "grad_norm": 1.4316396713256836, + "learning_rate": 5.157e-06, + "loss": 0.1429, + "step": 1722 + }, + { + "epoch": 1.4552364864864864, + "grad_norm": 1.800716519355774, + "learning_rate": 5.16e-06, + "loss": 0.2198, + "step": 1723 + }, + { + "epoch": 1.4560810810810811, + "grad_norm": 1.7588368654251099, + "learning_rate": 5.163e-06, + "loss": 0.2042, + "step": 1724 + }, + { + "epoch": 1.4569256756756757, + "grad_norm": 1.6161779165267944, + "learning_rate": 5.166e-06, + "loss": 0.2166, + "step": 1725 + }, + { + "epoch": 1.4577702702702702, + "grad_norm": 1.9350475072860718, + "learning_rate": 5.169e-06, + "loss": 0.1601, + "step": 1726 + }, + { + "epoch": 1.458614864864865, + "grad_norm": 1.880431890487671, + "learning_rate": 5.172e-06, + "loss": 0.1877, + "step": 1727 + }, + { + "epoch": 1.4594594594594594, + "grad_norm": 1.7019590139389038, + "learning_rate": 5.175e-06, + "loss": 0.1755, + "step": 1728 + }, + { + "epoch": 1.460304054054054, + "grad_norm": 1.4116965532302856, + "learning_rate": 5.178e-06, + "loss": 0.1536, + "step": 1729 + }, + { + "epoch": 1.4611486486486487, + "grad_norm": 1.624241828918457, + "learning_rate": 5.181e-06, + "loss": 0.2035, + "step": 1730 + }, + { + "epoch": 1.4619932432432432, + "grad_norm": 2.2962846755981445, + "learning_rate": 5.184000000000001e-06, + "loss": 0.2057, + "step": 1731 + }, + { + "epoch": 1.4628378378378377, + "grad_norm": 2.013277769088745, + "learning_rate": 5.1870000000000005e-06, + "loss": 0.177, + "step": 1732 + }, + { + "epoch": 1.4636824324324325, + "grad_norm": 1.8215001821517944, + "learning_rate": 5.1899999999999994e-06, + "loss": 0.2061, + "step": 1733 + }, + { + "epoch": 1.464527027027027, + "grad_norm": 2.0821073055267334, + "learning_rate": 5.193e-06, + "loss": 0.2028, + "step": 1734 + }, + { + "epoch": 1.4653716216216215, + "grad_norm": 9.864869117736816, + "learning_rate": 5.196e-06, + "loss": 1.0315, + "step": 1735 + }, + { + "epoch": 1.4662162162162162, + "grad_norm": 4.727660179138184, + "learning_rate": 5.1990000000000005e-06, + "loss": 0.5255, + "step": 1736 + }, + { + "epoch": 1.4670608108108107, + "grad_norm": 2.5688602924346924, + "learning_rate": 5.202e-06, + "loss": 0.3584, + "step": 1737 + }, + { + "epoch": 1.4679054054054055, + "grad_norm": 2.7495462894439697, + "learning_rate": 5.205e-06, + "loss": 0.2416, + "step": 1738 + }, + { + "epoch": 1.46875, + "grad_norm": 2.385784149169922, + "learning_rate": 5.208e-06, + "loss": 0.2062, + "step": 1739 + }, + { + "epoch": 1.4695945945945945, + "grad_norm": 1.7957216501235962, + "learning_rate": 5.211e-06, + "loss": 0.1535, + "step": 1740 + }, + { + "epoch": 1.4704391891891893, + "grad_norm": 1.1294934749603271, + "learning_rate": 5.214e-06, + "loss": 0.1584, + "step": 1741 + }, + { + "epoch": 1.4712837837837838, + "grad_norm": 0.980196475982666, + "learning_rate": 5.217e-06, + "loss": 0.1495, + "step": 1742 + }, + { + "epoch": 1.4721283783783785, + "grad_norm": 1.245204210281372, + "learning_rate": 5.22e-06, + "loss": 0.1738, + "step": 1743 + }, + { + "epoch": 1.472972972972973, + "grad_norm": 1.353797435760498, + "learning_rate": 5.223000000000001e-06, + "loss": 0.1794, + "step": 1744 + }, + { + "epoch": 1.4738175675675675, + "grad_norm": 1.308910608291626, + "learning_rate": 5.226e-06, + "loss": 0.1663, + "step": 1745 + }, + { + "epoch": 1.4746621621621623, + "grad_norm": 1.241871953010559, + "learning_rate": 5.229e-06, + "loss": 0.1572, + "step": 1746 + }, + { + "epoch": 1.4755067567567568, + "grad_norm": 1.2675310373306274, + "learning_rate": 5.232e-06, + "loss": 0.1247, + "step": 1747 + }, + { + "epoch": 1.4763513513513513, + "grad_norm": 1.3124409914016724, + "learning_rate": 5.235e-06, + "loss": 0.1603, + "step": 1748 + }, + { + "epoch": 1.477195945945946, + "grad_norm": 1.4828675985336304, + "learning_rate": 5.2380000000000005e-06, + "loss": 0.1544, + "step": 1749 + }, + { + "epoch": 1.4780405405405406, + "grad_norm": 1.2048252820968628, + "learning_rate": 5.241e-06, + "loss": 0.1402, + "step": 1750 + }, + { + "epoch": 1.478885135135135, + "grad_norm": 1.3324249982833862, + "learning_rate": 5.244e-06, + "loss": 0.1557, + "step": 1751 + }, + { + "epoch": 1.4797297297297298, + "grad_norm": 1.4925508499145508, + "learning_rate": 5.247e-06, + "loss": 0.1456, + "step": 1752 + }, + { + "epoch": 1.4805743243243243, + "grad_norm": 1.0661700963974, + "learning_rate": 5.25e-06, + "loss": 0.164, + "step": 1753 + }, + { + "epoch": 1.4814189189189189, + "grad_norm": 1.2397825717926025, + "learning_rate": 5.253e-06, + "loss": 0.185, + "step": 1754 + }, + { + "epoch": 1.4822635135135136, + "grad_norm": 1.1823170185089111, + "learning_rate": 5.256e-06, + "loss": 0.1645, + "step": 1755 + }, + { + "epoch": 1.4831081081081081, + "grad_norm": 1.3235700130462646, + "learning_rate": 5.259000000000001e-06, + "loss": 0.1556, + "step": 1756 + }, + { + "epoch": 1.4839527027027026, + "grad_norm": 1.1934698820114136, + "learning_rate": 5.262e-06, + "loss": 0.1727, + "step": 1757 + }, + { + "epoch": 1.4847972972972974, + "grad_norm": 1.8393796682357788, + "learning_rate": 5.2649999999999996e-06, + "loss": 0.1555, + "step": 1758 + }, + { + "epoch": 1.4856418918918919, + "grad_norm": 1.38296377658844, + "learning_rate": 5.268e-06, + "loss": 0.2395, + "step": 1759 + }, + { + "epoch": 1.4864864864864864, + "grad_norm": 1.3150092363357544, + "learning_rate": 5.271e-06, + "loss": 0.1661, + "step": 1760 + }, + { + "epoch": 1.4873310810810811, + "grad_norm": 1.1117826700210571, + "learning_rate": 5.274000000000001e-06, + "loss": 0.1623, + "step": 1761 + }, + { + "epoch": 1.4881756756756757, + "grad_norm": 1.592532753944397, + "learning_rate": 5.2770000000000005e-06, + "loss": 0.1705, + "step": 1762 + }, + { + "epoch": 1.4890202702702702, + "grad_norm": 1.3786741495132446, + "learning_rate": 5.279999999999999e-06, + "loss": 0.158, + "step": 1763 + }, + { + "epoch": 1.489864864864865, + "grad_norm": 2.23626708984375, + "learning_rate": 5.283e-06, + "loss": 0.2345, + "step": 1764 + }, + { + "epoch": 1.4907094594594594, + "grad_norm": 1.7825665473937988, + "learning_rate": 5.286e-06, + "loss": 0.2041, + "step": 1765 + }, + { + "epoch": 1.491554054054054, + "grad_norm": 1.242876648902893, + "learning_rate": 5.2890000000000005e-06, + "loss": 0.1733, + "step": 1766 + }, + { + "epoch": 1.4923986486486487, + "grad_norm": 1.7451236248016357, + "learning_rate": 5.292e-06, + "loss": 0.1871, + "step": 1767 + }, + { + "epoch": 1.4932432432432432, + "grad_norm": 2.3443620204925537, + "learning_rate": 5.295e-06, + "loss": 0.2171, + "step": 1768 + }, + { + "epoch": 1.4940878378378377, + "grad_norm": 1.3760885000228882, + "learning_rate": 5.298e-06, + "loss": 0.151, + "step": 1769 + }, + { + "epoch": 1.4949324324324325, + "grad_norm": 1.5156160593032837, + "learning_rate": 5.301e-06, + "loss": 0.1369, + "step": 1770 + }, + { + "epoch": 1.495777027027027, + "grad_norm": 1.4445114135742188, + "learning_rate": 5.304e-06, + "loss": 0.159, + "step": 1771 + }, + { + "epoch": 1.4966216216216215, + "grad_norm": 1.5173252820968628, + "learning_rate": 5.307e-06, + "loss": 0.1394, + "step": 1772 + }, + { + "epoch": 1.4974662162162162, + "grad_norm": 1.4055148363113403, + "learning_rate": 5.31e-06, + "loss": 0.2199, + "step": 1773 + }, + { + "epoch": 1.4983108108108107, + "grad_norm": 1.2944955825805664, + "learning_rate": 5.313000000000001e-06, + "loss": 0.1643, + "step": 1774 + }, + { + "epoch": 1.4991554054054055, + "grad_norm": 1.3175938129425049, + "learning_rate": 5.3160000000000004e-06, + "loss": 0.1642, + "step": 1775 + }, + { + "epoch": 1.5, + "grad_norm": 1.7875944375991821, + "learning_rate": 5.319e-06, + "loss": 0.1974, + "step": 1776 + }, + { + "epoch": 1.5008445945945947, + "grad_norm": 1.515864610671997, + "learning_rate": 5.322e-06, + "loss": 0.1773, + "step": 1777 + }, + { + "epoch": 1.501689189189189, + "grad_norm": 1.473175048828125, + "learning_rate": 5.325e-06, + "loss": 0.187, + "step": 1778 + }, + { + "epoch": 1.5025337837837838, + "grad_norm": 1.7597147226333618, + "learning_rate": 5.3280000000000005e-06, + "loss": 0.2042, + "step": 1779 + }, + { + "epoch": 1.5033783783783785, + "grad_norm": 1.8060210943222046, + "learning_rate": 5.331e-06, + "loss": 0.1807, + "step": 1780 + }, + { + "epoch": 1.504222972972973, + "grad_norm": 1.4538313150405884, + "learning_rate": 5.334000000000001e-06, + "loss": 0.1707, + "step": 1781 + }, + { + "epoch": 1.5050675675675675, + "grad_norm": 3.4785683155059814, + "learning_rate": 5.337e-06, + "loss": 0.2373, + "step": 1782 + }, + { + "epoch": 1.5059121621621623, + "grad_norm": 2.435424327850342, + "learning_rate": 5.34e-06, + "loss": 0.285, + "step": 1783 + }, + { + "epoch": 1.5067567567567568, + "grad_norm": 4.002475738525391, + "learning_rate": 5.343e-06, + "loss": 0.4163, + "step": 1784 + }, + { + "epoch": 1.5076013513513513, + "grad_norm": 6.342710971832275, + "learning_rate": 5.346e-06, + "loss": 1.0274, + "step": 1785 + }, + { + "epoch": 1.508445945945946, + "grad_norm": 3.843123435974121, + "learning_rate": 5.349e-06, + "loss": 0.5504, + "step": 1786 + }, + { + "epoch": 1.5092905405405406, + "grad_norm": 2.734795093536377, + "learning_rate": 5.352000000000001e-06, + "loss": 0.3153, + "step": 1787 + }, + { + "epoch": 1.510135135135135, + "grad_norm": 1.4879467487335205, + "learning_rate": 5.3549999999999996e-06, + "loss": 0.1863, + "step": 1788 + }, + { + "epoch": 1.5109797297297298, + "grad_norm": 5.234072208404541, + "learning_rate": 5.358e-06, + "loss": 0.2398, + "step": 1789 + }, + { + "epoch": 1.5118243243243243, + "grad_norm": 1.2426716089248657, + "learning_rate": 5.361e-06, + "loss": 0.1489, + "step": 1790 + }, + { + "epoch": 1.5126689189189189, + "grad_norm": 1.3855838775634766, + "learning_rate": 5.364e-06, + "loss": 0.2105, + "step": 1791 + }, + { + "epoch": 1.5135135135135136, + "grad_norm": 1.0607444047927856, + "learning_rate": 5.3670000000000005e-06, + "loss": 0.1287, + "step": 1792 + }, + { + "epoch": 1.5143581081081081, + "grad_norm": 1.52522873878479, + "learning_rate": 5.37e-06, + "loss": 0.1615, + "step": 1793 + }, + { + "epoch": 1.5152027027027026, + "grad_norm": 1.0820527076721191, + "learning_rate": 5.373e-06, + "loss": 0.1254, + "step": 1794 + }, + { + "epoch": 1.5160472972972974, + "grad_norm": 1.267735481262207, + "learning_rate": 5.376e-06, + "loss": 0.1698, + "step": 1795 + }, + { + "epoch": 1.5168918918918919, + "grad_norm": 1.4090099334716797, + "learning_rate": 5.379e-06, + "loss": 0.1762, + "step": 1796 + }, + { + "epoch": 1.5177364864864864, + "grad_norm": 1.5829359292984009, + "learning_rate": 5.382e-06, + "loss": 0.1645, + "step": 1797 + }, + { + "epoch": 1.5185810810810811, + "grad_norm": 1.2969056367874146, + "learning_rate": 5.385e-06, + "loss": 0.1628, + "step": 1798 + }, + { + "epoch": 1.5194256756756757, + "grad_norm": 1.5737413167953491, + "learning_rate": 5.388000000000001e-06, + "loss": 0.1463, + "step": 1799 + }, + { + "epoch": 1.5202702702702702, + "grad_norm": 1.3876036405563354, + "learning_rate": 5.391e-06, + "loss": 0.1532, + "step": 1800 + }, + { + "epoch": 1.521114864864865, + "grad_norm": 1.1821266412734985, + "learning_rate": 5.3939999999999995e-06, + "loss": 0.1473, + "step": 1801 + }, + { + "epoch": 1.5219594594594594, + "grad_norm": 1.33928644657135, + "learning_rate": 5.397e-06, + "loss": 0.1423, + "step": 1802 + }, + { + "epoch": 1.522804054054054, + "grad_norm": 1.3305950164794922, + "learning_rate": 5.4e-06, + "loss": 0.142, + "step": 1803 + }, + { + "epoch": 1.5236486486486487, + "grad_norm": 1.462319254875183, + "learning_rate": 5.403000000000001e-06, + "loss": 0.1877, + "step": 1804 + }, + { + "epoch": 1.5244932432432432, + "grad_norm": 1.1893945932388306, + "learning_rate": 5.406e-06, + "loss": 0.1814, + "step": 1805 + }, + { + "epoch": 1.5253378378378377, + "grad_norm": 1.229702353477478, + "learning_rate": 5.408999999999999e-06, + "loss": 0.1248, + "step": 1806 + }, + { + "epoch": 1.5261824324324325, + "grad_norm": 1.0888121128082275, + "learning_rate": 5.412e-06, + "loss": 0.1077, + "step": 1807 + }, + { + "epoch": 1.527027027027027, + "grad_norm": 1.171810507774353, + "learning_rate": 5.415e-06, + "loss": 0.1636, + "step": 1808 + }, + { + "epoch": 1.5278716216216215, + "grad_norm": 1.2258974313735962, + "learning_rate": 5.4180000000000005e-06, + "loss": 0.1339, + "step": 1809 + }, + { + "epoch": 1.5287162162162162, + "grad_norm": 1.3914068937301636, + "learning_rate": 5.421e-06, + "loss": 0.1642, + "step": 1810 + }, + { + "epoch": 1.529560810810811, + "grad_norm": 1.0833944082260132, + "learning_rate": 5.424e-06, + "loss": 0.1239, + "step": 1811 + }, + { + "epoch": 1.5304054054054053, + "grad_norm": 1.3970463275909424, + "learning_rate": 5.427e-06, + "loss": 0.1328, + "step": 1812 + }, + { + "epoch": 1.53125, + "grad_norm": 1.2050505876541138, + "learning_rate": 5.43e-06, + "loss": 0.1356, + "step": 1813 + }, + { + "epoch": 1.5320945945945947, + "grad_norm": 1.3698768615722656, + "learning_rate": 5.433e-06, + "loss": 0.1686, + "step": 1814 + }, + { + "epoch": 1.532939189189189, + "grad_norm": 1.4894375801086426, + "learning_rate": 5.436e-06, + "loss": 0.1495, + "step": 1815 + }, + { + "epoch": 1.5337837837837838, + "grad_norm": 1.3993009328842163, + "learning_rate": 5.439e-06, + "loss": 0.1844, + "step": 1816 + }, + { + "epoch": 1.5346283783783785, + "grad_norm": 1.27100670337677, + "learning_rate": 5.442000000000001e-06, + "loss": 0.134, + "step": 1817 + }, + { + "epoch": 1.535472972972973, + "grad_norm": 1.3380215167999268, + "learning_rate": 5.445e-06, + "loss": 0.1649, + "step": 1818 + }, + { + "epoch": 1.5363175675675675, + "grad_norm": 2.0315418243408203, + "learning_rate": 5.448e-06, + "loss": 0.1455, + "step": 1819 + }, + { + "epoch": 1.5371621621621623, + "grad_norm": 2.0095114707946777, + "learning_rate": 5.451e-06, + "loss": 0.205, + "step": 1820 + }, + { + "epoch": 1.5380067567567568, + "grad_norm": 1.5918735265731812, + "learning_rate": 5.454e-06, + "loss": 0.1944, + "step": 1821 + }, + { + "epoch": 1.5388513513513513, + "grad_norm": 1.6624912023544312, + "learning_rate": 5.4570000000000004e-06, + "loss": 0.1659, + "step": 1822 + }, + { + "epoch": 1.539695945945946, + "grad_norm": 1.6214181184768677, + "learning_rate": 5.46e-06, + "loss": 0.2035, + "step": 1823 + }, + { + "epoch": 1.5405405405405406, + "grad_norm": 1.9783599376678467, + "learning_rate": 5.463000000000001e-06, + "loss": 0.1728, + "step": 1824 + }, + { + "epoch": 1.541385135135135, + "grad_norm": 2.141246795654297, + "learning_rate": 5.466e-06, + "loss": 0.2067, + "step": 1825 + }, + { + "epoch": 1.5422297297297298, + "grad_norm": 1.4031541347503662, + "learning_rate": 5.469e-06, + "loss": 0.1973, + "step": 1826 + }, + { + "epoch": 1.5430743243243243, + "grad_norm": 1.5431256294250488, + "learning_rate": 5.472e-06, + "loss": 0.1611, + "step": 1827 + }, + { + "epoch": 1.5439189189189189, + "grad_norm": 1.7070610523223877, + "learning_rate": 5.475e-06, + "loss": 0.1757, + "step": 1828 + }, + { + "epoch": 1.5447635135135136, + "grad_norm": 1.5896567106246948, + "learning_rate": 5.478000000000001e-06, + "loss": 0.1382, + "step": 1829 + }, + { + "epoch": 1.5456081081081081, + "grad_norm": 1.6660511493682861, + "learning_rate": 5.4810000000000005e-06, + "loss": 0.1781, + "step": 1830 + }, + { + "epoch": 1.5464527027027026, + "grad_norm": 2.0485332012176514, + "learning_rate": 5.4839999999999995e-06, + "loss": 0.2239, + "step": 1831 + }, + { + "epoch": 1.5472972972972974, + "grad_norm": 1.8113595247268677, + "learning_rate": 5.487e-06, + "loss": 0.2251, + "step": 1832 + }, + { + "epoch": 1.5481418918918919, + "grad_norm": 2.0317542552948, + "learning_rate": 5.49e-06, + "loss": 0.2247, + "step": 1833 + }, + { + "epoch": 1.5489864864864864, + "grad_norm": 3.3907012939453125, + "learning_rate": 5.493000000000001e-06, + "loss": 0.3722, + "step": 1834 + }, + { + "epoch": 1.5498310810810811, + "grad_norm": 4.654313087463379, + "learning_rate": 5.496e-06, + "loss": 0.7757, + "step": 1835 + }, + { + "epoch": 1.5506756756756757, + "grad_norm": 3.4641096591949463, + "learning_rate": 5.499e-06, + "loss": 0.6915, + "step": 1836 + }, + { + "epoch": 1.5515202702702702, + "grad_norm": 3.089141607284546, + "learning_rate": 5.502e-06, + "loss": 0.3003, + "step": 1837 + }, + { + "epoch": 1.552364864864865, + "grad_norm": 3.510862112045288, + "learning_rate": 5.505e-06, + "loss": 0.2235, + "step": 1838 + }, + { + "epoch": 1.5532094594594594, + "grad_norm": 1.8659435510635376, + "learning_rate": 5.5080000000000005e-06, + "loss": 0.1587, + "step": 1839 + }, + { + "epoch": 1.554054054054054, + "grad_norm": 1.4730055332183838, + "learning_rate": 5.511e-06, + "loss": 0.1583, + "step": 1840 + }, + { + "epoch": 1.5548986486486487, + "grad_norm": 1.3563580513000488, + "learning_rate": 5.514e-06, + "loss": 0.1671, + "step": 1841 + }, + { + "epoch": 1.5557432432432432, + "grad_norm": 2.205695867538452, + "learning_rate": 5.517000000000001e-06, + "loss": 0.1694, + "step": 1842 + }, + { + "epoch": 1.5565878378378377, + "grad_norm": 1.7314438819885254, + "learning_rate": 5.52e-06, + "loss": 0.1881, + "step": 1843 + }, + { + "epoch": 1.5574324324324325, + "grad_norm": 1.0003999471664429, + "learning_rate": 5.523e-06, + "loss": 0.1257, + "step": 1844 + }, + { + "epoch": 1.558277027027027, + "grad_norm": 1.0979009866714478, + "learning_rate": 5.526e-06, + "loss": 0.146, + "step": 1845 + }, + { + "epoch": 1.5591216216216215, + "grad_norm": 1.223110318183899, + "learning_rate": 5.529e-06, + "loss": 0.1298, + "step": 1846 + }, + { + "epoch": 1.5599662162162162, + "grad_norm": 1.0498888492584229, + "learning_rate": 5.5320000000000006e-06, + "loss": 0.1074, + "step": 1847 + }, + { + "epoch": 1.560810810810811, + "grad_norm": 1.1568443775177002, + "learning_rate": 5.535e-06, + "loss": 0.1416, + "step": 1848 + }, + { + "epoch": 1.5616554054054053, + "grad_norm": 1.2317917346954346, + "learning_rate": 5.537999999999999e-06, + "loss": 0.1538, + "step": 1849 + }, + { + "epoch": 1.5625, + "grad_norm": 0.9998905658721924, + "learning_rate": 5.541e-06, + "loss": 0.1429, + "step": 1850 + }, + { + "epoch": 1.5633445945945947, + "grad_norm": 1.1288766860961914, + "learning_rate": 5.544e-06, + "loss": 0.1197, + "step": 1851 + }, + { + "epoch": 1.564189189189189, + "grad_norm": 1.2523422241210938, + "learning_rate": 5.547e-06, + "loss": 0.1638, + "step": 1852 + }, + { + "epoch": 1.5650337837837838, + "grad_norm": 1.1827731132507324, + "learning_rate": 5.55e-06, + "loss": 0.1418, + "step": 1853 + }, + { + "epoch": 1.5658783783783785, + "grad_norm": 1.3067511320114136, + "learning_rate": 5.553e-06, + "loss": 0.247, + "step": 1854 + }, + { + "epoch": 1.566722972972973, + "grad_norm": 1.1698919534683228, + "learning_rate": 5.556e-06, + "loss": 0.1503, + "step": 1855 + }, + { + "epoch": 1.5675675675675675, + "grad_norm": 5.2440667152404785, + "learning_rate": 5.559e-06, + "loss": 0.2185, + "step": 1856 + }, + { + "epoch": 1.5684121621621623, + "grad_norm": 1.0690189599990845, + "learning_rate": 5.562e-06, + "loss": 0.138, + "step": 1857 + }, + { + "epoch": 1.5692567567567568, + "grad_norm": 1.0397969484329224, + "learning_rate": 5.565e-06, + "loss": 0.1454, + "step": 1858 + }, + { + "epoch": 1.5701013513513513, + "grad_norm": 1.1338374614715576, + "learning_rate": 5.568e-06, + "loss": 0.1523, + "step": 1859 + }, + { + "epoch": 1.570945945945946, + "grad_norm": 1.1049154996871948, + "learning_rate": 5.5710000000000005e-06, + "loss": 0.1448, + "step": 1860 + }, + { + "epoch": 1.5717905405405406, + "grad_norm": 1.2855238914489746, + "learning_rate": 5.574e-06, + "loss": 0.1885, + "step": 1861 + }, + { + "epoch": 1.572635135135135, + "grad_norm": 1.0736238956451416, + "learning_rate": 5.577e-06, + "loss": 0.1292, + "step": 1862 + }, + { + "epoch": 1.5734797297297298, + "grad_norm": 1.2756259441375732, + "learning_rate": 5.58e-06, + "loss": 0.1929, + "step": 1863 + }, + { + "epoch": 1.5743243243243243, + "grad_norm": 1.4033399820327759, + "learning_rate": 5.583e-06, + "loss": 0.1591, + "step": 1864 + }, + { + "epoch": 1.5751689189189189, + "grad_norm": 1.601409912109375, + "learning_rate": 5.586e-06, + "loss": 0.214, + "step": 1865 + }, + { + "epoch": 1.5760135135135136, + "grad_norm": 1.1037006378173828, + "learning_rate": 5.589e-06, + "loss": 0.1307, + "step": 1866 + }, + { + "epoch": 1.5768581081081081, + "grad_norm": 1.787983775138855, + "learning_rate": 5.592000000000001e-06, + "loss": 0.1748, + "step": 1867 + }, + { + "epoch": 1.5777027027027026, + "grad_norm": 3.084714412689209, + "learning_rate": 5.595e-06, + "loss": 0.1505, + "step": 1868 + }, + { + "epoch": 1.5785472972972974, + "grad_norm": 1.4911261796951294, + "learning_rate": 5.598e-06, + "loss": 0.1877, + "step": 1869 + }, + { + "epoch": 1.5793918918918919, + "grad_norm": 1.7208868265151978, + "learning_rate": 5.601e-06, + "loss": 0.1983, + "step": 1870 + }, + { + "epoch": 1.5802364864864864, + "grad_norm": 1.1064574718475342, + "learning_rate": 5.604e-06, + "loss": 0.1201, + "step": 1871 + }, + { + "epoch": 1.5810810810810811, + "grad_norm": 1.520010232925415, + "learning_rate": 5.607000000000001e-06, + "loss": 0.1708, + "step": 1872 + }, + { + "epoch": 1.5819256756756757, + "grad_norm": 1.6282942295074463, + "learning_rate": 5.6100000000000005e-06, + "loss": 0.1823, + "step": 1873 + }, + { + "epoch": 1.5827702702702702, + "grad_norm": 1.3491960763931274, + "learning_rate": 5.6129999999999995e-06, + "loss": 0.1432, + "step": 1874 + }, + { + "epoch": 1.583614864864865, + "grad_norm": 1.5252851247787476, + "learning_rate": 5.616e-06, + "loss": 0.2007, + "step": 1875 + }, + { + "epoch": 1.5844594594594594, + "grad_norm": 1.748146653175354, + "learning_rate": 5.619e-06, + "loss": 0.1935, + "step": 1876 + }, + { + "epoch": 1.585304054054054, + "grad_norm": 1.5319974422454834, + "learning_rate": 5.6220000000000006e-06, + "loss": 0.1689, + "step": 1877 + }, + { + "epoch": 1.5861486486486487, + "grad_norm": 1.6152422428131104, + "learning_rate": 5.625e-06, + "loss": 0.1663, + "step": 1878 + }, + { + "epoch": 1.5869932432432432, + "grad_norm": 2.096973180770874, + "learning_rate": 5.628e-06, + "loss": 0.1647, + "step": 1879 + }, + { + "epoch": 1.5878378378378377, + "grad_norm": 1.7893593311309814, + "learning_rate": 5.631e-06, + "loss": 0.165, + "step": 1880 + }, + { + "epoch": 1.5886824324324325, + "grad_norm": 1.8532418012619019, + "learning_rate": 5.634e-06, + "loss": 0.1881, + "step": 1881 + }, + { + "epoch": 1.589527027027027, + "grad_norm": 1.9113237857818604, + "learning_rate": 5.637e-06, + "loss": 0.1876, + "step": 1882 + }, + { + "epoch": 1.5903716216216215, + "grad_norm": 2.0583059787750244, + "learning_rate": 5.64e-06, + "loss": 0.2157, + "step": 1883 + }, + { + "epoch": 1.5912162162162162, + "grad_norm": 2.9325084686279297, + "learning_rate": 5.643e-06, + "loss": 0.29, + "step": 1884 + }, + { + "epoch": 1.592060810810811, + "grad_norm": 14.57162857055664, + "learning_rate": 5.646000000000001e-06, + "loss": 0.8094, + "step": 1885 + }, + { + "epoch": 1.5929054054054053, + "grad_norm": 4.747203826904297, + "learning_rate": 5.649e-06, + "loss": 0.7243, + "step": 1886 + }, + { + "epoch": 1.59375, + "grad_norm": 1.576136827468872, + "learning_rate": 5.652e-06, + "loss": 0.2858, + "step": 1887 + }, + { + "epoch": 1.5945945945945947, + "grad_norm": 1.7460488080978394, + "learning_rate": 5.655e-06, + "loss": 0.2178, + "step": 1888 + }, + { + "epoch": 1.595439189189189, + "grad_norm": 2.358098268508911, + "learning_rate": 5.658e-06, + "loss": 0.2054, + "step": 1889 + }, + { + "epoch": 1.5962837837837838, + "grad_norm": 1.319964051246643, + "learning_rate": 5.6610000000000005e-06, + "loss": 0.1936, + "step": 1890 + }, + { + "epoch": 1.5971283783783785, + "grad_norm": 1.6012483835220337, + "learning_rate": 5.664e-06, + "loss": 0.1669, + "step": 1891 + }, + { + "epoch": 1.597972972972973, + "grad_norm": 1.0271233320236206, + "learning_rate": 5.667e-06, + "loss": 0.1382, + "step": 1892 + }, + { + "epoch": 1.5988175675675675, + "grad_norm": 1.177203893661499, + "learning_rate": 5.67e-06, + "loss": 0.1446, + "step": 1893 + }, + { + "epoch": 1.5996621621621623, + "grad_norm": 1.1970547437667847, + "learning_rate": 5.673e-06, + "loss": 0.1442, + "step": 1894 + }, + { + "epoch": 1.6005067567567568, + "grad_norm": 1.2234106063842773, + "learning_rate": 5.676e-06, + "loss": 0.1448, + "step": 1895 + }, + { + "epoch": 1.6013513513513513, + "grad_norm": 1.2582879066467285, + "learning_rate": 5.679e-06, + "loss": 0.1724, + "step": 1896 + }, + { + "epoch": 1.602195945945946, + "grad_norm": 1.2162714004516602, + "learning_rate": 5.682000000000001e-06, + "loss": 0.138, + "step": 1897 + }, + { + "epoch": 1.6030405405405406, + "grad_norm": 1.9920510053634644, + "learning_rate": 5.685e-06, + "loss": 0.1745, + "step": 1898 + }, + { + "epoch": 1.603885135135135, + "grad_norm": 1.1230909824371338, + "learning_rate": 5.688e-06, + "loss": 0.1265, + "step": 1899 + }, + { + "epoch": 1.6047297297297298, + "grad_norm": 1.6556646823883057, + "learning_rate": 5.691e-06, + "loss": 0.1622, + "step": 1900 + }, + { + "epoch": 1.6055743243243243, + "grad_norm": 1.2205569744110107, + "learning_rate": 5.694e-06, + "loss": 0.1603, + "step": 1901 + }, + { + "epoch": 1.6064189189189189, + "grad_norm": 1.329723834991455, + "learning_rate": 5.697000000000001e-06, + "loss": 0.1416, + "step": 1902 + }, + { + "epoch": 1.6072635135135136, + "grad_norm": 1.2290105819702148, + "learning_rate": 5.7000000000000005e-06, + "loss": 0.1769, + "step": 1903 + }, + { + "epoch": 1.6081081081081081, + "grad_norm": 1.3173389434814453, + "learning_rate": 5.703e-06, + "loss": 0.1601, + "step": 1904 + }, + { + "epoch": 1.6089527027027026, + "grad_norm": 1.3477675914764404, + "learning_rate": 5.706e-06, + "loss": 0.1655, + "step": 1905 + }, + { + "epoch": 1.6097972972972974, + "grad_norm": 1.3633004426956177, + "learning_rate": 5.709e-06, + "loss": 0.1359, + "step": 1906 + }, + { + "epoch": 1.6106418918918919, + "grad_norm": 1.4535424709320068, + "learning_rate": 5.7120000000000005e-06, + "loss": 0.1805, + "step": 1907 + }, + { + "epoch": 1.6114864864864864, + "grad_norm": 1.1093372106552124, + "learning_rate": 5.715e-06, + "loss": 0.1369, + "step": 1908 + }, + { + "epoch": 1.6123310810810811, + "grad_norm": 1.8192275762557983, + "learning_rate": 5.718e-06, + "loss": 0.173, + "step": 1909 + }, + { + "epoch": 1.6131756756756757, + "grad_norm": 1.0276073217391968, + "learning_rate": 5.721000000000001e-06, + "loss": 0.1242, + "step": 1910 + }, + { + "epoch": 1.6140202702702702, + "grad_norm": 1.3223265409469604, + "learning_rate": 5.724e-06, + "loss": 0.1651, + "step": 1911 + }, + { + "epoch": 1.614864864864865, + "grad_norm": 1.4675861597061157, + "learning_rate": 5.7269999999999995e-06, + "loss": 0.1742, + "step": 1912 + }, + { + "epoch": 1.6157094594594594, + "grad_norm": 1.2786974906921387, + "learning_rate": 5.73e-06, + "loss": 0.1265, + "step": 1913 + }, + { + "epoch": 1.616554054054054, + "grad_norm": 1.39027738571167, + "learning_rate": 5.733e-06, + "loss": 0.1685, + "step": 1914 + }, + { + "epoch": 1.6173986486486487, + "grad_norm": 1.614235281944275, + "learning_rate": 5.736000000000001e-06, + "loss": 0.1818, + "step": 1915 + }, + { + "epoch": 1.6182432432432432, + "grad_norm": 1.2132099866867065, + "learning_rate": 5.7390000000000004e-06, + "loss": 0.1324, + "step": 1916 + }, + { + "epoch": 1.6190878378378377, + "grad_norm": 1.4837857484817505, + "learning_rate": 5.741999999999999e-06, + "loss": 0.1789, + "step": 1917 + }, + { + "epoch": 1.6199324324324325, + "grad_norm": 1.4593403339385986, + "learning_rate": 5.745e-06, + "loss": 0.1579, + "step": 1918 + }, + { + "epoch": 1.620777027027027, + "grad_norm": 1.484617829322815, + "learning_rate": 5.748e-06, + "loss": 0.1799, + "step": 1919 + }, + { + "epoch": 1.6216216216216215, + "grad_norm": 1.2283076047897339, + "learning_rate": 5.7510000000000005e-06, + "loss": 0.1207, + "step": 1920 + }, + { + "epoch": 1.6224662162162162, + "grad_norm": 2.408100128173828, + "learning_rate": 5.754e-06, + "loss": 0.201, + "step": 1921 + }, + { + "epoch": 1.623310810810811, + "grad_norm": 1.7460942268371582, + "learning_rate": 5.757e-06, + "loss": 0.1515, + "step": 1922 + }, + { + "epoch": 1.6241554054054053, + "grad_norm": 1.5787410736083984, + "learning_rate": 5.76e-06, + "loss": 0.171, + "step": 1923 + }, + { + "epoch": 1.625, + "grad_norm": 1.5124914646148682, + "learning_rate": 5.763e-06, + "loss": 0.1304, + "step": 1924 + }, + { + "epoch": 1.6258445945945947, + "grad_norm": 1.7507193088531494, + "learning_rate": 5.766e-06, + "loss": 0.172, + "step": 1925 + }, + { + "epoch": 1.626689189189189, + "grad_norm": 1.560613989830017, + "learning_rate": 5.769e-06, + "loss": 0.1395, + "step": 1926 + }, + { + "epoch": 1.6275337837837838, + "grad_norm": 1.2297189235687256, + "learning_rate": 5.772e-06, + "loss": 0.1286, + "step": 1927 + }, + { + "epoch": 1.6283783783783785, + "grad_norm": 1.2642064094543457, + "learning_rate": 5.775000000000001e-06, + "loss": 0.1613, + "step": 1928 + }, + { + "epoch": 1.629222972972973, + "grad_norm": 2.9338157176971436, + "learning_rate": 5.7779999999999996e-06, + "loss": 0.1765, + "step": 1929 + }, + { + "epoch": 1.6300675675675675, + "grad_norm": 2.1085572242736816, + "learning_rate": 5.781e-06, + "loss": 0.2264, + "step": 1930 + }, + { + "epoch": 1.6309121621621623, + "grad_norm": 2.540832757949829, + "learning_rate": 5.784e-06, + "loss": 0.1845, + "step": 1931 + }, + { + "epoch": 1.6317567567567568, + "grad_norm": 2.569783926010132, + "learning_rate": 5.787e-06, + "loss": 0.2593, + "step": 1932 + }, + { + "epoch": 1.6326013513513513, + "grad_norm": 2.802905321121216, + "learning_rate": 5.7900000000000005e-06, + "loss": 0.2676, + "step": 1933 + }, + { + "epoch": 1.633445945945946, + "grad_norm": 3.5418014526367188, + "learning_rate": 5.793e-06, + "loss": 0.331, + "step": 1934 + }, + { + "epoch": 1.6342905405405406, + "grad_norm": 4.528469085693359, + "learning_rate": 5.796e-06, + "loss": 0.8412, + "step": 1935 + }, + { + "epoch": 1.635135135135135, + "grad_norm": 2.4498848915100098, + "learning_rate": 5.799e-06, + "loss": 0.4685, + "step": 1936 + }, + { + "epoch": 1.6359797297297298, + "grad_norm": 2.1326677799224854, + "learning_rate": 5.802e-06, + "loss": 0.2969, + "step": 1937 + }, + { + "epoch": 1.6368243243243243, + "grad_norm": 1.5092846155166626, + "learning_rate": 5.805e-06, + "loss": 0.212, + "step": 1938 + }, + { + "epoch": 1.6376689189189189, + "grad_norm": 1.3646063804626465, + "learning_rate": 5.808e-06, + "loss": 0.1658, + "step": 1939 + }, + { + "epoch": 1.6385135135135136, + "grad_norm": 1.4453192949295044, + "learning_rate": 5.811000000000001e-06, + "loss": 0.1683, + "step": 1940 + }, + { + "epoch": 1.6393581081081081, + "grad_norm": 1.1985071897506714, + "learning_rate": 5.814e-06, + "loss": 0.1654, + "step": 1941 + }, + { + "epoch": 1.6402027027027026, + "grad_norm": 1.53786301612854, + "learning_rate": 5.8169999999999995e-06, + "loss": 0.1651, + "step": 1942 + }, + { + "epoch": 1.6410472972972974, + "grad_norm": 0.8673379421234131, + "learning_rate": 5.82e-06, + "loss": 0.1367, + "step": 1943 + }, + { + "epoch": 1.6418918918918919, + "grad_norm": 1.8556796312332153, + "learning_rate": 5.823e-06, + "loss": 0.1396, + "step": 1944 + }, + { + "epoch": 1.6427364864864864, + "grad_norm": 1.165985107421875, + "learning_rate": 5.826000000000001e-06, + "loss": 0.1281, + "step": 1945 + }, + { + "epoch": 1.6435810810810811, + "grad_norm": 2.3738579750061035, + "learning_rate": 5.8290000000000004e-06, + "loss": 0.1345, + "step": 1946 + }, + { + "epoch": 1.6444256756756757, + "grad_norm": 1.1745798587799072, + "learning_rate": 5.832e-06, + "loss": 0.0907, + "step": 1947 + }, + { + "epoch": 1.6452702702702702, + "grad_norm": 1.276417851448059, + "learning_rate": 5.835e-06, + "loss": 0.1529, + "step": 1948 + }, + { + "epoch": 1.646114864864865, + "grad_norm": 1.1092883348464966, + "learning_rate": 5.838e-06, + "loss": 0.1151, + "step": 1949 + }, + { + "epoch": 1.6469594594594594, + "grad_norm": 1.2201752662658691, + "learning_rate": 5.8410000000000005e-06, + "loss": 0.1255, + "step": 1950 + }, + { + "epoch": 1.647804054054054, + "grad_norm": 1.6542696952819824, + "learning_rate": 5.844e-06, + "loss": 0.125, + "step": 1951 + }, + { + "epoch": 1.6486486486486487, + "grad_norm": 1.1680588722229004, + "learning_rate": 5.847e-06, + "loss": 0.1603, + "step": 1952 + }, + { + "epoch": 1.6494932432432432, + "grad_norm": 1.3396618366241455, + "learning_rate": 5.850000000000001e-06, + "loss": 0.1473, + "step": 1953 + }, + { + "epoch": 1.6503378378378377, + "grad_norm": 1.2439168691635132, + "learning_rate": 5.853e-06, + "loss": 0.1271, + "step": 1954 + }, + { + "epoch": 1.6511824324324325, + "grad_norm": 0.7169588208198547, + "learning_rate": 5.856e-06, + "loss": 0.0918, + "step": 1955 + }, + { + "epoch": 1.652027027027027, + "grad_norm": 0.8991798758506775, + "learning_rate": 5.859e-06, + "loss": 0.1087, + "step": 1956 + }, + { + "epoch": 1.6528716216216215, + "grad_norm": 1.8985179662704468, + "learning_rate": 5.862e-06, + "loss": 0.1442, + "step": 1957 + }, + { + "epoch": 1.6537162162162162, + "grad_norm": 1.1291759014129639, + "learning_rate": 5.865000000000001e-06, + "loss": 0.1265, + "step": 1958 + }, + { + "epoch": 1.654560810810811, + "grad_norm": 1.1920366287231445, + "learning_rate": 5.868e-06, + "loss": 0.1473, + "step": 1959 + }, + { + "epoch": 1.6554054054054053, + "grad_norm": 1.4136055707931519, + "learning_rate": 5.871e-06, + "loss": 0.1459, + "step": 1960 + }, + { + "epoch": 1.65625, + "grad_norm": 0.9832018613815308, + "learning_rate": 5.874e-06, + "loss": 0.1222, + "step": 1961 + }, + { + "epoch": 1.6570945945945947, + "grad_norm": 1.5730748176574707, + "learning_rate": 5.877e-06, + "loss": 0.1887, + "step": 1962 + }, + { + "epoch": 1.657939189189189, + "grad_norm": 1.1632428169250488, + "learning_rate": 5.8800000000000005e-06, + "loss": 0.1419, + "step": 1963 + }, + { + "epoch": 1.6587837837837838, + "grad_norm": 1.1682685613632202, + "learning_rate": 5.883e-06, + "loss": 0.1691, + "step": 1964 + }, + { + "epoch": 1.6596283783783785, + "grad_norm": 1.0931161642074585, + "learning_rate": 5.886000000000001e-06, + "loss": 0.1183, + "step": 1965 + }, + { + "epoch": 1.660472972972973, + "grad_norm": 1.4791364669799805, + "learning_rate": 5.889e-06, + "loss": 0.1526, + "step": 1966 + }, + { + "epoch": 1.6613175675675675, + "grad_norm": 1.596869945526123, + "learning_rate": 5.892e-06, + "loss": 0.162, + "step": 1967 + }, + { + "epoch": 1.6621621621621623, + "grad_norm": 1.6753650903701782, + "learning_rate": 5.895e-06, + "loss": 0.114, + "step": 1968 + }, + { + "epoch": 1.6630067567567568, + "grad_norm": 1.2241101264953613, + "learning_rate": 5.898e-06, + "loss": 0.1257, + "step": 1969 + }, + { + "epoch": 1.6638513513513513, + "grad_norm": 1.360020637512207, + "learning_rate": 5.901000000000001e-06, + "loss": 0.1749, + "step": 1970 + }, + { + "epoch": 1.664695945945946, + "grad_norm": 3.42240047454834, + "learning_rate": 5.9040000000000006e-06, + "loss": 0.2757, + "step": 1971 + }, + { + "epoch": 1.6655405405405406, + "grad_norm": 1.3324742317199707, + "learning_rate": 5.9069999999999995e-06, + "loss": 0.1239, + "step": 1972 + }, + { + "epoch": 1.666385135135135, + "grad_norm": 1.1489890813827515, + "learning_rate": 5.91e-06, + "loss": 0.128, + "step": 1973 + }, + { + "epoch": 1.6672297297297298, + "grad_norm": 1.442017674446106, + "learning_rate": 5.913e-06, + "loss": 0.1385, + "step": 1974 + }, + { + "epoch": 1.6680743243243243, + "grad_norm": 1.9183672666549683, + "learning_rate": 5.916e-06, + "loss": 0.1755, + "step": 1975 + }, + { + "epoch": 1.6689189189189189, + "grad_norm": 1.4974396228790283, + "learning_rate": 5.919e-06, + "loss": 0.1484, + "step": 1976 + }, + { + "epoch": 1.6697635135135136, + "grad_norm": 1.4198799133300781, + "learning_rate": 5.922e-06, + "loss": 0.1311, + "step": 1977 + }, + { + "epoch": 1.6706081081081081, + "grad_norm": 1.3429381847381592, + "learning_rate": 5.925e-06, + "loss": 0.1214, + "step": 1978 + }, + { + "epoch": 1.6714527027027026, + "grad_norm": 2.7442522048950195, + "learning_rate": 5.928e-06, + "loss": 0.2369, + "step": 1979 + }, + { + "epoch": 1.6722972972972974, + "grad_norm": 2.0746161937713623, + "learning_rate": 5.931e-06, + "loss": 0.2495, + "step": 1980 + }, + { + "epoch": 1.6731418918918919, + "grad_norm": 1.8421485424041748, + "learning_rate": 5.934e-06, + "loss": 0.2308, + "step": 1981 + }, + { + "epoch": 1.6739864864864864, + "grad_norm": 1.913034439086914, + "learning_rate": 5.937e-06, + "loss": 0.1491, + "step": 1982 + }, + { + "epoch": 1.6748310810810811, + "grad_norm": 2.5571956634521484, + "learning_rate": 5.940000000000001e-06, + "loss": 0.2262, + "step": 1983 + }, + { + "epoch": 1.6756756756756757, + "grad_norm": 2.672220468521118, + "learning_rate": 5.943e-06, + "loss": 0.243, + "step": 1984 + }, + { + "epoch": 1.6765202702702702, + "grad_norm": 3.8810880184173584, + "learning_rate": 5.9459999999999995e-06, + "loss": 0.6974, + "step": 1985 + }, + { + "epoch": 1.677364864864865, + "grad_norm": 2.4837851524353027, + "learning_rate": 5.949e-06, + "loss": 0.4325, + "step": 1986 + }, + { + "epoch": 1.6782094594594594, + "grad_norm": 3.891026496887207, + "learning_rate": 5.952e-06, + "loss": 0.2223, + "step": 1987 + }, + { + "epoch": 1.679054054054054, + "grad_norm": 1.2482508420944214, + "learning_rate": 5.955000000000001e-06, + "loss": 0.1477, + "step": 1988 + }, + { + "epoch": 1.6798986486486487, + "grad_norm": 1.3593889474868774, + "learning_rate": 5.958e-06, + "loss": 0.1942, + "step": 1989 + }, + { + "epoch": 1.6807432432432432, + "grad_norm": 1.938664436340332, + "learning_rate": 5.961e-06, + "loss": 0.1362, + "step": 1990 + }, + { + "epoch": 1.6815878378378377, + "grad_norm": 1.1833101511001587, + "learning_rate": 5.964e-06, + "loss": 0.1801, + "step": 1991 + }, + { + "epoch": 1.6824324324324325, + "grad_norm": 1.5023890733718872, + "learning_rate": 5.967e-06, + "loss": 0.1693, + "step": 1992 + }, + { + "epoch": 1.683277027027027, + "grad_norm": 1.0967732667922974, + "learning_rate": 5.9700000000000004e-06, + "loss": 0.1416, + "step": 1993 + }, + { + "epoch": 1.6841216216216215, + "grad_norm": 1.7777975797653198, + "learning_rate": 5.973e-06, + "loss": 0.1347, + "step": 1994 + }, + { + "epoch": 1.6849662162162162, + "grad_norm": 0.8855360150337219, + "learning_rate": 5.976e-06, + "loss": 0.1299, + "step": 1995 + }, + { + "epoch": 1.685810810810811, + "grad_norm": 1.0074701309204102, + "learning_rate": 5.979000000000001e-06, + "loss": 0.1159, + "step": 1996 + }, + { + "epoch": 1.6866554054054053, + "grad_norm": 0.9896659851074219, + "learning_rate": 5.982e-06, + "loss": 0.1159, + "step": 1997 + }, + { + "epoch": 1.6875, + "grad_norm": 1.2605328559875488, + "learning_rate": 5.985e-06, + "loss": 0.1286, + "step": 1998 + }, + { + "epoch": 1.6883445945945947, + "grad_norm": 1.1412729024887085, + "learning_rate": 5.988e-06, + "loss": 0.1427, + "step": 1999 + }, + { + "epoch": 1.689189189189189, + "grad_norm": 0.9548057913780212, + "learning_rate": 5.991e-06, + "loss": 0.1171, + "step": 2000 + }, + { + "epoch": 1.689189189189189, + "eval_cer": 0.15308007388964742, + "eval_loss": 0.6389010548591614, + "eval_runtime": 8.8034, + "eval_samples_per_second": 11.473, + "eval_steps_per_second": 0.454, + "eval_wer": 0.4184303350970018, + "step": 2000 + }, + { + "epoch": 1.6900337837837838, + "grad_norm": 1.1021534204483032, + "learning_rate": 5.9940000000000005e-06, + "loss": 0.1243, + "step": 2001 + }, + { + "epoch": 1.6908783783783785, + "grad_norm": 1.6326960325241089, + "learning_rate": 5.997e-06, + "loss": 0.1557, + "step": 2002 + }, + { + "epoch": 1.691722972972973, + "grad_norm": 1.4383702278137207, + "learning_rate": 6e-06, + "loss": 0.1212, + "step": 2003 + }, + { + "epoch": 1.6925675675675675, + "grad_norm": 1.2377560138702393, + "learning_rate": 6.003e-06, + "loss": 0.156, + "step": 2004 + }, + { + "epoch": 1.6934121621621623, + "grad_norm": 1.2715823650360107, + "learning_rate": 6.006e-06, + "loss": 0.143, + "step": 2005 + }, + { + "epoch": 1.6942567567567568, + "grad_norm": 1.2446091175079346, + "learning_rate": 6.009e-06, + "loss": 0.1143, + "step": 2006 + }, + { + "epoch": 1.6951013513513513, + "grad_norm": 1.6421846151351929, + "learning_rate": 6.012e-06, + "loss": 0.1836, + "step": 2007 + }, + { + "epoch": 1.695945945945946, + "grad_norm": 1.2848724126815796, + "learning_rate": 6.015000000000001e-06, + "loss": 0.1314, + "step": 2008 + }, + { + "epoch": 1.6967905405405406, + "grad_norm": 0.9601985812187195, + "learning_rate": 6.018e-06, + "loss": 0.0913, + "step": 2009 + }, + { + "epoch": 1.697635135135135, + "grad_norm": 1.3141627311706543, + "learning_rate": 6.021e-06, + "loss": 0.1682, + "step": 2010 + }, + { + "epoch": 1.6984797297297298, + "grad_norm": 1.222425103187561, + "learning_rate": 6.024e-06, + "loss": 0.1176, + "step": 2011 + }, + { + "epoch": 1.6993243243243243, + "grad_norm": 1.3296713829040527, + "learning_rate": 6.027e-06, + "loss": 0.1499, + "step": 2012 + }, + { + "epoch": 1.7001689189189189, + "grad_norm": 1.3633999824523926, + "learning_rate": 6.030000000000001e-06, + "loss": 0.133, + "step": 2013 + }, + { + "epoch": 1.7010135135135136, + "grad_norm": 1.309276819229126, + "learning_rate": 6.0330000000000005e-06, + "loss": 0.1452, + "step": 2014 + }, + { + "epoch": 1.7018581081081081, + "grad_norm": 1.1294827461242676, + "learning_rate": 6.0359999999999995e-06, + "loss": 0.1544, + "step": 2015 + }, + { + "epoch": 1.7027027027027026, + "grad_norm": 1.0290409326553345, + "learning_rate": 6.039e-06, + "loss": 0.1073, + "step": 2016 + }, + { + "epoch": 1.7035472972972974, + "grad_norm": 3.647226333618164, + "learning_rate": 6.042e-06, + "loss": 0.1535, + "step": 2017 + }, + { + "epoch": 1.7043918918918919, + "grad_norm": 1.3213340044021606, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.1308, + "step": 2018 + }, + { + "epoch": 1.7052364864864864, + "grad_norm": 1.4974024295806885, + "learning_rate": 6.048e-06, + "loss": 0.1531, + "step": 2019 + }, + { + "epoch": 1.7060810810810811, + "grad_norm": 1.3118715286254883, + "learning_rate": 6.051e-06, + "loss": 0.1454, + "step": 2020 + }, + { + "epoch": 1.7069256756756757, + "grad_norm": 1.6147431135177612, + "learning_rate": 6.054e-06, + "loss": 0.1668, + "step": 2021 + }, + { + "epoch": 1.7077702702702702, + "grad_norm": 2.090959310531616, + "learning_rate": 6.057e-06, + "loss": 0.1399, + "step": 2022 + }, + { + "epoch": 1.708614864864865, + "grad_norm": 1.717569351196289, + "learning_rate": 6.0600000000000004e-06, + "loss": 0.1353, + "step": 2023 + }, + { + "epoch": 1.7094594594594594, + "grad_norm": 2.336491346359253, + "learning_rate": 6.063e-06, + "loss": 0.1933, + "step": 2024 + }, + { + "epoch": 1.710304054054054, + "grad_norm": 1.2967239618301392, + "learning_rate": 6.066e-06, + "loss": 0.1174, + "step": 2025 + }, + { + "epoch": 1.7111486486486487, + "grad_norm": 1.640170693397522, + "learning_rate": 6.069000000000001e-06, + "loss": 0.1706, + "step": 2026 + }, + { + "epoch": 1.7119932432432432, + "grad_norm": 1.4562602043151855, + "learning_rate": 6.072e-06, + "loss": 0.1733, + "step": 2027 + }, + { + "epoch": 1.7128378378378377, + "grad_norm": 1.7936162948608398, + "learning_rate": 6.075e-06, + "loss": 0.2064, + "step": 2028 + }, + { + "epoch": 1.7136824324324325, + "grad_norm": 1.423949122428894, + "learning_rate": 6.078e-06, + "loss": 0.134, + "step": 2029 + }, + { + "epoch": 1.714527027027027, + "grad_norm": 2.4584813117980957, + "learning_rate": 6.081e-06, + "loss": 0.1763, + "step": 2030 + }, + { + "epoch": 1.7153716216216215, + "grad_norm": 1.2812120914459229, + "learning_rate": 6.0840000000000005e-06, + "loss": 0.141, + "step": 2031 + }, + { + "epoch": 1.7162162162162162, + "grad_norm": 2.644015073776245, + "learning_rate": 6.087e-06, + "loss": 0.2201, + "step": 2032 + }, + { + "epoch": 1.717060810810811, + "grad_norm": 2.5388083457946777, + "learning_rate": 6.090000000000001e-06, + "loss": 0.2085, + "step": 2033 + }, + { + "epoch": 1.7179054054054053, + "grad_norm": 9.905943870544434, + "learning_rate": 6.093e-06, + "loss": 0.2748, + "step": 2034 + }, + { + "epoch": 1.71875, + "grad_norm": 9.875591278076172, + "learning_rate": 6.096e-06, + "loss": 1.0468, + "step": 2035 + }, + { + "epoch": 1.7195945945945947, + "grad_norm": 53.16759490966797, + "learning_rate": 6.099e-06, + "loss": 0.4575, + "step": 2036 + }, + { + "epoch": 1.720439189189189, + "grad_norm": 1.8120124340057373, + "learning_rate": 6.102e-06, + "loss": 0.2327, + "step": 2037 + }, + { + "epoch": 1.7212837837837838, + "grad_norm": 1.8298662900924683, + "learning_rate": 6.105e-06, + "loss": 0.1999, + "step": 2038 + }, + { + "epoch": 1.7221283783783785, + "grad_norm": 1.9087481498718262, + "learning_rate": 6.108000000000001e-06, + "loss": 0.1745, + "step": 2039 + }, + { + "epoch": 1.722972972972973, + "grad_norm": 0.8406025767326355, + "learning_rate": 6.111e-06, + "loss": 0.1309, + "step": 2040 + }, + { + "epoch": 1.7238175675675675, + "grad_norm": 1.177677869796753, + "learning_rate": 6.114e-06, + "loss": 0.1847, + "step": 2041 + }, + { + "epoch": 1.7246621621621623, + "grad_norm": 1.2014291286468506, + "learning_rate": 6.117e-06, + "loss": 0.1718, + "step": 2042 + }, + { + "epoch": 1.7255067567567568, + "grad_norm": 5.695807933807373, + "learning_rate": 6.12e-06, + "loss": 0.1833, + "step": 2043 + }, + { + "epoch": 1.7263513513513513, + "grad_norm": 1.499327540397644, + "learning_rate": 6.1230000000000005e-06, + "loss": 0.1073, + "step": 2044 + }, + { + "epoch": 1.727195945945946, + "grad_norm": 0.9169625639915466, + "learning_rate": 6.126e-06, + "loss": 0.1404, + "step": 2045 + }, + { + "epoch": 1.7280405405405406, + "grad_norm": 1.0694313049316406, + "learning_rate": 6.129e-06, + "loss": 0.1318, + "step": 2046 + }, + { + "epoch": 1.728885135135135, + "grad_norm": 1.4382919073104858, + "learning_rate": 6.132e-06, + "loss": 0.1265, + "step": 2047 + }, + { + "epoch": 1.7297297297297298, + "grad_norm": 1.6150068044662476, + "learning_rate": 6.135e-06, + "loss": 0.1376, + "step": 2048 + }, + { + "epoch": 1.7305743243243243, + "grad_norm": 1.161498785018921, + "learning_rate": 6.138e-06, + "loss": 0.1046, + "step": 2049 + }, + { + "epoch": 1.7314189189189189, + "grad_norm": 1.4488911628723145, + "learning_rate": 6.141e-06, + "loss": 0.1392, + "step": 2050 + }, + { + "epoch": 1.7322635135135136, + "grad_norm": 1.0978926420211792, + "learning_rate": 6.144000000000001e-06, + "loss": 0.1299, + "step": 2051 + }, + { + "epoch": 1.7331081081081081, + "grad_norm": 2.7701144218444824, + "learning_rate": 6.147e-06, + "loss": 0.1743, + "step": 2052 + }, + { + "epoch": 1.7339527027027026, + "grad_norm": 1.2656372785568237, + "learning_rate": 6.1499999999999996e-06, + "loss": 0.1201, + "step": 2053 + }, + { + "epoch": 1.7347972972972974, + "grad_norm": 1.3013156652450562, + "learning_rate": 6.153e-06, + "loss": 0.1398, + "step": 2054 + }, + { + "epoch": 1.7356418918918919, + "grad_norm": 1.211892008781433, + "learning_rate": 6.156e-06, + "loss": 0.1319, + "step": 2055 + }, + { + "epoch": 1.7364864864864864, + "grad_norm": 1.4082523584365845, + "learning_rate": 6.159000000000001e-06, + "loss": 0.1251, + "step": 2056 + }, + { + "epoch": 1.7373310810810811, + "grad_norm": 1.3489588499069214, + "learning_rate": 6.1620000000000005e-06, + "loss": 0.1318, + "step": 2057 + }, + { + "epoch": 1.7381756756756757, + "grad_norm": 1.2169004678726196, + "learning_rate": 6.164999999999999e-06, + "loss": 0.1472, + "step": 2058 + }, + { + "epoch": 1.7390202702702702, + "grad_norm": 1.0765137672424316, + "learning_rate": 6.168e-06, + "loss": 0.1266, + "step": 2059 + }, + { + "epoch": 1.739864864864865, + "grad_norm": 2.1628668308258057, + "learning_rate": 6.171e-06, + "loss": 0.1557, + "step": 2060 + }, + { + "epoch": 1.7407094594594594, + "grad_norm": 1.1017036437988281, + "learning_rate": 6.1740000000000005e-06, + "loss": 0.1152, + "step": 2061 + }, + { + "epoch": 1.741554054054054, + "grad_norm": 1.2247732877731323, + "learning_rate": 6.177e-06, + "loss": 0.14, + "step": 2062 + }, + { + "epoch": 1.7423986486486487, + "grad_norm": 1.0472081899642944, + "learning_rate": 6.18e-06, + "loss": 0.1077, + "step": 2063 + }, + { + "epoch": 1.7432432432432432, + "grad_norm": 1.2408885955810547, + "learning_rate": 6.183e-06, + "loss": 0.1219, + "step": 2064 + }, + { + "epoch": 1.7440878378378377, + "grad_norm": 1.3630635738372803, + "learning_rate": 6.186e-06, + "loss": 0.1206, + "step": 2065 + }, + { + "epoch": 1.7449324324324325, + "grad_norm": 1.9301010370254517, + "learning_rate": 6.189e-06, + "loss": 0.1421, + "step": 2066 + }, + { + "epoch": 1.745777027027027, + "grad_norm": 1.0728249549865723, + "learning_rate": 6.192e-06, + "loss": 0.1308, + "step": 2067 + }, + { + "epoch": 1.7466216216216215, + "grad_norm": 1.1953643560409546, + "learning_rate": 6.195e-06, + "loss": 0.1333, + "step": 2068 + }, + { + "epoch": 1.7474662162162162, + "grad_norm": 1.4183740615844727, + "learning_rate": 6.198000000000001e-06, + "loss": 0.2036, + "step": 2069 + }, + { + "epoch": 1.748310810810811, + "grad_norm": 1.102730631828308, + "learning_rate": 6.201e-06, + "loss": 0.1345, + "step": 2070 + }, + { + "epoch": 1.7491554054054053, + "grad_norm": 4.564948081970215, + "learning_rate": 6.204e-06, + "loss": 0.2533, + "step": 2071 + }, + { + "epoch": 1.75, + "grad_norm": 1.182463526725769, + "learning_rate": 6.207e-06, + "loss": 0.1441, + "step": 2072 + }, + { + "epoch": 1.7508445945945947, + "grad_norm": 1.5100951194763184, + "learning_rate": 6.21e-06, + "loss": 0.162, + "step": 2073 + }, + { + "epoch": 1.751689189189189, + "grad_norm": 1.688610315322876, + "learning_rate": 6.2130000000000005e-06, + "loss": 0.1498, + "step": 2074 + }, + { + "epoch": 1.7525337837837838, + "grad_norm": 1.643114686012268, + "learning_rate": 6.216e-06, + "loss": 0.1755, + "step": 2075 + }, + { + "epoch": 1.7533783783783785, + "grad_norm": 1.4823931455612183, + "learning_rate": 6.219000000000001e-06, + "loss": 0.144, + "step": 2076 + }, + { + "epoch": 1.754222972972973, + "grad_norm": 1.1166726350784302, + "learning_rate": 6.222e-06, + "loss": 0.1266, + "step": 2077 + }, + { + "epoch": 1.7550675675675675, + "grad_norm": 1.5187420845031738, + "learning_rate": 6.225e-06, + "loss": 0.1879, + "step": 2078 + }, + { + "epoch": 1.7559121621621623, + "grad_norm": 1.469251036643982, + "learning_rate": 6.228e-06, + "loss": 0.1577, + "step": 2079 + }, + { + "epoch": 1.7567567567567568, + "grad_norm": 1.427419900894165, + "learning_rate": 6.231e-06, + "loss": 0.1565, + "step": 2080 + }, + { + "epoch": 1.7576013513513513, + "grad_norm": 2.0522918701171875, + "learning_rate": 6.234000000000001e-06, + "loss": 0.1695, + "step": 2081 + }, + { + "epoch": 1.758445945945946, + "grad_norm": 1.7113245725631714, + "learning_rate": 6.237000000000001e-06, + "loss": 0.1551, + "step": 2082 + }, + { + "epoch": 1.7592905405405406, + "grad_norm": 1.8578755855560303, + "learning_rate": 6.2399999999999995e-06, + "loss": 0.2123, + "step": 2083 + }, + { + "epoch": 1.760135135135135, + "grad_norm": 2.1960203647613525, + "learning_rate": 6.243e-06, + "loss": 0.2506, + "step": 2084 + }, + { + "epoch": 1.7609797297297298, + "grad_norm": 4.389821529388428, + "learning_rate": 6.246e-06, + "loss": 0.6336, + "step": 2085 + }, + { + "epoch": 1.7618243243243243, + "grad_norm": 3.1751110553741455, + "learning_rate": 6.249000000000001e-06, + "loss": 0.4293, + "step": 2086 + }, + { + "epoch": 1.7626689189189189, + "grad_norm": 2.245540142059326, + "learning_rate": 6.2520000000000004e-06, + "loss": 0.2657, + "step": 2087 + }, + { + "epoch": 1.7635135135135136, + "grad_norm": 2.3624536991119385, + "learning_rate": 6.255e-06, + "loss": 0.1678, + "step": 2088 + }, + { + "epoch": 1.7643581081081081, + "grad_norm": 1.8506691455841064, + "learning_rate": 6.258e-06, + "loss": 0.16, + "step": 2089 + }, + { + "epoch": 1.7652027027027026, + "grad_norm": 0.9752101302146912, + "learning_rate": 6.261e-06, + "loss": 0.1508, + "step": 2090 + }, + { + "epoch": 1.7660472972972974, + "grad_norm": 0.7899629473686218, + "learning_rate": 6.2640000000000005e-06, + "loss": 0.1299, + "step": 2091 + }, + { + "epoch": 1.7668918918918919, + "grad_norm": 1.0437506437301636, + "learning_rate": 6.267e-06, + "loss": 0.1147, + "step": 2092 + }, + { + "epoch": 1.7677364864864864, + "grad_norm": 1.0047173500061035, + "learning_rate": 6.27e-06, + "loss": 0.1411, + "step": 2093 + }, + { + "epoch": 1.7685810810810811, + "grad_norm": 1.0405603647232056, + "learning_rate": 6.273000000000001e-06, + "loss": 0.1187, + "step": 2094 + }, + { + "epoch": 1.7694256756756757, + "grad_norm": 0.9476592540740967, + "learning_rate": 6.276e-06, + "loss": 0.1169, + "step": 2095 + }, + { + "epoch": 1.7702702702702702, + "grad_norm": 2.339439868927002, + "learning_rate": 6.279e-06, + "loss": 0.1808, + "step": 2096 + }, + { + "epoch": 1.771114864864865, + "grad_norm": 0.9006716012954712, + "learning_rate": 6.282e-06, + "loss": 0.103, + "step": 2097 + }, + { + "epoch": 1.7719594594594594, + "grad_norm": 1.2728568315505981, + "learning_rate": 6.285e-06, + "loss": 0.1672, + "step": 2098 + }, + { + "epoch": 1.772804054054054, + "grad_norm": 1.0321969985961914, + "learning_rate": 6.288000000000001e-06, + "loss": 0.1204, + "step": 2099 + }, + { + "epoch": 1.7736486486486487, + "grad_norm": 1.0889111757278442, + "learning_rate": 6.291e-06, + "loss": 0.1091, + "step": 2100 + }, + { + "epoch": 1.7744932432432432, + "grad_norm": 1.0833550691604614, + "learning_rate": 6.293999999999999e-06, + "loss": 0.1189, + "step": 2101 + }, + { + "epoch": 1.7753378378378377, + "grad_norm": 1.0806214809417725, + "learning_rate": 6.297e-06, + "loss": 0.1213, + "step": 2102 + }, + { + "epoch": 1.7761824324324325, + "grad_norm": 1.200939416885376, + "learning_rate": 6.3e-06, + "loss": 0.1444, + "step": 2103 + }, + { + "epoch": 1.777027027027027, + "grad_norm": 0.8888285160064697, + "learning_rate": 6.3030000000000005e-06, + "loss": 0.1098, + "step": 2104 + }, + { + "epoch": 1.7778716216216215, + "grad_norm": 1.6914712190628052, + "learning_rate": 6.306e-06, + "loss": 0.1369, + "step": 2105 + }, + { + "epoch": 1.7787162162162162, + "grad_norm": 1.0007309913635254, + "learning_rate": 6.309e-06, + "loss": 0.1023, + "step": 2106 + }, + { + "epoch": 1.779560810810811, + "grad_norm": 1.1674100160598755, + "learning_rate": 6.312e-06, + "loss": 0.1528, + "step": 2107 + }, + { + "epoch": 1.7804054054054053, + "grad_norm": 1.0572506189346313, + "learning_rate": 6.315e-06, + "loss": 0.1256, + "step": 2108 + }, + { + "epoch": 1.78125, + "grad_norm": 1.2727535963058472, + "learning_rate": 6.318e-06, + "loss": 0.1181, + "step": 2109 + }, + { + "epoch": 1.7820945945945947, + "grad_norm": 0.9201080799102783, + "learning_rate": 6.321e-06, + "loss": 0.1314, + "step": 2110 + }, + { + "epoch": 1.782939189189189, + "grad_norm": 1.3739378452301025, + "learning_rate": 6.324e-06, + "loss": 0.1277, + "step": 2111 + }, + { + "epoch": 1.7837837837837838, + "grad_norm": 1.247348427772522, + "learning_rate": 6.327000000000001e-06, + "loss": 0.1447, + "step": 2112 + }, + { + "epoch": 1.7846283783783785, + "grad_norm": 1.1163101196289062, + "learning_rate": 6.3299999999999995e-06, + "loss": 0.1185, + "step": 2113 + }, + { + "epoch": 1.785472972972973, + "grad_norm": 1.577866792678833, + "learning_rate": 6.333e-06, + "loss": 0.1561, + "step": 2114 + }, + { + "epoch": 1.7863175675675675, + "grad_norm": 1.0657271146774292, + "learning_rate": 6.336e-06, + "loss": 0.1219, + "step": 2115 + }, + { + "epoch": 1.7871621621621623, + "grad_norm": 1.2733298540115356, + "learning_rate": 6.339e-06, + "loss": 0.1325, + "step": 2116 + }, + { + "epoch": 1.7880067567567568, + "grad_norm": 1.2610002756118774, + "learning_rate": 6.3420000000000004e-06, + "loss": 0.1294, + "step": 2117 + }, + { + "epoch": 1.7888513513513513, + "grad_norm": 1.176339864730835, + "learning_rate": 6.345e-06, + "loss": 0.1329, + "step": 2118 + }, + { + "epoch": 1.789695945945946, + "grad_norm": 1.4364200830459595, + "learning_rate": 6.348000000000001e-06, + "loss": 0.1165, + "step": 2119 + }, + { + "epoch": 1.7905405405405406, + "grad_norm": 1.1530662775039673, + "learning_rate": 6.351e-06, + "loss": 0.1069, + "step": 2120 + }, + { + "epoch": 1.791385135135135, + "grad_norm": 1.2441656589508057, + "learning_rate": 6.354e-06, + "loss": 0.1191, + "step": 2121 + }, + { + "epoch": 1.7922297297297298, + "grad_norm": 1.4068763256072998, + "learning_rate": 6.357e-06, + "loss": 0.1582, + "step": 2122 + }, + { + "epoch": 1.7930743243243243, + "grad_norm": 1.3868862390518188, + "learning_rate": 6.36e-06, + "loss": 0.1995, + "step": 2123 + }, + { + "epoch": 1.7939189189189189, + "grad_norm": 1.1738866567611694, + "learning_rate": 6.363000000000001e-06, + "loss": 0.1257, + "step": 2124 + }, + { + "epoch": 1.7947635135135136, + "grad_norm": 1.358820915222168, + "learning_rate": 6.3660000000000005e-06, + "loss": 0.1548, + "step": 2125 + }, + { + "epoch": 1.7956081081081081, + "grad_norm": 1.0812758207321167, + "learning_rate": 6.3689999999999995e-06, + "loss": 0.1356, + "step": 2126 + }, + { + "epoch": 1.7964527027027026, + "grad_norm": 1.606369137763977, + "learning_rate": 6.372e-06, + "loss": 0.195, + "step": 2127 + }, + { + "epoch": 1.7972972972972974, + "grad_norm": 1.7716666460037231, + "learning_rate": 6.375e-06, + "loss": 0.1778, + "step": 2128 + }, + { + "epoch": 1.7981418918918919, + "grad_norm": 1.3071073293685913, + "learning_rate": 6.378000000000001e-06, + "loss": 0.1378, + "step": 2129 + }, + { + "epoch": 1.7989864864864864, + "grad_norm": 1.992932677268982, + "learning_rate": 6.381e-06, + "loss": 0.1626, + "step": 2130 + }, + { + "epoch": 1.7998310810810811, + "grad_norm": 1.6383377313613892, + "learning_rate": 6.384e-06, + "loss": 0.1553, + "step": 2131 + }, + { + "epoch": 1.8006756756756757, + "grad_norm": 5.942189693450928, + "learning_rate": 6.387e-06, + "loss": 0.2302, + "step": 2132 + }, + { + "epoch": 1.8015202702702702, + "grad_norm": 2.2914321422576904, + "learning_rate": 6.39e-06, + "loss": 0.2207, + "step": 2133 + }, + { + "epoch": 1.802364864864865, + "grad_norm": 2.032013177871704, + "learning_rate": 6.3930000000000005e-06, + "loss": 0.2269, + "step": 2134 + }, + { + "epoch": 1.8032094594594594, + "grad_norm": 3.1153061389923096, + "learning_rate": 6.396e-06, + "loss": 0.7999, + "step": 2135 + }, + { + "epoch": 1.804054054054054, + "grad_norm": 2.585081100463867, + "learning_rate": 6.399e-06, + "loss": 0.4775, + "step": 2136 + }, + { + "epoch": 1.8048986486486487, + "grad_norm": 2.869380235671997, + "learning_rate": 6.402000000000001e-06, + "loss": 0.2456, + "step": 2137 + }, + { + "epoch": 1.8057432432432432, + "grad_norm": 1.8631556034088135, + "learning_rate": 6.405e-06, + "loss": 0.1848, + "step": 2138 + }, + { + "epoch": 1.8065878378378377, + "grad_norm": 1.052965521812439, + "learning_rate": 6.408e-06, + "loss": 0.1494, + "step": 2139 + }, + { + "epoch": 1.8074324324324325, + "grad_norm": 0.7998631000518799, + "learning_rate": 6.411e-06, + "loss": 0.1265, + "step": 2140 + }, + { + "epoch": 1.808277027027027, + "grad_norm": 1.0544301271438599, + "learning_rate": 6.414e-06, + "loss": 0.1487, + "step": 2141 + }, + { + "epoch": 1.8091216216216215, + "grad_norm": 0.9726176261901855, + "learning_rate": 6.4170000000000006e-06, + "loss": 0.1171, + "step": 2142 + }, + { + "epoch": 1.8099662162162162, + "grad_norm": 1.973771095275879, + "learning_rate": 6.42e-06, + "loss": 0.1549, + "step": 2143 + }, + { + "epoch": 1.810810810810811, + "grad_norm": 1.1337848901748657, + "learning_rate": 6.423e-06, + "loss": 0.1384, + "step": 2144 + }, + { + "epoch": 1.8116554054054053, + "grad_norm": 2.278548240661621, + "learning_rate": 6.426e-06, + "loss": 0.1464, + "step": 2145 + }, + { + "epoch": 1.8125, + "grad_norm": 1.292780876159668, + "learning_rate": 6.429e-06, + "loss": 0.1462, + "step": 2146 + }, + { + "epoch": 1.8133445945945947, + "grad_norm": 1.1544955968856812, + "learning_rate": 6.432e-06, + "loss": 0.1137, + "step": 2147 + }, + { + "epoch": 1.814189189189189, + "grad_norm": 1.2573612928390503, + "learning_rate": 6.435e-06, + "loss": 0.113, + "step": 2148 + }, + { + "epoch": 1.8150337837837838, + "grad_norm": 1.0040464401245117, + "learning_rate": 6.438000000000001e-06, + "loss": 0.1358, + "step": 2149 + }, + { + "epoch": 1.8158783783783785, + "grad_norm": 1.2700096368789673, + "learning_rate": 6.441e-06, + "loss": 0.1052, + "step": 2150 + }, + { + "epoch": 1.816722972972973, + "grad_norm": 1.1953891515731812, + "learning_rate": 6.444e-06, + "loss": 0.1395, + "step": 2151 + }, + { + "epoch": 1.8175675675675675, + "grad_norm": 0.9775186777114868, + "learning_rate": 6.447e-06, + "loss": 0.1228, + "step": 2152 + }, + { + "epoch": 1.8184121621621623, + "grad_norm": 1.1156736612319946, + "learning_rate": 6.45e-06, + "loss": 0.1156, + "step": 2153 + }, + { + "epoch": 1.8192567567567568, + "grad_norm": 1.2112460136413574, + "learning_rate": 6.453000000000001e-06, + "loss": 0.1285, + "step": 2154 + }, + { + "epoch": 1.8201013513513513, + "grad_norm": 0.8381634950637817, + "learning_rate": 6.4560000000000005e-06, + "loss": 0.1059, + "step": 2155 + }, + { + "epoch": 1.820945945945946, + "grad_norm": 1.1307711601257324, + "learning_rate": 6.4589999999999995e-06, + "loss": 0.1237, + "step": 2156 + }, + { + "epoch": 1.8217905405405406, + "grad_norm": 1.224463939666748, + "learning_rate": 6.462e-06, + "loss": 0.1217, + "step": 2157 + }, + { + "epoch": 1.822635135135135, + "grad_norm": 1.005128026008606, + "learning_rate": 6.465e-06, + "loss": 0.1402, + "step": 2158 + }, + { + "epoch": 1.8234797297297298, + "grad_norm": 1.0201618671417236, + "learning_rate": 6.468000000000001e-06, + "loss": 0.1187, + "step": 2159 + }, + { + "epoch": 1.8243243243243243, + "grad_norm": 1.0009511709213257, + "learning_rate": 6.471e-06, + "loss": 0.1348, + "step": 2160 + }, + { + "epoch": 1.8251689189189189, + "grad_norm": 1.185470461845398, + "learning_rate": 6.474e-06, + "loss": 0.1095, + "step": 2161 + }, + { + "epoch": 1.8260135135135136, + "grad_norm": 1.154400110244751, + "learning_rate": 6.477000000000001e-06, + "loss": 0.1216, + "step": 2162 + }, + { + "epoch": 1.8268581081081081, + "grad_norm": 1.4836293458938599, + "learning_rate": 6.48e-06, + "loss": 0.1506, + "step": 2163 + }, + { + "epoch": 1.8277027027027026, + "grad_norm": 1.0904505252838135, + "learning_rate": 6.483e-06, + "loss": 0.1303, + "step": 2164 + }, + { + "epoch": 1.8285472972972974, + "grad_norm": 0.9716386795043945, + "learning_rate": 6.486e-06, + "loss": 0.101, + "step": 2165 + }, + { + "epoch": 1.8293918918918919, + "grad_norm": 1.2873759269714355, + "learning_rate": 6.489e-06, + "loss": 0.1653, + "step": 2166 + }, + { + "epoch": 1.8302364864864864, + "grad_norm": 1.112714171409607, + "learning_rate": 6.492000000000001e-06, + "loss": 0.1494, + "step": 2167 + }, + { + "epoch": 1.8310810810810811, + "grad_norm": 1.4339066743850708, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.1407, + "step": 2168 + }, + { + "epoch": 1.8319256756756757, + "grad_norm": 1.160923719406128, + "learning_rate": 6.4979999999999994e-06, + "loss": 0.1294, + "step": 2169 + }, + { + "epoch": 1.8327702702702702, + "grad_norm": 1.261833667755127, + "learning_rate": 6.501e-06, + "loss": 0.1692, + "step": 2170 + }, + { + "epoch": 1.833614864864865, + "grad_norm": 1.0188966989517212, + "learning_rate": 6.504e-06, + "loss": 0.1424, + "step": 2171 + }, + { + "epoch": 1.8344594594594594, + "grad_norm": 1.3210488557815552, + "learning_rate": 6.5070000000000005e-06, + "loss": 0.148, + "step": 2172 + }, + { + "epoch": 1.835304054054054, + "grad_norm": 1.3301976919174194, + "learning_rate": 6.51e-06, + "loss": 0.1415, + "step": 2173 + }, + { + "epoch": 1.8361486486486487, + "grad_norm": 1.2153706550598145, + "learning_rate": 6.513e-06, + "loss": 0.1375, + "step": 2174 + }, + { + "epoch": 1.8369932432432432, + "grad_norm": 1.25640869140625, + "learning_rate": 6.516e-06, + "loss": 0.1338, + "step": 2175 + }, + { + "epoch": 1.8378378378378377, + "grad_norm": 1.1555848121643066, + "learning_rate": 6.519e-06, + "loss": 0.1559, + "step": 2176 + }, + { + "epoch": 1.8386824324324325, + "grad_norm": 1.1721701622009277, + "learning_rate": 6.522e-06, + "loss": 0.1398, + "step": 2177 + }, + { + "epoch": 1.839527027027027, + "grad_norm": 1.3286291360855103, + "learning_rate": 6.525e-06, + "loss": 0.1244, + "step": 2178 + }, + { + "epoch": 1.8403716216216215, + "grad_norm": 1.2205663919448853, + "learning_rate": 6.528e-06, + "loss": 0.1159, + "step": 2179 + }, + { + "epoch": 1.8412162162162162, + "grad_norm": 2.192399501800537, + "learning_rate": 6.531000000000001e-06, + "loss": 0.1997, + "step": 2180 + }, + { + "epoch": 1.842060810810811, + "grad_norm": 3.780648946762085, + "learning_rate": 6.534e-06, + "loss": 0.1939, + "step": 2181 + }, + { + "epoch": 1.8429054054054053, + "grad_norm": 1.7294989824295044, + "learning_rate": 6.537e-06, + "loss": 0.1957, + "step": 2182 + }, + { + "epoch": 1.84375, + "grad_norm": 2.067898750305176, + "learning_rate": 6.54e-06, + "loss": 0.2134, + "step": 2183 + }, + { + "epoch": 1.8445945945945947, + "grad_norm": 2.8137245178222656, + "learning_rate": 6.543e-06, + "loss": 0.1941, + "step": 2184 + }, + { + "epoch": 1.845439189189189, + "grad_norm": 6.156249523162842, + "learning_rate": 6.5460000000000005e-06, + "loss": 0.9052, + "step": 2185 + }, + { + "epoch": 1.8462837837837838, + "grad_norm": 1.8950800895690918, + "learning_rate": 6.549e-06, + "loss": 0.5826, + "step": 2186 + }, + { + "epoch": 1.8471283783783785, + "grad_norm": 2.757788896560669, + "learning_rate": 6.552e-06, + "loss": 0.341, + "step": 2187 + }, + { + "epoch": 1.847972972972973, + "grad_norm": 1.8498497009277344, + "learning_rate": 6.555e-06, + "loss": 0.2503, + "step": 2188 + }, + { + "epoch": 1.8488175675675675, + "grad_norm": 1.1124205589294434, + "learning_rate": 6.558e-06, + "loss": 0.1761, + "step": 2189 + }, + { + "epoch": 1.8496621621621623, + "grad_norm": 0.9287055730819702, + "learning_rate": 6.561e-06, + "loss": 0.1332, + "step": 2190 + }, + { + "epoch": 1.8505067567567568, + "grad_norm": 1.7292813062667847, + "learning_rate": 6.564e-06, + "loss": 0.2094, + "step": 2191 + }, + { + "epoch": 1.8513513513513513, + "grad_norm": 1.020951271057129, + "learning_rate": 6.567000000000001e-06, + "loss": 0.1266, + "step": 2192 + }, + { + "epoch": 1.852195945945946, + "grad_norm": 2.8043432235717773, + "learning_rate": 6.57e-06, + "loss": 0.1799, + "step": 2193 + }, + { + "epoch": 1.8530405405405406, + "grad_norm": 0.9260420203208923, + "learning_rate": 6.573e-06, + "loss": 0.1147, + "step": 2194 + }, + { + "epoch": 1.853885135135135, + "grad_norm": 1.1570398807525635, + "learning_rate": 6.576e-06, + "loss": 0.1313, + "step": 2195 + }, + { + "epoch": 1.8547297297297298, + "grad_norm": 0.9756576418876648, + "learning_rate": 6.579e-06, + "loss": 0.1045, + "step": 2196 + }, + { + "epoch": 1.8555743243243243, + "grad_norm": 0.9473168253898621, + "learning_rate": 6.582000000000001e-06, + "loss": 0.1307, + "step": 2197 + }, + { + "epoch": 1.8564189189189189, + "grad_norm": 1.0843729972839355, + "learning_rate": 6.5850000000000005e-06, + "loss": 0.1288, + "step": 2198 + }, + { + "epoch": 1.8572635135135136, + "grad_norm": 0.8414655923843384, + "learning_rate": 6.5879999999999994e-06, + "loss": 0.1155, + "step": 2199 + }, + { + "epoch": 1.8581081081081081, + "grad_norm": 1.086321234703064, + "learning_rate": 6.591e-06, + "loss": 0.1187, + "step": 2200 + }, + { + "epoch": 1.8589527027027026, + "grad_norm": 1.1596089601516724, + "learning_rate": 6.594e-06, + "loss": 0.0985, + "step": 2201 + }, + { + "epoch": 1.8597972972972974, + "grad_norm": 1.0688934326171875, + "learning_rate": 6.5970000000000005e-06, + "loss": 0.1121, + "step": 2202 + }, + { + "epoch": 1.8606418918918919, + "grad_norm": 1.0596575736999512, + "learning_rate": 6.6e-06, + "loss": 0.1162, + "step": 2203 + }, + { + "epoch": 1.8614864864864864, + "grad_norm": 2.10979962348938, + "learning_rate": 6.603e-06, + "loss": 0.106, + "step": 2204 + }, + { + "epoch": 1.8623310810810811, + "grad_norm": 1.106037974357605, + "learning_rate": 6.606000000000001e-06, + "loss": 0.1294, + "step": 2205 + }, + { + "epoch": 1.8631756756756757, + "grad_norm": 0.8989018797874451, + "learning_rate": 6.609e-06, + "loss": 0.1191, + "step": 2206 + }, + { + "epoch": 1.8640202702702702, + "grad_norm": 1.598931908607483, + "learning_rate": 6.612e-06, + "loss": 0.1084, + "step": 2207 + }, + { + "epoch": 1.864864864864865, + "grad_norm": 1.1082419157028198, + "learning_rate": 6.615e-06, + "loss": 0.1318, + "step": 2208 + }, + { + "epoch": 1.8657094594594594, + "grad_norm": 0.9610135555267334, + "learning_rate": 6.618e-06, + "loss": 0.0811, + "step": 2209 + }, + { + "epoch": 1.866554054054054, + "grad_norm": 1.252178430557251, + "learning_rate": 6.621000000000001e-06, + "loss": 0.1574, + "step": 2210 + }, + { + "epoch": 1.8673986486486487, + "grad_norm": 7.083797931671143, + "learning_rate": 6.6240000000000004e-06, + "loss": 0.1736, + "step": 2211 + }, + { + "epoch": 1.8682432432432432, + "grad_norm": 1.0479373931884766, + "learning_rate": 6.627e-06, + "loss": 0.1302, + "step": 2212 + }, + { + "epoch": 1.8690878378378377, + "grad_norm": 1.1833794116973877, + "learning_rate": 6.63e-06, + "loss": 0.1222, + "step": 2213 + }, + { + "epoch": 1.8699324324324325, + "grad_norm": 2.005720376968384, + "learning_rate": 6.633e-06, + "loss": 0.1393, + "step": 2214 + }, + { + "epoch": 1.870777027027027, + "grad_norm": 0.9508286118507385, + "learning_rate": 6.6360000000000005e-06, + "loss": 0.1195, + "step": 2215 + }, + { + "epoch": 1.8716216216216215, + "grad_norm": 1.4371155500411987, + "learning_rate": 6.639e-06, + "loss": 0.0985, + "step": 2216 + }, + { + "epoch": 1.8724662162162162, + "grad_norm": 1.314818024635315, + "learning_rate": 6.642000000000001e-06, + "loss": 0.1127, + "step": 2217 + }, + { + "epoch": 1.873310810810811, + "grad_norm": 1.7082796096801758, + "learning_rate": 6.645e-06, + "loss": 0.1208, + "step": 2218 + }, + { + "epoch": 1.8741554054054053, + "grad_norm": 1.2789206504821777, + "learning_rate": 6.648e-06, + "loss": 0.1323, + "step": 2219 + }, + { + "epoch": 1.875, + "grad_norm": 0.9151164293289185, + "learning_rate": 6.651e-06, + "loss": 0.1006, + "step": 2220 + }, + { + "epoch": 1.8758445945945947, + "grad_norm": 1.1321877241134644, + "learning_rate": 6.654e-06, + "loss": 0.1223, + "step": 2221 + }, + { + "epoch": 1.876689189189189, + "grad_norm": 1.440638542175293, + "learning_rate": 6.657e-06, + "loss": 0.1418, + "step": 2222 + }, + { + "epoch": 1.8775337837837838, + "grad_norm": 1.319799780845642, + "learning_rate": 6.660000000000001e-06, + "loss": 0.1452, + "step": 2223 + }, + { + "epoch": 1.8783783783783785, + "grad_norm": 1.3859444856643677, + "learning_rate": 6.6629999999999996e-06, + "loss": 0.122, + "step": 2224 + }, + { + "epoch": 1.879222972972973, + "grad_norm": 1.3997819423675537, + "learning_rate": 6.666e-06, + "loss": 0.1227, + "step": 2225 + }, + { + "epoch": 1.8800675675675675, + "grad_norm": 1.1035360097885132, + "learning_rate": 6.669e-06, + "loss": 0.0966, + "step": 2226 + }, + { + "epoch": 1.8809121621621623, + "grad_norm": 1.218550682067871, + "learning_rate": 6.672e-06, + "loss": 0.1339, + "step": 2227 + }, + { + "epoch": 1.8817567567567568, + "grad_norm": 1.3197805881500244, + "learning_rate": 6.6750000000000005e-06, + "loss": 0.1199, + "step": 2228 + }, + { + "epoch": 1.8826013513513513, + "grad_norm": 1.4777849912643433, + "learning_rate": 6.678e-06, + "loss": 0.1341, + "step": 2229 + }, + { + "epoch": 1.883445945945946, + "grad_norm": 1.956593632698059, + "learning_rate": 6.681e-06, + "loss": 0.1906, + "step": 2230 + }, + { + "epoch": 1.8842905405405406, + "grad_norm": 1.527190923690796, + "learning_rate": 6.684e-06, + "loss": 0.1647, + "step": 2231 + }, + { + "epoch": 1.885135135135135, + "grad_norm": 1.4182446002960205, + "learning_rate": 6.687e-06, + "loss": 0.178, + "step": 2232 + }, + { + "epoch": 1.8859797297297298, + "grad_norm": 1.779191493988037, + "learning_rate": 6.69e-06, + "loss": 0.2078, + "step": 2233 + }, + { + "epoch": 1.8868243243243243, + "grad_norm": 3.3239493370056152, + "learning_rate": 6.693e-06, + "loss": 0.2441, + "step": 2234 + }, + { + "epoch": 1.8876689189189189, + "grad_norm": 3.33569598197937, + "learning_rate": 6.696000000000001e-06, + "loss": 0.6746, + "step": 2235 + }, + { + "epoch": 1.8885135135135136, + "grad_norm": 2.7087926864624023, + "learning_rate": 6.699e-06, + "loss": 0.3632, + "step": 2236 + }, + { + "epoch": 1.8893581081081081, + "grad_norm": 1.592551589012146, + "learning_rate": 6.7019999999999995e-06, + "loss": 0.2263, + "step": 2237 + }, + { + "epoch": 1.8902027027027026, + "grad_norm": 1.1928666830062866, + "learning_rate": 6.705e-06, + "loss": 0.1768, + "step": 2238 + }, + { + "epoch": 1.8910472972972974, + "grad_norm": 1.0949774980545044, + "learning_rate": 6.708e-06, + "loss": 0.1607, + "step": 2239 + }, + { + "epoch": 1.8918918918918919, + "grad_norm": 0.9733695387840271, + "learning_rate": 6.711000000000001e-06, + "loss": 0.1099, + "step": 2240 + }, + { + "epoch": 1.8927364864864864, + "grad_norm": 1.7364087104797363, + "learning_rate": 6.7140000000000004e-06, + "loss": 0.1772, + "step": 2241 + }, + { + "epoch": 1.8935810810810811, + "grad_norm": 1.179868221282959, + "learning_rate": 6.716999999999999e-06, + "loss": 0.1228, + "step": 2242 + }, + { + "epoch": 1.8944256756756757, + "grad_norm": 1.1269148588180542, + "learning_rate": 6.72e-06, + "loss": 0.1264, + "step": 2243 + }, + { + "epoch": 1.8952702702702702, + "grad_norm": 0.9750103950500488, + "learning_rate": 6.723e-06, + "loss": 0.1159, + "step": 2244 + }, + { + "epoch": 1.896114864864865, + "grad_norm": 1.0395057201385498, + "learning_rate": 6.7260000000000005e-06, + "loss": 0.1068, + "step": 2245 + }, + { + "epoch": 1.8969594594594594, + "grad_norm": 0.8854040503501892, + "learning_rate": 6.729e-06, + "loss": 0.0936, + "step": 2246 + }, + { + "epoch": 1.897804054054054, + "grad_norm": 1.1774938106536865, + "learning_rate": 6.732e-06, + "loss": 0.1277, + "step": 2247 + }, + { + "epoch": 1.8986486486486487, + "grad_norm": 1.44331693649292, + "learning_rate": 6.735000000000001e-06, + "loss": 0.098, + "step": 2248 + }, + { + "epoch": 1.8994932432432432, + "grad_norm": 1.5801177024841309, + "learning_rate": 6.738e-06, + "loss": 0.1751, + "step": 2249 + }, + { + "epoch": 1.9003378378378377, + "grad_norm": 1.4890609979629517, + "learning_rate": 6.741e-06, + "loss": 0.1235, + "step": 2250 + }, + { + "epoch": 1.9011824324324325, + "grad_norm": 1.2619407176971436, + "learning_rate": 6.744e-06, + "loss": 0.1297, + "step": 2251 + }, + { + "epoch": 1.902027027027027, + "grad_norm": 1.3758633136749268, + "learning_rate": 6.747e-06, + "loss": 0.1005, + "step": 2252 + }, + { + "epoch": 1.9028716216216215, + "grad_norm": 0.9529840350151062, + "learning_rate": 6.750000000000001e-06, + "loss": 0.0926, + "step": 2253 + }, + { + "epoch": 1.9037162162162162, + "grad_norm": 2.113706588745117, + "learning_rate": 6.753e-06, + "loss": 0.1577, + "step": 2254 + }, + { + "epoch": 1.904560810810811, + "grad_norm": 1.4726420640945435, + "learning_rate": 6.756e-06, + "loss": 0.1153, + "step": 2255 + }, + { + "epoch": 1.9054054054054053, + "grad_norm": 1.4547163248062134, + "learning_rate": 6.759e-06, + "loss": 0.2164, + "step": 2256 + }, + { + "epoch": 1.90625, + "grad_norm": 1.3646327257156372, + "learning_rate": 6.762e-06, + "loss": 0.1514, + "step": 2257 + }, + { + "epoch": 1.9070945945945947, + "grad_norm": 0.7736602425575256, + "learning_rate": 6.7650000000000005e-06, + "loss": 0.0983, + "step": 2258 + }, + { + "epoch": 1.907939189189189, + "grad_norm": 1.0127463340759277, + "learning_rate": 6.768e-06, + "loss": 0.121, + "step": 2259 + }, + { + "epoch": 1.9087837837837838, + "grad_norm": 1.3630692958831787, + "learning_rate": 6.771000000000001e-06, + "loss": 0.1389, + "step": 2260 + }, + { + "epoch": 1.9096283783783785, + "grad_norm": 2.058675765991211, + "learning_rate": 6.774e-06, + "loss": 0.1354, + "step": 2261 + }, + { + "epoch": 1.910472972972973, + "grad_norm": 1.1486055850982666, + "learning_rate": 6.777e-06, + "loss": 0.1157, + "step": 2262 + }, + { + "epoch": 1.9113175675675675, + "grad_norm": 1.1683285236358643, + "learning_rate": 6.78e-06, + "loss": 0.1266, + "step": 2263 + }, + { + "epoch": 1.9121621621621623, + "grad_norm": 1.230776309967041, + "learning_rate": 6.783e-06, + "loss": 0.1391, + "step": 2264 + }, + { + "epoch": 1.9130067567567568, + "grad_norm": 1.1943597793579102, + "learning_rate": 6.786000000000001e-06, + "loss": 0.1362, + "step": 2265 + }, + { + "epoch": 1.9138513513513513, + "grad_norm": 1.1075196266174316, + "learning_rate": 6.7890000000000006e-06, + "loss": 0.1739, + "step": 2266 + }, + { + "epoch": 1.914695945945946, + "grad_norm": 1.0710245370864868, + "learning_rate": 6.7919999999999995e-06, + "loss": 0.101, + "step": 2267 + }, + { + "epoch": 1.9155405405405406, + "grad_norm": 1.4933388233184814, + "learning_rate": 6.795e-06, + "loss": 0.1401, + "step": 2268 + }, + { + "epoch": 1.916385135135135, + "grad_norm": 1.467674970626831, + "learning_rate": 6.798e-06, + "loss": 0.1142, + "step": 2269 + }, + { + "epoch": 1.9172297297297298, + "grad_norm": 1.0349066257476807, + "learning_rate": 6.801000000000001e-06, + "loss": 0.142, + "step": 2270 + }, + { + "epoch": 1.9180743243243243, + "grad_norm": 1.6008667945861816, + "learning_rate": 6.804e-06, + "loss": 0.1728, + "step": 2271 + }, + { + "epoch": 1.9189189189189189, + "grad_norm": 1.1407041549682617, + "learning_rate": 6.807e-06, + "loss": 0.1501, + "step": 2272 + }, + { + "epoch": 1.9197635135135136, + "grad_norm": 1.67826509475708, + "learning_rate": 6.81e-06, + "loss": 0.1194, + "step": 2273 + }, + { + "epoch": 1.9206081081081081, + "grad_norm": 1.3945963382720947, + "learning_rate": 6.813e-06, + "loss": 0.1471, + "step": 2274 + }, + { + "epoch": 1.9214527027027026, + "grad_norm": 1.3384573459625244, + "learning_rate": 6.8160000000000005e-06, + "loss": 0.1612, + "step": 2275 + }, + { + "epoch": 1.9222972972972974, + "grad_norm": 1.4098657369613647, + "learning_rate": 6.819e-06, + "loss": 0.1159, + "step": 2276 + }, + { + "epoch": 1.9231418918918919, + "grad_norm": 1.5431578159332275, + "learning_rate": 6.822e-06, + "loss": 0.1596, + "step": 2277 + }, + { + "epoch": 1.9239864864864864, + "grad_norm": 1.7742098569869995, + "learning_rate": 6.825000000000001e-06, + "loss": 0.1467, + "step": 2278 + }, + { + "epoch": 1.9248310810810811, + "grad_norm": 1.6378182172775269, + "learning_rate": 6.828e-06, + "loss": 0.1484, + "step": 2279 + }, + { + "epoch": 1.9256756756756757, + "grad_norm": 2.0025694370269775, + "learning_rate": 6.831e-06, + "loss": 0.2134, + "step": 2280 + }, + { + "epoch": 1.9265202702702702, + "grad_norm": 1.5955296754837036, + "learning_rate": 6.834e-06, + "loss": 0.1658, + "step": 2281 + }, + { + "epoch": 1.927364864864865, + "grad_norm": 1.303896427154541, + "learning_rate": 6.837e-06, + "loss": 0.1238, + "step": 2282 + }, + { + "epoch": 1.9282094594594594, + "grad_norm": 1.6215808391571045, + "learning_rate": 6.840000000000001e-06, + "loss": 0.2034, + "step": 2283 + }, + { + "epoch": 1.929054054054054, + "grad_norm": 2.9811301231384277, + "learning_rate": 6.843e-06, + "loss": 0.2296, + "step": 2284 + }, + { + "epoch": 1.9298986486486487, + "grad_norm": 4.20515251159668, + "learning_rate": 6.845999999999999e-06, + "loss": 0.6118, + "step": 2285 + }, + { + "epoch": 1.9307432432432432, + "grad_norm": 1.5078749656677246, + "learning_rate": 6.849e-06, + "loss": 0.3329, + "step": 2286 + }, + { + "epoch": 1.9315878378378377, + "grad_norm": 1.3817905187606812, + "learning_rate": 6.852e-06, + "loss": 0.2106, + "step": 2287 + }, + { + "epoch": 1.9324324324324325, + "grad_norm": 0.9480740427970886, + "learning_rate": 6.8550000000000004e-06, + "loss": 0.1558, + "step": 2288 + }, + { + "epoch": 1.933277027027027, + "grad_norm": 1.28413724899292, + "learning_rate": 6.858e-06, + "loss": 0.1439, + "step": 2289 + }, + { + "epoch": 1.9341216216216215, + "grad_norm": 1.0695288181304932, + "learning_rate": 6.861e-06, + "loss": 0.1435, + "step": 2290 + }, + { + "epoch": 1.9349662162162162, + "grad_norm": 0.8453320860862732, + "learning_rate": 6.864000000000001e-06, + "loss": 0.1382, + "step": 2291 + }, + { + "epoch": 1.935810810810811, + "grad_norm": 0.8742843270301819, + "learning_rate": 6.867e-06, + "loss": 0.1122, + "step": 2292 + }, + { + "epoch": 1.9366554054054053, + "grad_norm": 1.2053769826889038, + "learning_rate": 6.87e-06, + "loss": 0.1371, + "step": 2293 + }, + { + "epoch": 1.9375, + "grad_norm": 1.3301661014556885, + "learning_rate": 6.873e-06, + "loss": 0.1252, + "step": 2294 + }, + { + "epoch": 1.9383445945945947, + "grad_norm": 1.175487756729126, + "learning_rate": 6.876e-06, + "loss": 0.1369, + "step": 2295 + }, + { + "epoch": 1.939189189189189, + "grad_norm": 1.3227366209030151, + "learning_rate": 6.8790000000000005e-06, + "loss": 0.1292, + "step": 2296 + }, + { + "epoch": 1.9400337837837838, + "grad_norm": 1.0381426811218262, + "learning_rate": 6.882e-06, + "loss": 0.104, + "step": 2297 + }, + { + "epoch": 1.9408783783783785, + "grad_norm": 0.7632998824119568, + "learning_rate": 6.885e-06, + "loss": 0.1075, + "step": 2298 + }, + { + "epoch": 1.941722972972973, + "grad_norm": 0.9146774411201477, + "learning_rate": 6.888e-06, + "loss": 0.1089, + "step": 2299 + }, + { + "epoch": 1.9425675675675675, + "grad_norm": 1.1805763244628906, + "learning_rate": 6.891e-06, + "loss": 0.1396, + "step": 2300 + }, + { + "epoch": 1.9434121621621623, + "grad_norm": 1.0891902446746826, + "learning_rate": 6.894e-06, + "loss": 0.131, + "step": 2301 + }, + { + "epoch": 1.9442567567567568, + "grad_norm": 0.9072476029396057, + "learning_rate": 6.897e-06, + "loss": 0.1017, + "step": 2302 + }, + { + "epoch": 1.9451013513513513, + "grad_norm": 0.8811686635017395, + "learning_rate": 6.900000000000001e-06, + "loss": 0.1318, + "step": 2303 + }, + { + "epoch": 1.945945945945946, + "grad_norm": 0.9272662997245789, + "learning_rate": 6.903e-06, + "loss": 0.1419, + "step": 2304 + }, + { + "epoch": 1.9467905405405406, + "grad_norm": 0.8424041867256165, + "learning_rate": 6.906e-06, + "loss": 0.1059, + "step": 2305 + }, + { + "epoch": 1.947635135135135, + "grad_norm": 1.8654146194458008, + "learning_rate": 6.909e-06, + "loss": 0.1253, + "step": 2306 + }, + { + "epoch": 1.9484797297297298, + "grad_norm": 1.0455267429351807, + "learning_rate": 6.912e-06, + "loss": 0.1248, + "step": 2307 + }, + { + "epoch": 1.9493243243243243, + "grad_norm": 1.2317559719085693, + "learning_rate": 6.915000000000001e-06, + "loss": 0.1547, + "step": 2308 + }, + { + "epoch": 1.9501689189189189, + "grad_norm": 0.8135879039764404, + "learning_rate": 6.9180000000000005e-06, + "loss": 0.1061, + "step": 2309 + }, + { + "epoch": 1.9510135135135136, + "grad_norm": 1.1030696630477905, + "learning_rate": 6.9209999999999995e-06, + "loss": 0.1069, + "step": 2310 + }, + { + "epoch": 1.9518581081081081, + "grad_norm": 1.7793264389038086, + "learning_rate": 6.924e-06, + "loss": 0.129, + "step": 2311 + }, + { + "epoch": 1.9527027027027026, + "grad_norm": 0.7901331186294556, + "learning_rate": 6.927e-06, + "loss": 0.0991, + "step": 2312 + }, + { + "epoch": 1.9535472972972974, + "grad_norm": 0.988332986831665, + "learning_rate": 6.9300000000000006e-06, + "loss": 0.1206, + "step": 2313 + }, + { + "epoch": 1.9543918918918919, + "grad_norm": 2.7317943572998047, + "learning_rate": 6.933e-06, + "loss": 0.1681, + "step": 2314 + }, + { + "epoch": 1.9552364864864864, + "grad_norm": 1.178551197052002, + "learning_rate": 6.936e-06, + "loss": 0.1209, + "step": 2315 + }, + { + "epoch": 1.9560810810810811, + "grad_norm": 1.6271001100540161, + "learning_rate": 6.939e-06, + "loss": 0.1552, + "step": 2316 + }, + { + "epoch": 1.9569256756756757, + "grad_norm": 1.0523591041564941, + "learning_rate": 6.942e-06, + "loss": 0.1106, + "step": 2317 + }, + { + "epoch": 1.9577702702702702, + "grad_norm": 3.7056212425231934, + "learning_rate": 6.945e-06, + "loss": 0.1411, + "step": 2318 + }, + { + "epoch": 1.958614864864865, + "grad_norm": 0.8499191999435425, + "learning_rate": 6.948e-06, + "loss": 0.1004, + "step": 2319 + }, + { + "epoch": 1.9594594594594594, + "grad_norm": 1.1706290245056152, + "learning_rate": 6.951e-06, + "loss": 0.1467, + "step": 2320 + }, + { + "epoch": 1.960304054054054, + "grad_norm": 1.3724497556686401, + "learning_rate": 6.954000000000001e-06, + "loss": 0.1009, + "step": 2321 + }, + { + "epoch": 1.9611486486486487, + "grad_norm": 1.0529228448867798, + "learning_rate": 6.957e-06, + "loss": 0.1198, + "step": 2322 + }, + { + "epoch": 1.9619932432432432, + "grad_norm": 1.1640732288360596, + "learning_rate": 6.96e-06, + "loss": 0.1518, + "step": 2323 + }, + { + "epoch": 1.9628378378378377, + "grad_norm": 1.3173247575759888, + "learning_rate": 6.963e-06, + "loss": 0.1155, + "step": 2324 + }, + { + "epoch": 1.9636824324324325, + "grad_norm": 0.9539600014686584, + "learning_rate": 6.966e-06, + "loss": 0.0957, + "step": 2325 + }, + { + "epoch": 1.964527027027027, + "grad_norm": 1.2502638101577759, + "learning_rate": 6.9690000000000005e-06, + "loss": 0.1533, + "step": 2326 + }, + { + "epoch": 1.9653716216216215, + "grad_norm": 1.5120232105255127, + "learning_rate": 6.972e-06, + "loss": 0.1735, + "step": 2327 + }, + { + "epoch": 1.9662162162162162, + "grad_norm": 2.6156601905822754, + "learning_rate": 6.975e-06, + "loss": 0.1148, + "step": 2328 + }, + { + "epoch": 1.967060810810811, + "grad_norm": 1.1847797632217407, + "learning_rate": 6.978e-06, + "loss": 0.1451, + "step": 2329 + }, + { + "epoch": 1.9679054054054053, + "grad_norm": 1.1527228355407715, + "learning_rate": 6.981e-06, + "loss": 0.1379, + "step": 2330 + }, + { + "epoch": 1.96875, + "grad_norm": 1.3634190559387207, + "learning_rate": 6.984e-06, + "loss": 0.1341, + "step": 2331 + }, + { + "epoch": 1.9695945945945947, + "grad_norm": 1.3502929210662842, + "learning_rate": 6.987e-06, + "loss": 0.1603, + "step": 2332 + }, + { + "epoch": 1.970439189189189, + "grad_norm": 1.8066997528076172, + "learning_rate": 6.990000000000001e-06, + "loss": 0.1791, + "step": 2333 + }, + { + "epoch": 1.9712837837837838, + "grad_norm": 3.130704879760742, + "learning_rate": 6.993000000000001e-06, + "loss": 0.2918, + "step": 2334 + }, + { + "epoch": 1.9721283783783785, + "grad_norm": 3.272644519805908, + "learning_rate": 6.996e-06, + "loss": 0.6186, + "step": 2335 + }, + { + "epoch": 1.972972972972973, + "grad_norm": 1.4004939794540405, + "learning_rate": 6.999e-06, + "loss": 0.2364, + "step": 2336 + }, + { + "epoch": 1.9738175675675675, + "grad_norm": 1.5028772354125977, + "learning_rate": 7.002e-06, + "loss": 0.1601, + "step": 2337 + }, + { + "epoch": 1.9746621621621623, + "grad_norm": 1.0878822803497314, + "learning_rate": 7.005000000000001e-06, + "loss": 0.16, + "step": 2338 + }, + { + "epoch": 1.9755067567567568, + "grad_norm": 0.9565085172653198, + "learning_rate": 7.0080000000000005e-06, + "loss": 0.1228, + "step": 2339 + }, + { + "epoch": 1.9763513513513513, + "grad_norm": 0.8256822228431702, + "learning_rate": 7.011e-06, + "loss": 0.11, + "step": 2340 + }, + { + "epoch": 1.977195945945946, + "grad_norm": 1.1299891471862793, + "learning_rate": 7.014e-06, + "loss": 0.1577, + "step": 2341 + }, + { + "epoch": 1.9780405405405406, + "grad_norm": 0.9188127517700195, + "learning_rate": 7.017e-06, + "loss": 0.1112, + "step": 2342 + }, + { + "epoch": 1.978885135135135, + "grad_norm": 0.980688214302063, + "learning_rate": 7.0200000000000006e-06, + "loss": 0.095, + "step": 2343 + }, + { + "epoch": 1.9797297297297298, + "grad_norm": 0.9118990302085876, + "learning_rate": 7.023e-06, + "loss": 0.1053, + "step": 2344 + }, + { + "epoch": 1.9805743243243243, + "grad_norm": 1.191172480583191, + "learning_rate": 7.026e-06, + "loss": 0.1069, + "step": 2345 + }, + { + "epoch": 1.9814189189189189, + "grad_norm": 1.2310813665390015, + "learning_rate": 7.029000000000001e-06, + "loss": 0.143, + "step": 2346 + }, + { + "epoch": 1.9822635135135136, + "grad_norm": 0.7818002104759216, + "learning_rate": 7.032e-06, + "loss": 0.0953, + "step": 2347 + }, + { + "epoch": 1.9831081081081081, + "grad_norm": 1.3718023300170898, + "learning_rate": 7.0349999999999996e-06, + "loss": 0.1071, + "step": 2348 + }, + { + "epoch": 1.9839527027027026, + "grad_norm": 0.9540791511535645, + "learning_rate": 7.038e-06, + "loss": 0.1289, + "step": 2349 + }, + { + "epoch": 1.9847972972972974, + "grad_norm": 1.230220079421997, + "learning_rate": 7.041e-06, + "loss": 0.1049, + "step": 2350 + }, + { + "epoch": 1.9856418918918919, + "grad_norm": 1.2504544258117676, + "learning_rate": 7.044000000000001e-06, + "loss": 0.129, + "step": 2351 + }, + { + "epoch": 1.9864864864864864, + "grad_norm": 0.7837900519371033, + "learning_rate": 7.0470000000000005e-06, + "loss": 0.0981, + "step": 2352 + }, + { + "epoch": 1.9873310810810811, + "grad_norm": 1.003962755203247, + "learning_rate": 7.049999999999999e-06, + "loss": 0.1004, + "step": 2353 + }, + { + "epoch": 1.9881756756756757, + "grad_norm": 1.0295857191085815, + "learning_rate": 7.053e-06, + "loss": 0.1307, + "step": 2354 + }, + { + "epoch": 1.9890202702702702, + "grad_norm": 1.0591245889663696, + "learning_rate": 7.056e-06, + "loss": 0.1163, + "step": 2355 + }, + { + "epoch": 1.989864864864865, + "grad_norm": 0.9879153966903687, + "learning_rate": 7.0590000000000005e-06, + "loss": 0.1051, + "step": 2356 + }, + { + "epoch": 1.9907094594594594, + "grad_norm": 1.173729658126831, + "learning_rate": 7.062e-06, + "loss": 0.1544, + "step": 2357 + }, + { + "epoch": 1.991554054054054, + "grad_norm": 1.0108227729797363, + "learning_rate": 7.065e-06, + "loss": 0.0977, + "step": 2358 + }, + { + "epoch": 1.9923986486486487, + "grad_norm": 1.2876133918762207, + "learning_rate": 7.068e-06, + "loss": 0.122, + "step": 2359 + }, + { + "epoch": 1.9932432432432432, + "grad_norm": 1.114437222480774, + "learning_rate": 7.071e-06, + "loss": 0.1263, + "step": 2360 + }, + { + "epoch": 1.9940878378378377, + "grad_norm": 1.0937659740447998, + "learning_rate": 7.074e-06, + "loss": 0.0906, + "step": 2361 + }, + { + "epoch": 1.9949324324324325, + "grad_norm": 1.321212887763977, + "learning_rate": 7.077e-06, + "loss": 0.1502, + "step": 2362 + }, + { + "epoch": 1.995777027027027, + "grad_norm": 1.4911539554595947, + "learning_rate": 7.08e-06, + "loss": 0.1457, + "step": 2363 + }, + { + "epoch": 1.9966216216216215, + "grad_norm": 1.659818172454834, + "learning_rate": 7.083000000000001e-06, + "loss": 0.1639, + "step": 2364 + }, + { + "epoch": 1.9974662162162162, + "grad_norm": 1.2768394947052002, + "learning_rate": 7.086e-06, + "loss": 0.1465, + "step": 2365 + }, + { + "epoch": 1.998310810810811, + "grad_norm": 2.2949793338775635, + "learning_rate": 7.089e-06, + "loss": 0.1962, + "step": 2366 + }, + { + "epoch": 1.9991554054054053, + "grad_norm": 1.670772671699524, + "learning_rate": 7.092e-06, + "loss": 0.1756, + "step": 2367 + }, + { + "epoch": 2.0, + "grad_norm": 4.514655113220215, + "learning_rate": 7.095e-06, + "loss": 0.1994, + "step": 2368 + }, + { + "epoch": 2.0008445945945947, + "grad_norm": 2.311114549636841, + "learning_rate": 7.0980000000000005e-06, + "loss": 0.5492, + "step": 2369 + }, + { + "epoch": 2.001689189189189, + "grad_norm": 2.466826915740967, + "learning_rate": 7.101e-06, + "loss": 0.4747, + "step": 2370 + }, + { + "epoch": 2.0025337837837838, + "grad_norm": 1.2992451190948486, + "learning_rate": 7.104e-06, + "loss": 0.1652, + "step": 2371 + }, + { + "epoch": 2.0033783783783785, + "grad_norm": 1.3333684206008911, + "learning_rate": 7.107e-06, + "loss": 0.1404, + "step": 2372 + }, + { + "epoch": 2.004222972972973, + "grad_norm": 1.0439183712005615, + "learning_rate": 7.11e-06, + "loss": 0.1755, + "step": 2373 + }, + { + "epoch": 2.0050675675675675, + "grad_norm": 0.7572810649871826, + "learning_rate": 7.113e-06, + "loss": 0.1062, + "step": 2374 + }, + { + "epoch": 2.0059121621621623, + "grad_norm": 0.8596597909927368, + "learning_rate": 7.116e-06, + "loss": 0.1013, + "step": 2375 + }, + { + "epoch": 2.0067567567567566, + "grad_norm": 1.21437668800354, + "learning_rate": 7.119000000000001e-06, + "loss": 0.1384, + "step": 2376 + }, + { + "epoch": 2.0076013513513513, + "grad_norm": 1.332314133644104, + "learning_rate": 7.122000000000001e-06, + "loss": 0.1239, + "step": 2377 + }, + { + "epoch": 2.008445945945946, + "grad_norm": 0.966830313205719, + "learning_rate": 7.1249999999999995e-06, + "loss": 0.0938, + "step": 2378 + }, + { + "epoch": 2.0092905405405403, + "grad_norm": 0.9870560765266418, + "learning_rate": 7.128e-06, + "loss": 0.1117, + "step": 2379 + }, + { + "epoch": 2.010135135135135, + "grad_norm": 0.9069182276725769, + "learning_rate": 7.131e-06, + "loss": 0.1098, + "step": 2380 + }, + { + "epoch": 2.01097972972973, + "grad_norm": 1.4439316987991333, + "learning_rate": 7.134000000000001e-06, + "loss": 0.1141, + "step": 2381 + }, + { + "epoch": 2.011824324324324, + "grad_norm": 1.314219355583191, + "learning_rate": 7.1370000000000004e-06, + "loss": 0.1563, + "step": 2382 + }, + { + "epoch": 2.012668918918919, + "grad_norm": 0.7559779286384583, + "learning_rate": 7.14e-06, + "loss": 0.0894, + "step": 2383 + }, + { + "epoch": 2.0135135135135136, + "grad_norm": 0.837196946144104, + "learning_rate": 7.143e-06, + "loss": 0.1049, + "step": 2384 + }, + { + "epoch": 2.014358108108108, + "grad_norm": 1.1426730155944824, + "learning_rate": 7.146e-06, + "loss": 0.1161, + "step": 2385 + }, + { + "epoch": 2.0152027027027026, + "grad_norm": 0.7427102327346802, + "learning_rate": 7.1490000000000005e-06, + "loss": 0.1041, + "step": 2386 + }, + { + "epoch": 2.0160472972972974, + "grad_norm": 2.07391357421875, + "learning_rate": 7.152e-06, + "loss": 0.1538, + "step": 2387 + }, + { + "epoch": 2.016891891891892, + "grad_norm": 1.0067124366760254, + "learning_rate": 7.155e-06, + "loss": 0.1331, + "step": 2388 + }, + { + "epoch": 2.0177364864864864, + "grad_norm": 1.1468406915664673, + "learning_rate": 7.158000000000001e-06, + "loss": 0.1352, + "step": 2389 + }, + { + "epoch": 2.018581081081081, + "grad_norm": 0.9966523051261902, + "learning_rate": 7.161e-06, + "loss": 0.1363, + "step": 2390 + }, + { + "epoch": 2.019425675675676, + "grad_norm": 1.1009172201156616, + "learning_rate": 7.164e-06, + "loss": 0.1278, + "step": 2391 + }, + { + "epoch": 2.02027027027027, + "grad_norm": 0.8891368508338928, + "learning_rate": 7.167e-06, + "loss": 0.104, + "step": 2392 + }, + { + "epoch": 2.021114864864865, + "grad_norm": 0.9080232381820679, + "learning_rate": 7.17e-06, + "loss": 0.0912, + "step": 2393 + }, + { + "epoch": 2.0219594594594597, + "grad_norm": 0.8434339761734009, + "learning_rate": 7.173000000000001e-06, + "loss": 0.0881, + "step": 2394 + }, + { + "epoch": 2.022804054054054, + "grad_norm": 0.810663104057312, + "learning_rate": 7.176e-06, + "loss": 0.0814, + "step": 2395 + }, + { + "epoch": 2.0236486486486487, + "grad_norm": 0.9204426407814026, + "learning_rate": 7.179e-06, + "loss": 0.093, + "step": 2396 + }, + { + "epoch": 2.0244932432432434, + "grad_norm": 0.7813100814819336, + "learning_rate": 7.182e-06, + "loss": 0.0942, + "step": 2397 + }, + { + "epoch": 2.0253378378378377, + "grad_norm": 1.3153879642486572, + "learning_rate": 7.185e-06, + "loss": 0.1165, + "step": 2398 + }, + { + "epoch": 2.0261824324324325, + "grad_norm": 1.1084784269332886, + "learning_rate": 7.1880000000000005e-06, + "loss": 0.1257, + "step": 2399 + }, + { + "epoch": 2.027027027027027, + "grad_norm": 1.297277808189392, + "learning_rate": 7.191e-06, + "loss": 0.1204, + "step": 2400 + }, + { + "epoch": 2.0278716216216215, + "grad_norm": 1.059334397315979, + "learning_rate": 7.194000000000001e-06, + "loss": 0.1054, + "step": 2401 + }, + { + "epoch": 2.0287162162162162, + "grad_norm": 0.950145959854126, + "learning_rate": 7.197e-06, + "loss": 0.1146, + "step": 2402 + }, + { + "epoch": 2.029560810810811, + "grad_norm": 1.037108063697815, + "learning_rate": 7.2e-06, + "loss": 0.1184, + "step": 2403 + }, + { + "epoch": 2.0304054054054053, + "grad_norm": 1.0306638479232788, + "learning_rate": 7.203e-06, + "loss": 0.0893, + "step": 2404 + }, + { + "epoch": 2.03125, + "grad_norm": 1.0703511238098145, + "learning_rate": 7.206e-06, + "loss": 0.124, + "step": 2405 + }, + { + "epoch": 2.0320945945945947, + "grad_norm": 1.1708170175552368, + "learning_rate": 7.209000000000001e-06, + "loss": 0.1411, + "step": 2406 + }, + { + "epoch": 2.032939189189189, + "grad_norm": 0.9466145634651184, + "learning_rate": 7.2120000000000006e-06, + "loss": 0.1126, + "step": 2407 + }, + { + "epoch": 2.0337837837837838, + "grad_norm": 1.5248230695724487, + "learning_rate": 7.2149999999999995e-06, + "loss": 0.1607, + "step": 2408 + }, + { + "epoch": 2.0346283783783785, + "grad_norm": 1.487282395362854, + "learning_rate": 7.218e-06, + "loss": 0.1274, + "step": 2409 + }, + { + "epoch": 2.035472972972973, + "grad_norm": 1.3788272142410278, + "learning_rate": 7.221e-06, + "loss": 0.1408, + "step": 2410 + }, + { + "epoch": 2.0363175675675675, + "grad_norm": 1.6969337463378906, + "learning_rate": 7.224e-06, + "loss": 0.1724, + "step": 2411 + }, + { + "epoch": 2.0371621621621623, + "grad_norm": 1.2854942083358765, + "learning_rate": 7.2270000000000004e-06, + "loss": 0.1096, + "step": 2412 + }, + { + "epoch": 2.0380067567567566, + "grad_norm": 1.0247389078140259, + "learning_rate": 7.23e-06, + "loss": 0.1032, + "step": 2413 + }, + { + "epoch": 2.0388513513513513, + "grad_norm": 1.2279698848724365, + "learning_rate": 7.233e-06, + "loss": 0.1443, + "step": 2414 + }, + { + "epoch": 2.039695945945946, + "grad_norm": 1.1464968919754028, + "learning_rate": 7.236e-06, + "loss": 0.1332, + "step": 2415 + }, + { + "epoch": 2.0405405405405403, + "grad_norm": 1.613045334815979, + "learning_rate": 7.239e-06, + "loss": 0.157, + "step": 2416 + }, + { + "epoch": 2.041385135135135, + "grad_norm": 2.8196349143981934, + "learning_rate": 7.242e-06, + "loss": 0.216, + "step": 2417 + }, + { + "epoch": 2.04222972972973, + "grad_norm": 2.4164414405822754, + "learning_rate": 7.245e-06, + "loss": 0.2559, + "step": 2418 + }, + { + "epoch": 2.043074324324324, + "grad_norm": 2.494983673095703, + "learning_rate": 7.248000000000001e-06, + "loss": 0.7287, + "step": 2419 + }, + { + "epoch": 2.043918918918919, + "grad_norm": 2.364711284637451, + "learning_rate": 7.2510000000000005e-06, + "loss": 0.417, + "step": 2420 + }, + { + "epoch": 2.0447635135135136, + "grad_norm": 1.6013977527618408, + "learning_rate": 7.2539999999999995e-06, + "loss": 0.1977, + "step": 2421 + }, + { + "epoch": 2.045608108108108, + "grad_norm": 1.0372052192687988, + "learning_rate": 7.257e-06, + "loss": 0.1953, + "step": 2422 + }, + { + "epoch": 2.0464527027027026, + "grad_norm": 1.195057988166809, + "learning_rate": 7.26e-06, + "loss": 0.1244, + "step": 2423 + }, + { + "epoch": 2.0472972972972974, + "grad_norm": 0.7981372475624084, + "learning_rate": 7.263000000000001e-06, + "loss": 0.1182, + "step": 2424 + }, + { + "epoch": 2.048141891891892, + "grad_norm": 0.714902937412262, + "learning_rate": 7.266e-06, + "loss": 0.098, + "step": 2425 + }, + { + "epoch": 2.0489864864864864, + "grad_norm": 0.7969973683357239, + "learning_rate": 7.269e-06, + "loss": 0.0971, + "step": 2426 + }, + { + "epoch": 2.049831081081081, + "grad_norm": 0.9210981130599976, + "learning_rate": 7.272e-06, + "loss": 0.1316, + "step": 2427 + }, + { + "epoch": 2.050675675675676, + "grad_norm": 0.8299961090087891, + "learning_rate": 7.275e-06, + "loss": 0.1028, + "step": 2428 + }, + { + "epoch": 2.05152027027027, + "grad_norm": 0.9855493903160095, + "learning_rate": 7.2780000000000005e-06, + "loss": 0.1239, + "step": 2429 + }, + { + "epoch": 2.052364864864865, + "grad_norm": 1.2786064147949219, + "learning_rate": 7.281e-06, + "loss": 0.0906, + "step": 2430 + }, + { + "epoch": 2.0532094594594597, + "grad_norm": 0.885388195514679, + "learning_rate": 7.284e-06, + "loss": 0.109, + "step": 2431 + }, + { + "epoch": 2.054054054054054, + "grad_norm": 0.7469346523284912, + "learning_rate": 7.287000000000001e-06, + "loss": 0.0942, + "step": 2432 + }, + { + "epoch": 2.0548986486486487, + "grad_norm": 0.800622820854187, + "learning_rate": 7.29e-06, + "loss": 0.1079, + "step": 2433 + }, + { + "epoch": 2.0557432432432434, + "grad_norm": 7.488420486450195, + "learning_rate": 7.293e-06, + "loss": 0.0962, + "step": 2434 + }, + { + "epoch": 2.0565878378378377, + "grad_norm": 0.9331423044204712, + "learning_rate": 7.296e-06, + "loss": 0.0933, + "step": 2435 + }, + { + "epoch": 2.0574324324324325, + "grad_norm": 0.6714611053466797, + "learning_rate": 7.299e-06, + "loss": 0.091, + "step": 2436 + }, + { + "epoch": 2.058277027027027, + "grad_norm": 0.8082850575447083, + "learning_rate": 7.3020000000000006e-06, + "loss": 0.1005, + "step": 2437 + }, + { + "epoch": 2.0591216216216215, + "grad_norm": 1.4218839406967163, + "learning_rate": 7.305e-06, + "loss": 0.129, + "step": 2438 + }, + { + "epoch": 2.0599662162162162, + "grad_norm": 0.814913272857666, + "learning_rate": 7.308e-06, + "loss": 0.0959, + "step": 2439 + }, + { + "epoch": 2.060810810810811, + "grad_norm": 0.92442387342453, + "learning_rate": 7.311e-06, + "loss": 0.0944, + "step": 2440 + }, + { + "epoch": 2.0616554054054053, + "grad_norm": 0.8865571618080139, + "learning_rate": 7.314e-06, + "loss": 0.1095, + "step": 2441 + }, + { + "epoch": 2.0625, + "grad_norm": 1.2259958982467651, + "learning_rate": 7.317e-06, + "loss": 0.0958, + "step": 2442 + }, + { + "epoch": 2.0633445945945947, + "grad_norm": 0.9350430369377136, + "learning_rate": 7.32e-06, + "loss": 0.1077, + "step": 2443 + }, + { + "epoch": 2.064189189189189, + "grad_norm": 0.9889055490493774, + "learning_rate": 7.323000000000001e-06, + "loss": 0.1235, + "step": 2444 + }, + { + "epoch": 2.0650337837837838, + "grad_norm": 0.8435184359550476, + "learning_rate": 7.326e-06, + "loss": 0.11, + "step": 2445 + }, + { + "epoch": 2.0658783783783785, + "grad_norm": 0.8840857148170471, + "learning_rate": 7.329e-06, + "loss": 0.0956, + "step": 2446 + }, + { + "epoch": 2.066722972972973, + "grad_norm": 0.8926823139190674, + "learning_rate": 7.332e-06, + "loss": 0.0796, + "step": 2447 + }, + { + "epoch": 2.0675675675675675, + "grad_norm": 0.9767733812332153, + "learning_rate": 7.335e-06, + "loss": 0.1239, + "step": 2448 + }, + { + "epoch": 2.0684121621621623, + "grad_norm": 0.9129282236099243, + "learning_rate": 7.338000000000001e-06, + "loss": 0.1097, + "step": 2449 + }, + { + "epoch": 2.0692567567567566, + "grad_norm": 1.003250241279602, + "learning_rate": 7.3410000000000005e-06, + "loss": 0.1019, + "step": 2450 + }, + { + "epoch": 2.0701013513513513, + "grad_norm": 0.9469596743583679, + "learning_rate": 7.3439999999999995e-06, + "loss": 0.1089, + "step": 2451 + }, + { + "epoch": 2.070945945945946, + "grad_norm": 1.0129034519195557, + "learning_rate": 7.347e-06, + "loss": 0.1149, + "step": 2452 + }, + { + "epoch": 2.0717905405405403, + "grad_norm": 1.0088685750961304, + "learning_rate": 7.35e-06, + "loss": 0.0937, + "step": 2453 + }, + { + "epoch": 2.072635135135135, + "grad_norm": 1.1532288789749146, + "learning_rate": 7.353000000000001e-06, + "loss": 0.1229, + "step": 2454 + }, + { + "epoch": 2.07347972972973, + "grad_norm": 4.1493988037109375, + "learning_rate": 7.356e-06, + "loss": 0.2528, + "step": 2455 + }, + { + "epoch": 2.074324324324324, + "grad_norm": 2.2207469940185547, + "learning_rate": 7.359e-06, + "loss": 0.1139, + "step": 2456 + }, + { + "epoch": 2.075168918918919, + "grad_norm": 1.2350263595581055, + "learning_rate": 7.362e-06, + "loss": 0.1258, + "step": 2457 + }, + { + "epoch": 2.0760135135135136, + "grad_norm": 0.8838123679161072, + "learning_rate": 7.365e-06, + "loss": 0.0862, + "step": 2458 + }, + { + "epoch": 2.076858108108108, + "grad_norm": 1.4205037355422974, + "learning_rate": 7.3680000000000004e-06, + "loss": 0.1188, + "step": 2459 + }, + { + "epoch": 2.0777027027027026, + "grad_norm": 1.2058627605438232, + "learning_rate": 7.371e-06, + "loss": 0.115, + "step": 2460 + }, + { + "epoch": 2.0785472972972974, + "grad_norm": 1.2004715204238892, + "learning_rate": 7.374e-06, + "loss": 0.1357, + "step": 2461 + }, + { + "epoch": 2.079391891891892, + "grad_norm": 1.1603479385375977, + "learning_rate": 7.377000000000001e-06, + "loss": 0.1087, + "step": 2462 + }, + { + "epoch": 2.0802364864864864, + "grad_norm": 1.2560878992080688, + "learning_rate": 7.3800000000000005e-06, + "loss": 0.148, + "step": 2463 + }, + { + "epoch": 2.081081081081081, + "grad_norm": 1.449967622756958, + "learning_rate": 7.383e-06, + "loss": 0.1176, + "step": 2464 + }, + { + "epoch": 2.081925675675676, + "grad_norm": 1.3918594121932983, + "learning_rate": 7.386e-06, + "loss": 0.1458, + "step": 2465 + }, + { + "epoch": 2.08277027027027, + "grad_norm": 2.8300931453704834, + "learning_rate": 7.389e-06, + "loss": 0.1716, + "step": 2466 + }, + { + "epoch": 2.083614864864865, + "grad_norm": 1.5997779369354248, + "learning_rate": 7.3920000000000005e-06, + "loss": 0.1526, + "step": 2467 + }, + { + "epoch": 2.0844594594594597, + "grad_norm": 4.542311191558838, + "learning_rate": 7.395e-06, + "loss": 0.2216, + "step": 2468 + }, + { + "epoch": 2.085304054054054, + "grad_norm": 2.9128458499908447, + "learning_rate": 7.398000000000001e-06, + "loss": 0.6079, + "step": 2469 + }, + { + "epoch": 2.0861486486486487, + "grad_norm": 2.0944361686706543, + "learning_rate": 7.401e-06, + "loss": 0.3075, + "step": 2470 + }, + { + "epoch": 2.0869932432432434, + "grad_norm": 0.8755017518997192, + "learning_rate": 7.404e-06, + "loss": 0.1965, + "step": 2471 + }, + { + "epoch": 2.0878378378378377, + "grad_norm": 1.5601258277893066, + "learning_rate": 7.407e-06, + "loss": 0.1534, + "step": 2472 + }, + { + "epoch": 2.0886824324324325, + "grad_norm": 1.1276715993881226, + "learning_rate": 7.41e-06, + "loss": 0.1405, + "step": 2473 + }, + { + "epoch": 2.089527027027027, + "grad_norm": 0.8817822933197021, + "learning_rate": 7.413e-06, + "loss": 0.1003, + "step": 2474 + }, + { + "epoch": 2.0903716216216215, + "grad_norm": 0.73763507604599, + "learning_rate": 7.416000000000001e-06, + "loss": 0.1156, + "step": 2475 + }, + { + "epoch": 2.0912162162162162, + "grad_norm": 0.9773623943328857, + "learning_rate": 7.419e-06, + "loss": 0.1452, + "step": 2476 + }, + { + "epoch": 2.092060810810811, + "grad_norm": 0.8597537279129028, + "learning_rate": 7.422e-06, + "loss": 0.1148, + "step": 2477 + }, + { + "epoch": 2.0929054054054053, + "grad_norm": 0.8663530349731445, + "learning_rate": 7.425e-06, + "loss": 0.1078, + "step": 2478 + }, + { + "epoch": 2.09375, + "grad_norm": 0.915885865688324, + "learning_rate": 7.428e-06, + "loss": 0.0923, + "step": 2479 + }, + { + "epoch": 2.0945945945945947, + "grad_norm": 1.1584516763687134, + "learning_rate": 7.4310000000000005e-06, + "loss": 0.1018, + "step": 2480 + }, + { + "epoch": 2.095439189189189, + "grad_norm": 0.6770544052124023, + "learning_rate": 7.434e-06, + "loss": 0.0928, + "step": 2481 + }, + { + "epoch": 2.0962837837837838, + "grad_norm": 1.0571309328079224, + "learning_rate": 7.437e-06, + "loss": 0.0895, + "step": 2482 + }, + { + "epoch": 2.0971283783783785, + "grad_norm": 1.037299633026123, + "learning_rate": 7.44e-06, + "loss": 0.1149, + "step": 2483 + }, + { + "epoch": 2.097972972972973, + "grad_norm": 1.1423981189727783, + "learning_rate": 7.443e-06, + "loss": 0.1368, + "step": 2484 + }, + { + "epoch": 2.0988175675675675, + "grad_norm": 1.030254602432251, + "learning_rate": 7.446e-06, + "loss": 0.0973, + "step": 2485 + }, + { + "epoch": 2.0996621621621623, + "grad_norm": 0.7282869219779968, + "learning_rate": 7.449e-06, + "loss": 0.0906, + "step": 2486 + }, + { + "epoch": 2.1005067567567566, + "grad_norm": 0.8591020703315735, + "learning_rate": 7.452000000000001e-06, + "loss": 0.1298, + "step": 2487 + }, + { + "epoch": 2.1013513513513513, + "grad_norm": 0.834425151348114, + "learning_rate": 7.455e-06, + "loss": 0.0937, + "step": 2488 + }, + { + "epoch": 2.102195945945946, + "grad_norm": 0.9022700786590576, + "learning_rate": 7.4579999999999996e-06, + "loss": 0.1111, + "step": 2489 + }, + { + "epoch": 2.1030405405405403, + "grad_norm": 0.8270949125289917, + "learning_rate": 7.461e-06, + "loss": 0.0999, + "step": 2490 + }, + { + "epoch": 2.103885135135135, + "grad_norm": 1.1080458164215088, + "learning_rate": 7.464e-06, + "loss": 0.1405, + "step": 2491 + }, + { + "epoch": 2.10472972972973, + "grad_norm": 0.9050450921058655, + "learning_rate": 7.467000000000001e-06, + "loss": 0.1003, + "step": 2492 + }, + { + "epoch": 2.105574324324324, + "grad_norm": 1.094318151473999, + "learning_rate": 7.4700000000000005e-06, + "loss": 0.0935, + "step": 2493 + }, + { + "epoch": 2.106418918918919, + "grad_norm": 1.233775019645691, + "learning_rate": 7.4729999999999994e-06, + "loss": 0.1446, + "step": 2494 + }, + { + "epoch": 2.1072635135135136, + "grad_norm": 1.0716302394866943, + "learning_rate": 7.476e-06, + "loss": 0.1246, + "step": 2495 + }, + { + "epoch": 2.108108108108108, + "grad_norm": 0.8911319971084595, + "learning_rate": 7.479e-06, + "loss": 0.0842, + "step": 2496 + }, + { + "epoch": 2.1089527027027026, + "grad_norm": 1.020512342453003, + "learning_rate": 7.4820000000000005e-06, + "loss": 0.1062, + "step": 2497 + }, + { + "epoch": 2.1097972972972974, + "grad_norm": 1.2293068170547485, + "learning_rate": 7.485e-06, + "loss": 0.1578, + "step": 2498 + }, + { + "epoch": 2.110641891891892, + "grad_norm": 0.8151623010635376, + "learning_rate": 7.488e-06, + "loss": 0.1195, + "step": 2499 + }, + { + "epoch": 2.1114864864864864, + "grad_norm": 1.1510932445526123, + "learning_rate": 7.491e-06, + "loss": 0.0983, + "step": 2500 + }, + { + "epoch": 2.112331081081081, + "grad_norm": 1.437073826789856, + "learning_rate": 7.494e-06, + "loss": 0.1224, + "step": 2501 + }, + { + "epoch": 2.113175675675676, + "grad_norm": 0.8904203176498413, + "learning_rate": 7.497e-06, + "loss": 0.0995, + "step": 2502 + }, + { + "epoch": 2.11402027027027, + "grad_norm": 1.0161832571029663, + "learning_rate": 7.5e-06, + "loss": 0.1193, + "step": 2503 + }, + { + "epoch": 2.114864864864865, + "grad_norm": 1.0913541316986084, + "learning_rate": 7.503e-06, + "loss": 0.092, + "step": 2504 + }, + { + "epoch": 2.1157094594594597, + "grad_norm": 0.8860921859741211, + "learning_rate": 7.506e-06, + "loss": 0.0831, + "step": 2505 + }, + { + "epoch": 2.116554054054054, + "grad_norm": 0.893229603767395, + "learning_rate": 7.5090000000000004e-06, + "loss": 0.084, + "step": 2506 + }, + { + "epoch": 2.1173986486486487, + "grad_norm": 1.033858299255371, + "learning_rate": 7.512e-06, + "loss": 0.1004, + "step": 2507 + }, + { + "epoch": 2.1182432432432434, + "grad_norm": 1.2375006675720215, + "learning_rate": 7.515e-06, + "loss": 0.1517, + "step": 2508 + }, + { + "epoch": 2.1190878378378377, + "grad_norm": 1.3431038856506348, + "learning_rate": 7.518e-06, + "loss": 0.1198, + "step": 2509 + }, + { + "epoch": 2.1199324324324325, + "grad_norm": 1.1540991067886353, + "learning_rate": 7.521e-06, + "loss": 0.1254, + "step": 2510 + }, + { + "epoch": 2.120777027027027, + "grad_norm": 1.4451889991760254, + "learning_rate": 7.524000000000001e-06, + "loss": 0.1245, + "step": 2511 + }, + { + "epoch": 2.1216216216216215, + "grad_norm": 1.083866834640503, + "learning_rate": 7.527000000000001e-06, + "loss": 0.0999, + "step": 2512 + }, + { + "epoch": 2.1224662162162162, + "grad_norm": 1.2949641942977905, + "learning_rate": 7.53e-06, + "loss": 0.0982, + "step": 2513 + }, + { + "epoch": 2.123310810810811, + "grad_norm": 1.842560052871704, + "learning_rate": 7.533e-06, + "loss": 0.1311, + "step": 2514 + }, + { + "epoch": 2.1241554054054053, + "grad_norm": 1.2808175086975098, + "learning_rate": 7.5359999999999995e-06, + "loss": 0.1483, + "step": 2515 + }, + { + "epoch": 2.125, + "grad_norm": 1.61309015750885, + "learning_rate": 7.539000000000001e-06, + "loss": 0.1327, + "step": 2516 + }, + { + "epoch": 2.1258445945945947, + "grad_norm": 1.7452294826507568, + "learning_rate": 7.542000000000001e-06, + "loss": 0.1407, + "step": 2517 + }, + { + "epoch": 2.126689189189189, + "grad_norm": 3.4652953147888184, + "learning_rate": 7.545000000000001e-06, + "loss": 0.276, + "step": 2518 + }, + { + "epoch": 2.1275337837837838, + "grad_norm": 3.141605854034424, + "learning_rate": 7.5479999999999996e-06, + "loss": 0.6677, + "step": 2519 + }, + { + "epoch": 2.1283783783783785, + "grad_norm": 1.5490771532058716, + "learning_rate": 7.550999999999999e-06, + "loss": 0.3895, + "step": 2520 + }, + { + "epoch": 2.129222972972973, + "grad_norm": 1.1992745399475098, + "learning_rate": 7.554000000000001e-06, + "loss": 0.2195, + "step": 2521 + }, + { + "epoch": 2.1300675675675675, + "grad_norm": 0.834488570690155, + "learning_rate": 7.557000000000001e-06, + "loss": 0.1346, + "step": 2522 + }, + { + "epoch": 2.1309121621621623, + "grad_norm": 0.7696085572242737, + "learning_rate": 7.5600000000000005e-06, + "loss": 0.1393, + "step": 2523 + }, + { + "epoch": 2.1317567567567566, + "grad_norm": 0.8149336576461792, + "learning_rate": 7.563e-06, + "loss": 0.1099, + "step": 2524 + }, + { + "epoch": 2.1326013513513513, + "grad_norm": 0.8372048139572144, + "learning_rate": 7.565999999999999e-06, + "loss": 0.1504, + "step": 2525 + }, + { + "epoch": 2.133445945945946, + "grad_norm": 1.1562193632125854, + "learning_rate": 7.569000000000001e-06, + "loss": 0.1265, + "step": 2526 + }, + { + "epoch": 2.1342905405405403, + "grad_norm": 0.9477880597114563, + "learning_rate": 7.5720000000000005e-06, + "loss": 0.1076, + "step": 2527 + }, + { + "epoch": 2.135135135135135, + "grad_norm": 0.7642850279808044, + "learning_rate": 7.575e-06, + "loss": 0.083, + "step": 2528 + }, + { + "epoch": 2.13597972972973, + "grad_norm": 0.7542359828948975, + "learning_rate": 7.578e-06, + "loss": 0.0843, + "step": 2529 + }, + { + "epoch": 2.136824324324324, + "grad_norm": 0.8772742748260498, + "learning_rate": 7.581e-06, + "loss": 0.1221, + "step": 2530 + }, + { + "epoch": 2.137668918918919, + "grad_norm": 0.6799959540367126, + "learning_rate": 7.5840000000000006e-06, + "loss": 0.0925, + "step": 2531 + }, + { + "epoch": 2.1385135135135136, + "grad_norm": 1.9989432096481323, + "learning_rate": 7.587e-06, + "loss": 0.113, + "step": 2532 + }, + { + "epoch": 2.139358108108108, + "grad_norm": 1.0933516025543213, + "learning_rate": 7.59e-06, + "loss": 0.1262, + "step": 2533 + }, + { + "epoch": 2.1402027027027026, + "grad_norm": 0.8949605822563171, + "learning_rate": 7.593e-06, + "loss": 0.0934, + "step": 2534 + }, + { + "epoch": 2.1410472972972974, + "grad_norm": 1.1295379400253296, + "learning_rate": 7.596e-06, + "loss": 0.1356, + "step": 2535 + }, + { + "epoch": 2.141891891891892, + "grad_norm": 1.080066442489624, + "learning_rate": 7.599000000000001e-06, + "loss": 0.106, + "step": 2536 + }, + { + "epoch": 2.1427364864864864, + "grad_norm": 1.0788980722427368, + "learning_rate": 7.602e-06, + "loss": 0.1031, + "step": 2537 + }, + { + "epoch": 2.143581081081081, + "grad_norm": 0.9111500382423401, + "learning_rate": 7.605e-06, + "loss": 0.0964, + "step": 2538 + }, + { + "epoch": 2.144425675675676, + "grad_norm": 0.730889618396759, + "learning_rate": 7.608e-06, + "loss": 0.0745, + "step": 2539 + }, + { + "epoch": 2.14527027027027, + "grad_norm": 0.8864085674285889, + "learning_rate": 7.611e-06, + "loss": 0.1032, + "step": 2540 + }, + { + "epoch": 2.146114864864865, + "grad_norm": 0.8444825410842896, + "learning_rate": 7.614000000000001e-06, + "loss": 0.1024, + "step": 2541 + }, + { + "epoch": 2.1469594594594597, + "grad_norm": 0.925871729850769, + "learning_rate": 7.617000000000001e-06, + "loss": 0.1087, + "step": 2542 + }, + { + "epoch": 2.147804054054054, + "grad_norm": 0.9951529502868652, + "learning_rate": 7.62e-06, + "loss": 0.0952, + "step": 2543 + }, + { + "epoch": 2.1486486486486487, + "grad_norm": 0.9829926490783691, + "learning_rate": 7.623e-06, + "loss": 0.0972, + "step": 2544 + }, + { + "epoch": 2.1494932432432434, + "grad_norm": 1.103360891342163, + "learning_rate": 7.6259999999999995e-06, + "loss": 0.1009, + "step": 2545 + }, + { + "epoch": 2.1503378378378377, + "grad_norm": 1.0112600326538086, + "learning_rate": 7.629000000000001e-06, + "loss": 0.1002, + "step": 2546 + }, + { + "epoch": 2.1511824324324325, + "grad_norm": 0.8843089938163757, + "learning_rate": 7.632e-06, + "loss": 0.0773, + "step": 2547 + }, + { + "epoch": 2.152027027027027, + "grad_norm": 1.1789799928665161, + "learning_rate": 7.635e-06, + "loss": 0.127, + "step": 2548 + }, + { + "epoch": 2.1528716216216215, + "grad_norm": 1.3580968379974365, + "learning_rate": 7.638e-06, + "loss": 0.1013, + "step": 2549 + }, + { + "epoch": 2.1537162162162162, + "grad_norm": 1.425561547279358, + "learning_rate": 7.641e-06, + "loss": 0.1292, + "step": 2550 + }, + { + "epoch": 2.154560810810811, + "grad_norm": 1.1304305791854858, + "learning_rate": 7.644000000000002e-06, + "loss": 0.1246, + "step": 2551 + }, + { + "epoch": 2.1554054054054053, + "grad_norm": 1.1653343439102173, + "learning_rate": 7.647000000000001e-06, + "loss": 0.1239, + "step": 2552 + }, + { + "epoch": 2.15625, + "grad_norm": 3.2050178050994873, + "learning_rate": 7.65e-06, + "loss": 0.2168, + "step": 2553 + }, + { + "epoch": 2.1570945945945947, + "grad_norm": 1.1617985963821411, + "learning_rate": 7.653e-06, + "loss": 0.1411, + "step": 2554 + }, + { + "epoch": 2.157939189189189, + "grad_norm": 0.8314923644065857, + "learning_rate": 7.656e-06, + "loss": 0.0816, + "step": 2555 + }, + { + "epoch": 2.1587837837837838, + "grad_norm": 0.9761619567871094, + "learning_rate": 7.659e-06, + "loss": 0.1035, + "step": 2556 + }, + { + "epoch": 2.1596283783783785, + "grad_norm": 1.045493245124817, + "learning_rate": 7.662e-06, + "loss": 0.1156, + "step": 2557 + }, + { + "epoch": 2.160472972972973, + "grad_norm": 0.9542612433433533, + "learning_rate": 7.665e-06, + "loss": 0.1369, + "step": 2558 + }, + { + "epoch": 2.1613175675675675, + "grad_norm": 1.125159740447998, + "learning_rate": 7.668e-06, + "loss": 0.1207, + "step": 2559 + }, + { + "epoch": 2.1621621621621623, + "grad_norm": 1.433834195137024, + "learning_rate": 7.671e-06, + "loss": 0.1245, + "step": 2560 + }, + { + "epoch": 2.1630067567567566, + "grad_norm": 1.1769260168075562, + "learning_rate": 7.674000000000001e-06, + "loss": 0.1306, + "step": 2561 + }, + { + "epoch": 2.1638513513513513, + "grad_norm": 1.1935228109359741, + "learning_rate": 7.677000000000001e-06, + "loss": 0.1057, + "step": 2562 + }, + { + "epoch": 2.164695945945946, + "grad_norm": 0.8676669001579285, + "learning_rate": 7.680000000000001e-06, + "loss": 0.1143, + "step": 2563 + }, + { + "epoch": 2.1655405405405403, + "grad_norm": 1.0807814598083496, + "learning_rate": 7.683e-06, + "loss": 0.0932, + "step": 2564 + }, + { + "epoch": 2.166385135135135, + "grad_norm": 1.219175100326538, + "learning_rate": 7.685999999999999e-06, + "loss": 0.1046, + "step": 2565 + }, + { + "epoch": 2.16722972972973, + "grad_norm": 1.582105278968811, + "learning_rate": 7.688999999999999e-06, + "loss": 0.1287, + "step": 2566 + }, + { + "epoch": 2.168074324324324, + "grad_norm": 1.3416062593460083, + "learning_rate": 7.692e-06, + "loss": 0.1214, + "step": 2567 + }, + { + "epoch": 2.168918918918919, + "grad_norm": 1.5590007305145264, + "learning_rate": 7.695e-06, + "loss": 0.1721, + "step": 2568 + }, + { + "epoch": 2.1697635135135136, + "grad_norm": 5.607897758483887, + "learning_rate": 7.698e-06, + "loss": 0.7733, + "step": 2569 + }, + { + "epoch": 2.170608108108108, + "grad_norm": 3.8319947719573975, + "learning_rate": 7.701e-06, + "loss": 0.3344, + "step": 2570 + }, + { + "epoch": 2.1714527027027026, + "grad_norm": 1.28493332862854, + "learning_rate": 7.704e-06, + "loss": 0.2449, + "step": 2571 + }, + { + "epoch": 2.1722972972972974, + "grad_norm": 1.115098237991333, + "learning_rate": 7.707000000000001e-06, + "loss": 0.1445, + "step": 2572 + }, + { + "epoch": 2.173141891891892, + "grad_norm": 0.861830472946167, + "learning_rate": 7.71e-06, + "loss": 0.1325, + "step": 2573 + }, + { + "epoch": 2.1739864864864864, + "grad_norm": 0.8973118662834167, + "learning_rate": 7.713e-06, + "loss": 0.1468, + "step": 2574 + }, + { + "epoch": 2.174831081081081, + "grad_norm": 2.733534574508667, + "learning_rate": 7.716e-06, + "loss": 0.1065, + "step": 2575 + }, + { + "epoch": 2.175675675675676, + "grad_norm": 1.9642083644866943, + "learning_rate": 7.719e-06, + "loss": 0.1066, + "step": 2576 + }, + { + "epoch": 2.17652027027027, + "grad_norm": 0.9541019201278687, + "learning_rate": 7.722e-06, + "loss": 0.1054, + "step": 2577 + }, + { + "epoch": 2.177364864864865, + "grad_norm": 3.6972007751464844, + "learning_rate": 7.725e-06, + "loss": 0.1249, + "step": 2578 + }, + { + "epoch": 2.1782094594594597, + "grad_norm": 2.580733299255371, + "learning_rate": 7.728e-06, + "loss": 0.1178, + "step": 2579 + }, + { + "epoch": 2.179054054054054, + "grad_norm": 0.7466360926628113, + "learning_rate": 7.731e-06, + "loss": 0.0864, + "step": 2580 + }, + { + "epoch": 2.1798986486486487, + "grad_norm": 1.031697154045105, + "learning_rate": 7.733999999999999e-06, + "loss": 0.1157, + "step": 2581 + }, + { + "epoch": 2.1807432432432434, + "grad_norm": 0.8405641317367554, + "learning_rate": 7.737e-06, + "loss": 0.1223, + "step": 2582 + }, + { + "epoch": 2.1815878378378377, + "grad_norm": 0.9979385733604431, + "learning_rate": 7.74e-06, + "loss": 0.0949, + "step": 2583 + }, + { + "epoch": 2.1824324324324325, + "grad_norm": 4.636630535125732, + "learning_rate": 7.743e-06, + "loss": 0.1208, + "step": 2584 + }, + { + "epoch": 2.183277027027027, + "grad_norm": 1.0172531604766846, + "learning_rate": 7.746e-06, + "loss": 0.1043, + "step": 2585 + }, + { + "epoch": 2.1841216216216215, + "grad_norm": 0.6751869320869446, + "learning_rate": 7.749e-06, + "loss": 0.0993, + "step": 2586 + }, + { + "epoch": 2.1849662162162162, + "grad_norm": 0.8238723278045654, + "learning_rate": 7.752000000000001e-06, + "loss": 0.1019, + "step": 2587 + }, + { + "epoch": 2.185810810810811, + "grad_norm": 0.990664005279541, + "learning_rate": 7.755000000000001e-06, + "loss": 0.1136, + "step": 2588 + }, + { + "epoch": 2.1866554054054053, + "grad_norm": 0.8416265845298767, + "learning_rate": 7.758000000000001e-06, + "loss": 0.0863, + "step": 2589 + }, + { + "epoch": 2.1875, + "grad_norm": 0.8971896767616272, + "learning_rate": 7.760999999999999e-06, + "loss": 0.096, + "step": 2590 + }, + { + "epoch": 2.1883445945945947, + "grad_norm": 0.9510109424591064, + "learning_rate": 7.763999999999999e-06, + "loss": 0.1077, + "step": 2591 + }, + { + "epoch": 2.189189189189189, + "grad_norm": 0.823485255241394, + "learning_rate": 7.767e-06, + "loss": 0.1027, + "step": 2592 + }, + { + "epoch": 2.1900337837837838, + "grad_norm": 0.767302930355072, + "learning_rate": 7.77e-06, + "loss": 0.0914, + "step": 2593 + }, + { + "epoch": 2.1908783783783785, + "grad_norm": 0.9800236225128174, + "learning_rate": 7.773e-06, + "loss": 0.112, + "step": 2594 + }, + { + "epoch": 2.191722972972973, + "grad_norm": 0.8976082801818848, + "learning_rate": 7.776e-06, + "loss": 0.094, + "step": 2595 + }, + { + "epoch": 2.1925675675675675, + "grad_norm": 0.9200021028518677, + "learning_rate": 7.779e-06, + "loss": 0.1055, + "step": 2596 + }, + { + "epoch": 2.1934121621621623, + "grad_norm": 1.0984203815460205, + "learning_rate": 7.782000000000001e-06, + "loss": 0.1079, + "step": 2597 + }, + { + "epoch": 2.1942567567567566, + "grad_norm": 0.9375830292701721, + "learning_rate": 7.785000000000001e-06, + "loss": 0.0859, + "step": 2598 + }, + { + "epoch": 2.1951013513513513, + "grad_norm": 0.7727369070053101, + "learning_rate": 7.788e-06, + "loss": 0.0783, + "step": 2599 + }, + { + "epoch": 2.195945945945946, + "grad_norm": 1.1611179113388062, + "learning_rate": 7.791e-06, + "loss": 0.0915, + "step": 2600 + }, + { + "epoch": 2.1967905405405403, + "grad_norm": 0.9479565620422363, + "learning_rate": 7.794e-06, + "loss": 0.0974, + "step": 2601 + }, + { + "epoch": 2.197635135135135, + "grad_norm": 0.8062677383422852, + "learning_rate": 7.797e-06, + "loss": 0.0939, + "step": 2602 + }, + { + "epoch": 2.19847972972973, + "grad_norm": 1.3339941501617432, + "learning_rate": 7.8e-06, + "loss": 0.1382, + "step": 2603 + }, + { + "epoch": 2.199324324324324, + "grad_norm": 1.113738775253296, + "learning_rate": 7.803e-06, + "loss": 0.1332, + "step": 2604 + }, + { + "epoch": 2.200168918918919, + "grad_norm": 0.9150586128234863, + "learning_rate": 7.806e-06, + "loss": 0.1135, + "step": 2605 + }, + { + "epoch": 2.2010135135135136, + "grad_norm": 1.0492433309555054, + "learning_rate": 7.809e-06, + "loss": 0.1046, + "step": 2606 + }, + { + "epoch": 2.201858108108108, + "grad_norm": 0.9371888041496277, + "learning_rate": 7.812e-06, + "loss": 0.0743, + "step": 2607 + }, + { + "epoch": 2.2027027027027026, + "grad_norm": 2.2903687953948975, + "learning_rate": 7.815e-06, + "loss": 0.1458, + "step": 2608 + }, + { + "epoch": 2.2035472972972974, + "grad_norm": 1.0072691440582275, + "learning_rate": 7.818e-06, + "loss": 0.092, + "step": 2609 + }, + { + "epoch": 2.204391891891892, + "grad_norm": 1.654893398284912, + "learning_rate": 7.821e-06, + "loss": 0.1121, + "step": 2610 + }, + { + "epoch": 2.2052364864864864, + "grad_norm": 0.9794138073921204, + "learning_rate": 7.824e-06, + "loss": 0.0932, + "step": 2611 + }, + { + "epoch": 2.206081081081081, + "grad_norm": 1.1821868419647217, + "learning_rate": 7.827000000000001e-06, + "loss": 0.1207, + "step": 2612 + }, + { + "epoch": 2.206925675675676, + "grad_norm": 1.3433302640914917, + "learning_rate": 7.830000000000001e-06, + "loss": 0.1147, + "step": 2613 + }, + { + "epoch": 2.20777027027027, + "grad_norm": 1.510356068611145, + "learning_rate": 7.833e-06, + "loss": 0.1678, + "step": 2614 + }, + { + "epoch": 2.208614864864865, + "grad_norm": 1.17335844039917, + "learning_rate": 7.836e-06, + "loss": 0.1271, + "step": 2615 + }, + { + "epoch": 2.2094594594594597, + "grad_norm": 1.4660074710845947, + "learning_rate": 7.838999999999999e-06, + "loss": 0.1506, + "step": 2616 + }, + { + "epoch": 2.210304054054054, + "grad_norm": 3.0228333473205566, + "learning_rate": 7.842e-06, + "loss": 0.3099, + "step": 2617 + }, + { + "epoch": 2.2111486486486487, + "grad_norm": 2.0910873413085938, + "learning_rate": 7.845e-06, + "loss": 0.2307, + "step": 2618 + }, + { + "epoch": 2.2119932432432434, + "grad_norm": 3.3535335063934326, + "learning_rate": 7.848e-06, + "loss": 0.5903, + "step": 2619 + }, + { + "epoch": 2.2128378378378377, + "grad_norm": 2.9510514736175537, + "learning_rate": 7.851e-06, + "loss": 0.3475, + "step": 2620 + }, + { + "epoch": 2.2136824324324325, + "grad_norm": 1.0739821195602417, + "learning_rate": 7.854e-06, + "loss": 0.1912, + "step": 2621 + }, + { + "epoch": 2.214527027027027, + "grad_norm": 0.7077192068099976, + "learning_rate": 7.857000000000001e-06, + "loss": 0.1369, + "step": 2622 + }, + { + "epoch": 2.2153716216216215, + "grad_norm": 1.8063815832138062, + "learning_rate": 7.860000000000001e-06, + "loss": 0.1261, + "step": 2623 + }, + { + "epoch": 2.2162162162162162, + "grad_norm": 4.852212905883789, + "learning_rate": 7.863e-06, + "loss": 0.128, + "step": 2624 + }, + { + "epoch": 2.217060810810811, + "grad_norm": 1.0120763778686523, + "learning_rate": 7.866e-06, + "loss": 0.1052, + "step": 2625 + }, + { + "epoch": 2.2179054054054053, + "grad_norm": 1.065406322479248, + "learning_rate": 7.868999999999999e-06, + "loss": 0.1277, + "step": 2626 + }, + { + "epoch": 2.21875, + "grad_norm": 0.9561744332313538, + "learning_rate": 7.872e-06, + "loss": 0.0973, + "step": 2627 + }, + { + "epoch": 2.2195945945945947, + "grad_norm": 1.240104079246521, + "learning_rate": 7.875e-06, + "loss": 0.1052, + "step": 2628 + }, + { + "epoch": 2.220439189189189, + "grad_norm": 1.5193520784378052, + "learning_rate": 7.878e-06, + "loss": 0.1144, + "step": 2629 + }, + { + "epoch": 2.2212837837837838, + "grad_norm": 1.2150659561157227, + "learning_rate": 7.881e-06, + "loss": 0.109, + "step": 2630 + }, + { + "epoch": 2.2221283783783785, + "grad_norm": 0.9558464288711548, + "learning_rate": 7.884e-06, + "loss": 0.0978, + "step": 2631 + }, + { + "epoch": 2.222972972972973, + "grad_norm": 1.0746357440948486, + "learning_rate": 7.887000000000001e-06, + "loss": 0.0944, + "step": 2632 + }, + { + "epoch": 2.2238175675675675, + "grad_norm": 6.144338607788086, + "learning_rate": 7.89e-06, + "loss": 0.1123, + "step": 2633 + }, + { + "epoch": 2.2246621621621623, + "grad_norm": 2.8782958984375, + "learning_rate": 7.893e-06, + "loss": 0.0896, + "step": 2634 + }, + { + "epoch": 2.2255067567567566, + "grad_norm": 0.8737870454788208, + "learning_rate": 7.896e-06, + "loss": 0.0809, + "step": 2635 + }, + { + "epoch": 2.2263513513513513, + "grad_norm": 1.1396763324737549, + "learning_rate": 7.899e-06, + "loss": 0.1022, + "step": 2636 + }, + { + "epoch": 2.227195945945946, + "grad_norm": 0.8646081686019897, + "learning_rate": 7.902000000000002e-06, + "loss": 0.0738, + "step": 2637 + }, + { + "epoch": 2.2280405405405403, + "grad_norm": 0.91187584400177, + "learning_rate": 7.905000000000001e-06, + "loss": 0.0916, + "step": 2638 + }, + { + "epoch": 2.228885135135135, + "grad_norm": 0.776462733745575, + "learning_rate": 7.908e-06, + "loss": 0.091, + "step": 2639 + }, + { + "epoch": 2.22972972972973, + "grad_norm": 1.241607666015625, + "learning_rate": 7.911e-06, + "loss": 0.101, + "step": 2640 + }, + { + "epoch": 2.230574324324324, + "grad_norm": 0.9980747103691101, + "learning_rate": 7.913999999999999e-06, + "loss": 0.117, + "step": 2641 + }, + { + "epoch": 2.231418918918919, + "grad_norm": 1.773608922958374, + "learning_rate": 7.917e-06, + "loss": 0.1862, + "step": 2642 + }, + { + "epoch": 2.2322635135135136, + "grad_norm": 1.0816127061843872, + "learning_rate": 7.92e-06, + "loss": 0.1249, + "step": 2643 + }, + { + "epoch": 2.233108108108108, + "grad_norm": 0.7111592292785645, + "learning_rate": 7.923e-06, + "loss": 0.0793, + "step": 2644 + }, + { + "epoch": 2.2339527027027026, + "grad_norm": 1.3400459289550781, + "learning_rate": 7.926e-06, + "loss": 0.1391, + "step": 2645 + }, + { + "epoch": 2.2347972972972974, + "grad_norm": 0.9135341048240662, + "learning_rate": 7.929e-06, + "loss": 0.0881, + "step": 2646 + }, + { + "epoch": 2.235641891891892, + "grad_norm": 0.8017033934593201, + "learning_rate": 7.932000000000001e-06, + "loss": 0.0966, + "step": 2647 + }, + { + "epoch": 2.2364864864864864, + "grad_norm": 0.9671177864074707, + "learning_rate": 7.935000000000001e-06, + "loss": 0.1088, + "step": 2648 + }, + { + "epoch": 2.237331081081081, + "grad_norm": 0.9669617414474487, + "learning_rate": 7.938000000000001e-06, + "loss": 0.0885, + "step": 2649 + }, + { + "epoch": 2.238175675675676, + "grad_norm": 0.798901379108429, + "learning_rate": 7.941e-06, + "loss": 0.1034, + "step": 2650 + }, + { + "epoch": 2.23902027027027, + "grad_norm": 1.0171564817428589, + "learning_rate": 7.943999999999999e-06, + "loss": 0.1023, + "step": 2651 + }, + { + "epoch": 2.239864864864865, + "grad_norm": 1.3991564512252808, + "learning_rate": 7.947e-06, + "loss": 0.0913, + "step": 2652 + }, + { + "epoch": 2.2407094594594597, + "grad_norm": 1.1625481843948364, + "learning_rate": 7.95e-06, + "loss": 0.1027, + "step": 2653 + }, + { + "epoch": 2.241554054054054, + "grad_norm": 0.9092011451721191, + "learning_rate": 7.953e-06, + "loss": 0.0962, + "step": 2654 + }, + { + "epoch": 2.2423986486486487, + "grad_norm": 1.112504005432129, + "learning_rate": 7.956e-06, + "loss": 0.095, + "step": 2655 + }, + { + "epoch": 2.2432432432432434, + "grad_norm": 1.3686498403549194, + "learning_rate": 7.959e-06, + "loss": 0.1405, + "step": 2656 + }, + { + "epoch": 2.2440878378378377, + "grad_norm": 1.5349023342132568, + "learning_rate": 7.962000000000001e-06, + "loss": 0.079, + "step": 2657 + }, + { + "epoch": 2.2449324324324325, + "grad_norm": 1.433780312538147, + "learning_rate": 7.965e-06, + "loss": 0.098, + "step": 2658 + }, + { + "epoch": 2.245777027027027, + "grad_norm": 1.039743185043335, + "learning_rate": 7.968e-06, + "loss": 0.1074, + "step": 2659 + }, + { + "epoch": 2.2466216216216215, + "grad_norm": 0.9581905007362366, + "learning_rate": 7.971e-06, + "loss": 0.0931, + "step": 2660 + }, + { + "epoch": 2.2474662162162162, + "grad_norm": 0.9505957961082458, + "learning_rate": 7.974e-06, + "loss": 0.0949, + "step": 2661 + }, + { + "epoch": 2.248310810810811, + "grad_norm": 1.190417766571045, + "learning_rate": 7.977000000000002e-06, + "loss": 0.1085, + "step": 2662 + }, + { + "epoch": 2.2491554054054053, + "grad_norm": 1.3247454166412354, + "learning_rate": 7.98e-06, + "loss": 0.1232, + "step": 2663 + }, + { + "epoch": 2.25, + "grad_norm": 1.5240061283111572, + "learning_rate": 7.983e-06, + "loss": 0.1414, + "step": 2664 + }, + { + "epoch": 2.2508445945945947, + "grad_norm": 1.5020310878753662, + "learning_rate": 7.986e-06, + "loss": 0.1437, + "step": 2665 + }, + { + "epoch": 2.251689189189189, + "grad_norm": 1.5662449598312378, + "learning_rate": 7.989e-06, + "loss": 0.151, + "step": 2666 + }, + { + "epoch": 2.2525337837837838, + "grad_norm": 2.115277051925659, + "learning_rate": 7.992e-06, + "loss": 0.1484, + "step": 2667 + }, + { + "epoch": 2.2533783783783785, + "grad_norm": 3.2109591960906982, + "learning_rate": 7.995e-06, + "loss": 0.1508, + "step": 2668 + }, + { + "epoch": 2.254222972972973, + "grad_norm": 3.4447550773620605, + "learning_rate": 7.998e-06, + "loss": 0.6564, + "step": 2669 + }, + { + "epoch": 2.2550675675675675, + "grad_norm": 2.718384027481079, + "learning_rate": 8.001e-06, + "loss": 0.5088, + "step": 2670 + }, + { + "epoch": 2.2559121621621623, + "grad_norm": 1.961025357246399, + "learning_rate": 8.004e-06, + "loss": 0.2281, + "step": 2671 + }, + { + "epoch": 2.2567567567567566, + "grad_norm": 7.5757527351379395, + "learning_rate": 8.007000000000001e-06, + "loss": 0.1236, + "step": 2672 + }, + { + "epoch": 2.2576013513513513, + "grad_norm": 1.2286871671676636, + "learning_rate": 8.010000000000001e-06, + "loss": 0.1382, + "step": 2673 + }, + { + "epoch": 2.258445945945946, + "grad_norm": 1.0124045610427856, + "learning_rate": 8.013000000000001e-06, + "loss": 0.1132, + "step": 2674 + }, + { + "epoch": 2.2592905405405403, + "grad_norm": 1.2788454294204712, + "learning_rate": 8.016e-06, + "loss": 0.0961, + "step": 2675 + }, + { + "epoch": 2.260135135135135, + "grad_norm": 1.2783410549163818, + "learning_rate": 8.018999999999999e-06, + "loss": 0.1066, + "step": 2676 + }, + { + "epoch": 2.26097972972973, + "grad_norm": 1.1290473937988281, + "learning_rate": 8.022e-06, + "loss": 0.1044, + "step": 2677 + }, + { + "epoch": 2.2618243243243246, + "grad_norm": 1.8135031461715698, + "learning_rate": 8.025e-06, + "loss": 0.0784, + "step": 2678 + }, + { + "epoch": 2.262668918918919, + "grad_norm": 0.7140819430351257, + "learning_rate": 8.028e-06, + "loss": 0.0745, + "step": 2679 + }, + { + "epoch": 2.2635135135135136, + "grad_norm": 0.8922361731529236, + "learning_rate": 8.031e-06, + "loss": 0.0819, + "step": 2680 + }, + { + "epoch": 2.264358108108108, + "grad_norm": 0.9589013457298279, + "learning_rate": 8.034e-06, + "loss": 0.1091, + "step": 2681 + }, + { + "epoch": 2.2652027027027026, + "grad_norm": 1.0460244417190552, + "learning_rate": 8.037000000000001e-06, + "loss": 0.0879, + "step": 2682 + }, + { + "epoch": 2.2660472972972974, + "grad_norm": 1.0182983875274658, + "learning_rate": 8.040000000000001e-06, + "loss": 0.1076, + "step": 2683 + }, + { + "epoch": 2.266891891891892, + "grad_norm": 0.8952338695526123, + "learning_rate": 8.043e-06, + "loss": 0.1047, + "step": 2684 + }, + { + "epoch": 2.2677364864864864, + "grad_norm": 1.0370891094207764, + "learning_rate": 8.046e-06, + "loss": 0.1097, + "step": 2685 + }, + { + "epoch": 2.268581081081081, + "grad_norm": 0.9598262310028076, + "learning_rate": 8.049e-06, + "loss": 0.086, + "step": 2686 + }, + { + "epoch": 2.2694256756756754, + "grad_norm": 0.9878175258636475, + "learning_rate": 8.052000000000002e-06, + "loss": 0.0986, + "step": 2687 + }, + { + "epoch": 2.27027027027027, + "grad_norm": 0.8768099546432495, + "learning_rate": 8.055e-06, + "loss": 0.1001, + "step": 2688 + }, + { + "epoch": 2.271114864864865, + "grad_norm": 0.8740895390510559, + "learning_rate": 8.058e-06, + "loss": 0.103, + "step": 2689 + }, + { + "epoch": 2.2719594594594597, + "grad_norm": 0.9545888304710388, + "learning_rate": 8.061e-06, + "loss": 0.0786, + "step": 2690 + }, + { + "epoch": 2.272804054054054, + "grad_norm": 1.1333321332931519, + "learning_rate": 8.064e-06, + "loss": 0.099, + "step": 2691 + }, + { + "epoch": 2.2736486486486487, + "grad_norm": 0.9183576703071594, + "learning_rate": 8.067e-06, + "loss": 0.1047, + "step": 2692 + }, + { + "epoch": 2.2744932432432434, + "grad_norm": 0.8504357933998108, + "learning_rate": 8.07e-06, + "loss": 0.0837, + "step": 2693 + }, + { + "epoch": 2.2753378378378377, + "grad_norm": 1.1326122283935547, + "learning_rate": 8.073e-06, + "loss": 0.095, + "step": 2694 + }, + { + "epoch": 2.2761824324324325, + "grad_norm": 1.1189815998077393, + "learning_rate": 8.076e-06, + "loss": 0.1607, + "step": 2695 + }, + { + "epoch": 2.277027027027027, + "grad_norm": 0.9831520915031433, + "learning_rate": 8.079e-06, + "loss": 0.1095, + "step": 2696 + }, + { + "epoch": 2.2778716216216215, + "grad_norm": 0.8771715760231018, + "learning_rate": 8.082e-06, + "loss": 0.0929, + "step": 2697 + }, + { + "epoch": 2.2787162162162162, + "grad_norm": 0.7680870294570923, + "learning_rate": 8.085000000000001e-06, + "loss": 0.0899, + "step": 2698 + }, + { + "epoch": 2.279560810810811, + "grad_norm": 0.8084818720817566, + "learning_rate": 8.088000000000001e-06, + "loss": 0.1037, + "step": 2699 + }, + { + "epoch": 2.2804054054054053, + "grad_norm": 0.9247319102287292, + "learning_rate": 8.091e-06, + "loss": 0.1112, + "step": 2700 + }, + { + "epoch": 2.28125, + "grad_norm": 1.06620454788208, + "learning_rate": 8.093999999999999e-06, + "loss": 0.1006, + "step": 2701 + }, + { + "epoch": 2.2820945945945947, + "grad_norm": 1.0395015478134155, + "learning_rate": 8.096999999999999e-06, + "loss": 0.0883, + "step": 2702 + }, + { + "epoch": 2.282939189189189, + "grad_norm": 0.8615615367889404, + "learning_rate": 8.1e-06, + "loss": 0.0877, + "step": 2703 + }, + { + "epoch": 2.2837837837837838, + "grad_norm": 1.6651079654693604, + "learning_rate": 8.103e-06, + "loss": 0.1095, + "step": 2704 + }, + { + "epoch": 2.2846283783783785, + "grad_norm": 1.0647108554840088, + "learning_rate": 8.106e-06, + "loss": 0.1237, + "step": 2705 + }, + { + "epoch": 2.285472972972973, + "grad_norm": 1.4106805324554443, + "learning_rate": 8.109e-06, + "loss": 0.0903, + "step": 2706 + }, + { + "epoch": 2.2863175675675675, + "grad_norm": 0.9897668957710266, + "learning_rate": 8.112e-06, + "loss": 0.1094, + "step": 2707 + }, + { + "epoch": 2.2871621621621623, + "grad_norm": 1.639352560043335, + "learning_rate": 8.115000000000001e-06, + "loss": 0.1251, + "step": 2708 + }, + { + "epoch": 2.2880067567567566, + "grad_norm": 1.0216383934020996, + "learning_rate": 8.118000000000001e-06, + "loss": 0.1024, + "step": 2709 + }, + { + "epoch": 2.2888513513513513, + "grad_norm": 1.0540739297866821, + "learning_rate": 8.121e-06, + "loss": 0.1019, + "step": 2710 + }, + { + "epoch": 2.289695945945946, + "grad_norm": 1.293045997619629, + "learning_rate": 8.124e-06, + "loss": 0.1187, + "step": 2711 + }, + { + "epoch": 2.2905405405405403, + "grad_norm": 1.0544886589050293, + "learning_rate": 8.126999999999999e-06, + "loss": 0.1217, + "step": 2712 + }, + { + "epoch": 2.291385135135135, + "grad_norm": 1.2298638820648193, + "learning_rate": 8.13e-06, + "loss": 0.1276, + "step": 2713 + }, + { + "epoch": 2.29222972972973, + "grad_norm": 1.0725370645523071, + "learning_rate": 8.133e-06, + "loss": 0.106, + "step": 2714 + }, + { + "epoch": 2.2930743243243246, + "grad_norm": 0.922572672367096, + "learning_rate": 8.136e-06, + "loss": 0.0943, + "step": 2715 + }, + { + "epoch": 2.293918918918919, + "grad_norm": 1.3496730327606201, + "learning_rate": 8.139e-06, + "loss": 0.1258, + "step": 2716 + }, + { + "epoch": 2.2947635135135136, + "grad_norm": 1.362781047821045, + "learning_rate": 8.142e-06, + "loss": 0.1404, + "step": 2717 + }, + { + "epoch": 2.295608108108108, + "grad_norm": 2.9579150676727295, + "learning_rate": 8.145e-06, + "loss": 0.174, + "step": 2718 + }, + { + "epoch": 2.2964527027027026, + "grad_norm": 3.4858529567718506, + "learning_rate": 8.148e-06, + "loss": 0.5305, + "step": 2719 + }, + { + "epoch": 2.2972972972972974, + "grad_norm": 1.5865174531936646, + "learning_rate": 8.151e-06, + "loss": 0.265, + "step": 2720 + }, + { + "epoch": 2.298141891891892, + "grad_norm": 1.7080172300338745, + "learning_rate": 8.154e-06, + "loss": 0.2237, + "step": 2721 + }, + { + "epoch": 2.2989864864864864, + "grad_norm": 1.2579463720321655, + "learning_rate": 8.157e-06, + "loss": 0.106, + "step": 2722 + }, + { + "epoch": 2.299831081081081, + "grad_norm": 1.3976686000823975, + "learning_rate": 8.160000000000001e-06, + "loss": 0.1216, + "step": 2723 + }, + { + "epoch": 2.3006756756756754, + "grad_norm": 1.0131584405899048, + "learning_rate": 8.163000000000001e-06, + "loss": 0.1158, + "step": 2724 + }, + { + "epoch": 2.30152027027027, + "grad_norm": 0.7634105682373047, + "learning_rate": 8.166e-06, + "loss": 0.0972, + "step": 2725 + }, + { + "epoch": 2.302364864864865, + "grad_norm": 1.0273528099060059, + "learning_rate": 8.169e-06, + "loss": 0.0983, + "step": 2726 + }, + { + "epoch": 2.3032094594594597, + "grad_norm": 0.9842183589935303, + "learning_rate": 8.171999999999999e-06, + "loss": 0.1114, + "step": 2727 + }, + { + "epoch": 2.304054054054054, + "grad_norm": 0.750312328338623, + "learning_rate": 8.175e-06, + "loss": 0.0712, + "step": 2728 + }, + { + "epoch": 2.3048986486486487, + "grad_norm": 0.8210047483444214, + "learning_rate": 8.178e-06, + "loss": 0.0776, + "step": 2729 + }, + { + "epoch": 2.3057432432432434, + "grad_norm": 0.896247923374176, + "learning_rate": 8.181e-06, + "loss": 0.113, + "step": 2730 + }, + { + "epoch": 2.3065878378378377, + "grad_norm": 0.874345064163208, + "learning_rate": 8.184e-06, + "loss": 0.1138, + "step": 2731 + }, + { + "epoch": 2.3074324324324325, + "grad_norm": 0.8903641700744629, + "learning_rate": 8.187e-06, + "loss": 0.0958, + "step": 2732 + }, + { + "epoch": 2.308277027027027, + "grad_norm": 0.8338792324066162, + "learning_rate": 8.190000000000001e-06, + "loss": 0.0683, + "step": 2733 + }, + { + "epoch": 2.3091216216216215, + "grad_norm": 4.342731475830078, + "learning_rate": 8.193000000000001e-06, + "loss": 0.0885, + "step": 2734 + }, + { + "epoch": 2.3099662162162162, + "grad_norm": 1.0704848766326904, + "learning_rate": 8.196e-06, + "loss": 0.1125, + "step": 2735 + }, + { + "epoch": 2.310810810810811, + "grad_norm": 2.166290760040283, + "learning_rate": 8.199e-06, + "loss": 0.0798, + "step": 2736 + }, + { + "epoch": 2.3116554054054053, + "grad_norm": 0.8103851675987244, + "learning_rate": 8.201999999999999e-06, + "loss": 0.0916, + "step": 2737 + }, + { + "epoch": 2.3125, + "grad_norm": 0.8793722987174988, + "learning_rate": 8.205e-06, + "loss": 0.1131, + "step": 2738 + }, + { + "epoch": 2.3133445945945947, + "grad_norm": 1.4591559171676636, + "learning_rate": 8.208e-06, + "loss": 0.1198, + "step": 2739 + }, + { + "epoch": 2.314189189189189, + "grad_norm": 2.451566457748413, + "learning_rate": 8.211e-06, + "loss": 0.1104, + "step": 2740 + }, + { + "epoch": 2.3150337837837838, + "grad_norm": 0.6792970895767212, + "learning_rate": 8.214e-06, + "loss": 0.0763, + "step": 2741 + }, + { + "epoch": 2.3158783783783785, + "grad_norm": 0.9662771821022034, + "learning_rate": 8.217e-06, + "loss": 0.092, + "step": 2742 + }, + { + "epoch": 2.316722972972973, + "grad_norm": 1.4709724187850952, + "learning_rate": 8.220000000000001e-06, + "loss": 0.0925, + "step": 2743 + }, + { + "epoch": 2.3175675675675675, + "grad_norm": 0.7626177072525024, + "learning_rate": 8.223e-06, + "loss": 0.105, + "step": 2744 + }, + { + "epoch": 2.3184121621621623, + "grad_norm": 1.0890642404556274, + "learning_rate": 8.226e-06, + "loss": 0.0975, + "step": 2745 + }, + { + "epoch": 2.3192567567567566, + "grad_norm": 0.8476321697235107, + "learning_rate": 8.229e-06, + "loss": 0.071, + "step": 2746 + }, + { + "epoch": 2.3201013513513513, + "grad_norm": 0.9988681674003601, + "learning_rate": 8.232e-06, + "loss": 0.1039, + "step": 2747 + }, + { + "epoch": 2.320945945945946, + "grad_norm": 1.0009852647781372, + "learning_rate": 8.235000000000002e-06, + "loss": 0.0968, + "step": 2748 + }, + { + "epoch": 2.3217905405405403, + "grad_norm": 0.8751891255378723, + "learning_rate": 8.238e-06, + "loss": 0.0998, + "step": 2749 + }, + { + "epoch": 2.322635135135135, + "grad_norm": 1.210060477256775, + "learning_rate": 8.241e-06, + "loss": 0.1224, + "step": 2750 + }, + { + "epoch": 2.32347972972973, + "grad_norm": 0.9190838932991028, + "learning_rate": 8.244e-06, + "loss": 0.0664, + "step": 2751 + }, + { + "epoch": 2.3243243243243246, + "grad_norm": 1.0880740880966187, + "learning_rate": 8.246999999999999e-06, + "loss": 0.1033, + "step": 2752 + }, + { + "epoch": 2.325168918918919, + "grad_norm": 0.7935388088226318, + "learning_rate": 8.25e-06, + "loss": 0.0973, + "step": 2753 + }, + { + "epoch": 2.3260135135135136, + "grad_norm": 1.0482298135757446, + "learning_rate": 8.253e-06, + "loss": 0.1291, + "step": 2754 + }, + { + "epoch": 2.326858108108108, + "grad_norm": 1.1992202997207642, + "learning_rate": 8.256e-06, + "loss": 0.1211, + "step": 2755 + }, + { + "epoch": 2.3277027027027026, + "grad_norm": 0.949406623840332, + "learning_rate": 8.259e-06, + "loss": 0.0914, + "step": 2756 + }, + { + "epoch": 2.3285472972972974, + "grad_norm": 0.7134950160980225, + "learning_rate": 8.262e-06, + "loss": 0.0792, + "step": 2757 + }, + { + "epoch": 2.329391891891892, + "grad_norm": 0.9578901529312134, + "learning_rate": 8.265000000000001e-06, + "loss": 0.0724, + "step": 2758 + }, + { + "epoch": 2.3302364864864864, + "grad_norm": 1.0454238653182983, + "learning_rate": 8.268000000000001e-06, + "loss": 0.1162, + "step": 2759 + }, + { + "epoch": 2.331081081081081, + "grad_norm": 1.005331039428711, + "learning_rate": 8.271000000000001e-06, + "loss": 0.1054, + "step": 2760 + }, + { + "epoch": 2.3319256756756754, + "grad_norm": 1.1921268701553345, + "learning_rate": 8.274e-06, + "loss": 0.1299, + "step": 2761 + }, + { + "epoch": 2.33277027027027, + "grad_norm": 1.2583194971084595, + "learning_rate": 8.276999999999999e-06, + "loss": 0.145, + "step": 2762 + }, + { + "epoch": 2.333614864864865, + "grad_norm": 1.100099802017212, + "learning_rate": 8.28e-06, + "loss": 0.1292, + "step": 2763 + }, + { + "epoch": 2.3344594594594597, + "grad_norm": 1.2286498546600342, + "learning_rate": 8.283e-06, + "loss": 0.1357, + "step": 2764 + }, + { + "epoch": 2.335304054054054, + "grad_norm": 1.334963083267212, + "learning_rate": 8.286e-06, + "loss": 0.1331, + "step": 2765 + }, + { + "epoch": 2.3361486486486487, + "grad_norm": 1.4435138702392578, + "learning_rate": 8.289e-06, + "loss": 0.1253, + "step": 2766 + }, + { + "epoch": 2.3369932432432434, + "grad_norm": 1.648334264755249, + "learning_rate": 8.292e-06, + "loss": 0.139, + "step": 2767 + }, + { + "epoch": 2.3378378378378377, + "grad_norm": 1.9227856397628784, + "learning_rate": 8.295000000000001e-06, + "loss": 0.193, + "step": 2768 + }, + { + "epoch": 2.3386824324324325, + "grad_norm": 3.788438558578491, + "learning_rate": 8.298000000000001e-06, + "loss": 0.7005, + "step": 2769 + }, + { + "epoch": 2.339527027027027, + "grad_norm": 12.033964157104492, + "learning_rate": 8.301e-06, + "loss": 0.593, + "step": 2770 + }, + { + "epoch": 2.3403716216216215, + "grad_norm": 2.222529888153076, + "learning_rate": 8.304e-06, + "loss": 0.2899, + "step": 2771 + }, + { + "epoch": 2.3412162162162162, + "grad_norm": 1.7972967624664307, + "learning_rate": 8.307e-06, + "loss": 0.1815, + "step": 2772 + }, + { + "epoch": 2.342060810810811, + "grad_norm": 1.5719980001449585, + "learning_rate": 8.310000000000002e-06, + "loss": 0.1023, + "step": 2773 + }, + { + "epoch": 2.3429054054054053, + "grad_norm": 0.9047634601593018, + "learning_rate": 8.313e-06, + "loss": 0.1343, + "step": 2774 + }, + { + "epoch": 2.34375, + "grad_norm": 0.7445411682128906, + "learning_rate": 8.316e-06, + "loss": 0.1031, + "step": 2775 + }, + { + "epoch": 2.3445945945945947, + "grad_norm": 0.9807145595550537, + "learning_rate": 8.319e-06, + "loss": 0.1132, + "step": 2776 + }, + { + "epoch": 2.345439189189189, + "grad_norm": 0.8879412412643433, + "learning_rate": 8.322e-06, + "loss": 0.11, + "step": 2777 + }, + { + "epoch": 2.3462837837837838, + "grad_norm": 0.8261304497718811, + "learning_rate": 8.325e-06, + "loss": 0.101, + "step": 2778 + }, + { + "epoch": 2.3471283783783785, + "grad_norm": 1.021418571472168, + "learning_rate": 8.328e-06, + "loss": 0.1192, + "step": 2779 + }, + { + "epoch": 2.347972972972973, + "grad_norm": 0.7942743897438049, + "learning_rate": 8.331e-06, + "loss": 0.1076, + "step": 2780 + }, + { + "epoch": 2.3488175675675675, + "grad_norm": 0.9024556875228882, + "learning_rate": 8.334e-06, + "loss": 0.1107, + "step": 2781 + }, + { + "epoch": 2.3496621621621623, + "grad_norm": 1.1336417198181152, + "learning_rate": 8.337e-06, + "loss": 0.1232, + "step": 2782 + }, + { + "epoch": 2.3505067567567566, + "grad_norm": 0.7442722320556641, + "learning_rate": 8.340000000000001e-06, + "loss": 0.0928, + "step": 2783 + }, + { + "epoch": 2.3513513513513513, + "grad_norm": 0.8218228816986084, + "learning_rate": 8.343000000000001e-06, + "loss": 0.0914, + "step": 2784 + }, + { + "epoch": 2.352195945945946, + "grad_norm": 0.8733848333358765, + "learning_rate": 8.346000000000001e-06, + "loss": 0.0991, + "step": 2785 + }, + { + "epoch": 2.3530405405405403, + "grad_norm": 0.7552855610847473, + "learning_rate": 8.349e-06, + "loss": 0.101, + "step": 2786 + }, + { + "epoch": 2.353885135135135, + "grad_norm": 0.6517903804779053, + "learning_rate": 8.351999999999999e-06, + "loss": 0.0726, + "step": 2787 + }, + { + "epoch": 2.35472972972973, + "grad_norm": 0.9354560375213623, + "learning_rate": 8.355e-06, + "loss": 0.1098, + "step": 2788 + }, + { + "epoch": 2.3555743243243246, + "grad_norm": 1.0363095998764038, + "learning_rate": 8.358e-06, + "loss": 0.1161, + "step": 2789 + }, + { + "epoch": 2.356418918918919, + "grad_norm": 0.9048130512237549, + "learning_rate": 8.361e-06, + "loss": 0.0954, + "step": 2790 + }, + { + "epoch": 2.3572635135135136, + "grad_norm": 7.103671550750732, + "learning_rate": 8.364e-06, + "loss": 0.0833, + "step": 2791 + }, + { + "epoch": 2.358108108108108, + "grad_norm": 0.9376493096351624, + "learning_rate": 8.367e-06, + "loss": 0.0935, + "step": 2792 + }, + { + "epoch": 2.3589527027027026, + "grad_norm": 0.9662079811096191, + "learning_rate": 8.370000000000001e-06, + "loss": 0.1274, + "step": 2793 + }, + { + "epoch": 2.3597972972972974, + "grad_norm": 1.0749459266662598, + "learning_rate": 8.373000000000001e-06, + "loss": 0.0795, + "step": 2794 + }, + { + "epoch": 2.360641891891892, + "grad_norm": 0.9866902828216553, + "learning_rate": 8.376e-06, + "loss": 0.1005, + "step": 2795 + }, + { + "epoch": 2.3614864864864864, + "grad_norm": 1.4175554513931274, + "learning_rate": 8.379e-06, + "loss": 0.0972, + "step": 2796 + }, + { + "epoch": 2.362331081081081, + "grad_norm": 0.8805397152900696, + "learning_rate": 8.382e-06, + "loss": 0.0919, + "step": 2797 + }, + { + "epoch": 2.3631756756756754, + "grad_norm": 0.7775231599807739, + "learning_rate": 8.385e-06, + "loss": 0.0716, + "step": 2798 + }, + { + "epoch": 2.36402027027027, + "grad_norm": 1.127677083015442, + "learning_rate": 8.388e-06, + "loss": 0.0722, + "step": 2799 + }, + { + "epoch": 2.364864864864865, + "grad_norm": 1.2382715940475464, + "learning_rate": 8.391e-06, + "loss": 0.1288, + "step": 2800 + }, + { + "epoch": 2.3657094594594597, + "grad_norm": 0.9251129627227783, + "learning_rate": 8.394e-06, + "loss": 0.0691, + "step": 2801 + }, + { + "epoch": 2.366554054054054, + "grad_norm": 0.8146246671676636, + "learning_rate": 8.397e-06, + "loss": 0.088, + "step": 2802 + }, + { + "epoch": 2.3673986486486487, + "grad_norm": 0.877099871635437, + "learning_rate": 8.400000000000001e-06, + "loss": 0.0994, + "step": 2803 + }, + { + "epoch": 2.3682432432432434, + "grad_norm": 1.0455164909362793, + "learning_rate": 8.403e-06, + "loss": 0.1111, + "step": 2804 + }, + { + "epoch": 2.3690878378378377, + "grad_norm": 1.0068439245224, + "learning_rate": 8.406e-06, + "loss": 0.1093, + "step": 2805 + }, + { + "epoch": 2.3699324324324325, + "grad_norm": 0.8505437970161438, + "learning_rate": 8.409e-06, + "loss": 0.0989, + "step": 2806 + }, + { + "epoch": 2.370777027027027, + "grad_norm": 1.0476340055465698, + "learning_rate": 8.412e-06, + "loss": 0.121, + "step": 2807 + }, + { + "epoch": 2.3716216216216215, + "grad_norm": 1.4054499864578247, + "learning_rate": 8.415000000000002e-06, + "loss": 0.1482, + "step": 2808 + }, + { + "epoch": 2.3724662162162162, + "grad_norm": 1.3297404050827026, + "learning_rate": 8.418000000000001e-06, + "loss": 0.1156, + "step": 2809 + }, + { + "epoch": 2.373310810810811, + "grad_norm": 1.204207420349121, + "learning_rate": 8.421000000000001e-06, + "loss": 0.1096, + "step": 2810 + }, + { + "epoch": 2.3741554054054053, + "grad_norm": 0.8064358234405518, + "learning_rate": 8.424e-06, + "loss": 0.0852, + "step": 2811 + }, + { + "epoch": 2.375, + "grad_norm": 1.222520112991333, + "learning_rate": 8.426999999999999e-06, + "loss": 0.1061, + "step": 2812 + }, + { + "epoch": 2.3758445945945947, + "grad_norm": 0.9922530055046082, + "learning_rate": 8.43e-06, + "loss": 0.1051, + "step": 2813 + }, + { + "epoch": 2.376689189189189, + "grad_norm": 1.0725584030151367, + "learning_rate": 8.433e-06, + "loss": 0.0912, + "step": 2814 + }, + { + "epoch": 2.3775337837837838, + "grad_norm": 1.3701987266540527, + "learning_rate": 8.436e-06, + "loss": 0.1504, + "step": 2815 + }, + { + "epoch": 2.3783783783783785, + "grad_norm": 1.7596153020858765, + "learning_rate": 8.439e-06, + "loss": 0.1754, + "step": 2816 + }, + { + "epoch": 2.379222972972973, + "grad_norm": 1.4774951934814453, + "learning_rate": 8.442e-06, + "loss": 0.1733, + "step": 2817 + }, + { + "epoch": 2.3800675675675675, + "grad_norm": 2.366161346435547, + "learning_rate": 8.445e-06, + "loss": 0.3092, + "step": 2818 + }, + { + "epoch": 2.3809121621621623, + "grad_norm": 2.4383397102355957, + "learning_rate": 8.448000000000001e-06, + "loss": 0.6315, + "step": 2819 + }, + { + "epoch": 2.3817567567567566, + "grad_norm": 2.760587692260742, + "learning_rate": 8.451000000000001e-06, + "loss": 0.3616, + "step": 2820 + }, + { + "epoch": 2.3826013513513513, + "grad_norm": 1.1149334907531738, + "learning_rate": 8.454e-06, + "loss": 0.1936, + "step": 2821 + }, + { + "epoch": 2.383445945945946, + "grad_norm": 1.1735246181488037, + "learning_rate": 8.457e-06, + "loss": 0.1233, + "step": 2822 + }, + { + "epoch": 2.3842905405405403, + "grad_norm": 0.9356536269187927, + "learning_rate": 8.459999999999999e-06, + "loss": 0.1101, + "step": 2823 + }, + { + "epoch": 2.385135135135135, + "grad_norm": 0.890210747718811, + "learning_rate": 8.463e-06, + "loss": 0.1283, + "step": 2824 + }, + { + "epoch": 2.38597972972973, + "grad_norm": 0.8064762353897095, + "learning_rate": 8.466e-06, + "loss": 0.0975, + "step": 2825 + }, + { + "epoch": 2.3868243243243246, + "grad_norm": 1.0413124561309814, + "learning_rate": 8.469e-06, + "loss": 0.0885, + "step": 2826 + }, + { + "epoch": 2.387668918918919, + "grad_norm": 0.799299418926239, + "learning_rate": 8.472e-06, + "loss": 0.0969, + "step": 2827 + }, + { + "epoch": 2.3885135135135136, + "grad_norm": 0.8139575123786926, + "learning_rate": 8.475e-06, + "loss": 0.0953, + "step": 2828 + }, + { + "epoch": 2.389358108108108, + "grad_norm": 0.9388561844825745, + "learning_rate": 8.478e-06, + "loss": 0.1041, + "step": 2829 + }, + { + "epoch": 2.3902027027027026, + "grad_norm": 0.6105353236198425, + "learning_rate": 8.481e-06, + "loss": 0.0556, + "step": 2830 + }, + { + "epoch": 2.3910472972972974, + "grad_norm": 0.9432845711708069, + "learning_rate": 8.484e-06, + "loss": 0.1105, + "step": 2831 + }, + { + "epoch": 2.391891891891892, + "grad_norm": 0.8108059167861938, + "learning_rate": 8.487e-06, + "loss": 0.0954, + "step": 2832 + }, + { + "epoch": 2.3927364864864864, + "grad_norm": 0.7514955997467041, + "learning_rate": 8.49e-06, + "loss": 0.085, + "step": 2833 + }, + { + "epoch": 2.393581081081081, + "grad_norm": 1.466284155845642, + "learning_rate": 8.493000000000002e-06, + "loss": 0.0888, + "step": 2834 + }, + { + "epoch": 2.3944256756756754, + "grad_norm": 0.790127694606781, + "learning_rate": 8.496e-06, + "loss": 0.0883, + "step": 2835 + }, + { + "epoch": 2.39527027027027, + "grad_norm": 0.8282477259635925, + "learning_rate": 8.499e-06, + "loss": 0.1124, + "step": 2836 + }, + { + "epoch": 2.396114864864865, + "grad_norm": 1.3115862607955933, + "learning_rate": 8.502e-06, + "loss": 0.091, + "step": 2837 + }, + { + "epoch": 2.3969594594594597, + "grad_norm": 0.905318558216095, + "learning_rate": 8.504999999999999e-06, + "loss": 0.0674, + "step": 2838 + }, + { + "epoch": 2.397804054054054, + "grad_norm": 1.0962927341461182, + "learning_rate": 8.508e-06, + "loss": 0.1066, + "step": 2839 + }, + { + "epoch": 2.3986486486486487, + "grad_norm": 0.645790696144104, + "learning_rate": 8.511e-06, + "loss": 0.0807, + "step": 2840 + }, + { + "epoch": 2.3994932432432434, + "grad_norm": 1.1231127977371216, + "learning_rate": 8.514e-06, + "loss": 0.1057, + "step": 2841 + }, + { + "epoch": 2.4003378378378377, + "grad_norm": 1.1298805475234985, + "learning_rate": 8.517e-06, + "loss": 0.124, + "step": 2842 + }, + { + "epoch": 2.4011824324324325, + "grad_norm": 0.924505352973938, + "learning_rate": 8.52e-06, + "loss": 0.1054, + "step": 2843 + }, + { + "epoch": 2.402027027027027, + "grad_norm": 0.7989174723625183, + "learning_rate": 8.523000000000001e-06, + "loss": 0.0861, + "step": 2844 + }, + { + "epoch": 2.4028716216216215, + "grad_norm": 0.6569541692733765, + "learning_rate": 8.526000000000001e-06, + "loss": 0.0719, + "step": 2845 + }, + { + "epoch": 2.4037162162162162, + "grad_norm": 0.9616947770118713, + "learning_rate": 8.529e-06, + "loss": 0.1072, + "step": 2846 + }, + { + "epoch": 2.404560810810811, + "grad_norm": 0.719377338886261, + "learning_rate": 8.532e-06, + "loss": 0.0772, + "step": 2847 + }, + { + "epoch": 2.4054054054054053, + "grad_norm": 1.0482758283615112, + "learning_rate": 8.534999999999999e-06, + "loss": 0.098, + "step": 2848 + }, + { + "epoch": 2.40625, + "grad_norm": 0.7354392409324646, + "learning_rate": 8.538e-06, + "loss": 0.0913, + "step": 2849 + }, + { + "epoch": 2.4070945945945947, + "grad_norm": 0.821740984916687, + "learning_rate": 8.541e-06, + "loss": 0.1076, + "step": 2850 + }, + { + "epoch": 2.407939189189189, + "grad_norm": 0.8968009948730469, + "learning_rate": 8.544e-06, + "loss": 0.0699, + "step": 2851 + }, + { + "epoch": 2.4087837837837838, + "grad_norm": 1.785891056060791, + "learning_rate": 8.547e-06, + "loss": 0.1441, + "step": 2852 + }, + { + "epoch": 2.4096283783783785, + "grad_norm": 0.7926217317581177, + "learning_rate": 8.55e-06, + "loss": 0.084, + "step": 2853 + }, + { + "epoch": 2.410472972972973, + "grad_norm": 1.1815468072891235, + "learning_rate": 8.553000000000001e-06, + "loss": 0.0972, + "step": 2854 + }, + { + "epoch": 2.4113175675675675, + "grad_norm": 7.660136699676514, + "learning_rate": 8.556e-06, + "loss": 0.117, + "step": 2855 + }, + { + "epoch": 2.4121621621621623, + "grad_norm": 0.9802360534667969, + "learning_rate": 8.559e-06, + "loss": 0.09, + "step": 2856 + }, + { + "epoch": 2.4130067567567566, + "grad_norm": 1.00425386428833, + "learning_rate": 8.562e-06, + "loss": 0.0926, + "step": 2857 + }, + { + "epoch": 2.4138513513513513, + "grad_norm": 1.2201215028762817, + "learning_rate": 8.565e-06, + "loss": 0.1124, + "step": 2858 + }, + { + "epoch": 2.414695945945946, + "grad_norm": 0.9951536655426025, + "learning_rate": 8.568000000000002e-06, + "loss": 0.1025, + "step": 2859 + }, + { + "epoch": 2.4155405405405403, + "grad_norm": 1.5076555013656616, + "learning_rate": 8.571e-06, + "loss": 0.0948, + "step": 2860 + }, + { + "epoch": 2.416385135135135, + "grad_norm": 0.8916618227958679, + "learning_rate": 8.574e-06, + "loss": 0.0804, + "step": 2861 + }, + { + "epoch": 2.41722972972973, + "grad_norm": 0.8951012492179871, + "learning_rate": 8.577e-06, + "loss": 0.0855, + "step": 2862 + }, + { + "epoch": 2.4180743243243246, + "grad_norm": 1.4681857824325562, + "learning_rate": 8.58e-06, + "loss": 0.1194, + "step": 2863 + }, + { + "epoch": 2.418918918918919, + "grad_norm": 1.1519291400909424, + "learning_rate": 8.583e-06, + "loss": 0.0925, + "step": 2864 + }, + { + "epoch": 2.4197635135135136, + "grad_norm": 1.112414836883545, + "learning_rate": 8.586e-06, + "loss": 0.114, + "step": 2865 + }, + { + "epoch": 2.420608108108108, + "grad_norm": 1.9506072998046875, + "learning_rate": 8.589e-06, + "loss": 0.1333, + "step": 2866 + }, + { + "epoch": 2.4214527027027026, + "grad_norm": 1.7907384634017944, + "learning_rate": 8.592e-06, + "loss": 0.1561, + "step": 2867 + }, + { + "epoch": 2.4222972972972974, + "grad_norm": 2.3397557735443115, + "learning_rate": 8.595e-06, + "loss": 0.1929, + "step": 2868 + }, + { + "epoch": 2.423141891891892, + "grad_norm": 1.8698058128356934, + "learning_rate": 8.598000000000001e-06, + "loss": 0.5371, + "step": 2869 + }, + { + "epoch": 2.4239864864864864, + "grad_norm": 2.284301519393921, + "learning_rate": 8.601000000000001e-06, + "loss": 0.3167, + "step": 2870 + }, + { + "epoch": 2.424831081081081, + "grad_norm": 1.0193742513656616, + "learning_rate": 8.604000000000001e-06, + "loss": 0.1603, + "step": 2871 + }, + { + "epoch": 2.4256756756756754, + "grad_norm": 1.076301097869873, + "learning_rate": 8.606999999999999e-06, + "loss": 0.1483, + "step": 2872 + }, + { + "epoch": 2.42652027027027, + "grad_norm": 0.9411473870277405, + "learning_rate": 8.609999999999999e-06, + "loss": 0.1281, + "step": 2873 + }, + { + "epoch": 2.427364864864865, + "grad_norm": 0.7834156155586243, + "learning_rate": 8.613e-06, + "loss": 0.092, + "step": 2874 + }, + { + "epoch": 2.4282094594594597, + "grad_norm": 0.9364745616912842, + "learning_rate": 8.616e-06, + "loss": 0.1018, + "step": 2875 + }, + { + "epoch": 2.429054054054054, + "grad_norm": 1.1748449802398682, + "learning_rate": 8.619e-06, + "loss": 0.092, + "step": 2876 + }, + { + "epoch": 2.4298986486486487, + "grad_norm": 0.7195788025856018, + "learning_rate": 8.622e-06, + "loss": 0.091, + "step": 2877 + }, + { + "epoch": 2.4307432432432434, + "grad_norm": 0.8245499730110168, + "learning_rate": 8.625e-06, + "loss": 0.1057, + "step": 2878 + }, + { + "epoch": 2.4315878378378377, + "grad_norm": 0.9551841020584106, + "learning_rate": 8.628000000000001e-06, + "loss": 0.0918, + "step": 2879 + }, + { + "epoch": 2.4324324324324325, + "grad_norm": 1.142669677734375, + "learning_rate": 8.631000000000001e-06, + "loss": 0.1145, + "step": 2880 + }, + { + "epoch": 2.433277027027027, + "grad_norm": 0.9088649153709412, + "learning_rate": 8.634e-06, + "loss": 0.0794, + "step": 2881 + }, + { + "epoch": 2.4341216216216215, + "grad_norm": 0.9298703670501709, + "learning_rate": 8.637e-06, + "loss": 0.1715, + "step": 2882 + }, + { + "epoch": 2.4349662162162162, + "grad_norm": 0.972898542881012, + "learning_rate": 8.64e-06, + "loss": 0.0909, + "step": 2883 + }, + { + "epoch": 2.435810810810811, + "grad_norm": 0.8709102272987366, + "learning_rate": 8.643e-06, + "loss": 0.1078, + "step": 2884 + }, + { + "epoch": 2.4366554054054053, + "grad_norm": 0.7588875889778137, + "learning_rate": 8.646e-06, + "loss": 0.0681, + "step": 2885 + }, + { + "epoch": 2.4375, + "grad_norm": 0.9594582915306091, + "learning_rate": 8.649e-06, + "loss": 0.1214, + "step": 2886 + }, + { + "epoch": 2.4383445945945947, + "grad_norm": 0.9574998617172241, + "learning_rate": 8.652e-06, + "loss": 0.1021, + "step": 2887 + }, + { + "epoch": 2.439189189189189, + "grad_norm": 0.8277511596679688, + "learning_rate": 8.655e-06, + "loss": 0.0979, + "step": 2888 + }, + { + "epoch": 2.4400337837837838, + "grad_norm": 1.087310791015625, + "learning_rate": 8.658e-06, + "loss": 0.1378, + "step": 2889 + }, + { + "epoch": 2.4408783783783785, + "grad_norm": 0.8375915288925171, + "learning_rate": 8.661e-06, + "loss": 0.0718, + "step": 2890 + }, + { + "epoch": 2.441722972972973, + "grad_norm": 0.8849912285804749, + "learning_rate": 8.664e-06, + "loss": 0.0836, + "step": 2891 + }, + { + "epoch": 2.4425675675675675, + "grad_norm": 0.8716853260993958, + "learning_rate": 8.667e-06, + "loss": 0.0727, + "step": 2892 + }, + { + "epoch": 2.4434121621621623, + "grad_norm": 0.9286633729934692, + "learning_rate": 8.67e-06, + "loss": 0.1527, + "step": 2893 + }, + { + "epoch": 2.4442567567567566, + "grad_norm": 1.1235134601593018, + "learning_rate": 8.673000000000001e-06, + "loss": 0.0939, + "step": 2894 + }, + { + "epoch": 2.4451013513513513, + "grad_norm": 0.8685243725776672, + "learning_rate": 8.676000000000001e-06, + "loss": 0.0954, + "step": 2895 + }, + { + "epoch": 2.445945945945946, + "grad_norm": 1.0323994159698486, + "learning_rate": 8.679000000000001e-06, + "loss": 0.0702, + "step": 2896 + }, + { + "epoch": 2.4467905405405403, + "grad_norm": 0.9405861496925354, + "learning_rate": 8.682e-06, + "loss": 0.104, + "step": 2897 + }, + { + "epoch": 2.447635135135135, + "grad_norm": 0.9808124899864197, + "learning_rate": 8.684999999999999e-06, + "loss": 0.1136, + "step": 2898 + }, + { + "epoch": 2.44847972972973, + "grad_norm": 0.9854186177253723, + "learning_rate": 8.688e-06, + "loss": 0.1062, + "step": 2899 + }, + { + "epoch": 2.4493243243243246, + "grad_norm": 0.9208290576934814, + "learning_rate": 8.691e-06, + "loss": 0.085, + "step": 2900 + }, + { + "epoch": 2.450168918918919, + "grad_norm": 0.7536394596099854, + "learning_rate": 8.694e-06, + "loss": 0.069, + "step": 2901 + }, + { + "epoch": 2.4510135135135136, + "grad_norm": 1.6512197256088257, + "learning_rate": 8.697e-06, + "loss": 0.2302, + "step": 2902 + }, + { + "epoch": 2.451858108108108, + "grad_norm": 1.1971168518066406, + "learning_rate": 8.7e-06, + "loss": 0.0963, + "step": 2903 + }, + { + "epoch": 2.4527027027027026, + "grad_norm": 1.1825448274612427, + "learning_rate": 8.703000000000001e-06, + "loss": 0.0852, + "step": 2904 + }, + { + "epoch": 2.4535472972972974, + "grad_norm": 1.318536400794983, + "learning_rate": 8.706000000000001e-06, + "loss": 0.1263, + "step": 2905 + }, + { + "epoch": 2.454391891891892, + "grad_norm": 1.2530208826065063, + "learning_rate": 8.709e-06, + "loss": 0.1322, + "step": 2906 + }, + { + "epoch": 2.4552364864864864, + "grad_norm": 0.7944090962409973, + "learning_rate": 8.712e-06, + "loss": 0.0899, + "step": 2907 + }, + { + "epoch": 2.456081081081081, + "grad_norm": 2.448282480239868, + "learning_rate": 8.715e-06, + "loss": 0.128, + "step": 2908 + }, + { + "epoch": 2.4569256756756754, + "grad_norm": 0.8319694399833679, + "learning_rate": 8.718e-06, + "loss": 0.085, + "step": 2909 + }, + { + "epoch": 2.45777027027027, + "grad_norm": 1.0446240901947021, + "learning_rate": 8.721e-06, + "loss": 0.1068, + "step": 2910 + }, + { + "epoch": 2.458614864864865, + "grad_norm": 1.0938113927841187, + "learning_rate": 8.724e-06, + "loss": 0.1071, + "step": 2911 + }, + { + "epoch": 2.4594594594594597, + "grad_norm": 1.288773536682129, + "learning_rate": 8.727e-06, + "loss": 0.1341, + "step": 2912 + }, + { + "epoch": 2.460304054054054, + "grad_norm": 0.8630824685096741, + "learning_rate": 8.73e-06, + "loss": 0.0939, + "step": 2913 + }, + { + "epoch": 2.4611486486486487, + "grad_norm": 1.299561858177185, + "learning_rate": 8.733000000000001e-06, + "loss": 0.1398, + "step": 2914 + }, + { + "epoch": 2.4619932432432434, + "grad_norm": 1.5301774740219116, + "learning_rate": 8.736e-06, + "loss": 0.1537, + "step": 2915 + }, + { + "epoch": 2.4628378378378377, + "grad_norm": 1.7208234071731567, + "learning_rate": 8.739e-06, + "loss": 0.1723, + "step": 2916 + }, + { + "epoch": 2.4636824324324325, + "grad_norm": 1.7319529056549072, + "learning_rate": 8.742e-06, + "loss": 0.1411, + "step": 2917 + }, + { + "epoch": 2.464527027027027, + "grad_norm": 2.236510753631592, + "learning_rate": 8.745e-06, + "loss": 0.202, + "step": 2918 + }, + { + "epoch": 2.4653716216216215, + "grad_norm": 4.065610885620117, + "learning_rate": 8.748000000000002e-06, + "loss": 0.6145, + "step": 2919 + }, + { + "epoch": 2.4662162162162162, + "grad_norm": 1.7841122150421143, + "learning_rate": 8.751000000000001e-06, + "loss": 0.3407, + "step": 2920 + }, + { + "epoch": 2.467060810810811, + "grad_norm": 2.124269723892212, + "learning_rate": 8.754e-06, + "loss": 0.2276, + "step": 2921 + }, + { + "epoch": 2.4679054054054053, + "grad_norm": 1.3373024463653564, + "learning_rate": 8.757e-06, + "loss": 0.1288, + "step": 2922 + }, + { + "epoch": 2.46875, + "grad_norm": 0.7354568243026733, + "learning_rate": 8.759999999999999e-06, + "loss": 0.111, + "step": 2923 + }, + { + "epoch": 2.4695945945945947, + "grad_norm": 0.8875673413276672, + "learning_rate": 8.763e-06, + "loss": 0.1058, + "step": 2924 + }, + { + "epoch": 2.470439189189189, + "grad_norm": 1.6896717548370361, + "learning_rate": 8.766e-06, + "loss": 0.1091, + "step": 2925 + }, + { + "epoch": 2.4712837837837838, + "grad_norm": 0.7328709363937378, + "learning_rate": 8.769e-06, + "loss": 0.0812, + "step": 2926 + }, + { + "epoch": 2.4721283783783785, + "grad_norm": 0.632072389125824, + "learning_rate": 8.772e-06, + "loss": 0.0732, + "step": 2927 + }, + { + "epoch": 2.472972972972973, + "grad_norm": 1.2364897727966309, + "learning_rate": 8.775e-06, + "loss": 0.1089, + "step": 2928 + }, + { + "epoch": 2.4738175675675675, + "grad_norm": 0.82864910364151, + "learning_rate": 8.778000000000001e-06, + "loss": 0.1211, + "step": 2929 + }, + { + "epoch": 2.4746621621621623, + "grad_norm": 0.8414250016212463, + "learning_rate": 8.781000000000001e-06, + "loss": 0.1015, + "step": 2930 + }, + { + "epoch": 2.4755067567567566, + "grad_norm": 0.851396918296814, + "learning_rate": 8.784000000000001e-06, + "loss": 0.0993, + "step": 2931 + }, + { + "epoch": 2.4763513513513513, + "grad_norm": 0.9236922860145569, + "learning_rate": 8.787e-06, + "loss": 0.0979, + "step": 2932 + }, + { + "epoch": 2.477195945945946, + "grad_norm": 0.9173739552497864, + "learning_rate": 8.79e-06, + "loss": 0.1265, + "step": 2933 + }, + { + "epoch": 2.4780405405405403, + "grad_norm": 0.8319063782691956, + "learning_rate": 8.793e-06, + "loss": 0.0942, + "step": 2934 + }, + { + "epoch": 2.478885135135135, + "grad_norm": 0.7446277737617493, + "learning_rate": 8.796e-06, + "loss": 0.0702, + "step": 2935 + }, + { + "epoch": 2.47972972972973, + "grad_norm": 0.6952183246612549, + "learning_rate": 8.799e-06, + "loss": 0.0999, + "step": 2936 + }, + { + "epoch": 2.4805743243243246, + "grad_norm": 0.8430996537208557, + "learning_rate": 8.802e-06, + "loss": 0.0986, + "step": 2937 + }, + { + "epoch": 2.481418918918919, + "grad_norm": 0.7625723481178284, + "learning_rate": 8.805e-06, + "loss": 0.0856, + "step": 2938 + }, + { + "epoch": 2.4822635135135136, + "grad_norm": 0.9338132739067078, + "learning_rate": 8.808000000000001e-06, + "loss": 0.1823, + "step": 2939 + }, + { + "epoch": 2.483108108108108, + "grad_norm": 0.5598660111427307, + "learning_rate": 8.811000000000001e-06, + "loss": 0.0756, + "step": 2940 + }, + { + "epoch": 2.4839527027027026, + "grad_norm": 0.5924811959266663, + "learning_rate": 8.814e-06, + "loss": 0.0689, + "step": 2941 + }, + { + "epoch": 2.4847972972972974, + "grad_norm": 0.7426071166992188, + "learning_rate": 8.817e-06, + "loss": 0.0755, + "step": 2942 + }, + { + "epoch": 2.485641891891892, + "grad_norm": 0.923914909362793, + "learning_rate": 8.82e-06, + "loss": 0.0969, + "step": 2943 + }, + { + "epoch": 2.4864864864864864, + "grad_norm": 0.9543959498405457, + "learning_rate": 8.823e-06, + "loss": 0.0897, + "step": 2944 + }, + { + "epoch": 2.487331081081081, + "grad_norm": 1.014148235321045, + "learning_rate": 8.826000000000002e-06, + "loss": 0.0952, + "step": 2945 + }, + { + "epoch": 2.4881756756756754, + "grad_norm": 1.285049319267273, + "learning_rate": 8.829e-06, + "loss": 0.1084, + "step": 2946 + }, + { + "epoch": 2.48902027027027, + "grad_norm": 0.848284900188446, + "learning_rate": 8.832e-06, + "loss": 0.0804, + "step": 2947 + }, + { + "epoch": 2.489864864864865, + "grad_norm": 1.008657693862915, + "learning_rate": 8.835e-06, + "loss": 0.1049, + "step": 2948 + }, + { + "epoch": 2.4907094594594597, + "grad_norm": 0.928043007850647, + "learning_rate": 8.837999999999999e-06, + "loss": 0.0663, + "step": 2949 + }, + { + "epoch": 2.491554054054054, + "grad_norm": 0.8995162844657898, + "learning_rate": 8.841e-06, + "loss": 0.0727, + "step": 2950 + }, + { + "epoch": 2.4923986486486487, + "grad_norm": 0.9136526584625244, + "learning_rate": 8.844e-06, + "loss": 0.106, + "step": 2951 + }, + { + "epoch": 2.4932432432432434, + "grad_norm": 1.300165057182312, + "learning_rate": 8.847e-06, + "loss": 0.0951, + "step": 2952 + }, + { + "epoch": 2.4940878378378377, + "grad_norm": 1.0379160642623901, + "learning_rate": 8.85e-06, + "loss": 0.1119, + "step": 2953 + }, + { + "epoch": 2.4949324324324325, + "grad_norm": 1.125101089477539, + "learning_rate": 8.853e-06, + "loss": 0.1041, + "step": 2954 + }, + { + "epoch": 2.495777027027027, + "grad_norm": 1.1737334728240967, + "learning_rate": 8.856000000000001e-06, + "loss": 0.1115, + "step": 2955 + }, + { + "epoch": 2.4966216216216215, + "grad_norm": 1.0009346008300781, + "learning_rate": 8.859000000000001e-06, + "loss": 0.0945, + "step": 2956 + }, + { + "epoch": 2.4974662162162162, + "grad_norm": 1.078938364982605, + "learning_rate": 8.862000000000001e-06, + "loss": 0.0962, + "step": 2957 + }, + { + "epoch": 2.498310810810811, + "grad_norm": 1.0300195217132568, + "learning_rate": 8.864999999999999e-06, + "loss": 0.0724, + "step": 2958 + }, + { + "epoch": 2.4991554054054053, + "grad_norm": 0.9784924387931824, + "learning_rate": 8.867999999999999e-06, + "loss": 0.1029, + "step": 2959 + }, + { + "epoch": 2.5, + "grad_norm": 1.3057109117507935, + "learning_rate": 8.871e-06, + "loss": 0.1121, + "step": 2960 + }, + { + "epoch": 2.5008445945945947, + "grad_norm": 1.1841657161712646, + "learning_rate": 8.874e-06, + "loss": 0.089, + "step": 2961 + }, + { + "epoch": 2.501689189189189, + "grad_norm": 1.1495676040649414, + "learning_rate": 8.877e-06, + "loss": 0.1184, + "step": 2962 + }, + { + "epoch": 2.5025337837837838, + "grad_norm": 1.2267069816589355, + "learning_rate": 8.88e-06, + "loss": 0.1287, + "step": 2963 + }, + { + "epoch": 2.5033783783783785, + "grad_norm": 1.3278133869171143, + "learning_rate": 8.883e-06, + "loss": 0.1101, + "step": 2964 + }, + { + "epoch": 2.504222972972973, + "grad_norm": 1.4213019609451294, + "learning_rate": 8.886000000000001e-06, + "loss": 0.1416, + "step": 2965 + }, + { + "epoch": 2.5050675675675675, + "grad_norm": 1.726180911064148, + "learning_rate": 8.889e-06, + "loss": 0.1566, + "step": 2966 + }, + { + "epoch": 2.5059121621621623, + "grad_norm": 1.8157976865768433, + "learning_rate": 8.892e-06, + "loss": 0.136, + "step": 2967 + }, + { + "epoch": 2.506756756756757, + "grad_norm": 2.8164169788360596, + "learning_rate": 8.895e-06, + "loss": 0.2789, + "step": 2968 + }, + { + "epoch": 2.5076013513513513, + "grad_norm": 4.4367241859436035, + "learning_rate": 8.898e-06, + "loss": 0.6497, + "step": 2969 + }, + { + "epoch": 2.508445945945946, + "grad_norm": 1.6342594623565674, + "learning_rate": 8.901e-06, + "loss": 0.4274, + "step": 2970 + }, + { + "epoch": 2.5092905405405403, + "grad_norm": 1.204917073249817, + "learning_rate": 8.904e-06, + "loss": 0.1855, + "step": 2971 + }, + { + "epoch": 2.510135135135135, + "grad_norm": 1.194904088973999, + "learning_rate": 8.907e-06, + "loss": 0.1564, + "step": 2972 + }, + { + "epoch": 2.51097972972973, + "grad_norm": 0.8861545920372009, + "learning_rate": 8.91e-06, + "loss": 0.1094, + "step": 2973 + }, + { + "epoch": 2.5118243243243246, + "grad_norm": 0.7558777332305908, + "learning_rate": 8.913e-06, + "loss": 0.1072, + "step": 2974 + }, + { + "epoch": 2.512668918918919, + "grad_norm": 0.9224202632904053, + "learning_rate": 8.916e-06, + "loss": 0.1196, + "step": 2975 + }, + { + "epoch": 2.5135135135135136, + "grad_norm": 0.8170168995857239, + "learning_rate": 8.919e-06, + "loss": 0.0994, + "step": 2976 + }, + { + "epoch": 2.514358108108108, + "grad_norm": 1.1848114728927612, + "learning_rate": 8.922e-06, + "loss": 0.1098, + "step": 2977 + }, + { + "epoch": 2.5152027027027026, + "grad_norm": 0.7149918079376221, + "learning_rate": 8.925e-06, + "loss": 0.0687, + "step": 2978 + }, + { + "epoch": 2.5160472972972974, + "grad_norm": 0.8510966300964355, + "learning_rate": 8.928e-06, + "loss": 0.0936, + "step": 2979 + }, + { + "epoch": 2.516891891891892, + "grad_norm": 7.53773832321167, + "learning_rate": 8.931000000000001e-06, + "loss": 0.1242, + "step": 2980 + }, + { + "epoch": 2.5177364864864864, + "grad_norm": 0.7217527627944946, + "learning_rate": 8.934000000000001e-06, + "loss": 0.0817, + "step": 2981 + }, + { + "epoch": 2.518581081081081, + "grad_norm": 0.7856106758117676, + "learning_rate": 8.937000000000001e-06, + "loss": 0.0888, + "step": 2982 + }, + { + "epoch": 2.5194256756756754, + "grad_norm": 0.8920230865478516, + "learning_rate": 8.939999999999999e-06, + "loss": 0.0982, + "step": 2983 + }, + { + "epoch": 2.52027027027027, + "grad_norm": 2.026050567626953, + "learning_rate": 8.942999999999999e-06, + "loss": 0.115, + "step": 2984 + }, + { + "epoch": 2.521114864864865, + "grad_norm": 0.8263787627220154, + "learning_rate": 8.946e-06, + "loss": 0.0922, + "step": 2985 + }, + { + "epoch": 2.5219594594594597, + "grad_norm": 0.7917515635490417, + "learning_rate": 8.949e-06, + "loss": 0.0764, + "step": 2986 + }, + { + "epoch": 2.522804054054054, + "grad_norm": 0.7050510048866272, + "learning_rate": 8.952e-06, + "loss": 0.0894, + "step": 2987 + }, + { + "epoch": 2.5236486486486487, + "grad_norm": 0.8589829802513123, + "learning_rate": 8.955e-06, + "loss": 0.0784, + "step": 2988 + }, + { + "epoch": 2.524493243243243, + "grad_norm": 0.6674860715866089, + "learning_rate": 8.958e-06, + "loss": 0.0737, + "step": 2989 + }, + { + "epoch": 2.5253378378378377, + "grad_norm": 0.836300790309906, + "learning_rate": 8.961000000000001e-06, + "loss": 0.0825, + "step": 2990 + }, + { + "epoch": 2.5261824324324325, + "grad_norm": 0.8888970613479614, + "learning_rate": 8.964000000000001e-06, + "loss": 0.0776, + "step": 2991 + }, + { + "epoch": 2.527027027027027, + "grad_norm": 0.776516318321228, + "learning_rate": 8.967e-06, + "loss": 0.0798, + "step": 2992 + }, + { + "epoch": 2.5278716216216215, + "grad_norm": 0.905535101890564, + "learning_rate": 8.97e-06, + "loss": 0.101, + "step": 2993 + }, + { + "epoch": 2.5287162162162162, + "grad_norm": 0.8232070803642273, + "learning_rate": 8.973e-06, + "loss": 0.0773, + "step": 2994 + }, + { + "epoch": 2.529560810810811, + "grad_norm": 1.4771836996078491, + "learning_rate": 8.976e-06, + "loss": 0.1174, + "step": 2995 + }, + { + "epoch": 2.5304054054054053, + "grad_norm": 1.1963733434677124, + "learning_rate": 8.979e-06, + "loss": 0.069, + "step": 2996 + }, + { + "epoch": 2.53125, + "grad_norm": 0.9679632782936096, + "learning_rate": 8.982e-06, + "loss": 0.0735, + "step": 2997 + }, + { + "epoch": 2.5320945945945947, + "grad_norm": 0.8736961483955383, + "learning_rate": 8.985e-06, + "loss": 0.0904, + "step": 2998 + }, + { + "epoch": 2.532939189189189, + "grad_norm": 1.092555046081543, + "learning_rate": 8.988e-06, + "loss": 0.0889, + "step": 2999 + }, + { + "epoch": 2.5337837837837838, + "grad_norm": 0.9118308424949646, + "learning_rate": 8.991e-06, + "loss": 0.0837, + "step": 3000 + }, + { + "epoch": 2.5337837837837838, + "eval_cer": 0.13950686691831982, + "eval_loss": 0.5226371884346008, + "eval_runtime": 8.2819, + "eval_samples_per_second": 12.195, + "eval_steps_per_second": 0.483, + "eval_wer": 0.38051146384479717, + "step": 3000 + }, + { + "epoch": 2.5346283783783785, + "grad_norm": 0.9162833094596863, + "learning_rate": 8.994e-06, + "loss": 0.0967, + "step": 3001 + }, + { + "epoch": 2.535472972972973, + "grad_norm": 0.8497509360313416, + "learning_rate": 8.997e-06, + "loss": 0.093, + "step": 3002 + }, + { + "epoch": 2.5363175675675675, + "grad_norm": 1.1725261211395264, + "learning_rate": 9e-06, + "loss": 0.0631, + "step": 3003 + }, + { + "epoch": 2.5371621621621623, + "grad_norm": 0.775682270526886, + "learning_rate": 9.003e-06, + "loss": 0.0666, + "step": 3004 + }, + { + "epoch": 2.538006756756757, + "grad_norm": 0.8870813846588135, + "learning_rate": 9.006000000000002e-06, + "loss": 0.081, + "step": 3005 + }, + { + "epoch": 2.5388513513513513, + "grad_norm": 1.313116192817688, + "learning_rate": 9.009000000000001e-06, + "loss": 0.0808, + "step": 3006 + }, + { + "epoch": 2.539695945945946, + "grad_norm": 1.1857577562332153, + "learning_rate": 9.012e-06, + "loss": 0.1405, + "step": 3007 + }, + { + "epoch": 2.5405405405405403, + "grad_norm": 0.6406274437904358, + "learning_rate": 9.015e-06, + "loss": 0.0544, + "step": 3008 + }, + { + "epoch": 2.541385135135135, + "grad_norm": 1.223819613456726, + "learning_rate": 9.017999999999999e-06, + "loss": 0.1264, + "step": 3009 + }, + { + "epoch": 2.54222972972973, + "grad_norm": 1.0267047882080078, + "learning_rate": 9.021e-06, + "loss": 0.0932, + "step": 3010 + }, + { + "epoch": 2.5430743243243246, + "grad_norm": 1.6685376167297363, + "learning_rate": 9.024e-06, + "loss": 0.1076, + "step": 3011 + }, + { + "epoch": 2.543918918918919, + "grad_norm": 1.258824110031128, + "learning_rate": 9.027e-06, + "loss": 0.1493, + "step": 3012 + }, + { + "epoch": 2.5447635135135136, + "grad_norm": 0.8450575470924377, + "learning_rate": 9.03e-06, + "loss": 0.0995, + "step": 3013 + }, + { + "epoch": 2.545608108108108, + "grad_norm": 1.4941140413284302, + "learning_rate": 9.033e-06, + "loss": 0.149, + "step": 3014 + }, + { + "epoch": 2.5464527027027026, + "grad_norm": 0.9788738489151001, + "learning_rate": 9.036000000000001e-06, + "loss": 0.094, + "step": 3015 + }, + { + "epoch": 2.5472972972972974, + "grad_norm": 1.093908429145813, + "learning_rate": 9.039000000000001e-06, + "loss": 0.1288, + "step": 3016 + }, + { + "epoch": 2.548141891891892, + "grad_norm": 1.1313217878341675, + "learning_rate": 9.042e-06, + "loss": 0.1372, + "step": 3017 + }, + { + "epoch": 2.5489864864864864, + "grad_norm": 2.1244664192199707, + "learning_rate": 9.045e-06, + "loss": 0.1881, + "step": 3018 + }, + { + "epoch": 2.549831081081081, + "grad_norm": 8.630959510803223, + "learning_rate": 9.048e-06, + "loss": 0.5808, + "step": 3019 + }, + { + "epoch": 2.5506756756756754, + "grad_norm": 3.0916895866394043, + "learning_rate": 9.051e-06, + "loss": 0.4703, + "step": 3020 + }, + { + "epoch": 2.55152027027027, + "grad_norm": 1.1021511554718018, + "learning_rate": 9.054e-06, + "loss": 0.1424, + "step": 3021 + }, + { + "epoch": 2.552364864864865, + "grad_norm": 1.2049497365951538, + "learning_rate": 9.057e-06, + "loss": 0.1265, + "step": 3022 + }, + { + "epoch": 2.5532094594594597, + "grad_norm": 0.9926026463508606, + "learning_rate": 9.06e-06, + "loss": 0.1083, + "step": 3023 + }, + { + "epoch": 2.554054054054054, + "grad_norm": 0.6301072239875793, + "learning_rate": 9.063e-06, + "loss": 0.1026, + "step": 3024 + }, + { + "epoch": 2.5548986486486487, + "grad_norm": 0.6901852488517761, + "learning_rate": 9.066000000000001e-06, + "loss": 0.0901, + "step": 3025 + }, + { + "epoch": 2.555743243243243, + "grad_norm": 0.7182523608207703, + "learning_rate": 9.069e-06, + "loss": 0.0834, + "step": 3026 + }, + { + "epoch": 2.5565878378378377, + "grad_norm": 0.7607287764549255, + "learning_rate": 9.072e-06, + "loss": 0.0844, + "step": 3027 + }, + { + "epoch": 2.5574324324324325, + "grad_norm": 0.8595036268234253, + "learning_rate": 9.075e-06, + "loss": 0.0894, + "step": 3028 + }, + { + "epoch": 2.558277027027027, + "grad_norm": 0.8946948051452637, + "learning_rate": 9.078e-06, + "loss": 0.1009, + "step": 3029 + }, + { + "epoch": 2.5591216216216215, + "grad_norm": 0.8657110929489136, + "learning_rate": 9.081000000000002e-06, + "loss": 0.0896, + "step": 3030 + }, + { + "epoch": 2.5599662162162162, + "grad_norm": 1.1402548551559448, + "learning_rate": 9.084000000000001e-06, + "loss": 0.1099, + "step": 3031 + }, + { + "epoch": 2.560810810810811, + "grad_norm": 0.6646518707275391, + "learning_rate": 9.087e-06, + "loss": 0.0741, + "step": 3032 + }, + { + "epoch": 2.5616554054054053, + "grad_norm": 0.6865001916885376, + "learning_rate": 9.09e-06, + "loss": 0.0741, + "step": 3033 + }, + { + "epoch": 2.5625, + "grad_norm": 0.9100553393363953, + "learning_rate": 9.093e-06, + "loss": 0.0851, + "step": 3034 + }, + { + "epoch": 2.5633445945945947, + "grad_norm": 0.7991436123847961, + "learning_rate": 9.096e-06, + "loss": 0.0865, + "step": 3035 + }, + { + "epoch": 2.564189189189189, + "grad_norm": 0.8225799202919006, + "learning_rate": 9.099e-06, + "loss": 0.0812, + "step": 3036 + }, + { + "epoch": 2.5650337837837838, + "grad_norm": 0.7815044522285461, + "learning_rate": 9.102e-06, + "loss": 0.0806, + "step": 3037 + }, + { + "epoch": 2.5658783783783785, + "grad_norm": 0.8073715567588806, + "learning_rate": 9.105e-06, + "loss": 0.0944, + "step": 3038 + }, + { + "epoch": 2.566722972972973, + "grad_norm": 0.7157806158065796, + "learning_rate": 9.108e-06, + "loss": 0.084, + "step": 3039 + }, + { + "epoch": 2.5675675675675675, + "grad_norm": 3.812443256378174, + "learning_rate": 9.111000000000001e-06, + "loss": 0.1112, + "step": 3040 + }, + { + "epoch": 2.5684121621621623, + "grad_norm": 0.7147251963615417, + "learning_rate": 9.114000000000001e-06, + "loss": 0.0814, + "step": 3041 + }, + { + "epoch": 2.569256756756757, + "grad_norm": 0.7224287986755371, + "learning_rate": 9.117000000000001e-06, + "loss": 0.0742, + "step": 3042 + }, + { + "epoch": 2.5701013513513513, + "grad_norm": 1.069077491760254, + "learning_rate": 9.12e-06, + "loss": 0.0882, + "step": 3043 + }, + { + "epoch": 2.570945945945946, + "grad_norm": 1.8496891260147095, + "learning_rate": 9.122999999999999e-06, + "loss": 0.1007, + "step": 3044 + }, + { + "epoch": 2.5717905405405403, + "grad_norm": 0.8030915260314941, + "learning_rate": 9.126e-06, + "loss": 0.088, + "step": 3045 + }, + { + "epoch": 2.572635135135135, + "grad_norm": 0.9948773384094238, + "learning_rate": 9.129e-06, + "loss": 0.0932, + "step": 3046 + }, + { + "epoch": 2.57347972972973, + "grad_norm": 0.7437900304794312, + "learning_rate": 9.132e-06, + "loss": 0.0687, + "step": 3047 + }, + { + "epoch": 2.5743243243243246, + "grad_norm": 0.8735523223876953, + "learning_rate": 9.135e-06, + "loss": 0.0706, + "step": 3048 + }, + { + "epoch": 2.575168918918919, + "grad_norm": 1.0688130855560303, + "learning_rate": 9.138e-06, + "loss": 0.1047, + "step": 3049 + }, + { + "epoch": 2.5760135135135136, + "grad_norm": 1.28206467628479, + "learning_rate": 9.141000000000001e-06, + "loss": 0.0994, + "step": 3050 + }, + { + "epoch": 2.576858108108108, + "grad_norm": 1.1718533039093018, + "learning_rate": 9.144000000000001e-06, + "loss": 0.0916, + "step": 3051 + }, + { + "epoch": 2.5777027027027026, + "grad_norm": 1.021514654159546, + "learning_rate": 9.147e-06, + "loss": 0.1206, + "step": 3052 + }, + { + "epoch": 2.5785472972972974, + "grad_norm": 1.1707422733306885, + "learning_rate": 9.15e-06, + "loss": 0.0963, + "step": 3053 + }, + { + "epoch": 2.579391891891892, + "grad_norm": 0.6414626836776733, + "learning_rate": 9.153e-06, + "loss": 0.0579, + "step": 3054 + }, + { + "epoch": 2.5802364864864864, + "grad_norm": 1.2017567157745361, + "learning_rate": 9.156000000000002e-06, + "loss": 0.1357, + "step": 3055 + }, + { + "epoch": 2.581081081081081, + "grad_norm": 0.9682698249816895, + "learning_rate": 9.159e-06, + "loss": 0.0843, + "step": 3056 + }, + { + "epoch": 2.5819256756756754, + "grad_norm": 1.2223471403121948, + "learning_rate": 9.162e-06, + "loss": 0.1031, + "step": 3057 + }, + { + "epoch": 2.58277027027027, + "grad_norm": 1.2612581253051758, + "learning_rate": 9.165e-06, + "loss": 0.1241, + "step": 3058 + }, + { + "epoch": 2.583614864864865, + "grad_norm": 1.1903626918792725, + "learning_rate": 9.168e-06, + "loss": 0.1051, + "step": 3059 + }, + { + "epoch": 2.5844594594594597, + "grad_norm": 1.8411277532577515, + "learning_rate": 9.171e-06, + "loss": 0.1048, + "step": 3060 + }, + { + "epoch": 2.585304054054054, + "grad_norm": 1.2108343839645386, + "learning_rate": 9.174e-06, + "loss": 0.0893, + "step": 3061 + }, + { + "epoch": 2.5861486486486487, + "grad_norm": 2.209320068359375, + "learning_rate": 9.177e-06, + "loss": 0.1824, + "step": 3062 + }, + { + "epoch": 2.586993243243243, + "grad_norm": 0.9402831196784973, + "learning_rate": 9.18e-06, + "loss": 0.1024, + "step": 3063 + }, + { + "epoch": 2.5878378378378377, + "grad_norm": 1.949992060661316, + "learning_rate": 9.183e-06, + "loss": 0.1423, + "step": 3064 + }, + { + "epoch": 2.5886824324324325, + "grad_norm": 1.0762749910354614, + "learning_rate": 9.186000000000001e-06, + "loss": 0.0928, + "step": 3065 + }, + { + "epoch": 2.589527027027027, + "grad_norm": 1.811893343925476, + "learning_rate": 9.189000000000001e-06, + "loss": 0.1394, + "step": 3066 + }, + { + "epoch": 2.5903716216216215, + "grad_norm": 3.866811990737915, + "learning_rate": 9.192000000000001e-06, + "loss": 0.1893, + "step": 3067 + }, + { + "epoch": 2.5912162162162162, + "grad_norm": 3.685944080352783, + "learning_rate": 9.195000000000001e-06, + "loss": 0.243, + "step": 3068 + }, + { + "epoch": 2.592060810810811, + "grad_norm": 4.052498817443848, + "learning_rate": 9.197999999999999e-06, + "loss": 0.5013, + "step": 3069 + }, + { + "epoch": 2.5929054054054053, + "grad_norm": 1.1307772397994995, + "learning_rate": 9.200999999999999e-06, + "loss": 0.3107, + "step": 3070 + }, + { + "epoch": 2.59375, + "grad_norm": 1.0133105516433716, + "learning_rate": 9.204e-06, + "loss": 0.1799, + "step": 3071 + }, + { + "epoch": 2.5945945945945947, + "grad_norm": 0.992694616317749, + "learning_rate": 9.207e-06, + "loss": 0.1264, + "step": 3072 + }, + { + "epoch": 2.595439189189189, + "grad_norm": 0.8877226114273071, + "learning_rate": 9.21e-06, + "loss": 0.1067, + "step": 3073 + }, + { + "epoch": 2.5962837837837838, + "grad_norm": 0.8103659749031067, + "learning_rate": 9.213e-06, + "loss": 0.1078, + "step": 3074 + }, + { + "epoch": 2.5971283783783785, + "grad_norm": 0.833867609500885, + "learning_rate": 9.216e-06, + "loss": 0.0916, + "step": 3075 + }, + { + "epoch": 2.597972972972973, + "grad_norm": 0.7871026992797852, + "learning_rate": 9.219000000000001e-06, + "loss": 0.1018, + "step": 3076 + }, + { + "epoch": 2.5988175675675675, + "grad_norm": 1.162632942199707, + "learning_rate": 9.222e-06, + "loss": 0.0834, + "step": 3077 + }, + { + "epoch": 2.5996621621621623, + "grad_norm": 0.619968056678772, + "learning_rate": 9.225e-06, + "loss": 0.0696, + "step": 3078 + }, + { + "epoch": 2.600506756756757, + "grad_norm": 0.7788549661636353, + "learning_rate": 9.228e-06, + "loss": 0.0815, + "step": 3079 + }, + { + "epoch": 2.6013513513513513, + "grad_norm": 1.3386147022247314, + "learning_rate": 9.231e-06, + "loss": 0.0886, + "step": 3080 + }, + { + "epoch": 2.602195945945946, + "grad_norm": 0.8041108846664429, + "learning_rate": 9.234e-06, + "loss": 0.07, + "step": 3081 + }, + { + "epoch": 2.6030405405405403, + "grad_norm": 5.900571346282959, + "learning_rate": 9.237e-06, + "loss": 0.0718, + "step": 3082 + }, + { + "epoch": 2.603885135135135, + "grad_norm": 2.7883989810943604, + "learning_rate": 9.24e-06, + "loss": 0.0905, + "step": 3083 + }, + { + "epoch": 2.60472972972973, + "grad_norm": 0.9655504822731018, + "learning_rate": 9.243e-06, + "loss": 0.0689, + "step": 3084 + }, + { + "epoch": 2.6055743243243246, + "grad_norm": 0.7069668173789978, + "learning_rate": 9.246e-06, + "loss": 0.0664, + "step": 3085 + }, + { + "epoch": 2.606418918918919, + "grad_norm": 0.8877211213111877, + "learning_rate": 9.249e-06, + "loss": 0.08, + "step": 3086 + }, + { + "epoch": 2.6072635135135136, + "grad_norm": 0.8519203662872314, + "learning_rate": 9.252e-06, + "loss": 0.0872, + "step": 3087 + }, + { + "epoch": 2.608108108108108, + "grad_norm": 1.2231011390686035, + "learning_rate": 9.255e-06, + "loss": 0.0786, + "step": 3088 + }, + { + "epoch": 2.6089527027027026, + "grad_norm": 2.307565450668335, + "learning_rate": 9.258e-06, + "loss": 0.0727, + "step": 3089 + }, + { + "epoch": 2.6097972972972974, + "grad_norm": 0.983696699142456, + "learning_rate": 9.261e-06, + "loss": 0.095, + "step": 3090 + }, + { + "epoch": 2.610641891891892, + "grad_norm": 0.8014037013053894, + "learning_rate": 9.264000000000001e-06, + "loss": 0.0951, + "step": 3091 + }, + { + "epoch": 2.6114864864864864, + "grad_norm": 0.7155590653419495, + "learning_rate": 9.267000000000001e-06, + "loss": 0.0816, + "step": 3092 + }, + { + "epoch": 2.612331081081081, + "grad_norm": 0.788916826248169, + "learning_rate": 9.27e-06, + "loss": 0.0979, + "step": 3093 + }, + { + "epoch": 2.6131756756756754, + "grad_norm": 1.0007777214050293, + "learning_rate": 9.272999999999999e-06, + "loss": 0.1182, + "step": 3094 + }, + { + "epoch": 2.61402027027027, + "grad_norm": 0.7749466300010681, + "learning_rate": 9.275999999999999e-06, + "loss": 0.0931, + "step": 3095 + }, + { + "epoch": 2.614864864864865, + "grad_norm": 1.7186449766159058, + "learning_rate": 9.279e-06, + "loss": 0.1106, + "step": 3096 + }, + { + "epoch": 2.6157094594594597, + "grad_norm": 0.8358545899391174, + "learning_rate": 9.282e-06, + "loss": 0.0771, + "step": 3097 + }, + { + "epoch": 2.616554054054054, + "grad_norm": 0.8723780512809753, + "learning_rate": 9.285e-06, + "loss": 0.0892, + "step": 3098 + }, + { + "epoch": 2.6173986486486487, + "grad_norm": 1.0027308464050293, + "learning_rate": 9.288e-06, + "loss": 0.0796, + "step": 3099 + }, + { + "epoch": 2.618243243243243, + "grad_norm": 1.0099806785583496, + "learning_rate": 9.291e-06, + "loss": 0.1096, + "step": 3100 + }, + { + "epoch": 2.6190878378378377, + "grad_norm": 0.6206059455871582, + "learning_rate": 9.294000000000001e-06, + "loss": 0.0601, + "step": 3101 + }, + { + "epoch": 2.6199324324324325, + "grad_norm": 2.180076837539673, + "learning_rate": 9.297000000000001e-06, + "loss": 0.0911, + "step": 3102 + }, + { + "epoch": 2.620777027027027, + "grad_norm": 0.9132147431373596, + "learning_rate": 9.3e-06, + "loss": 0.0846, + "step": 3103 + }, + { + "epoch": 2.6216216216216215, + "grad_norm": 0.7329359650611877, + "learning_rate": 9.303e-06, + "loss": 0.0524, + "step": 3104 + }, + { + "epoch": 2.6224662162162162, + "grad_norm": 1.136372685432434, + "learning_rate": 9.306e-06, + "loss": 0.0919, + "step": 3105 + }, + { + "epoch": 2.623310810810811, + "grad_norm": 1.4163060188293457, + "learning_rate": 9.309e-06, + "loss": 0.0919, + "step": 3106 + }, + { + "epoch": 2.6241554054054053, + "grad_norm": 0.9821440577507019, + "learning_rate": 9.312e-06, + "loss": 0.1037, + "step": 3107 + }, + { + "epoch": 2.625, + "grad_norm": 1.393478274345398, + "learning_rate": 9.315e-06, + "loss": 0.1174, + "step": 3108 + }, + { + "epoch": 2.6258445945945947, + "grad_norm": 1.1217910051345825, + "learning_rate": 9.318e-06, + "loss": 0.0842, + "step": 3109 + }, + { + "epoch": 2.626689189189189, + "grad_norm": 1.1321532726287842, + "learning_rate": 9.321e-06, + "loss": 0.085, + "step": 3110 + }, + { + "epoch": 2.6275337837837838, + "grad_norm": 0.8538263440132141, + "learning_rate": 9.324000000000001e-06, + "loss": 0.1002, + "step": 3111 + }, + { + "epoch": 2.6283783783783785, + "grad_norm": 1.1662925481796265, + "learning_rate": 9.327e-06, + "loss": 0.1254, + "step": 3112 + }, + { + "epoch": 2.629222972972973, + "grad_norm": 1.7226786613464355, + "learning_rate": 9.33e-06, + "loss": 0.1134, + "step": 3113 + }, + { + "epoch": 2.6300675675675675, + "grad_norm": 0.9765592217445374, + "learning_rate": 9.333e-06, + "loss": 0.1105, + "step": 3114 + }, + { + "epoch": 2.6309121621621623, + "grad_norm": 1.0647834539413452, + "learning_rate": 9.336e-06, + "loss": 0.1415, + "step": 3115 + }, + { + "epoch": 2.631756756756757, + "grad_norm": 1.4259389638900757, + "learning_rate": 9.339000000000002e-06, + "loss": 0.1075, + "step": 3116 + }, + { + "epoch": 2.6326013513513513, + "grad_norm": 2.0419275760650635, + "learning_rate": 9.342000000000001e-06, + "loss": 0.148, + "step": 3117 + }, + { + "epoch": 2.633445945945946, + "grad_norm": 2.2206618785858154, + "learning_rate": 9.345e-06, + "loss": 0.2263, + "step": 3118 + }, + { + "epoch": 2.6342905405405403, + "grad_norm": 1.9513893127441406, + "learning_rate": 9.348e-06, + "loss": 0.4957, + "step": 3119 + }, + { + "epoch": 2.635135135135135, + "grad_norm": 1.1040061712265015, + "learning_rate": 9.350999999999999e-06, + "loss": 0.2992, + "step": 3120 + }, + { + "epoch": 2.63597972972973, + "grad_norm": 0.6796560883522034, + "learning_rate": 9.354e-06, + "loss": 0.1371, + "step": 3121 + }, + { + "epoch": 2.6368243243243246, + "grad_norm": 0.837579607963562, + "learning_rate": 9.357e-06, + "loss": 0.1524, + "step": 3122 + }, + { + "epoch": 2.637668918918919, + "grad_norm": 1.1762983798980713, + "learning_rate": 9.36e-06, + "loss": 0.0897, + "step": 3123 + }, + { + "epoch": 2.6385135135135136, + "grad_norm": 1.445678472518921, + "learning_rate": 9.363e-06, + "loss": 0.127, + "step": 3124 + }, + { + "epoch": 2.639358108108108, + "grad_norm": 0.761118471622467, + "learning_rate": 9.366e-06, + "loss": 0.0945, + "step": 3125 + }, + { + "epoch": 2.6402027027027026, + "grad_norm": 0.6809384226799011, + "learning_rate": 9.369000000000001e-06, + "loss": 0.0774, + "step": 3126 + }, + { + "epoch": 2.6410472972972974, + "grad_norm": 0.7949889898300171, + "learning_rate": 9.372000000000001e-06, + "loss": 0.099, + "step": 3127 + }, + { + "epoch": 2.641891891891892, + "grad_norm": 0.7836605906486511, + "learning_rate": 9.375000000000001e-06, + "loss": 0.0963, + "step": 3128 + }, + { + "epoch": 2.6427364864864864, + "grad_norm": 0.6784588694572449, + "learning_rate": 9.378e-06, + "loss": 0.0859, + "step": 3129 + }, + { + "epoch": 2.643581081081081, + "grad_norm": 0.7351099252700806, + "learning_rate": 9.380999999999999e-06, + "loss": 0.07, + "step": 3130 + }, + { + "epoch": 2.6444256756756754, + "grad_norm": 0.6589672565460205, + "learning_rate": 9.384e-06, + "loss": 0.0742, + "step": 3131 + }, + { + "epoch": 2.64527027027027, + "grad_norm": 0.48088839650154114, + "learning_rate": 9.387e-06, + "loss": 0.0587, + "step": 3132 + }, + { + "epoch": 2.646114864864865, + "grad_norm": 0.7032947540283203, + "learning_rate": 9.39e-06, + "loss": 0.0844, + "step": 3133 + }, + { + "epoch": 2.6469594594594597, + "grad_norm": 1.0969263315200806, + "learning_rate": 9.393e-06, + "loss": 0.1046, + "step": 3134 + }, + { + "epoch": 2.647804054054054, + "grad_norm": 0.7268163561820984, + "learning_rate": 9.396e-06, + "loss": 0.0685, + "step": 3135 + }, + { + "epoch": 2.6486486486486487, + "grad_norm": 0.9192765951156616, + "learning_rate": 9.399000000000001e-06, + "loss": 0.0958, + "step": 3136 + }, + { + "epoch": 2.649493243243243, + "grad_norm": 0.747870147228241, + "learning_rate": 9.402e-06, + "loss": 0.0948, + "step": 3137 + }, + { + "epoch": 2.6503378378378377, + "grad_norm": 1.101033329963684, + "learning_rate": 9.405e-06, + "loss": 0.1059, + "step": 3138 + }, + { + "epoch": 2.6511824324324325, + "grad_norm": 0.7259149551391602, + "learning_rate": 9.408e-06, + "loss": 0.0831, + "step": 3139 + }, + { + "epoch": 2.652027027027027, + "grad_norm": 0.7201205492019653, + "learning_rate": 9.411e-06, + "loss": 0.0876, + "step": 3140 + }, + { + "epoch": 2.6528716216216215, + "grad_norm": 0.7761155962944031, + "learning_rate": 9.414000000000002e-06, + "loss": 0.0947, + "step": 3141 + }, + { + "epoch": 2.6537162162162162, + "grad_norm": 0.9711384773254395, + "learning_rate": 9.417e-06, + "loss": 0.1286, + "step": 3142 + }, + { + "epoch": 2.654560810810811, + "grad_norm": 0.7221491932868958, + "learning_rate": 9.42e-06, + "loss": 0.0684, + "step": 3143 + }, + { + "epoch": 2.6554054054054053, + "grad_norm": 1.031461477279663, + "learning_rate": 9.423e-06, + "loss": 0.1003, + "step": 3144 + }, + { + "epoch": 2.65625, + "grad_norm": 0.7079136371612549, + "learning_rate": 9.426e-06, + "loss": 0.0862, + "step": 3145 + }, + { + "epoch": 2.6570945945945947, + "grad_norm": 0.7784949541091919, + "learning_rate": 9.429e-06, + "loss": 0.092, + "step": 3146 + }, + { + "epoch": 2.657939189189189, + "grad_norm": 0.7869747877120972, + "learning_rate": 9.432e-06, + "loss": 0.094, + "step": 3147 + }, + { + "epoch": 2.6587837837837838, + "grad_norm": 0.9629301428794861, + "learning_rate": 9.435e-06, + "loss": 0.0822, + "step": 3148 + }, + { + "epoch": 2.6596283783783785, + "grad_norm": 1.1387587785720825, + "learning_rate": 9.438e-06, + "loss": 0.0883, + "step": 3149 + }, + { + "epoch": 2.660472972972973, + "grad_norm": 0.6170675754547119, + "learning_rate": 9.441e-06, + "loss": 0.072, + "step": 3150 + }, + { + "epoch": 2.6613175675675675, + "grad_norm": 0.8044098019599915, + "learning_rate": 9.444000000000001e-06, + "loss": 0.0806, + "step": 3151 + }, + { + "epoch": 2.6621621621621623, + "grad_norm": 1.1386693716049194, + "learning_rate": 9.447000000000001e-06, + "loss": 0.0868, + "step": 3152 + }, + { + "epoch": 2.663006756756757, + "grad_norm": 0.9133210182189941, + "learning_rate": 9.450000000000001e-06, + "loss": 0.1091, + "step": 3153 + }, + { + "epoch": 2.6638513513513513, + "grad_norm": 0.9379620552062988, + "learning_rate": 9.453e-06, + "loss": 0.0932, + "step": 3154 + }, + { + "epoch": 2.664695945945946, + "grad_norm": 0.6876785755157471, + "learning_rate": 9.455999999999999e-06, + "loss": 0.0735, + "step": 3155 + }, + { + "epoch": 2.6655405405405403, + "grad_norm": 1.0956438779830933, + "learning_rate": 9.459e-06, + "loss": 0.0965, + "step": 3156 + }, + { + "epoch": 2.666385135135135, + "grad_norm": 0.9038230180740356, + "learning_rate": 9.462e-06, + "loss": 0.0729, + "step": 3157 + }, + { + "epoch": 2.66722972972973, + "grad_norm": 0.9221885204315186, + "learning_rate": 9.465e-06, + "loss": 0.1255, + "step": 3158 + }, + { + "epoch": 2.6680743243243246, + "grad_norm": 1.162077784538269, + "learning_rate": 9.468e-06, + "loss": 0.1118, + "step": 3159 + }, + { + "epoch": 2.668918918918919, + "grad_norm": 1.1854987144470215, + "learning_rate": 9.471e-06, + "loss": 0.0996, + "step": 3160 + }, + { + "epoch": 2.6697635135135136, + "grad_norm": 1.1305537223815918, + "learning_rate": 9.474000000000001e-06, + "loss": 0.1056, + "step": 3161 + }, + { + "epoch": 2.670608108108108, + "grad_norm": 1.1932629346847534, + "learning_rate": 9.477000000000001e-06, + "loss": 0.1072, + "step": 3162 + }, + { + "epoch": 2.6714527027027026, + "grad_norm": 1.1229028701782227, + "learning_rate": 9.48e-06, + "loss": 0.1377, + "step": 3163 + }, + { + "epoch": 2.6722972972972974, + "grad_norm": 1.1437885761260986, + "learning_rate": 9.483e-06, + "loss": 0.1244, + "step": 3164 + }, + { + "epoch": 2.673141891891892, + "grad_norm": 1.3316013813018799, + "learning_rate": 9.486e-06, + "loss": 0.1053, + "step": 3165 + }, + { + "epoch": 2.6739864864864864, + "grad_norm": 1.3423207998275757, + "learning_rate": 9.489000000000002e-06, + "loss": 0.1711, + "step": 3166 + }, + { + "epoch": 2.674831081081081, + "grad_norm": 2.2671926021575928, + "learning_rate": 9.492e-06, + "loss": 0.1762, + "step": 3167 + }, + { + "epoch": 2.6756756756756754, + "grad_norm": 2.2747974395751953, + "learning_rate": 9.495e-06, + "loss": 0.2474, + "step": 3168 + }, + { + "epoch": 2.67652027027027, + "grad_norm": 1.7111480236053467, + "learning_rate": 9.498e-06, + "loss": 0.5804, + "step": 3169 + }, + { + "epoch": 2.677364864864865, + "grad_norm": 1.5526338815689087, + "learning_rate": 9.501e-06, + "loss": 0.3605, + "step": 3170 + }, + { + "epoch": 2.6782094594594597, + "grad_norm": 1.1236683130264282, + "learning_rate": 9.504e-06, + "loss": 0.1578, + "step": 3171 + }, + { + "epoch": 2.679054054054054, + "grad_norm": 1.5862770080566406, + "learning_rate": 9.507e-06, + "loss": 0.1187, + "step": 3172 + }, + { + "epoch": 2.6798986486486487, + "grad_norm": 0.905281126499176, + "learning_rate": 9.51e-06, + "loss": 0.0966, + "step": 3173 + }, + { + "epoch": 2.680743243243243, + "grad_norm": 0.6318177580833435, + "learning_rate": 9.513e-06, + "loss": 0.0971, + "step": 3174 + }, + { + "epoch": 2.6815878378378377, + "grad_norm": 0.9492335915565491, + "learning_rate": 9.516e-06, + "loss": 0.1101, + "step": 3175 + }, + { + "epoch": 2.6824324324324325, + "grad_norm": 0.855905294418335, + "learning_rate": 9.519000000000002e-06, + "loss": 0.0881, + "step": 3176 + }, + { + "epoch": 2.683277027027027, + "grad_norm": 0.7462963461875916, + "learning_rate": 9.522000000000001e-06, + "loss": 0.0798, + "step": 3177 + }, + { + "epoch": 2.6841216216216215, + "grad_norm": 1.6483701467514038, + "learning_rate": 9.525000000000001e-06, + "loss": 0.1102, + "step": 3178 + }, + { + "epoch": 2.6849662162162162, + "grad_norm": 1.0039476156234741, + "learning_rate": 9.528e-06, + "loss": 0.0796, + "step": 3179 + }, + { + "epoch": 2.685810810810811, + "grad_norm": 0.9204238653182983, + "learning_rate": 9.530999999999999e-06, + "loss": 0.1064, + "step": 3180 + }, + { + "epoch": 2.6866554054054053, + "grad_norm": 0.7994691729545593, + "learning_rate": 9.534e-06, + "loss": 0.0633, + "step": 3181 + }, + { + "epoch": 2.6875, + "grad_norm": 0.9419965744018555, + "learning_rate": 9.537e-06, + "loss": 0.1074, + "step": 3182 + }, + { + "epoch": 2.6883445945945947, + "grad_norm": 0.8896024823188782, + "learning_rate": 9.54e-06, + "loss": 0.0845, + "step": 3183 + }, + { + "epoch": 2.689189189189189, + "grad_norm": 0.6563376188278198, + "learning_rate": 9.543e-06, + "loss": 0.0718, + "step": 3184 + }, + { + "epoch": 2.6900337837837838, + "grad_norm": 1.0952047109603882, + "learning_rate": 9.546e-06, + "loss": 0.0755, + "step": 3185 + }, + { + "epoch": 2.6908783783783785, + "grad_norm": 0.7403814196586609, + "learning_rate": 9.549000000000001e-06, + "loss": 0.0911, + "step": 3186 + }, + { + "epoch": 2.691722972972973, + "grad_norm": 0.7286202907562256, + "learning_rate": 9.552000000000001e-06, + "loss": 0.0892, + "step": 3187 + }, + { + "epoch": 2.6925675675675675, + "grad_norm": 0.8633921146392822, + "learning_rate": 9.555e-06, + "loss": 0.0924, + "step": 3188 + }, + { + "epoch": 2.6934121621621623, + "grad_norm": 0.7203841209411621, + "learning_rate": 9.558e-06, + "loss": 0.085, + "step": 3189 + }, + { + "epoch": 2.694256756756757, + "grad_norm": 2.0385689735412598, + "learning_rate": 9.561e-06, + "loss": 0.1051, + "step": 3190 + }, + { + "epoch": 2.6951013513513513, + "grad_norm": 0.989281952381134, + "learning_rate": 9.564e-06, + "loss": 0.0886, + "step": 3191 + }, + { + "epoch": 2.695945945945946, + "grad_norm": 0.5388309955596924, + "learning_rate": 9.567e-06, + "loss": 0.0753, + "step": 3192 + }, + { + "epoch": 2.6967905405405403, + "grad_norm": 0.7755555510520935, + "learning_rate": 9.57e-06, + "loss": 0.0814, + "step": 3193 + }, + { + "epoch": 2.697635135135135, + "grad_norm": 0.8103413581848145, + "learning_rate": 9.573e-06, + "loss": 0.0693, + "step": 3194 + }, + { + "epoch": 2.69847972972973, + "grad_norm": 1.022438645362854, + "learning_rate": 9.576e-06, + "loss": 0.0833, + "step": 3195 + }, + { + "epoch": 2.6993243243243246, + "grad_norm": 0.9201300740242004, + "learning_rate": 9.579e-06, + "loss": 0.0924, + "step": 3196 + }, + { + "epoch": 2.700168918918919, + "grad_norm": 0.9922829866409302, + "learning_rate": 9.582e-06, + "loss": 0.1125, + "step": 3197 + }, + { + "epoch": 2.7010135135135136, + "grad_norm": 1.4418340921401978, + "learning_rate": 9.585e-06, + "loss": 0.0934, + "step": 3198 + }, + { + "epoch": 2.701858108108108, + "grad_norm": 1.3023523092269897, + "learning_rate": 9.588e-06, + "loss": 0.1208, + "step": 3199 + }, + { + "epoch": 2.7027027027027026, + "grad_norm": 0.8171996474266052, + "learning_rate": 9.591e-06, + "loss": 0.0728, + "step": 3200 + }, + { + "epoch": 2.7035472972972974, + "grad_norm": 0.9413707256317139, + "learning_rate": 9.594e-06, + "loss": 0.1079, + "step": 3201 + }, + { + "epoch": 2.704391891891892, + "grad_norm": 0.6834462881088257, + "learning_rate": 9.597000000000001e-06, + "loss": 0.0532, + "step": 3202 + }, + { + "epoch": 2.7052364864864864, + "grad_norm": 0.8536336421966553, + "learning_rate": 9.600000000000001e-06, + "loss": 0.0853, + "step": 3203 + }, + { + "epoch": 2.706081081081081, + "grad_norm": 0.9527955055236816, + "learning_rate": 9.603e-06, + "loss": 0.1011, + "step": 3204 + }, + { + "epoch": 2.7069256756756754, + "grad_norm": 0.83083575963974, + "learning_rate": 9.606e-06, + "loss": 0.072, + "step": 3205 + }, + { + "epoch": 2.70777027027027, + "grad_norm": 1.0471270084381104, + "learning_rate": 9.608999999999999e-06, + "loss": 0.0995, + "step": 3206 + }, + { + "epoch": 2.708614864864865, + "grad_norm": 0.8046274185180664, + "learning_rate": 9.612e-06, + "loss": 0.0849, + "step": 3207 + }, + { + "epoch": 2.7094594594594597, + "grad_norm": 3.405776262283325, + "learning_rate": 9.615e-06, + "loss": 0.1309, + "step": 3208 + }, + { + "epoch": 2.710304054054054, + "grad_norm": 1.2264858484268188, + "learning_rate": 9.618e-06, + "loss": 0.0984, + "step": 3209 + }, + { + "epoch": 2.7111486486486487, + "grad_norm": 1.155169129371643, + "learning_rate": 9.621e-06, + "loss": 0.1138, + "step": 3210 + }, + { + "epoch": 2.711993243243243, + "grad_norm": 0.8534510731697083, + "learning_rate": 9.624e-06, + "loss": 0.0923, + "step": 3211 + }, + { + "epoch": 2.7128378378378377, + "grad_norm": 1.345462441444397, + "learning_rate": 9.627000000000001e-06, + "loss": 0.0937, + "step": 3212 + }, + { + "epoch": 2.7136824324324325, + "grad_norm": 1.198759913444519, + "learning_rate": 9.630000000000001e-06, + "loss": 0.1211, + "step": 3213 + }, + { + "epoch": 2.714527027027027, + "grad_norm": 1.1089823246002197, + "learning_rate": 9.633e-06, + "loss": 0.1105, + "step": 3214 + }, + { + "epoch": 2.7153716216216215, + "grad_norm": 1.0939542055130005, + "learning_rate": 9.636e-06, + "loss": 0.0997, + "step": 3215 + }, + { + "epoch": 2.7162162162162162, + "grad_norm": 1.5748969316482544, + "learning_rate": 9.638999999999999e-06, + "loss": 0.1498, + "step": 3216 + }, + { + "epoch": 2.717060810810811, + "grad_norm": 1.5078169107437134, + "learning_rate": 9.642e-06, + "loss": 0.1346, + "step": 3217 + }, + { + "epoch": 2.7179054054054053, + "grad_norm": 1.8760366439819336, + "learning_rate": 9.645e-06, + "loss": 0.1591, + "step": 3218 + }, + { + "epoch": 2.71875, + "grad_norm": 2.8373374938964844, + "learning_rate": 9.648e-06, + "loss": 0.6683, + "step": 3219 + }, + { + "epoch": 2.7195945945945947, + "grad_norm": 1.4421802759170532, + "learning_rate": 9.651e-06, + "loss": 0.3584, + "step": 3220 + }, + { + "epoch": 2.720439189189189, + "grad_norm": 1.1873387098312378, + "learning_rate": 9.654e-06, + "loss": 0.2485, + "step": 3221 + }, + { + "epoch": 2.7212837837837838, + "grad_norm": 1.2631088495254517, + "learning_rate": 9.657000000000001e-06, + "loss": 0.1208, + "step": 3222 + }, + { + "epoch": 2.7221283783783785, + "grad_norm": 0.8740649223327637, + "learning_rate": 9.66e-06, + "loss": 0.1211, + "step": 3223 + }, + { + "epoch": 2.722972972972973, + "grad_norm": 0.8374167084693909, + "learning_rate": 9.663e-06, + "loss": 0.1219, + "step": 3224 + }, + { + "epoch": 2.7238175675675675, + "grad_norm": 0.5563183426856995, + "learning_rate": 9.666e-06, + "loss": 0.0678, + "step": 3225 + }, + { + "epoch": 2.7246621621621623, + "grad_norm": 0.81444251537323, + "learning_rate": 9.669e-06, + "loss": 0.0878, + "step": 3226 + }, + { + "epoch": 2.725506756756757, + "grad_norm": 0.6933479309082031, + "learning_rate": 9.672000000000002e-06, + "loss": 0.0799, + "step": 3227 + }, + { + "epoch": 2.7263513513513513, + "grad_norm": 1.5713711977005005, + "learning_rate": 9.675e-06, + "loss": 0.0979, + "step": 3228 + }, + { + "epoch": 2.727195945945946, + "grad_norm": 1.0597630739212036, + "learning_rate": 9.678e-06, + "loss": 0.0961, + "step": 3229 + }, + { + "epoch": 2.7280405405405403, + "grad_norm": 0.7917459607124329, + "learning_rate": 9.681e-06, + "loss": 0.0751, + "step": 3230 + }, + { + "epoch": 2.728885135135135, + "grad_norm": 0.8058605790138245, + "learning_rate": 9.683999999999999e-06, + "loss": 0.081, + "step": 3231 + }, + { + "epoch": 2.72972972972973, + "grad_norm": 1.310734510421753, + "learning_rate": 9.687e-06, + "loss": 0.0749, + "step": 3232 + }, + { + "epoch": 2.7305743243243246, + "grad_norm": 0.6979947090148926, + "learning_rate": 9.69e-06, + "loss": 0.0761, + "step": 3233 + }, + { + "epoch": 2.731418918918919, + "grad_norm": 0.806571364402771, + "learning_rate": 9.693e-06, + "loss": 0.0727, + "step": 3234 + }, + { + "epoch": 2.7322635135135136, + "grad_norm": 0.7980080842971802, + "learning_rate": 9.696e-06, + "loss": 0.07, + "step": 3235 + }, + { + "epoch": 2.733108108108108, + "grad_norm": 0.7042094469070435, + "learning_rate": 9.699e-06, + "loss": 0.0716, + "step": 3236 + }, + { + "epoch": 2.7339527027027026, + "grad_norm": 0.9573781490325928, + "learning_rate": 9.702000000000001e-06, + "loss": 0.0964, + "step": 3237 + }, + { + "epoch": 2.7347972972972974, + "grad_norm": 0.9820629358291626, + "learning_rate": 9.705000000000001e-06, + "loss": 0.0985, + "step": 3238 + }, + { + "epoch": 2.735641891891892, + "grad_norm": 0.9164472222328186, + "learning_rate": 9.708000000000001e-06, + "loss": 0.0847, + "step": 3239 + }, + { + "epoch": 2.7364864864864864, + "grad_norm": 0.8627683520317078, + "learning_rate": 9.711e-06, + "loss": 0.0999, + "step": 3240 + }, + { + "epoch": 2.737331081081081, + "grad_norm": 1.0392229557037354, + "learning_rate": 9.713999999999999e-06, + "loss": 0.1512, + "step": 3241 + }, + { + "epoch": 2.7381756756756754, + "grad_norm": 0.8341125249862671, + "learning_rate": 9.717e-06, + "loss": 0.0965, + "step": 3242 + }, + { + "epoch": 2.73902027027027, + "grad_norm": 0.9071174263954163, + "learning_rate": 9.72e-06, + "loss": 0.0756, + "step": 3243 + }, + { + "epoch": 2.739864864864865, + "grad_norm": 0.8742268681526184, + "learning_rate": 9.723e-06, + "loss": 0.0942, + "step": 3244 + }, + { + "epoch": 2.7407094594594597, + "grad_norm": 1.147253155708313, + "learning_rate": 9.726e-06, + "loss": 0.1151, + "step": 3245 + }, + { + "epoch": 2.741554054054054, + "grad_norm": 0.5735572576522827, + "learning_rate": 9.729e-06, + "loss": 0.0527, + "step": 3246 + }, + { + "epoch": 2.7423986486486487, + "grad_norm": 0.8264737725257874, + "learning_rate": 9.732000000000001e-06, + "loss": 0.0966, + "step": 3247 + }, + { + "epoch": 2.743243243243243, + "grad_norm": 1.2649604082107544, + "learning_rate": 9.735e-06, + "loss": 0.1224, + "step": 3248 + }, + { + "epoch": 2.7440878378378377, + "grad_norm": 1.072597622871399, + "learning_rate": 9.738e-06, + "loss": 0.0794, + "step": 3249 + }, + { + "epoch": 2.7449324324324325, + "grad_norm": 0.8894184231758118, + "learning_rate": 9.741e-06, + "loss": 0.0892, + "step": 3250 + }, + { + "epoch": 2.745777027027027, + "grad_norm": 1.2136585712432861, + "learning_rate": 9.744e-06, + "loss": 0.1016, + "step": 3251 + }, + { + "epoch": 2.7466216216216215, + "grad_norm": 0.7999183535575867, + "learning_rate": 9.747000000000002e-06, + "loss": 0.0711, + "step": 3252 + }, + { + "epoch": 2.7474662162162162, + "grad_norm": 0.9031201004981995, + "learning_rate": 9.75e-06, + "loss": 0.0989, + "step": 3253 + }, + { + "epoch": 2.748310810810811, + "grad_norm": 0.832575798034668, + "learning_rate": 9.753e-06, + "loss": 0.0715, + "step": 3254 + }, + { + "epoch": 2.7491554054054053, + "grad_norm": 0.9419900178909302, + "learning_rate": 9.756e-06, + "loss": 0.1211, + "step": 3255 + }, + { + "epoch": 2.75, + "grad_norm": 0.7830213904380798, + "learning_rate": 9.759e-06, + "loss": 0.0807, + "step": 3256 + }, + { + "epoch": 2.7508445945945947, + "grad_norm": 0.7946850657463074, + "learning_rate": 9.762e-06, + "loss": 0.0815, + "step": 3257 + }, + { + "epoch": 2.751689189189189, + "grad_norm": 0.8308147192001343, + "learning_rate": 9.765e-06, + "loss": 0.098, + "step": 3258 + }, + { + "epoch": 2.7525337837837838, + "grad_norm": 1.0822819471359253, + "learning_rate": 9.768e-06, + "loss": 0.1029, + "step": 3259 + }, + { + "epoch": 2.7533783783783785, + "grad_norm": 1.0891026258468628, + "learning_rate": 9.771e-06, + "loss": 0.0898, + "step": 3260 + }, + { + "epoch": 2.754222972972973, + "grad_norm": 0.8523608446121216, + "learning_rate": 9.774e-06, + "loss": 0.1033, + "step": 3261 + }, + { + "epoch": 2.7550675675675675, + "grad_norm": 1.3845748901367188, + "learning_rate": 9.777000000000001e-06, + "loss": 0.1233, + "step": 3262 + }, + { + "epoch": 2.7559121621621623, + "grad_norm": 1.3049918413162231, + "learning_rate": 9.780000000000001e-06, + "loss": 0.1219, + "step": 3263 + }, + { + "epoch": 2.756756756756757, + "grad_norm": 1.3296433687210083, + "learning_rate": 9.783000000000001e-06, + "loss": 0.1023, + "step": 3264 + }, + { + "epoch": 2.7576013513513513, + "grad_norm": 1.4397697448730469, + "learning_rate": 9.785999999999999e-06, + "loss": 0.1201, + "step": 3265 + }, + { + "epoch": 2.758445945945946, + "grad_norm": 0.941826581954956, + "learning_rate": 9.788999999999999e-06, + "loss": 0.0875, + "step": 3266 + }, + { + "epoch": 2.7592905405405403, + "grad_norm": 1.7214038372039795, + "learning_rate": 9.792e-06, + "loss": 0.1339, + "step": 3267 + }, + { + "epoch": 2.760135135135135, + "grad_norm": 1.7376880645751953, + "learning_rate": 9.795e-06, + "loss": 0.1806, + "step": 3268 + }, + { + "epoch": 2.76097972972973, + "grad_norm": 2.991950035095215, + "learning_rate": 9.798e-06, + "loss": 0.5537, + "step": 3269 + }, + { + "epoch": 2.7618243243243246, + "grad_norm": 1.5179173946380615, + "learning_rate": 9.801e-06, + "loss": 0.3426, + "step": 3270 + }, + { + "epoch": 2.762668918918919, + "grad_norm": 1.1111342906951904, + "learning_rate": 9.804e-06, + "loss": 0.1893, + "step": 3271 + }, + { + "epoch": 2.7635135135135136, + "grad_norm": 0.8692975640296936, + "learning_rate": 9.807000000000001e-06, + "loss": 0.1366, + "step": 3272 + }, + { + "epoch": 2.764358108108108, + "grad_norm": 3.2981889247894287, + "learning_rate": 9.810000000000001e-06, + "loss": 0.0723, + "step": 3273 + }, + { + "epoch": 2.7652027027027026, + "grad_norm": 1.0286628007888794, + "learning_rate": 9.813e-06, + "loss": 0.0922, + "step": 3274 + }, + { + "epoch": 2.7660472972972974, + "grad_norm": 0.6240686178207397, + "learning_rate": 9.816e-06, + "loss": 0.076, + "step": 3275 + }, + { + "epoch": 2.766891891891892, + "grad_norm": 0.7388156056404114, + "learning_rate": 9.819e-06, + "loss": 0.0923, + "step": 3276 + }, + { + "epoch": 2.7677364864864864, + "grad_norm": 0.7214952111244202, + "learning_rate": 9.822e-06, + "loss": 0.0824, + "step": 3277 + }, + { + "epoch": 2.768581081081081, + "grad_norm": 1.0133520364761353, + "learning_rate": 9.825e-06, + "loss": 0.1089, + "step": 3278 + }, + { + "epoch": 2.7694256756756754, + "grad_norm": 0.8612502217292786, + "learning_rate": 9.828e-06, + "loss": 0.0985, + "step": 3279 + }, + { + "epoch": 2.77027027027027, + "grad_norm": 1.0002906322479248, + "learning_rate": 9.831e-06, + "loss": 0.0892, + "step": 3280 + }, + { + "epoch": 2.771114864864865, + "grad_norm": 0.861675500869751, + "learning_rate": 9.834e-06, + "loss": 0.0907, + "step": 3281 + }, + { + "epoch": 2.7719594594594597, + "grad_norm": 0.671066403388977, + "learning_rate": 9.837000000000001e-06, + "loss": 0.0636, + "step": 3282 + }, + { + "epoch": 2.772804054054054, + "grad_norm": 0.7711736559867859, + "learning_rate": 9.84e-06, + "loss": 0.07, + "step": 3283 + }, + { + "epoch": 2.7736486486486487, + "grad_norm": 1.3627318143844604, + "learning_rate": 9.843e-06, + "loss": 0.1336, + "step": 3284 + }, + { + "epoch": 2.774493243243243, + "grad_norm": 0.9104006886482239, + "learning_rate": 9.846e-06, + "loss": 0.1127, + "step": 3285 + }, + { + "epoch": 2.7753378378378377, + "grad_norm": 0.8128175735473633, + "learning_rate": 9.849e-06, + "loss": 0.1003, + "step": 3286 + }, + { + "epoch": 2.7761824324324325, + "grad_norm": 0.6402392983436584, + "learning_rate": 9.852000000000002e-06, + "loss": 0.0788, + "step": 3287 + }, + { + "epoch": 2.777027027027027, + "grad_norm": 0.7193582057952881, + "learning_rate": 9.855000000000001e-06, + "loss": 0.0703, + "step": 3288 + }, + { + "epoch": 2.7778716216216215, + "grad_norm": 0.9033893942832947, + "learning_rate": 9.858000000000001e-06, + "loss": 0.1006, + "step": 3289 + }, + { + "epoch": 2.7787162162162162, + "grad_norm": 0.8662302494049072, + "learning_rate": 9.861e-06, + "loss": 0.1005, + "step": 3290 + }, + { + "epoch": 2.779560810810811, + "grad_norm": 0.7612871527671814, + "learning_rate": 9.863999999999999e-06, + "loss": 0.0788, + "step": 3291 + }, + { + "epoch": 2.7804054054054053, + "grad_norm": 0.6161317229270935, + "learning_rate": 9.867e-06, + "loss": 0.0747, + "step": 3292 + }, + { + "epoch": 2.78125, + "grad_norm": 0.8395415544509888, + "learning_rate": 9.87e-06, + "loss": 0.0966, + "step": 3293 + }, + { + "epoch": 2.7820945945945947, + "grad_norm": 0.7243384718894958, + "learning_rate": 9.873e-06, + "loss": 0.0717, + "step": 3294 + }, + { + "epoch": 2.782939189189189, + "grad_norm": 1.046683430671692, + "learning_rate": 9.876e-06, + "loss": 0.0817, + "step": 3295 + }, + { + "epoch": 2.7837837837837838, + "grad_norm": 0.6792428493499756, + "learning_rate": 9.879e-06, + "loss": 0.0789, + "step": 3296 + }, + { + "epoch": 2.7846283783783785, + "grad_norm": 0.687483012676239, + "learning_rate": 9.882000000000001e-06, + "loss": 0.0716, + "step": 3297 + }, + { + "epoch": 2.785472972972973, + "grad_norm": 0.7949536442756653, + "learning_rate": 9.885000000000001e-06, + "loss": 0.0752, + "step": 3298 + }, + { + "epoch": 2.7863175675675675, + "grad_norm": 0.946696937084198, + "learning_rate": 9.888000000000001e-06, + "loss": 0.0935, + "step": 3299 + }, + { + "epoch": 2.7871621621621623, + "grad_norm": 1.8461703062057495, + "learning_rate": 9.891e-06, + "loss": 0.1268, + "step": 3300 + }, + { + "epoch": 2.788006756756757, + "grad_norm": 0.7919331192970276, + "learning_rate": 9.894e-06, + "loss": 0.0851, + "step": 3301 + }, + { + "epoch": 2.7888513513513513, + "grad_norm": 0.7708006501197815, + "learning_rate": 9.897e-06, + "loss": 0.0971, + "step": 3302 + }, + { + "epoch": 2.789695945945946, + "grad_norm": 1.36503005027771, + "learning_rate": 9.9e-06, + "loss": 0.1219, + "step": 3303 + }, + { + "epoch": 2.7905405405405403, + "grad_norm": 0.6696494817733765, + "learning_rate": 9.903e-06, + "loss": 0.0681, + "step": 3304 + }, + { + "epoch": 2.791385135135135, + "grad_norm": 0.8407897353172302, + "learning_rate": 9.906e-06, + "loss": 0.1014, + "step": 3305 + }, + { + "epoch": 2.79222972972973, + "grad_norm": 0.8491315245628357, + "learning_rate": 9.909e-06, + "loss": 0.1077, + "step": 3306 + }, + { + "epoch": 2.7930743243243246, + "grad_norm": 1.5238490104675293, + "learning_rate": 9.912000000000001e-06, + "loss": 0.1253, + "step": 3307 + }, + { + "epoch": 2.793918918918919, + "grad_norm": 0.8607914447784424, + "learning_rate": 9.915e-06, + "loss": 0.0967, + "step": 3308 + }, + { + "epoch": 2.7947635135135136, + "grad_norm": 0.9944217205047607, + "learning_rate": 9.918e-06, + "loss": 0.1084, + "step": 3309 + }, + { + "epoch": 2.795608108108108, + "grad_norm": 0.8241388201713562, + "learning_rate": 9.921e-06, + "loss": 0.0794, + "step": 3310 + }, + { + "epoch": 2.7964527027027026, + "grad_norm": 1.0703574419021606, + "learning_rate": 9.924e-06, + "loss": 0.103, + "step": 3311 + }, + { + "epoch": 2.7972972972972974, + "grad_norm": 1.3662360906600952, + "learning_rate": 9.927000000000002e-06, + "loss": 0.1268, + "step": 3312 + }, + { + "epoch": 2.798141891891892, + "grad_norm": 1.247449517250061, + "learning_rate": 9.930000000000001e-06, + "loss": 0.1316, + "step": 3313 + }, + { + "epoch": 2.7989864864864864, + "grad_norm": 1.909847378730774, + "learning_rate": 9.933e-06, + "loss": 0.1963, + "step": 3314 + }, + { + "epoch": 2.799831081081081, + "grad_norm": 1.3722666501998901, + "learning_rate": 9.936e-06, + "loss": 0.1322, + "step": 3315 + }, + { + "epoch": 2.8006756756756754, + "grad_norm": 1.2628569602966309, + "learning_rate": 9.939e-06, + "loss": 0.1189, + "step": 3316 + }, + { + "epoch": 2.80152027027027, + "grad_norm": 1.8682262897491455, + "learning_rate": 9.941999999999999e-06, + "loss": 0.1588, + "step": 3317 + }, + { + "epoch": 2.802364864864865, + "grad_norm": 2.7939419746398926, + "learning_rate": 9.945e-06, + "loss": 0.2097, + "step": 3318 + }, + { + "epoch": 2.8032094594594597, + "grad_norm": 2.404261350631714, + "learning_rate": 9.948e-06, + "loss": 0.4637, + "step": 3319 + }, + { + "epoch": 2.804054054054054, + "grad_norm": 1.0136052370071411, + "learning_rate": 9.951e-06, + "loss": 0.2716, + "step": 3320 + }, + { + "epoch": 2.8048986486486487, + "grad_norm": 1.6014478206634521, + "learning_rate": 9.954e-06, + "loss": 0.1475, + "step": 3321 + }, + { + "epoch": 2.805743243243243, + "grad_norm": 0.9165838360786438, + "learning_rate": 9.957e-06, + "loss": 0.1049, + "step": 3322 + }, + { + "epoch": 2.8065878378378377, + "grad_norm": 0.6828792691230774, + "learning_rate": 9.960000000000001e-06, + "loss": 0.0752, + "step": 3323 + }, + { + "epoch": 2.8074324324324325, + "grad_norm": 0.6978644132614136, + "learning_rate": 9.963000000000001e-06, + "loss": 0.0945, + "step": 3324 + }, + { + "epoch": 2.808277027027027, + "grad_norm": 0.7203534245491028, + "learning_rate": 9.966e-06, + "loss": 0.0825, + "step": 3325 + }, + { + "epoch": 2.8091216216216215, + "grad_norm": 0.690492570400238, + "learning_rate": 9.969e-06, + "loss": 0.088, + "step": 3326 + }, + { + "epoch": 2.8099662162162162, + "grad_norm": 0.5402799844741821, + "learning_rate": 9.971999999999999e-06, + "loss": 0.0725, + "step": 3327 + }, + { + "epoch": 2.810810810810811, + "grad_norm": 0.657356858253479, + "learning_rate": 9.975e-06, + "loss": 0.0907, + "step": 3328 + }, + { + "epoch": 2.8116554054054053, + "grad_norm": 0.7542479634284973, + "learning_rate": 9.978e-06, + "loss": 0.0789, + "step": 3329 + }, + { + "epoch": 2.8125, + "grad_norm": 0.921737015247345, + "learning_rate": 9.981e-06, + "loss": 0.1149, + "step": 3330 + }, + { + "epoch": 2.8133445945945947, + "grad_norm": 0.800340473651886, + "learning_rate": 9.984e-06, + "loss": 0.0968, + "step": 3331 + }, + { + "epoch": 2.814189189189189, + "grad_norm": 0.6008946299552917, + "learning_rate": 9.987e-06, + "loss": 0.0615, + "step": 3332 + }, + { + "epoch": 2.8150337837837838, + "grad_norm": 0.8296480774879456, + "learning_rate": 9.990000000000001e-06, + "loss": 0.0868, + "step": 3333 + }, + { + "epoch": 2.8158783783783785, + "grad_norm": 1.6744046211242676, + "learning_rate": 9.993e-06, + "loss": 0.0847, + "step": 3334 + }, + { + "epoch": 2.816722972972973, + "grad_norm": 0.5229650735855103, + "learning_rate": 9.996e-06, + "loss": 0.0565, + "step": 3335 + }, + { + "epoch": 2.8175675675675675, + "grad_norm": 0.791536271572113, + "learning_rate": 9.999e-06, + "loss": 0.0642, + "step": 3336 + }, + { + "epoch": 2.8184121621621623, + "grad_norm": 0.6474236249923706, + "learning_rate": 1.0002e-05, + "loss": 0.0667, + "step": 3337 + }, + { + "epoch": 2.819256756756757, + "grad_norm": 0.6491873264312744, + "learning_rate": 1.0005000000000002e-05, + "loss": 0.0686, + "step": 3338 + }, + { + "epoch": 2.8201013513513513, + "grad_norm": 0.5232314467430115, + "learning_rate": 1.0008e-05, + "loss": 0.0586, + "step": 3339 + }, + { + "epoch": 2.820945945945946, + "grad_norm": 0.7744026184082031, + "learning_rate": 1.0011e-05, + "loss": 0.0754, + "step": 3340 + }, + { + "epoch": 2.8217905405405403, + "grad_norm": 0.7474740743637085, + "learning_rate": 1.0014e-05, + "loss": 0.0615, + "step": 3341 + }, + { + "epoch": 2.822635135135135, + "grad_norm": 0.7615523934364319, + "learning_rate": 1.0016999999999999e-05, + "loss": 0.079, + "step": 3342 + }, + { + "epoch": 2.82347972972973, + "grad_norm": 1.2011457681655884, + "learning_rate": 1.002e-05, + "loss": 0.0846, + "step": 3343 + }, + { + "epoch": 2.8243243243243246, + "grad_norm": 0.9678772687911987, + "learning_rate": 1.0023e-05, + "loss": 0.087, + "step": 3344 + }, + { + "epoch": 2.825168918918919, + "grad_norm": 0.9442117810249329, + "learning_rate": 1.0026e-05, + "loss": 0.0869, + "step": 3345 + }, + { + "epoch": 2.8260135135135136, + "grad_norm": 0.7850094437599182, + "learning_rate": 1.0029e-05, + "loss": 0.075, + "step": 3346 + }, + { + "epoch": 2.826858108108108, + "grad_norm": 1.1055577993392944, + "learning_rate": 1.0032e-05, + "loss": 0.1163, + "step": 3347 + }, + { + "epoch": 2.8277027027027026, + "grad_norm": 0.9588947296142578, + "learning_rate": 1.0035000000000001e-05, + "loss": 0.0622, + "step": 3348 + }, + { + "epoch": 2.8285472972972974, + "grad_norm": 0.8478729128837585, + "learning_rate": 1.0038000000000001e-05, + "loss": 0.0647, + "step": 3349 + }, + { + "epoch": 2.829391891891892, + "grad_norm": 0.7079030871391296, + "learning_rate": 1.0041000000000001e-05, + "loss": 0.0985, + "step": 3350 + }, + { + "epoch": 2.8302364864864864, + "grad_norm": 1.6497864723205566, + "learning_rate": 1.0043999999999999e-05, + "loss": 0.1057, + "step": 3351 + }, + { + "epoch": 2.831081081081081, + "grad_norm": 0.7958265542984009, + "learning_rate": 1.0046999999999999e-05, + "loss": 0.096, + "step": 3352 + }, + { + "epoch": 2.8319256756756754, + "grad_norm": 0.8283199667930603, + "learning_rate": 1.005e-05, + "loss": 0.0888, + "step": 3353 + }, + { + "epoch": 2.83277027027027, + "grad_norm": 1.1724004745483398, + "learning_rate": 1.0053e-05, + "loss": 0.0875, + "step": 3354 + }, + { + "epoch": 2.833614864864865, + "grad_norm": 0.8301272392272949, + "learning_rate": 1.0056e-05, + "loss": 0.0932, + "step": 3355 + }, + { + "epoch": 2.8344594594594597, + "grad_norm": 0.6782865524291992, + "learning_rate": 1.0059e-05, + "loss": 0.0813, + "step": 3356 + }, + { + "epoch": 2.835304054054054, + "grad_norm": 1.2694246768951416, + "learning_rate": 1.0062e-05, + "loss": 0.1096, + "step": 3357 + }, + { + "epoch": 2.8361486486486487, + "grad_norm": 0.9070749878883362, + "learning_rate": 1.0065000000000001e-05, + "loss": 0.0888, + "step": 3358 + }, + { + "epoch": 2.836993243243243, + "grad_norm": 1.0518608093261719, + "learning_rate": 1.0068e-05, + "loss": 0.0796, + "step": 3359 + }, + { + "epoch": 2.8378378378378377, + "grad_norm": 0.8050525784492493, + "learning_rate": 1.0071e-05, + "loss": 0.0859, + "step": 3360 + }, + { + "epoch": 2.8386824324324325, + "grad_norm": 0.8248040080070496, + "learning_rate": 1.0074e-05, + "loss": 0.0832, + "step": 3361 + }, + { + "epoch": 2.839527027027027, + "grad_norm": 1.1178282499313354, + "learning_rate": 1.0077e-05, + "loss": 0.093, + "step": 3362 + }, + { + "epoch": 2.8403716216216215, + "grad_norm": 1.042966365814209, + "learning_rate": 1.008e-05, + "loss": 0.1151, + "step": 3363 + }, + { + "epoch": 2.8412162162162162, + "grad_norm": 1.4087938070297241, + "learning_rate": 1.0083e-05, + "loss": 0.1222, + "step": 3364 + }, + { + "epoch": 2.842060810810811, + "grad_norm": 1.499961256980896, + "learning_rate": 1.0086e-05, + "loss": 0.1097, + "step": 3365 + }, + { + "epoch": 2.8429054054054053, + "grad_norm": 1.4397423267364502, + "learning_rate": 1.0089e-05, + "loss": 0.1268, + "step": 3366 + }, + { + "epoch": 2.84375, + "grad_norm": 1.7848503589630127, + "learning_rate": 1.0092e-05, + "loss": 0.1589, + "step": 3367 + }, + { + "epoch": 2.8445945945945947, + "grad_norm": 2.4163219928741455, + "learning_rate": 1.0095e-05, + "loss": 0.1993, + "step": 3368 + }, + { + "epoch": 2.845439189189189, + "grad_norm": 1.6376619338989258, + "learning_rate": 1.0098e-05, + "loss": 0.5895, + "step": 3369 + }, + { + "epoch": 2.8462837837837838, + "grad_norm": 1.5046560764312744, + "learning_rate": 1.0101e-05, + "loss": 0.3086, + "step": 3370 + }, + { + "epoch": 2.8471283783783785, + "grad_norm": 0.8911560773849487, + "learning_rate": 1.0104e-05, + "loss": 0.1699, + "step": 3371 + }, + { + "epoch": 2.847972972972973, + "grad_norm": 0.9801562428474426, + "learning_rate": 1.0107e-05, + "loss": 0.1354, + "step": 3372 + }, + { + "epoch": 2.8488175675675675, + "grad_norm": 0.7977813482284546, + "learning_rate": 1.0110000000000001e-05, + "loss": 0.1575, + "step": 3373 + }, + { + "epoch": 2.8496621621621623, + "grad_norm": 0.7181064486503601, + "learning_rate": 1.0113000000000001e-05, + "loss": 0.0736, + "step": 3374 + }, + { + "epoch": 2.850506756756757, + "grad_norm": 0.8270705938339233, + "learning_rate": 1.0116000000000001e-05, + "loss": 0.0974, + "step": 3375 + }, + { + "epoch": 2.8513513513513513, + "grad_norm": 0.6918227672576904, + "learning_rate": 1.0119e-05, + "loss": 0.0854, + "step": 3376 + }, + { + "epoch": 2.852195945945946, + "grad_norm": 0.6811494827270508, + "learning_rate": 1.0121999999999999e-05, + "loss": 0.0783, + "step": 3377 + }, + { + "epoch": 2.8530405405405403, + "grad_norm": 0.6788722276687622, + "learning_rate": 1.0125e-05, + "loss": 0.0806, + "step": 3378 + }, + { + "epoch": 2.853885135135135, + "grad_norm": 0.6731625199317932, + "learning_rate": 1.0128e-05, + "loss": 0.0669, + "step": 3379 + }, + { + "epoch": 2.85472972972973, + "grad_norm": 0.6544719934463501, + "learning_rate": 1.0131e-05, + "loss": 0.0672, + "step": 3380 + }, + { + "epoch": 2.8555743243243246, + "grad_norm": 0.5565752387046814, + "learning_rate": 1.0134e-05, + "loss": 0.0639, + "step": 3381 + }, + { + "epoch": 2.856418918918919, + "grad_norm": 0.8388181328773499, + "learning_rate": 1.0137e-05, + "loss": 0.0936, + "step": 3382 + }, + { + "epoch": 2.8572635135135136, + "grad_norm": 0.6378377079963684, + "learning_rate": 1.0140000000000001e-05, + "loss": 0.0698, + "step": 3383 + }, + { + "epoch": 2.858108108108108, + "grad_norm": 0.6488668322563171, + "learning_rate": 1.0143000000000001e-05, + "loss": 0.0798, + "step": 3384 + }, + { + "epoch": 2.8589527027027026, + "grad_norm": 0.8208922147750854, + "learning_rate": 1.0146e-05, + "loss": 0.0724, + "step": 3385 + }, + { + "epoch": 2.8597972972972974, + "grad_norm": 0.7392420172691345, + "learning_rate": 1.0149e-05, + "loss": 0.0798, + "step": 3386 + }, + { + "epoch": 2.860641891891892, + "grad_norm": 0.6909862160682678, + "learning_rate": 1.0152e-05, + "loss": 0.0618, + "step": 3387 + }, + { + "epoch": 2.8614864864864864, + "grad_norm": 0.8742467761039734, + "learning_rate": 1.0155e-05, + "loss": 0.0784, + "step": 3388 + }, + { + "epoch": 2.862331081081081, + "grad_norm": 0.8129295706748962, + "learning_rate": 1.0158e-05, + "loss": 0.1128, + "step": 3389 + }, + { + "epoch": 2.8631756756756754, + "grad_norm": 0.8960872888565063, + "learning_rate": 1.0161e-05, + "loss": 0.0771, + "step": 3390 + }, + { + "epoch": 2.86402027027027, + "grad_norm": 0.7842119336128235, + "learning_rate": 1.0164e-05, + "loss": 0.0705, + "step": 3391 + }, + { + "epoch": 2.864864864864865, + "grad_norm": 0.516377329826355, + "learning_rate": 1.0167e-05, + "loss": 0.0659, + "step": 3392 + }, + { + "epoch": 2.8657094594594597, + "grad_norm": 0.8201648592948914, + "learning_rate": 1.0170000000000001e-05, + "loss": 0.0816, + "step": 3393 + }, + { + "epoch": 2.866554054054054, + "grad_norm": 1.7040419578552246, + "learning_rate": 1.0173e-05, + "loss": 0.1342, + "step": 3394 + }, + { + "epoch": 2.8673986486486487, + "grad_norm": 0.7005295157432556, + "learning_rate": 1.0176e-05, + "loss": 0.0669, + "step": 3395 + }, + { + "epoch": 2.868243243243243, + "grad_norm": 1.2946598529815674, + "learning_rate": 1.0179e-05, + "loss": 0.0861, + "step": 3396 + }, + { + "epoch": 2.8690878378378377, + "grad_norm": 0.9481815695762634, + "learning_rate": 1.0182e-05, + "loss": 0.0982, + "step": 3397 + }, + { + "epoch": 2.8699324324324325, + "grad_norm": 0.9576536417007446, + "learning_rate": 1.0185000000000002e-05, + "loss": 0.089, + "step": 3398 + }, + { + "epoch": 2.870777027027027, + "grad_norm": 0.6497342586517334, + "learning_rate": 1.0188000000000001e-05, + "loss": 0.0785, + "step": 3399 + }, + { + "epoch": 2.8716216216216215, + "grad_norm": 0.9415722489356995, + "learning_rate": 1.0191e-05, + "loss": 0.0952, + "step": 3400 + }, + { + "epoch": 2.8724662162162162, + "grad_norm": 0.7059310078620911, + "learning_rate": 1.0194e-05, + "loss": 0.094, + "step": 3401 + }, + { + "epoch": 2.873310810810811, + "grad_norm": 1.4102792739868164, + "learning_rate": 1.0196999999999999e-05, + "loss": 0.0993, + "step": 3402 + }, + { + "epoch": 2.8741554054054053, + "grad_norm": 0.6360430121421814, + "learning_rate": 1.02e-05, + "loss": 0.0734, + "step": 3403 + }, + { + "epoch": 2.875, + "grad_norm": 0.7320536375045776, + "learning_rate": 1.0203e-05, + "loss": 0.0655, + "step": 3404 + }, + { + "epoch": 2.8758445945945947, + "grad_norm": 0.9888918399810791, + "learning_rate": 1.0206e-05, + "loss": 0.1029, + "step": 3405 + }, + { + "epoch": 2.876689189189189, + "grad_norm": 0.6454761624336243, + "learning_rate": 1.0209e-05, + "loss": 0.0687, + "step": 3406 + }, + { + "epoch": 2.8775337837837838, + "grad_norm": 0.6487717032432556, + "learning_rate": 1.0212e-05, + "loss": 0.0569, + "step": 3407 + }, + { + "epoch": 2.8783783783783785, + "grad_norm": 1.1305099725723267, + "learning_rate": 1.0215000000000001e-05, + "loss": 0.0871, + "step": 3408 + }, + { + "epoch": 2.879222972972973, + "grad_norm": 1.1993130445480347, + "learning_rate": 1.0218000000000001e-05, + "loss": 0.096, + "step": 3409 + }, + { + "epoch": 2.8800675675675675, + "grad_norm": 0.8427866697311401, + "learning_rate": 1.0221000000000001e-05, + "loss": 0.0843, + "step": 3410 + }, + { + "epoch": 2.8809121621621623, + "grad_norm": 1.3009892702102661, + "learning_rate": 1.0224e-05, + "loss": 0.1004, + "step": 3411 + }, + { + "epoch": 2.881756756756757, + "grad_norm": 1.5449780225753784, + "learning_rate": 1.0227e-05, + "loss": 0.0971, + "step": 3412 + }, + { + "epoch": 2.8826013513513513, + "grad_norm": 1.4384641647338867, + "learning_rate": 1.023e-05, + "loss": 0.1203, + "step": 3413 + }, + { + "epoch": 2.883445945945946, + "grad_norm": 0.8775342702865601, + "learning_rate": 1.0233e-05, + "loss": 0.1085, + "step": 3414 + }, + { + "epoch": 2.8842905405405403, + "grad_norm": 1.1378270387649536, + "learning_rate": 1.0236e-05, + "loss": 0.1047, + "step": 3415 + }, + { + "epoch": 2.885135135135135, + "grad_norm": 0.7883628606796265, + "learning_rate": 1.0239e-05, + "loss": 0.0753, + "step": 3416 + }, + { + "epoch": 2.88597972972973, + "grad_norm": 1.658412218093872, + "learning_rate": 1.0242e-05, + "loss": 0.1552, + "step": 3417 + }, + { + "epoch": 2.8868243243243246, + "grad_norm": 1.5751179456710815, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.1407, + "step": 3418 + }, + { + "epoch": 2.887668918918919, + "grad_norm": 1.7975126504898071, + "learning_rate": 1.0248e-05, + "loss": 0.5142, + "step": 3419 + }, + { + "epoch": 2.8885135135135136, + "grad_norm": 1.1899195909500122, + "learning_rate": 1.0251e-05, + "loss": 0.3058, + "step": 3420 + }, + { + "epoch": 2.889358108108108, + "grad_norm": 0.9114620089530945, + "learning_rate": 1.0254e-05, + "loss": 0.1927, + "step": 3421 + }, + { + "epoch": 2.8902027027027026, + "grad_norm": 0.8076234459877014, + "learning_rate": 1.0257e-05, + "loss": 0.173, + "step": 3422 + }, + { + "epoch": 2.8910472972972974, + "grad_norm": 0.8724126219749451, + "learning_rate": 1.0260000000000002e-05, + "loss": 0.1249, + "step": 3423 + }, + { + "epoch": 2.891891891891892, + "grad_norm": 0.6291051506996155, + "learning_rate": 1.0263000000000002e-05, + "loss": 0.0811, + "step": 3424 + }, + { + "epoch": 2.8927364864864864, + "grad_norm": 0.7430391907691956, + "learning_rate": 1.0266e-05, + "loss": 0.1023, + "step": 3425 + }, + { + "epoch": 2.893581081081081, + "grad_norm": 0.8421326875686646, + "learning_rate": 1.0269e-05, + "loss": 0.117, + "step": 3426 + }, + { + "epoch": 2.8944256756756754, + "grad_norm": 0.6976498961448669, + "learning_rate": 1.0272e-05, + "loss": 0.1128, + "step": 3427 + }, + { + "epoch": 2.89527027027027, + "grad_norm": 0.8315488696098328, + "learning_rate": 1.0275e-05, + "loss": 0.1052, + "step": 3428 + }, + { + "epoch": 2.896114864864865, + "grad_norm": 0.8098863363265991, + "learning_rate": 1.0278e-05, + "loss": 0.0822, + "step": 3429 + }, + { + "epoch": 2.8969594594594597, + "grad_norm": 0.6690912246704102, + "learning_rate": 1.0281e-05, + "loss": 0.0936, + "step": 3430 + }, + { + "epoch": 2.897804054054054, + "grad_norm": 0.6136941313743591, + "learning_rate": 1.0284e-05, + "loss": 0.0748, + "step": 3431 + }, + { + "epoch": 2.8986486486486487, + "grad_norm": 0.7135041952133179, + "learning_rate": 1.0287e-05, + "loss": 0.0726, + "step": 3432 + }, + { + "epoch": 2.899493243243243, + "grad_norm": 0.8507094979286194, + "learning_rate": 1.0290000000000001e-05, + "loss": 0.073, + "step": 3433 + }, + { + "epoch": 2.9003378378378377, + "grad_norm": 0.6009098887443542, + "learning_rate": 1.0293000000000001e-05, + "loss": 0.0724, + "step": 3434 + }, + { + "epoch": 2.9011824324324325, + "grad_norm": 1.0162639617919922, + "learning_rate": 1.0296000000000001e-05, + "loss": 0.0919, + "step": 3435 + }, + { + "epoch": 2.902027027027027, + "grad_norm": 0.7270508408546448, + "learning_rate": 1.0299e-05, + "loss": 0.0814, + "step": 3436 + }, + { + "epoch": 2.9028716216216215, + "grad_norm": 0.834464430809021, + "learning_rate": 1.0301999999999999e-05, + "loss": 0.1126, + "step": 3437 + }, + { + "epoch": 2.9037162162162162, + "grad_norm": 0.7493636012077332, + "learning_rate": 1.0305e-05, + "loss": 0.0986, + "step": 3438 + }, + { + "epoch": 2.904560810810811, + "grad_norm": 0.769964873790741, + "learning_rate": 1.0308e-05, + "loss": 0.081, + "step": 3439 + }, + { + "epoch": 2.9054054054054053, + "grad_norm": 0.6033273339271545, + "learning_rate": 1.0311e-05, + "loss": 0.0752, + "step": 3440 + }, + { + "epoch": 2.90625, + "grad_norm": 0.669103741645813, + "learning_rate": 1.0314e-05, + "loss": 0.0796, + "step": 3441 + }, + { + "epoch": 2.9070945945945947, + "grad_norm": 0.7212713360786438, + "learning_rate": 1.0317e-05, + "loss": 0.074, + "step": 3442 + }, + { + "epoch": 2.907939189189189, + "grad_norm": 0.8450162410736084, + "learning_rate": 1.032e-05, + "loss": 0.0687, + "step": 3443 + }, + { + "epoch": 2.9087837837837838, + "grad_norm": 0.8931389451026917, + "learning_rate": 1.0323000000000001e-05, + "loss": 0.0761, + "step": 3444 + }, + { + "epoch": 2.9096283783783785, + "grad_norm": 0.7757656574249268, + "learning_rate": 1.0326e-05, + "loss": 0.0849, + "step": 3445 + }, + { + "epoch": 2.910472972972973, + "grad_norm": 1.066131591796875, + "learning_rate": 1.0329e-05, + "loss": 0.1164, + "step": 3446 + }, + { + "epoch": 2.9113175675675675, + "grad_norm": 0.9389185309410095, + "learning_rate": 1.0332e-05, + "loss": 0.084, + "step": 3447 + }, + { + "epoch": 2.9121621621621623, + "grad_norm": 1.4284038543701172, + "learning_rate": 1.0335e-05, + "loss": 0.1271, + "step": 3448 + }, + { + "epoch": 2.913006756756757, + "grad_norm": 0.58078533411026, + "learning_rate": 1.0338e-05, + "loss": 0.0551, + "step": 3449 + }, + { + "epoch": 2.9138513513513513, + "grad_norm": 1.0042170286178589, + "learning_rate": 1.0341e-05, + "loss": 0.1023, + "step": 3450 + }, + { + "epoch": 2.914695945945946, + "grad_norm": 0.5930121541023254, + "learning_rate": 1.0344e-05, + "loss": 0.0707, + "step": 3451 + }, + { + "epoch": 2.9155405405405403, + "grad_norm": 0.9095163345336914, + "learning_rate": 1.0347e-05, + "loss": 0.0839, + "step": 3452 + }, + { + "epoch": 2.916385135135135, + "grad_norm": 1.0387717485427856, + "learning_rate": 1.035e-05, + "loss": 0.0871, + "step": 3453 + }, + { + "epoch": 2.91722972972973, + "grad_norm": 0.6514203548431396, + "learning_rate": 1.0353e-05, + "loss": 0.0632, + "step": 3454 + }, + { + "epoch": 2.9180743243243246, + "grad_norm": 0.963912308216095, + "learning_rate": 1.0356e-05, + "loss": 0.0986, + "step": 3455 + }, + { + "epoch": 2.918918918918919, + "grad_norm": 0.8278841972351074, + "learning_rate": 1.0359e-05, + "loss": 0.0659, + "step": 3456 + }, + { + "epoch": 2.9197635135135136, + "grad_norm": 0.8898023962974548, + "learning_rate": 1.0362e-05, + "loss": 0.0893, + "step": 3457 + }, + { + "epoch": 2.920608108108108, + "grad_norm": 0.9282484650611877, + "learning_rate": 1.0365e-05, + "loss": 0.0846, + "step": 3458 + }, + { + "epoch": 2.9214527027027026, + "grad_norm": 3.156475305557251, + "learning_rate": 1.0368000000000001e-05, + "loss": 0.1184, + "step": 3459 + }, + { + "epoch": 2.9222972972972974, + "grad_norm": 0.9231650829315186, + "learning_rate": 1.0371000000000001e-05, + "loss": 0.09, + "step": 3460 + }, + { + "epoch": 2.923141891891892, + "grad_norm": 0.8800138831138611, + "learning_rate": 1.0374000000000001e-05, + "loss": 0.101, + "step": 3461 + }, + { + "epoch": 2.9239864864864864, + "grad_norm": 1.0366672277450562, + "learning_rate": 1.0376999999999999e-05, + "loss": 0.1254, + "step": 3462 + }, + { + "epoch": 2.924831081081081, + "grad_norm": 0.8471466898918152, + "learning_rate": 1.0379999999999999e-05, + "loss": 0.0761, + "step": 3463 + }, + { + "epoch": 2.9256756756756754, + "grad_norm": 1.607799768447876, + "learning_rate": 1.0383e-05, + "loss": 0.1228, + "step": 3464 + }, + { + "epoch": 2.92652027027027, + "grad_norm": 1.1490716934204102, + "learning_rate": 1.0386e-05, + "loss": 0.0835, + "step": 3465 + }, + { + "epoch": 2.927364864864865, + "grad_norm": 1.5762444734573364, + "learning_rate": 1.0389e-05, + "loss": 0.1741, + "step": 3466 + }, + { + "epoch": 2.9282094594594597, + "grad_norm": 1.391764760017395, + "learning_rate": 1.0392e-05, + "loss": 0.1492, + "step": 3467 + }, + { + "epoch": 2.929054054054054, + "grad_norm": 1.253175973892212, + "learning_rate": 1.0395e-05, + "loss": 0.1389, + "step": 3468 + }, + { + "epoch": 2.9298986486486487, + "grad_norm": 2.1961023807525635, + "learning_rate": 1.0398000000000001e-05, + "loss": 0.5859, + "step": 3469 + }, + { + "epoch": 2.930743243243243, + "grad_norm": 3.3960893154144287, + "learning_rate": 1.0401000000000001e-05, + "loss": 0.3401, + "step": 3470 + }, + { + "epoch": 2.9315878378378377, + "grad_norm": 0.9170448780059814, + "learning_rate": 1.0404e-05, + "loss": 0.1305, + "step": 3471 + }, + { + "epoch": 2.9324324324324325, + "grad_norm": 0.7001641988754272, + "learning_rate": 1.0407e-05, + "loss": 0.0903, + "step": 3472 + }, + { + "epoch": 2.933277027027027, + "grad_norm": 1.004518985748291, + "learning_rate": 1.041e-05, + "loss": 0.1211, + "step": 3473 + }, + { + "epoch": 2.9341216216216215, + "grad_norm": 0.5295295119285583, + "learning_rate": 1.0413e-05, + "loss": 0.0757, + "step": 3474 + }, + { + "epoch": 2.9349662162162162, + "grad_norm": 0.7906724810600281, + "learning_rate": 1.0416e-05, + "loss": 0.1233, + "step": 3475 + }, + { + "epoch": 2.935810810810811, + "grad_norm": 0.7852138876914978, + "learning_rate": 1.0419e-05, + "loss": 0.0844, + "step": 3476 + }, + { + "epoch": 2.9366554054054053, + "grad_norm": 0.7858715057373047, + "learning_rate": 1.0422e-05, + "loss": 0.0986, + "step": 3477 + }, + { + "epoch": 2.9375, + "grad_norm": 0.7068582773208618, + "learning_rate": 1.0425e-05, + "loss": 0.0905, + "step": 3478 + }, + { + "epoch": 2.9383445945945947, + "grad_norm": 0.5905661582946777, + "learning_rate": 1.0428e-05, + "loss": 0.0632, + "step": 3479 + }, + { + "epoch": 2.939189189189189, + "grad_norm": 0.6125189661979675, + "learning_rate": 1.0431e-05, + "loss": 0.0601, + "step": 3480 + }, + { + "epoch": 2.9400337837837838, + "grad_norm": 1.0887250900268555, + "learning_rate": 1.0434e-05, + "loss": 0.0766, + "step": 3481 + }, + { + "epoch": 2.9408783783783785, + "grad_norm": 0.7410288453102112, + "learning_rate": 1.0437e-05, + "loss": 0.0819, + "step": 3482 + }, + { + "epoch": 2.941722972972973, + "grad_norm": 0.7250632047653198, + "learning_rate": 1.044e-05, + "loss": 0.0671, + "step": 3483 + }, + { + "epoch": 2.9425675675675675, + "grad_norm": 0.6763442158699036, + "learning_rate": 1.0443000000000001e-05, + "loss": 0.0694, + "step": 3484 + }, + { + "epoch": 2.9434121621621623, + "grad_norm": 0.8462119102478027, + "learning_rate": 1.0446000000000001e-05, + "loss": 0.111, + "step": 3485 + }, + { + "epoch": 2.944256756756757, + "grad_norm": 0.7139792442321777, + "learning_rate": 1.0449e-05, + "loss": 0.0887, + "step": 3486 + }, + { + "epoch": 2.9451013513513513, + "grad_norm": 0.7319249510765076, + "learning_rate": 1.0452e-05, + "loss": 0.0667, + "step": 3487 + }, + { + "epoch": 2.945945945945946, + "grad_norm": 0.5384945273399353, + "learning_rate": 1.0454999999999999e-05, + "loss": 0.0555, + "step": 3488 + }, + { + "epoch": 2.9467905405405403, + "grad_norm": 0.5964032411575317, + "learning_rate": 1.0458e-05, + "loss": 0.0634, + "step": 3489 + }, + { + "epoch": 2.947635135135135, + "grad_norm": 0.6884199976921082, + "learning_rate": 1.0461e-05, + "loss": 0.0675, + "step": 3490 + }, + { + "epoch": 2.94847972972973, + "grad_norm": 0.7979791760444641, + "learning_rate": 1.0464e-05, + "loss": 0.0968, + "step": 3491 + }, + { + "epoch": 2.9493243243243246, + "grad_norm": 0.7932937145233154, + "learning_rate": 1.0467e-05, + "loss": 0.1165, + "step": 3492 + }, + { + "epoch": 2.950168918918919, + "grad_norm": 0.6749271154403687, + "learning_rate": 1.047e-05, + "loss": 0.0642, + "step": 3493 + }, + { + "epoch": 2.9510135135135136, + "grad_norm": 0.6862701177597046, + "learning_rate": 1.0473000000000001e-05, + "loss": 0.0802, + "step": 3494 + }, + { + "epoch": 2.951858108108108, + "grad_norm": 0.874779999256134, + "learning_rate": 1.0476000000000001e-05, + "loss": 0.0891, + "step": 3495 + }, + { + "epoch": 2.9527027027027026, + "grad_norm": 1.000909447669983, + "learning_rate": 1.0479e-05, + "loss": 0.0895, + "step": 3496 + }, + { + "epoch": 2.9535472972972974, + "grad_norm": 0.5520820021629333, + "learning_rate": 1.0482e-05, + "loss": 0.0527, + "step": 3497 + }, + { + "epoch": 2.954391891891892, + "grad_norm": 1.013612151145935, + "learning_rate": 1.0485e-05, + "loss": 0.0915, + "step": 3498 + }, + { + "epoch": 2.9552364864864864, + "grad_norm": 0.6926777958869934, + "learning_rate": 1.0488e-05, + "loss": 0.0803, + "step": 3499 + }, + { + "epoch": 2.956081081081081, + "grad_norm": 0.8165414929389954, + "learning_rate": 1.0491e-05, + "loss": 0.1029, + "step": 3500 + }, + { + "epoch": 2.9569256756756754, + "grad_norm": 0.8850348591804504, + "learning_rate": 1.0494e-05, + "loss": 0.0937, + "step": 3501 + }, + { + "epoch": 2.95777027027027, + "grad_norm": 1.4627050161361694, + "learning_rate": 1.0497e-05, + "loss": 0.1124, + "step": 3502 + }, + { + "epoch": 2.958614864864865, + "grad_norm": 0.8144027590751648, + "learning_rate": 1.05e-05, + "loss": 0.0963, + "step": 3503 + }, + { + "epoch": 2.9594594594594597, + "grad_norm": 0.6915841698646545, + "learning_rate": 1.0503000000000001e-05, + "loss": 0.0766, + "step": 3504 + }, + { + "epoch": 2.960304054054054, + "grad_norm": 1.0030858516693115, + "learning_rate": 1.0506e-05, + "loss": 0.1393, + "step": 3505 + }, + { + "epoch": 2.9611486486486487, + "grad_norm": 0.724112868309021, + "learning_rate": 1.0509e-05, + "loss": 0.0979, + "step": 3506 + }, + { + "epoch": 2.961993243243243, + "grad_norm": 0.7846480011940002, + "learning_rate": 1.0512e-05, + "loss": 0.0901, + "step": 3507 + }, + { + "epoch": 2.9628378378378377, + "grad_norm": 0.7791213989257812, + "learning_rate": 1.0515e-05, + "loss": 0.0852, + "step": 3508 + }, + { + "epoch": 2.9636824324324325, + "grad_norm": 1.1541637182235718, + "learning_rate": 1.0518000000000002e-05, + "loss": 0.1005, + "step": 3509 + }, + { + "epoch": 2.964527027027027, + "grad_norm": 0.7436721920967102, + "learning_rate": 1.0521000000000001e-05, + "loss": 0.0812, + "step": 3510 + }, + { + "epoch": 2.9653716216216215, + "grad_norm": 0.7285637259483337, + "learning_rate": 1.0524e-05, + "loss": 0.0763, + "step": 3511 + }, + { + "epoch": 2.9662162162162162, + "grad_norm": 1.0141345262527466, + "learning_rate": 1.0527e-05, + "loss": 0.094, + "step": 3512 + }, + { + "epoch": 2.967060810810811, + "grad_norm": 1.330171823501587, + "learning_rate": 1.0529999999999999e-05, + "loss": 0.1013, + "step": 3513 + }, + { + "epoch": 2.9679054054054053, + "grad_norm": 0.944522500038147, + "learning_rate": 1.0533e-05, + "loss": 0.0858, + "step": 3514 + }, + { + "epoch": 2.96875, + "grad_norm": 1.1692100763320923, + "learning_rate": 1.0536e-05, + "loss": 0.1123, + "step": 3515 + }, + { + "epoch": 2.9695945945945947, + "grad_norm": 1.7135335206985474, + "learning_rate": 1.0539e-05, + "loss": 0.169, + "step": 3516 + }, + { + "epoch": 2.970439189189189, + "grad_norm": 1.5520994663238525, + "learning_rate": 1.0542e-05, + "loss": 0.1492, + "step": 3517 + }, + { + "epoch": 2.9712837837837838, + "grad_norm": 1.647302508354187, + "learning_rate": 1.0545e-05, + "loss": 0.178, + "step": 3518 + }, + { + "epoch": 2.9721283783783785, + "grad_norm": 2.4644577503204346, + "learning_rate": 1.0548000000000001e-05, + "loss": 0.493, + "step": 3519 + }, + { + "epoch": 2.972972972972973, + "grad_norm": 1.2664122581481934, + "learning_rate": 1.0551000000000001e-05, + "loss": 0.2137, + "step": 3520 + }, + { + "epoch": 2.9738175675675675, + "grad_norm": 0.8122864365577698, + "learning_rate": 1.0554000000000001e-05, + "loss": 0.1291, + "step": 3521 + }, + { + "epoch": 2.9746621621621623, + "grad_norm": 0.9360572695732117, + "learning_rate": 1.0557e-05, + "loss": 0.115, + "step": 3522 + }, + { + "epoch": 2.975506756756757, + "grad_norm": 0.5972496867179871, + "learning_rate": 1.0559999999999999e-05, + "loss": 0.0919, + "step": 3523 + }, + { + "epoch": 2.9763513513513513, + "grad_norm": 0.7249741554260254, + "learning_rate": 1.0563e-05, + "loss": 0.0789, + "step": 3524 + }, + { + "epoch": 2.977195945945946, + "grad_norm": 0.6787363886833191, + "learning_rate": 1.0566e-05, + "loss": 0.0893, + "step": 3525 + }, + { + "epoch": 2.9780405405405403, + "grad_norm": 0.8924059867858887, + "learning_rate": 1.0569e-05, + "loss": 0.1151, + "step": 3526 + }, + { + "epoch": 2.978885135135135, + "grad_norm": 0.9737341403961182, + "learning_rate": 1.0572e-05, + "loss": 0.0838, + "step": 3527 + }, + { + "epoch": 2.97972972972973, + "grad_norm": 0.7725181579589844, + "learning_rate": 1.0575e-05, + "loss": 0.077, + "step": 3528 + }, + { + "epoch": 2.9805743243243246, + "grad_norm": 0.7803391814231873, + "learning_rate": 1.0578000000000001e-05, + "loss": 0.0766, + "step": 3529 + }, + { + "epoch": 2.981418918918919, + "grad_norm": 0.8163331747055054, + "learning_rate": 1.0581e-05, + "loss": 0.0784, + "step": 3530 + }, + { + "epoch": 2.9822635135135136, + "grad_norm": 0.6827102303504944, + "learning_rate": 1.0584e-05, + "loss": 0.0617, + "step": 3531 + }, + { + "epoch": 2.983108108108108, + "grad_norm": 0.7189823389053345, + "learning_rate": 1.0587e-05, + "loss": 0.0711, + "step": 3532 + }, + { + "epoch": 2.9839527027027026, + "grad_norm": 0.7950515151023865, + "learning_rate": 1.059e-05, + "loss": 0.0664, + "step": 3533 + }, + { + "epoch": 2.9847972972972974, + "grad_norm": 0.8320746421813965, + "learning_rate": 1.0593000000000002e-05, + "loss": 0.0982, + "step": 3534 + }, + { + "epoch": 2.985641891891892, + "grad_norm": 0.9675417542457581, + "learning_rate": 1.0596e-05, + "loss": 0.0831, + "step": 3535 + }, + { + "epoch": 2.9864864864864864, + "grad_norm": 0.8788098692893982, + "learning_rate": 1.0599e-05, + "loss": 0.0713, + "step": 3536 + }, + { + "epoch": 2.987331081081081, + "grad_norm": 0.9957739114761353, + "learning_rate": 1.0602e-05, + "loss": 0.1085, + "step": 3537 + }, + { + "epoch": 2.9881756756756754, + "grad_norm": 0.7743231654167175, + "learning_rate": 1.0605e-05, + "loss": 0.0841, + "step": 3538 + }, + { + "epoch": 2.98902027027027, + "grad_norm": 0.7211588025093079, + "learning_rate": 1.0608e-05, + "loss": 0.0647, + "step": 3539 + }, + { + "epoch": 2.989864864864865, + "grad_norm": 0.8070222735404968, + "learning_rate": 1.0611e-05, + "loss": 0.0859, + "step": 3540 + }, + { + "epoch": 2.9907094594594597, + "grad_norm": 0.9206874966621399, + "learning_rate": 1.0614e-05, + "loss": 0.0838, + "step": 3541 + }, + { + "epoch": 2.991554054054054, + "grad_norm": 1.466158390045166, + "learning_rate": 1.0617e-05, + "loss": 0.1424, + "step": 3542 + }, + { + "epoch": 2.9923986486486487, + "grad_norm": 0.9903789162635803, + "learning_rate": 1.062e-05, + "loss": 0.0814, + "step": 3543 + }, + { + "epoch": 2.993243243243243, + "grad_norm": 0.9994624853134155, + "learning_rate": 1.0623000000000001e-05, + "loss": 0.0707, + "step": 3544 + }, + { + "epoch": 2.9940878378378377, + "grad_norm": 0.9449057579040527, + "learning_rate": 1.0626000000000001e-05, + "loss": 0.1135, + "step": 3545 + }, + { + "epoch": 2.9949324324324325, + "grad_norm": 1.3281055688858032, + "learning_rate": 1.0629000000000001e-05, + "loss": 0.1077, + "step": 3546 + }, + { + "epoch": 2.995777027027027, + "grad_norm": 0.8572678565979004, + "learning_rate": 1.0632000000000001e-05, + "loss": 0.085, + "step": 3547 + }, + { + "epoch": 2.9966216216216215, + "grad_norm": 1.144513487815857, + "learning_rate": 1.0634999999999999e-05, + "loss": 0.1177, + "step": 3548 + }, + { + "epoch": 2.9974662162162162, + "grad_norm": 1.2517569065093994, + "learning_rate": 1.0638e-05, + "loss": 0.1086, + "step": 3549 + }, + { + "epoch": 2.998310810810811, + "grad_norm": 1.6033227443695068, + "learning_rate": 1.0641e-05, + "loss": 0.1063, + "step": 3550 + }, + { + "epoch": 2.9991554054054053, + "grad_norm": 1.0003184080123901, + "learning_rate": 1.0644e-05, + "loss": 0.1136, + "step": 3551 + }, + { + "epoch": 3.0, + "grad_norm": 2.116436719894409, + "learning_rate": 1.0647e-05, + "loss": 0.1838, + "step": 3552 + }, + { + "epoch": 3.0008445945945947, + "grad_norm": 11.225780487060547, + "learning_rate": 1.065e-05, + "loss": 0.6159, + "step": 3553 + }, + { + "epoch": 3.001689189189189, + "grad_norm": 1.3245024681091309, + "learning_rate": 1.0653000000000001e-05, + "loss": 0.4169, + "step": 3554 + }, + { + "epoch": 3.0025337837837838, + "grad_norm": 1.1535181999206543, + "learning_rate": 1.0656000000000001e-05, + "loss": 0.1989, + "step": 3555 + }, + { + "epoch": 3.0033783783783785, + "grad_norm": 0.8594576716423035, + "learning_rate": 1.0659e-05, + "loss": 0.1196, + "step": 3556 + }, + { + "epoch": 3.004222972972973, + "grad_norm": 0.6592747569084167, + "learning_rate": 1.0662e-05, + "loss": 0.0831, + "step": 3557 + }, + { + "epoch": 3.0050675675675675, + "grad_norm": 0.6606388688087463, + "learning_rate": 1.0665e-05, + "loss": 0.0983, + "step": 3558 + }, + { + "epoch": 3.0059121621621623, + "grad_norm": 0.6036314368247986, + "learning_rate": 1.0668000000000002e-05, + "loss": 0.0825, + "step": 3559 + }, + { + "epoch": 3.0067567567567566, + "grad_norm": 0.7002618312835693, + "learning_rate": 1.0671e-05, + "loss": 0.0958, + "step": 3560 + }, + { + "epoch": 3.0076013513513513, + "grad_norm": 0.6643809080123901, + "learning_rate": 1.0674e-05, + "loss": 0.0701, + "step": 3561 + }, + { + "epoch": 3.008445945945946, + "grad_norm": 0.6779912114143372, + "learning_rate": 1.0677e-05, + "loss": 0.092, + "step": 3562 + }, + { + "epoch": 3.0092905405405403, + "grad_norm": 0.5765418410301208, + "learning_rate": 1.068e-05, + "loss": 0.0754, + "step": 3563 + }, + { + "epoch": 3.010135135135135, + "grad_norm": 0.8048807978630066, + "learning_rate": 1.0683000000000001e-05, + "loss": 0.0874, + "step": 3564 + }, + { + "epoch": 3.01097972972973, + "grad_norm": 0.7539348006248474, + "learning_rate": 1.0686e-05, + "loss": 0.0783, + "step": 3565 + }, + { + "epoch": 3.011824324324324, + "grad_norm": 0.646503210067749, + "learning_rate": 1.0689e-05, + "loss": 0.0629, + "step": 3566 + }, + { + "epoch": 3.012668918918919, + "grad_norm": 0.9699810743331909, + "learning_rate": 1.0692e-05, + "loss": 0.0852, + "step": 3567 + }, + { + "epoch": 3.0135135135135136, + "grad_norm": 0.5924709439277649, + "learning_rate": 1.0695e-05, + "loss": 0.065, + "step": 3568 + }, + { + "epoch": 3.014358108108108, + "grad_norm": 0.5934966206550598, + "learning_rate": 1.0698e-05, + "loss": 0.0671, + "step": 3569 + }, + { + "epoch": 3.0152027027027026, + "grad_norm": 0.8413817286491394, + "learning_rate": 1.0701000000000001e-05, + "loss": 0.1129, + "step": 3570 + }, + { + "epoch": 3.0160472972972974, + "grad_norm": 0.7114881873130798, + "learning_rate": 1.0704000000000001e-05, + "loss": 0.0605, + "step": 3571 + }, + { + "epoch": 3.016891891891892, + "grad_norm": 0.9058358669281006, + "learning_rate": 1.0707e-05, + "loss": 0.0729, + "step": 3572 + }, + { + "epoch": 3.0177364864864864, + "grad_norm": 0.6633976697921753, + "learning_rate": 1.0709999999999999e-05, + "loss": 0.0712, + "step": 3573 + }, + { + "epoch": 3.018581081081081, + "grad_norm": 0.6108345985412598, + "learning_rate": 1.0712999999999999e-05, + "loss": 0.0743, + "step": 3574 + }, + { + "epoch": 3.019425675675676, + "grad_norm": 0.6011790037155151, + "learning_rate": 1.0716e-05, + "loss": 0.0706, + "step": 3575 + }, + { + "epoch": 3.02027027027027, + "grad_norm": 0.5192442536354065, + "learning_rate": 1.0719e-05, + "loss": 0.0593, + "step": 3576 + }, + { + "epoch": 3.021114864864865, + "grad_norm": 0.752047598361969, + "learning_rate": 1.0722e-05, + "loss": 0.1092, + "step": 3577 + }, + { + "epoch": 3.0219594594594597, + "grad_norm": 0.898767352104187, + "learning_rate": 1.0725e-05, + "loss": 0.0818, + "step": 3578 + }, + { + "epoch": 3.022804054054054, + "grad_norm": 0.6953643560409546, + "learning_rate": 1.0728e-05, + "loss": 0.0816, + "step": 3579 + }, + { + "epoch": 3.0236486486486487, + "grad_norm": 0.8058777451515198, + "learning_rate": 1.0731000000000001e-05, + "loss": 0.0931, + "step": 3580 + }, + { + "epoch": 3.0244932432432434, + "grad_norm": 0.8262551426887512, + "learning_rate": 1.0734000000000001e-05, + "loss": 0.0726, + "step": 3581 + }, + { + "epoch": 3.0253378378378377, + "grad_norm": 0.7942811250686646, + "learning_rate": 1.0737e-05, + "loss": 0.0742, + "step": 3582 + }, + { + "epoch": 3.0261824324324325, + "grad_norm": 1.0904316902160645, + "learning_rate": 1.074e-05, + "loss": 0.1019, + "step": 3583 + }, + { + "epoch": 3.027027027027027, + "grad_norm": 1.001868724822998, + "learning_rate": 1.0743e-05, + "loss": 0.0772, + "step": 3584 + }, + { + "epoch": 3.0278716216216215, + "grad_norm": 0.7346161007881165, + "learning_rate": 1.0746e-05, + "loss": 0.0616, + "step": 3585 + }, + { + "epoch": 3.0287162162162162, + "grad_norm": 0.9123615026473999, + "learning_rate": 1.0749e-05, + "loss": 0.0714, + "step": 3586 + }, + { + "epoch": 3.029560810810811, + "grad_norm": 0.66813725233078, + "learning_rate": 1.0752e-05, + "loss": 0.0676, + "step": 3587 + }, + { + "epoch": 3.0304054054054053, + "grad_norm": 1.0674751996994019, + "learning_rate": 1.0755e-05, + "loss": 0.119, + "step": 3588 + }, + { + "epoch": 3.03125, + "grad_norm": 0.76346355676651, + "learning_rate": 1.0758e-05, + "loss": 0.0551, + "step": 3589 + }, + { + "epoch": 3.0320945945945947, + "grad_norm": 0.9412923455238342, + "learning_rate": 1.0761e-05, + "loss": 0.0768, + "step": 3590 + }, + { + "epoch": 3.032939189189189, + "grad_norm": 0.7069803476333618, + "learning_rate": 1.0764e-05, + "loss": 0.0532, + "step": 3591 + }, + { + "epoch": 3.0337837837837838, + "grad_norm": 1.0111650228500366, + "learning_rate": 1.0767e-05, + "loss": 0.0847, + "step": 3592 + }, + { + "epoch": 3.0346283783783785, + "grad_norm": 1.06943941116333, + "learning_rate": 1.077e-05, + "loss": 0.1041, + "step": 3593 + }, + { + "epoch": 3.035472972972973, + "grad_norm": 0.8973060846328735, + "learning_rate": 1.0773e-05, + "loss": 0.0806, + "step": 3594 + }, + { + "epoch": 3.0363175675675675, + "grad_norm": 0.8547393679618835, + "learning_rate": 1.0776000000000002e-05, + "loss": 0.0572, + "step": 3595 + }, + { + "epoch": 3.0371621621621623, + "grad_norm": 1.1819339990615845, + "learning_rate": 1.0779000000000001e-05, + "loss": 0.1289, + "step": 3596 + }, + { + "epoch": 3.0380067567567566, + "grad_norm": 0.8364986777305603, + "learning_rate": 1.0782e-05, + "loss": 0.1055, + "step": 3597 + }, + { + "epoch": 3.0388513513513513, + "grad_norm": 1.019155740737915, + "learning_rate": 1.0785e-05, + "loss": 0.09, + "step": 3598 + }, + { + "epoch": 3.039695945945946, + "grad_norm": 1.0948272943496704, + "learning_rate": 1.0787999999999999e-05, + "loss": 0.1049, + "step": 3599 + }, + { + "epoch": 3.0405405405405403, + "grad_norm": 2.050034761428833, + "learning_rate": 1.0791e-05, + "loss": 0.1219, + "step": 3600 + }, + { + "epoch": 3.041385135135135, + "grad_norm": 1.1419891119003296, + "learning_rate": 1.0794e-05, + "loss": 0.1375, + "step": 3601 + }, + { + "epoch": 3.04222972972973, + "grad_norm": 1.9792919158935547, + "learning_rate": 1.0797e-05, + "loss": 0.2024, + "step": 3602 + }, + { + "epoch": 3.043074324324324, + "grad_norm": 4.378783702850342, + "learning_rate": 1.08e-05, + "loss": 0.5184, + "step": 3603 + }, + { + "epoch": 3.043918918918919, + "grad_norm": 1.5521899461746216, + "learning_rate": 1.0803e-05, + "loss": 0.2519, + "step": 3604 + }, + { + "epoch": 3.0447635135135136, + "grad_norm": 1.2789806127548218, + "learning_rate": 1.0806000000000001e-05, + "loss": 0.1478, + "step": 3605 + }, + { + "epoch": 3.045608108108108, + "grad_norm": 0.9645568132400513, + "learning_rate": 1.0809000000000001e-05, + "loss": 0.1006, + "step": 3606 + }, + { + "epoch": 3.0464527027027026, + "grad_norm": 0.6151700615882874, + "learning_rate": 1.0812e-05, + "loss": 0.0696, + "step": 3607 + }, + { + "epoch": 3.0472972972972974, + "grad_norm": 1.0448510646820068, + "learning_rate": 1.0815e-05, + "loss": 0.0864, + "step": 3608 + }, + { + "epoch": 3.048141891891892, + "grad_norm": 0.7523443102836609, + "learning_rate": 1.0817999999999999e-05, + "loss": 0.0958, + "step": 3609 + }, + { + "epoch": 3.0489864864864864, + "grad_norm": 0.7632469534873962, + "learning_rate": 1.0821e-05, + "loss": 0.076, + "step": 3610 + }, + { + "epoch": 3.049831081081081, + "grad_norm": 0.7097877264022827, + "learning_rate": 1.0824e-05, + "loss": 0.0838, + "step": 3611 + }, + { + "epoch": 3.050675675675676, + "grad_norm": 1.001352071762085, + "learning_rate": 1.0827e-05, + "loss": 0.0926, + "step": 3612 + }, + { + "epoch": 3.05152027027027, + "grad_norm": 0.7003563046455383, + "learning_rate": 1.083e-05, + "loss": 0.0643, + "step": 3613 + }, + { + "epoch": 3.052364864864865, + "grad_norm": 0.8328469395637512, + "learning_rate": 1.0833e-05, + "loss": 0.107, + "step": 3614 + }, + { + "epoch": 3.0532094594594597, + "grad_norm": 0.8504636287689209, + "learning_rate": 1.0836000000000001e-05, + "loss": 0.0887, + "step": 3615 + }, + { + "epoch": 3.054054054054054, + "grad_norm": 0.6359783411026001, + "learning_rate": 1.0839e-05, + "loss": 0.0682, + "step": 3616 + }, + { + "epoch": 3.0548986486486487, + "grad_norm": 0.670621931552887, + "learning_rate": 1.0842e-05, + "loss": 0.0698, + "step": 3617 + }, + { + "epoch": 3.0557432432432434, + "grad_norm": 0.5755274295806885, + "learning_rate": 1.0845e-05, + "loss": 0.057, + "step": 3618 + }, + { + "epoch": 3.0565878378378377, + "grad_norm": 0.716833233833313, + "learning_rate": 1.0848e-05, + "loss": 0.0685, + "step": 3619 + }, + { + "epoch": 3.0574324324324325, + "grad_norm": 0.981069028377533, + "learning_rate": 1.0851000000000002e-05, + "loss": 0.0556, + "step": 3620 + }, + { + "epoch": 3.058277027027027, + "grad_norm": 0.7877779603004456, + "learning_rate": 1.0854e-05, + "loss": 0.0819, + "step": 3621 + }, + { + "epoch": 3.0591216216216215, + "grad_norm": 0.7331547737121582, + "learning_rate": 1.0857e-05, + "loss": 0.0803, + "step": 3622 + }, + { + "epoch": 3.0599662162162162, + "grad_norm": 0.5572734475135803, + "learning_rate": 1.086e-05, + "loss": 0.0509, + "step": 3623 + }, + { + "epoch": 3.060810810810811, + "grad_norm": 0.5971097350120544, + "learning_rate": 1.0863e-05, + "loss": 0.0434, + "step": 3624 + }, + { + "epoch": 3.0616554054054053, + "grad_norm": 0.6559865474700928, + "learning_rate": 1.0866e-05, + "loss": 0.0642, + "step": 3625 + }, + { + "epoch": 3.0625, + "grad_norm": 0.5825645327568054, + "learning_rate": 1.0869e-05, + "loss": 0.0591, + "step": 3626 + }, + { + "epoch": 3.0633445945945947, + "grad_norm": 0.6807399988174438, + "learning_rate": 1.0872e-05, + "loss": 0.0664, + "step": 3627 + }, + { + "epoch": 3.064189189189189, + "grad_norm": 0.9542753100395203, + "learning_rate": 1.0875e-05, + "loss": 0.0882, + "step": 3628 + }, + { + "epoch": 3.0650337837837838, + "grad_norm": 0.7552712559700012, + "learning_rate": 1.0878e-05, + "loss": 0.0743, + "step": 3629 + }, + { + "epoch": 3.0658783783783785, + "grad_norm": 0.8178406357765198, + "learning_rate": 1.0881000000000001e-05, + "loss": 0.0687, + "step": 3630 + }, + { + "epoch": 3.066722972972973, + "grad_norm": 0.9446877837181091, + "learning_rate": 1.0884000000000001e-05, + "loss": 0.0802, + "step": 3631 + }, + { + "epoch": 3.0675675675675675, + "grad_norm": 0.9334259629249573, + "learning_rate": 1.0887000000000001e-05, + "loss": 0.0895, + "step": 3632 + }, + { + "epoch": 3.0684121621621623, + "grad_norm": 1.3159092664718628, + "learning_rate": 1.089e-05, + "loss": 0.0719, + "step": 3633 + }, + { + "epoch": 3.0692567567567566, + "grad_norm": 1.7431622743606567, + "learning_rate": 1.0892999999999999e-05, + "loss": 0.1045, + "step": 3634 + }, + { + "epoch": 3.0701013513513513, + "grad_norm": 0.9419763684272766, + "learning_rate": 1.0896e-05, + "loss": 0.082, + "step": 3635 + }, + { + "epoch": 3.070945945945946, + "grad_norm": 0.9032451510429382, + "learning_rate": 1.0899e-05, + "loss": 0.1053, + "step": 3636 + }, + { + "epoch": 3.0717905405405403, + "grad_norm": 0.7759376168251038, + "learning_rate": 1.0902e-05, + "loss": 0.0854, + "step": 3637 + }, + { + "epoch": 3.072635135135135, + "grad_norm": 0.8675012588500977, + "learning_rate": 1.0905e-05, + "loss": 0.1039, + "step": 3638 + }, + { + "epoch": 3.07347972972973, + "grad_norm": 0.8619109392166138, + "learning_rate": 1.0908e-05, + "loss": 0.0953, + "step": 3639 + }, + { + "epoch": 3.074324324324324, + "grad_norm": 0.9257604479789734, + "learning_rate": 1.0911000000000001e-05, + "loss": 0.0905, + "step": 3640 + }, + { + "epoch": 3.075168918918919, + "grad_norm": 0.8921952843666077, + "learning_rate": 1.0914000000000001e-05, + "loss": 0.0757, + "step": 3641 + }, + { + "epoch": 3.0760135135135136, + "grad_norm": 0.9607582092285156, + "learning_rate": 1.0917e-05, + "loss": 0.0802, + "step": 3642 + }, + { + "epoch": 3.076858108108108, + "grad_norm": 0.7629030346870422, + "learning_rate": 1.092e-05, + "loss": 0.0884, + "step": 3643 + }, + { + "epoch": 3.0777027027027026, + "grad_norm": 0.8855393528938293, + "learning_rate": 1.0923e-05, + "loss": 0.102, + "step": 3644 + }, + { + "epoch": 3.0785472972972974, + "grad_norm": 1.0264332294464111, + "learning_rate": 1.0926000000000002e-05, + "loss": 0.0958, + "step": 3645 + }, + { + "epoch": 3.079391891891892, + "grad_norm": 0.9198261499404907, + "learning_rate": 1.0929e-05, + "loss": 0.1027, + "step": 3646 + }, + { + "epoch": 3.0802364864864864, + "grad_norm": 0.9738942980766296, + "learning_rate": 1.0932e-05, + "loss": 0.0935, + "step": 3647 + }, + { + "epoch": 3.081081081081081, + "grad_norm": 0.9465426802635193, + "learning_rate": 1.0935e-05, + "loss": 0.0906, + "step": 3648 + }, + { + "epoch": 3.081925675675676, + "grad_norm": 1.1518913507461548, + "learning_rate": 1.0938e-05, + "loss": 0.1114, + "step": 3649 + }, + { + "epoch": 3.08277027027027, + "grad_norm": 0.9869148135185242, + "learning_rate": 1.0941e-05, + "loss": 0.0977, + "step": 3650 + }, + { + "epoch": 3.083614864864865, + "grad_norm": 1.3653839826583862, + "learning_rate": 1.0944e-05, + "loss": 0.1307, + "step": 3651 + }, + { + "epoch": 3.0844594594594597, + "grad_norm": 2.2286956310272217, + "learning_rate": 1.0947e-05, + "loss": 0.1637, + "step": 3652 + }, + { + "epoch": 3.085304054054054, + "grad_norm": 1.8065658807754517, + "learning_rate": 1.095e-05, + "loss": 0.4987, + "step": 3653 + }, + { + "epoch": 3.0861486486486487, + "grad_norm": 1.4988881349563599, + "learning_rate": 1.0953e-05, + "loss": 0.3062, + "step": 3654 + }, + { + "epoch": 3.0869932432432434, + "grad_norm": 1.0046073198318481, + "learning_rate": 1.0956000000000001e-05, + "loss": 0.1376, + "step": 3655 + }, + { + "epoch": 3.0878378378378377, + "grad_norm": 0.8041070103645325, + "learning_rate": 1.0959000000000001e-05, + "loss": 0.1019, + "step": 3656 + }, + { + "epoch": 3.0886824324324325, + "grad_norm": 0.6025729775428772, + "learning_rate": 1.0962000000000001e-05, + "loss": 0.0922, + "step": 3657 + }, + { + "epoch": 3.089527027027027, + "grad_norm": 0.7268189191818237, + "learning_rate": 1.0965e-05, + "loss": 0.0899, + "step": 3658 + }, + { + "epoch": 3.0903716216216215, + "grad_norm": 0.857214093208313, + "learning_rate": 1.0967999999999999e-05, + "loss": 0.1047, + "step": 3659 + }, + { + "epoch": 3.0912162162162162, + "grad_norm": 0.7368853688240051, + "learning_rate": 1.0971e-05, + "loss": 0.069, + "step": 3660 + }, + { + "epoch": 3.092060810810811, + "grad_norm": 0.8651418685913086, + "learning_rate": 1.0974e-05, + "loss": 0.0929, + "step": 3661 + }, + { + "epoch": 3.0929054054054053, + "grad_norm": 0.69999098777771, + "learning_rate": 1.0977e-05, + "loss": 0.0967, + "step": 3662 + }, + { + "epoch": 3.09375, + "grad_norm": 0.7281485795974731, + "learning_rate": 1.098e-05, + "loss": 0.0893, + "step": 3663 + }, + { + "epoch": 3.0945945945945947, + "grad_norm": 0.9266915917396545, + "learning_rate": 1.0983e-05, + "loss": 0.085, + "step": 3664 + }, + { + "epoch": 3.095439189189189, + "grad_norm": 0.8384146094322205, + "learning_rate": 1.0986000000000001e-05, + "loss": 0.068, + "step": 3665 + }, + { + "epoch": 3.0962837837837838, + "grad_norm": 0.8588875532150269, + "learning_rate": 1.0989000000000001e-05, + "loss": 0.0896, + "step": 3666 + }, + { + "epoch": 3.0971283783783785, + "grad_norm": 0.5701978206634521, + "learning_rate": 1.0992e-05, + "loss": 0.0637, + "step": 3667 + }, + { + "epoch": 3.097972972972973, + "grad_norm": 0.8045923113822937, + "learning_rate": 1.0995e-05, + "loss": 0.0773, + "step": 3668 + }, + { + "epoch": 3.0988175675675675, + "grad_norm": 0.9825679063796997, + "learning_rate": 1.0998e-05, + "loss": 0.1024, + "step": 3669 + }, + { + "epoch": 3.0996621621621623, + "grad_norm": 0.6409597396850586, + "learning_rate": 1.1001e-05, + "loss": 0.0661, + "step": 3670 + }, + { + "epoch": 3.1005067567567566, + "grad_norm": 0.7695608139038086, + "learning_rate": 1.1004e-05, + "loss": 0.073, + "step": 3671 + }, + { + "epoch": 3.1013513513513513, + "grad_norm": 0.8074352145195007, + "learning_rate": 1.1007e-05, + "loss": 0.07, + "step": 3672 + }, + { + "epoch": 3.102195945945946, + "grad_norm": 0.9447247385978699, + "learning_rate": 1.101e-05, + "loss": 0.0852, + "step": 3673 + }, + { + "epoch": 3.1030405405405403, + "grad_norm": 0.5857628583908081, + "learning_rate": 1.1013e-05, + "loss": 0.0515, + "step": 3674 + }, + { + "epoch": 3.103885135135135, + "grad_norm": 0.6366325616836548, + "learning_rate": 1.1016000000000001e-05, + "loss": 0.0749, + "step": 3675 + }, + { + "epoch": 3.10472972972973, + "grad_norm": 0.8834441304206848, + "learning_rate": 1.1019e-05, + "loss": 0.0993, + "step": 3676 + }, + { + "epoch": 3.105574324324324, + "grad_norm": 0.9313024282455444, + "learning_rate": 1.1022e-05, + "loss": 0.0576, + "step": 3677 + }, + { + "epoch": 3.106418918918919, + "grad_norm": 1.2278683185577393, + "learning_rate": 1.1025e-05, + "loss": 0.0709, + "step": 3678 + }, + { + "epoch": 3.1072635135135136, + "grad_norm": 0.6049045324325562, + "learning_rate": 1.1028e-05, + "loss": 0.0589, + "step": 3679 + }, + { + "epoch": 3.108108108108108, + "grad_norm": 0.7848204970359802, + "learning_rate": 1.1031000000000002e-05, + "loss": 0.0713, + "step": 3680 + }, + { + "epoch": 3.1089527027027026, + "grad_norm": 0.8024513125419617, + "learning_rate": 1.1034000000000001e-05, + "loss": 0.0718, + "step": 3681 + }, + { + "epoch": 3.1097972972972974, + "grad_norm": 0.7589991688728333, + "learning_rate": 1.1037000000000001e-05, + "loss": 0.0694, + "step": 3682 + }, + { + "epoch": 3.110641891891892, + "grad_norm": 0.9074271321296692, + "learning_rate": 1.104e-05, + "loss": 0.0888, + "step": 3683 + }, + { + "epoch": 3.1114864864864864, + "grad_norm": 0.5652978420257568, + "learning_rate": 1.1042999999999999e-05, + "loss": 0.0701, + "step": 3684 + }, + { + "epoch": 3.112331081081081, + "grad_norm": 0.7094019055366516, + "learning_rate": 1.1046e-05, + "loss": 0.0605, + "step": 3685 + }, + { + "epoch": 3.113175675675676, + "grad_norm": 0.7883692383766174, + "learning_rate": 1.1049e-05, + "loss": 0.0644, + "step": 3686 + }, + { + "epoch": 3.11402027027027, + "grad_norm": 0.8649086356163025, + "learning_rate": 1.1052e-05, + "loss": 0.0737, + "step": 3687 + }, + { + "epoch": 3.114864864864865, + "grad_norm": 0.8321612477302551, + "learning_rate": 1.1055e-05, + "loss": 0.0605, + "step": 3688 + }, + { + "epoch": 3.1157094594594597, + "grad_norm": 0.7582159042358398, + "learning_rate": 1.1058e-05, + "loss": 0.0769, + "step": 3689 + }, + { + "epoch": 3.116554054054054, + "grad_norm": 1.0734578371047974, + "learning_rate": 1.1061000000000001e-05, + "loss": 0.0702, + "step": 3690 + }, + { + "epoch": 3.1173986486486487, + "grad_norm": 1.014235496520996, + "learning_rate": 1.1064000000000001e-05, + "loss": 0.088, + "step": 3691 + }, + { + "epoch": 3.1182432432432434, + "grad_norm": 0.9570469856262207, + "learning_rate": 1.1067000000000001e-05, + "loss": 0.0799, + "step": 3692 + }, + { + "epoch": 3.1190878378378377, + "grad_norm": 0.8083419799804688, + "learning_rate": 1.107e-05, + "loss": 0.0638, + "step": 3693 + }, + { + "epoch": 3.1199324324324325, + "grad_norm": 0.7937059998512268, + "learning_rate": 1.1073e-05, + "loss": 0.0788, + "step": 3694 + }, + { + "epoch": 3.120777027027027, + "grad_norm": 0.8874779939651489, + "learning_rate": 1.1075999999999999e-05, + "loss": 0.0747, + "step": 3695 + }, + { + "epoch": 3.1216216216216215, + "grad_norm": 2.2312028408050537, + "learning_rate": 1.1079e-05, + "loss": 0.1816, + "step": 3696 + }, + { + "epoch": 3.1224662162162162, + "grad_norm": 0.9712106585502625, + "learning_rate": 1.1082e-05, + "loss": 0.0721, + "step": 3697 + }, + { + "epoch": 3.123310810810811, + "grad_norm": 1.571279764175415, + "learning_rate": 1.1085e-05, + "loss": 0.0955, + "step": 3698 + }, + { + "epoch": 3.1241554054054053, + "grad_norm": 0.8882010579109192, + "learning_rate": 1.1088e-05, + "loss": 0.0785, + "step": 3699 + }, + { + "epoch": 3.125, + "grad_norm": 1.0180093050003052, + "learning_rate": 1.1091e-05, + "loss": 0.0836, + "step": 3700 + }, + { + "epoch": 3.1258445945945947, + "grad_norm": 1.3789808750152588, + "learning_rate": 1.1094e-05, + "loss": 0.1308, + "step": 3701 + }, + { + "epoch": 3.126689189189189, + "grad_norm": 1.6676141023635864, + "learning_rate": 1.1097e-05, + "loss": 0.1401, + "step": 3702 + }, + { + "epoch": 3.1275337837837838, + "grad_norm": 2.7210886478424072, + "learning_rate": 1.11e-05, + "loss": 0.5759, + "step": 3703 + }, + { + "epoch": 3.1283783783783785, + "grad_norm": 1.4152138233184814, + "learning_rate": 1.1103e-05, + "loss": 0.3249, + "step": 3704 + }, + { + "epoch": 3.129222972972973, + "grad_norm": 2.0483860969543457, + "learning_rate": 1.1106e-05, + "loss": 0.2138, + "step": 3705 + }, + { + "epoch": 3.1300675675675675, + "grad_norm": 1.3880891799926758, + "learning_rate": 1.1109000000000002e-05, + "loss": 0.1147, + "step": 3706 + }, + { + "epoch": 3.1309121621621623, + "grad_norm": 0.9551907777786255, + "learning_rate": 1.1112e-05, + "loss": 0.0991, + "step": 3707 + }, + { + "epoch": 3.1317567567567566, + "grad_norm": 0.6467481851577759, + "learning_rate": 1.1115e-05, + "loss": 0.0868, + "step": 3708 + }, + { + "epoch": 3.1326013513513513, + "grad_norm": 0.7139769196510315, + "learning_rate": 1.1118e-05, + "loss": 0.0791, + "step": 3709 + }, + { + "epoch": 3.133445945945946, + "grad_norm": 0.663169801235199, + "learning_rate": 1.1120999999999999e-05, + "loss": 0.0796, + "step": 3710 + }, + { + "epoch": 3.1342905405405403, + "grad_norm": 1.0923709869384766, + "learning_rate": 1.1124e-05, + "loss": 0.067, + "step": 3711 + }, + { + "epoch": 3.135135135135135, + "grad_norm": 0.7665802240371704, + "learning_rate": 1.1127e-05, + "loss": 0.0895, + "step": 3712 + }, + { + "epoch": 3.13597972972973, + "grad_norm": 0.6028164029121399, + "learning_rate": 1.113e-05, + "loss": 0.0596, + "step": 3713 + }, + { + "epoch": 3.136824324324324, + "grad_norm": 0.7728317379951477, + "learning_rate": 1.1133e-05, + "loss": 0.0695, + "step": 3714 + }, + { + "epoch": 3.137668918918919, + "grad_norm": 0.5495914220809937, + "learning_rate": 1.1136e-05, + "loss": 0.0637, + "step": 3715 + }, + { + "epoch": 3.1385135135135136, + "grad_norm": 0.6262581944465637, + "learning_rate": 1.1139000000000001e-05, + "loss": 0.0598, + "step": 3716 + }, + { + "epoch": 3.139358108108108, + "grad_norm": 0.7711997628211975, + "learning_rate": 1.1142000000000001e-05, + "loss": 0.0685, + "step": 3717 + }, + { + "epoch": 3.1402027027027026, + "grad_norm": 0.7135451436042786, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0871, + "step": 3718 + }, + { + "epoch": 3.1410472972972974, + "grad_norm": 1.7577394247055054, + "learning_rate": 1.1148e-05, + "loss": 0.0757, + "step": 3719 + }, + { + "epoch": 3.141891891891892, + "grad_norm": 0.7981122732162476, + "learning_rate": 1.1150999999999999e-05, + "loss": 0.0651, + "step": 3720 + }, + { + "epoch": 3.1427364864864864, + "grad_norm": 0.5425859093666077, + "learning_rate": 1.1154e-05, + "loss": 0.059, + "step": 3721 + }, + { + "epoch": 3.143581081081081, + "grad_norm": 0.8012245893478394, + "learning_rate": 1.1157e-05, + "loss": 0.0662, + "step": 3722 + }, + { + "epoch": 3.144425675675676, + "grad_norm": 0.7537211775779724, + "learning_rate": 1.116e-05, + "loss": 0.0985, + "step": 3723 + }, + { + "epoch": 3.14527027027027, + "grad_norm": 0.6126032471656799, + "learning_rate": 1.1163e-05, + "loss": 0.0462, + "step": 3724 + }, + { + "epoch": 3.146114864864865, + "grad_norm": 0.7441989779472351, + "learning_rate": 1.1166e-05, + "loss": 0.083, + "step": 3725 + }, + { + "epoch": 3.1469594594594597, + "grad_norm": 0.5869786143302917, + "learning_rate": 1.1169000000000001e-05, + "loss": 0.0813, + "step": 3726 + }, + { + "epoch": 3.147804054054054, + "grad_norm": 0.5742023587226868, + "learning_rate": 1.1172e-05, + "loss": 0.0484, + "step": 3727 + }, + { + "epoch": 3.1486486486486487, + "grad_norm": 0.5866444706916809, + "learning_rate": 1.1175e-05, + "loss": 0.0456, + "step": 3728 + }, + { + "epoch": 3.1494932432432434, + "grad_norm": 0.621372640132904, + "learning_rate": 1.1178e-05, + "loss": 0.0654, + "step": 3729 + }, + { + "epoch": 3.1503378378378377, + "grad_norm": 0.7453851699829102, + "learning_rate": 1.1181e-05, + "loss": 0.0964, + "step": 3730 + }, + { + "epoch": 3.1511824324324325, + "grad_norm": 1.1524243354797363, + "learning_rate": 1.1184000000000002e-05, + "loss": 0.0687, + "step": 3731 + }, + { + "epoch": 3.152027027027027, + "grad_norm": 0.786373496055603, + "learning_rate": 1.1187e-05, + "loss": 0.1081, + "step": 3732 + }, + { + "epoch": 3.1528716216216215, + "grad_norm": 0.7284941673278809, + "learning_rate": 1.119e-05, + "loss": 0.0752, + "step": 3733 + }, + { + "epoch": 3.1537162162162162, + "grad_norm": 0.6761149168014526, + "learning_rate": 1.1193e-05, + "loss": 0.0676, + "step": 3734 + }, + { + "epoch": 3.154560810810811, + "grad_norm": 0.5835392475128174, + "learning_rate": 1.1196e-05, + "loss": 0.051, + "step": 3735 + }, + { + "epoch": 3.1554054054054053, + "grad_norm": 0.7846307754516602, + "learning_rate": 1.1199e-05, + "loss": 0.0538, + "step": 3736 + }, + { + "epoch": 3.15625, + "grad_norm": 0.6573534607887268, + "learning_rate": 1.1202e-05, + "loss": 0.0529, + "step": 3737 + }, + { + "epoch": 3.1570945945945947, + "grad_norm": 1.102984070777893, + "learning_rate": 1.1205e-05, + "loss": 0.0754, + "step": 3738 + }, + { + "epoch": 3.157939189189189, + "grad_norm": 0.7870478630065918, + "learning_rate": 1.1208e-05, + "loss": 0.0631, + "step": 3739 + }, + { + "epoch": 3.1587837837837838, + "grad_norm": 0.8790913224220276, + "learning_rate": 1.1211e-05, + "loss": 0.0595, + "step": 3740 + }, + { + "epoch": 3.1596283783783785, + "grad_norm": 0.9435224533081055, + "learning_rate": 1.1214000000000001e-05, + "loss": 0.0593, + "step": 3741 + }, + { + "epoch": 3.160472972972973, + "grad_norm": 1.1707398891448975, + "learning_rate": 1.1217000000000001e-05, + "loss": 0.0968, + "step": 3742 + }, + { + "epoch": 3.1613175675675675, + "grad_norm": 1.0928279161453247, + "learning_rate": 1.1220000000000001e-05, + "loss": 0.0785, + "step": 3743 + }, + { + "epoch": 3.1621621621621623, + "grad_norm": 0.9053871631622314, + "learning_rate": 1.1222999999999999e-05, + "loss": 0.0711, + "step": 3744 + }, + { + "epoch": 3.1630067567567566, + "grad_norm": 0.882350742816925, + "learning_rate": 1.1225999999999999e-05, + "loss": 0.0802, + "step": 3745 + }, + { + "epoch": 3.1638513513513513, + "grad_norm": 0.7731640934944153, + "learning_rate": 1.1229e-05, + "loss": 0.0434, + "step": 3746 + }, + { + "epoch": 3.164695945945946, + "grad_norm": 1.4263683557510376, + "learning_rate": 1.1232e-05, + "loss": 0.0874, + "step": 3747 + }, + { + "epoch": 3.1655405405405403, + "grad_norm": 2.9981136322021484, + "learning_rate": 1.1235e-05, + "loss": 0.091, + "step": 3748 + }, + { + "epoch": 3.166385135135135, + "grad_norm": 1.1123855113983154, + "learning_rate": 1.1238e-05, + "loss": 0.1114, + "step": 3749 + }, + { + "epoch": 3.16722972972973, + "grad_norm": 1.0812574625015259, + "learning_rate": 1.1241e-05, + "loss": 0.092, + "step": 3750 + }, + { + "epoch": 3.168074324324324, + "grad_norm": 1.4202059507369995, + "learning_rate": 1.1244000000000001e-05, + "loss": 0.1471, + "step": 3751 + }, + { + "epoch": 3.168918918918919, + "grad_norm": 1.988275170326233, + "learning_rate": 1.1247000000000001e-05, + "loss": 0.1429, + "step": 3752 + }, + { + "epoch": 3.1697635135135136, + "grad_norm": 1.5390491485595703, + "learning_rate": 1.125e-05, + "loss": 0.4964, + "step": 3753 + }, + { + "epoch": 3.170608108108108, + "grad_norm": 2.5891830921173096, + "learning_rate": 1.1253e-05, + "loss": 0.2683, + "step": 3754 + }, + { + "epoch": 3.1714527027027026, + "grad_norm": 0.7069483995437622, + "learning_rate": 1.1256e-05, + "loss": 0.1373, + "step": 3755 + }, + { + "epoch": 3.1722972972972974, + "grad_norm": 0.6842374801635742, + "learning_rate": 1.1259e-05, + "loss": 0.0944, + "step": 3756 + }, + { + "epoch": 3.173141891891892, + "grad_norm": 0.7677066326141357, + "learning_rate": 1.1262e-05, + "loss": 0.1048, + "step": 3757 + }, + { + "epoch": 3.1739864864864864, + "grad_norm": 0.5426579713821411, + "learning_rate": 1.1265e-05, + "loss": 0.0731, + "step": 3758 + }, + { + "epoch": 3.174831081081081, + "grad_norm": 0.9403834342956543, + "learning_rate": 1.1268e-05, + "loss": 0.1299, + "step": 3759 + }, + { + "epoch": 3.175675675675676, + "grad_norm": 0.6819109916687012, + "learning_rate": 1.1271e-05, + "loss": 0.0667, + "step": 3760 + }, + { + "epoch": 3.17652027027027, + "grad_norm": 0.570276141166687, + "learning_rate": 1.1274e-05, + "loss": 0.0619, + "step": 3761 + }, + { + "epoch": 3.177364864864865, + "grad_norm": 0.7134559154510498, + "learning_rate": 1.1277e-05, + "loss": 0.0616, + "step": 3762 + }, + { + "epoch": 3.1782094594594597, + "grad_norm": 0.5062280297279358, + "learning_rate": 1.128e-05, + "loss": 0.0503, + "step": 3763 + }, + { + "epoch": 3.179054054054054, + "grad_norm": 0.5771971940994263, + "learning_rate": 1.1283e-05, + "loss": 0.0479, + "step": 3764 + }, + { + "epoch": 3.1798986486486487, + "grad_norm": 0.4340534210205078, + "learning_rate": 1.1286e-05, + "loss": 0.0454, + "step": 3765 + }, + { + "epoch": 3.1807432432432434, + "grad_norm": 0.9779492020606995, + "learning_rate": 1.1289000000000002e-05, + "loss": 0.1064, + "step": 3766 + }, + { + "epoch": 3.1815878378378377, + "grad_norm": 0.7829322218894958, + "learning_rate": 1.1292000000000001e-05, + "loss": 0.0945, + "step": 3767 + }, + { + "epoch": 3.1824324324324325, + "grad_norm": 0.6747731566429138, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0501, + "step": 3768 + }, + { + "epoch": 3.183277027027027, + "grad_norm": 1.0284833908081055, + "learning_rate": 1.1298e-05, + "loss": 0.0858, + "step": 3769 + }, + { + "epoch": 3.1841216216216215, + "grad_norm": 0.6583215594291687, + "learning_rate": 1.1300999999999999e-05, + "loss": 0.0653, + "step": 3770 + }, + { + "epoch": 3.1849662162162162, + "grad_norm": 0.6376005411148071, + "learning_rate": 1.1304e-05, + "loss": 0.0583, + "step": 3771 + }, + { + "epoch": 3.185810810810811, + "grad_norm": 0.5899355411529541, + "learning_rate": 1.1307e-05, + "loss": 0.0669, + "step": 3772 + }, + { + "epoch": 3.1866554054054053, + "grad_norm": 0.7558696866035461, + "learning_rate": 1.131e-05, + "loss": 0.0568, + "step": 3773 + }, + { + "epoch": 3.1875, + "grad_norm": 0.55085688829422, + "learning_rate": 1.1313e-05, + "loss": 0.0611, + "step": 3774 + }, + { + "epoch": 3.1883445945945947, + "grad_norm": 0.723311722278595, + "learning_rate": 1.1316e-05, + "loss": 0.079, + "step": 3775 + }, + { + "epoch": 3.189189189189189, + "grad_norm": 0.6713302731513977, + "learning_rate": 1.1319000000000001e-05, + "loss": 0.0715, + "step": 3776 + }, + { + "epoch": 3.1900337837837838, + "grad_norm": 0.6702399849891663, + "learning_rate": 1.1322000000000001e-05, + "loss": 0.0609, + "step": 3777 + }, + { + "epoch": 3.1908783783783785, + "grad_norm": 0.7730740308761597, + "learning_rate": 1.1325e-05, + "loss": 0.0731, + "step": 3778 + }, + { + "epoch": 3.191722972972973, + "grad_norm": 0.700680136680603, + "learning_rate": 1.1328e-05, + "loss": 0.0856, + "step": 3779 + }, + { + "epoch": 3.1925675675675675, + "grad_norm": 0.6314879059791565, + "learning_rate": 1.1331e-05, + "loss": 0.0545, + "step": 3780 + }, + { + "epoch": 3.1934121621621623, + "grad_norm": 0.6535696983337402, + "learning_rate": 1.1334e-05, + "loss": 0.07, + "step": 3781 + }, + { + "epoch": 3.1942567567567566, + "grad_norm": 0.7561300992965698, + "learning_rate": 1.1337e-05, + "loss": 0.0717, + "step": 3782 + }, + { + "epoch": 3.1951013513513513, + "grad_norm": 0.6263580322265625, + "learning_rate": 1.134e-05, + "loss": 0.0486, + "step": 3783 + }, + { + "epoch": 3.195945945945946, + "grad_norm": 0.7801764011383057, + "learning_rate": 1.1343e-05, + "loss": 0.0731, + "step": 3784 + }, + { + "epoch": 3.1967905405405403, + "grad_norm": 0.736094057559967, + "learning_rate": 1.1346e-05, + "loss": 0.0715, + "step": 3785 + }, + { + "epoch": 3.197635135135135, + "grad_norm": 0.6020583510398865, + "learning_rate": 1.1349000000000001e-05, + "loss": 0.0697, + "step": 3786 + }, + { + "epoch": 3.19847972972973, + "grad_norm": 0.7818613052368164, + "learning_rate": 1.1352e-05, + "loss": 0.0751, + "step": 3787 + }, + { + "epoch": 3.199324324324324, + "grad_norm": 0.6944395303726196, + "learning_rate": 1.1355e-05, + "loss": 0.052, + "step": 3788 + }, + { + "epoch": 3.200168918918919, + "grad_norm": 0.9127331972122192, + "learning_rate": 1.1358e-05, + "loss": 0.0959, + "step": 3789 + }, + { + "epoch": 3.2010135135135136, + "grad_norm": 1.3201539516448975, + "learning_rate": 1.1361e-05, + "loss": 0.0801, + "step": 3790 + }, + { + "epoch": 3.201858108108108, + "grad_norm": 0.8911144733428955, + "learning_rate": 1.1364000000000002e-05, + "loss": 0.09, + "step": 3791 + }, + { + "epoch": 3.2027027027027026, + "grad_norm": 0.9401916861534119, + "learning_rate": 1.1367000000000001e-05, + "loss": 0.0893, + "step": 3792 + }, + { + "epoch": 3.2035472972972974, + "grad_norm": 0.8002727031707764, + "learning_rate": 1.137e-05, + "loss": 0.0763, + "step": 3793 + }, + { + "epoch": 3.204391891891892, + "grad_norm": 0.9019563794136047, + "learning_rate": 1.1373e-05, + "loss": 0.0877, + "step": 3794 + }, + { + "epoch": 3.2052364864864864, + "grad_norm": 0.9363192915916443, + "learning_rate": 1.1376e-05, + "loss": 0.1043, + "step": 3795 + }, + { + "epoch": 3.206081081081081, + "grad_norm": 0.8437493443489075, + "learning_rate": 1.1379e-05, + "loss": 0.083, + "step": 3796 + }, + { + "epoch": 3.206925675675676, + "grad_norm": 0.5984321236610413, + "learning_rate": 1.1382e-05, + "loss": 0.055, + "step": 3797 + }, + { + "epoch": 3.20777027027027, + "grad_norm": 1.1239888668060303, + "learning_rate": 1.1385e-05, + "loss": 0.1115, + "step": 3798 + }, + { + "epoch": 3.208614864864865, + "grad_norm": 0.7856543660163879, + "learning_rate": 1.1388e-05, + "loss": 0.0727, + "step": 3799 + }, + { + "epoch": 3.2094594594594597, + "grad_norm": 1.062475323677063, + "learning_rate": 1.1391e-05, + "loss": 0.1089, + "step": 3800 + }, + { + "epoch": 3.210304054054054, + "grad_norm": 1.2144137620925903, + "learning_rate": 1.1394000000000001e-05, + "loss": 0.1048, + "step": 3801 + }, + { + "epoch": 3.2111486486486487, + "grad_norm": 2.190474510192871, + "learning_rate": 1.1397000000000001e-05, + "loss": 0.2511, + "step": 3802 + }, + { + "epoch": 3.2119932432432434, + "grad_norm": 1.42459237575531, + "learning_rate": 1.1400000000000001e-05, + "loss": 0.4295, + "step": 3803 + }, + { + "epoch": 3.2128378378378377, + "grad_norm": 1.5330619812011719, + "learning_rate": 1.1403e-05, + "loss": 0.3885, + "step": 3804 + }, + { + "epoch": 3.2136824324324325, + "grad_norm": 0.9580733776092529, + "learning_rate": 1.1406e-05, + "loss": 0.1416, + "step": 3805 + }, + { + "epoch": 3.214527027027027, + "grad_norm": 0.9929201006889343, + "learning_rate": 1.1409e-05, + "loss": 0.1129, + "step": 3806 + }, + { + "epoch": 3.2153716216216215, + "grad_norm": 0.8468766808509827, + "learning_rate": 1.1412e-05, + "loss": 0.1564, + "step": 3807 + }, + { + "epoch": 3.2162162162162162, + "grad_norm": 0.5783662796020508, + "learning_rate": 1.1415e-05, + "loss": 0.0773, + "step": 3808 + }, + { + "epoch": 3.217060810810811, + "grad_norm": 0.5401755571365356, + "learning_rate": 1.1418e-05, + "loss": 0.0641, + "step": 3809 + }, + { + "epoch": 3.2179054054054053, + "grad_norm": 0.8073268532752991, + "learning_rate": 1.1421e-05, + "loss": 0.0838, + "step": 3810 + }, + { + "epoch": 3.21875, + "grad_norm": 0.7739666104316711, + "learning_rate": 1.1424000000000001e-05, + "loss": 0.084, + "step": 3811 + }, + { + "epoch": 3.2195945945945947, + "grad_norm": 0.7098476886749268, + "learning_rate": 1.1427000000000001e-05, + "loss": 0.0727, + "step": 3812 + }, + { + "epoch": 3.220439189189189, + "grad_norm": 0.8683897256851196, + "learning_rate": 1.143e-05, + "loss": 0.0909, + "step": 3813 + }, + { + "epoch": 3.2212837837837838, + "grad_norm": 0.6999081969261169, + "learning_rate": 1.1433e-05, + "loss": 0.0687, + "step": 3814 + }, + { + "epoch": 3.2221283783783785, + "grad_norm": 0.6832756400108337, + "learning_rate": 1.1436e-05, + "loss": 0.0631, + "step": 3815 + }, + { + "epoch": 3.222972972972973, + "grad_norm": 0.8634663224220276, + "learning_rate": 1.1439e-05, + "loss": 0.0588, + "step": 3816 + }, + { + "epoch": 3.2238175675675675, + "grad_norm": 0.7122857570648193, + "learning_rate": 1.1442000000000002e-05, + "loss": 0.0752, + "step": 3817 + }, + { + "epoch": 3.2246621621621623, + "grad_norm": 0.8201393485069275, + "learning_rate": 1.1445e-05, + "loss": 0.0874, + "step": 3818 + }, + { + "epoch": 3.2255067567567566, + "grad_norm": 0.7611166834831238, + "learning_rate": 1.1448e-05, + "loss": 0.0695, + "step": 3819 + }, + { + "epoch": 3.2263513513513513, + "grad_norm": 0.5429022312164307, + "learning_rate": 1.1451e-05, + "loss": 0.0553, + "step": 3820 + }, + { + "epoch": 3.227195945945946, + "grad_norm": 0.6387404799461365, + "learning_rate": 1.1453999999999999e-05, + "loss": 0.081, + "step": 3821 + }, + { + "epoch": 3.2280405405405403, + "grad_norm": 0.8896227478981018, + "learning_rate": 1.1457e-05, + "loss": 0.0846, + "step": 3822 + }, + { + "epoch": 3.228885135135135, + "grad_norm": 0.6150014996528625, + "learning_rate": 1.146e-05, + "loss": 0.0609, + "step": 3823 + }, + { + "epoch": 3.22972972972973, + "grad_norm": 0.6245673298835754, + "learning_rate": 1.1463e-05, + "loss": 0.0591, + "step": 3824 + }, + { + "epoch": 3.230574324324324, + "grad_norm": 0.806188702583313, + "learning_rate": 1.1466e-05, + "loss": 0.0825, + "step": 3825 + }, + { + "epoch": 3.231418918918919, + "grad_norm": 0.7328096032142639, + "learning_rate": 1.1469e-05, + "loss": 0.0807, + "step": 3826 + }, + { + "epoch": 3.2322635135135136, + "grad_norm": 0.8584875464439392, + "learning_rate": 1.1472000000000001e-05, + "loss": 0.0895, + "step": 3827 + }, + { + "epoch": 3.233108108108108, + "grad_norm": 0.5320190191268921, + "learning_rate": 1.1475000000000001e-05, + "loss": 0.0589, + "step": 3828 + }, + { + "epoch": 3.2339527027027026, + "grad_norm": 0.90260249376297, + "learning_rate": 1.1478000000000001e-05, + "loss": 0.0779, + "step": 3829 + }, + { + "epoch": 3.2347972972972974, + "grad_norm": 0.7117367386817932, + "learning_rate": 1.1480999999999999e-05, + "loss": 0.0565, + "step": 3830 + }, + { + "epoch": 3.235641891891892, + "grad_norm": 1.0133790969848633, + "learning_rate": 1.1483999999999999e-05, + "loss": 0.0565, + "step": 3831 + }, + { + "epoch": 3.2364864864864864, + "grad_norm": 0.6671521067619324, + "learning_rate": 1.1487e-05, + "loss": 0.0556, + "step": 3832 + }, + { + "epoch": 3.237331081081081, + "grad_norm": 0.9337318539619446, + "learning_rate": 1.149e-05, + "loss": 0.0849, + "step": 3833 + }, + { + "epoch": 3.238175675675676, + "grad_norm": 1.6361979246139526, + "learning_rate": 1.1493e-05, + "loss": 0.1165, + "step": 3834 + }, + { + "epoch": 3.23902027027027, + "grad_norm": 0.6205429434776306, + "learning_rate": 1.1496e-05, + "loss": 0.0652, + "step": 3835 + }, + { + "epoch": 3.239864864864865, + "grad_norm": 0.755202054977417, + "learning_rate": 1.1499e-05, + "loss": 0.0786, + "step": 3836 + }, + { + "epoch": 3.2407094594594597, + "grad_norm": 0.7422866821289062, + "learning_rate": 1.1502000000000001e-05, + "loss": 0.0672, + "step": 3837 + }, + { + "epoch": 3.241554054054054, + "grad_norm": 0.7144129872322083, + "learning_rate": 1.1505e-05, + "loss": 0.0667, + "step": 3838 + }, + { + "epoch": 3.2423986486486487, + "grad_norm": 0.7396690845489502, + "learning_rate": 1.1508e-05, + "loss": 0.0647, + "step": 3839 + }, + { + "epoch": 3.2432432432432434, + "grad_norm": 0.8047087788581848, + "learning_rate": 1.1511e-05, + "loss": 0.0591, + "step": 3840 + }, + { + "epoch": 3.2440878378378377, + "grad_norm": 0.6390584707260132, + "learning_rate": 1.1514e-05, + "loss": 0.0721, + "step": 3841 + }, + { + "epoch": 3.2449324324324325, + "grad_norm": 0.8906922340393066, + "learning_rate": 1.1517e-05, + "loss": 0.0837, + "step": 3842 + }, + { + "epoch": 3.245777027027027, + "grad_norm": 0.7728691697120667, + "learning_rate": 1.152e-05, + "loss": 0.0559, + "step": 3843 + }, + { + "epoch": 3.2466216216216215, + "grad_norm": 0.8290387392044067, + "learning_rate": 1.1523e-05, + "loss": 0.0796, + "step": 3844 + }, + { + "epoch": 3.2474662162162162, + "grad_norm": 3.1602916717529297, + "learning_rate": 1.1526e-05, + "loss": 0.1139, + "step": 3845 + }, + { + "epoch": 3.248310810810811, + "grad_norm": 0.6938696503639221, + "learning_rate": 1.1529e-05, + "loss": 0.0784, + "step": 3846 + }, + { + "epoch": 3.2491554054054053, + "grad_norm": 1.2991608381271362, + "learning_rate": 1.1532e-05, + "loss": 0.1013, + "step": 3847 + }, + { + "epoch": 3.25, + "grad_norm": 0.7596430778503418, + "learning_rate": 1.1535e-05, + "loss": 0.0854, + "step": 3848 + }, + { + "epoch": 3.2508445945945947, + "grad_norm": 0.9953436255455017, + "learning_rate": 1.1538e-05, + "loss": 0.0824, + "step": 3849 + }, + { + "epoch": 3.251689189189189, + "grad_norm": 1.0915911197662354, + "learning_rate": 1.1541e-05, + "loss": 0.0938, + "step": 3850 + }, + { + "epoch": 3.2525337837837838, + "grad_norm": 1.4545097351074219, + "learning_rate": 1.1544e-05, + "loss": 0.1354, + "step": 3851 + }, + { + "epoch": 3.2533783783783785, + "grad_norm": 1.8647878170013428, + "learning_rate": 1.1547000000000001e-05, + "loss": 0.1296, + "step": 3852 + }, + { + "epoch": 3.254222972972973, + "grad_norm": 1.9705537557601929, + "learning_rate": 1.1550000000000001e-05, + "loss": 0.4881, + "step": 3853 + }, + { + "epoch": 3.2550675675675675, + "grad_norm": 2.0931708812713623, + "learning_rate": 1.1553000000000001e-05, + "loss": 0.2765, + "step": 3854 + }, + { + "epoch": 3.2559121621621623, + "grad_norm": 0.6787639856338501, + "learning_rate": 1.1555999999999999e-05, + "loss": 0.17, + "step": 3855 + }, + { + "epoch": 3.2567567567567566, + "grad_norm": 0.7816166877746582, + "learning_rate": 1.1558999999999999e-05, + "loss": 0.119, + "step": 3856 + }, + { + "epoch": 3.2576013513513513, + "grad_norm": 0.6921269297599792, + "learning_rate": 1.1562e-05, + "loss": 0.0964, + "step": 3857 + }, + { + "epoch": 3.258445945945946, + "grad_norm": 0.5692390203475952, + "learning_rate": 1.1565e-05, + "loss": 0.0549, + "step": 3858 + }, + { + "epoch": 3.2592905405405403, + "grad_norm": 0.5352365970611572, + "learning_rate": 1.1568e-05, + "loss": 0.0721, + "step": 3859 + }, + { + "epoch": 3.260135135135135, + "grad_norm": 0.7562052607536316, + "learning_rate": 1.1571e-05, + "loss": 0.0664, + "step": 3860 + }, + { + "epoch": 3.26097972972973, + "grad_norm": 0.5131056308746338, + "learning_rate": 1.1574e-05, + "loss": 0.0577, + "step": 3861 + }, + { + "epoch": 3.2618243243243246, + "grad_norm": 0.5517268776893616, + "learning_rate": 1.1577000000000001e-05, + "loss": 0.0759, + "step": 3862 + }, + { + "epoch": 3.262668918918919, + "grad_norm": 0.6397984623908997, + "learning_rate": 1.1580000000000001e-05, + "loss": 0.0721, + "step": 3863 + }, + { + "epoch": 3.2635135135135136, + "grad_norm": 0.6941590309143066, + "learning_rate": 1.1583e-05, + "loss": 0.0828, + "step": 3864 + }, + { + "epoch": 3.264358108108108, + "grad_norm": 0.9059235453605652, + "learning_rate": 1.1586e-05, + "loss": 0.0618, + "step": 3865 + }, + { + "epoch": 3.2652027027027026, + "grad_norm": 0.5904377698898315, + "learning_rate": 1.1589e-05, + "loss": 0.0995, + "step": 3866 + }, + { + "epoch": 3.2660472972972974, + "grad_norm": 0.6090877056121826, + "learning_rate": 1.1592e-05, + "loss": 0.0686, + "step": 3867 + }, + { + "epoch": 3.266891891891892, + "grad_norm": 0.7964613437652588, + "learning_rate": 1.1595e-05, + "loss": 0.0627, + "step": 3868 + }, + { + "epoch": 3.2677364864864864, + "grad_norm": 0.6899588704109192, + "learning_rate": 1.1598e-05, + "loss": 0.0645, + "step": 3869 + }, + { + "epoch": 3.268581081081081, + "grad_norm": 1.198974370956421, + "learning_rate": 1.1601e-05, + "loss": 0.0878, + "step": 3870 + }, + { + "epoch": 3.2694256756756754, + "grad_norm": 0.8339559435844421, + "learning_rate": 1.1604e-05, + "loss": 0.0605, + "step": 3871 + }, + { + "epoch": 3.27027027027027, + "grad_norm": 0.6743698716163635, + "learning_rate": 1.1607000000000001e-05, + "loss": 0.0745, + "step": 3872 + }, + { + "epoch": 3.271114864864865, + "grad_norm": 0.4749554991722107, + "learning_rate": 1.161e-05, + "loss": 0.0392, + "step": 3873 + }, + { + "epoch": 3.2719594594594597, + "grad_norm": 1.0463296175003052, + "learning_rate": 1.1613e-05, + "loss": 0.1305, + "step": 3874 + }, + { + "epoch": 3.272804054054054, + "grad_norm": 1.0882588624954224, + "learning_rate": 1.1616e-05, + "loss": 0.0912, + "step": 3875 + }, + { + "epoch": 3.2736486486486487, + "grad_norm": 0.658528745174408, + "learning_rate": 1.1619e-05, + "loss": 0.0619, + "step": 3876 + }, + { + "epoch": 3.2744932432432434, + "grad_norm": 0.5665156841278076, + "learning_rate": 1.1622000000000002e-05, + "loss": 0.0576, + "step": 3877 + }, + { + "epoch": 3.2753378378378377, + "grad_norm": 0.959863007068634, + "learning_rate": 1.1625000000000001e-05, + "loss": 0.1118, + "step": 3878 + }, + { + "epoch": 3.2761824324324325, + "grad_norm": 0.5063533782958984, + "learning_rate": 1.1628e-05, + "loss": 0.0439, + "step": 3879 + }, + { + "epoch": 3.277027027027027, + "grad_norm": 0.6300864219665527, + "learning_rate": 1.1631e-05, + "loss": 0.0716, + "step": 3880 + }, + { + "epoch": 3.2778716216216215, + "grad_norm": 0.5921394228935242, + "learning_rate": 1.1633999999999999e-05, + "loss": 0.0639, + "step": 3881 + }, + { + "epoch": 3.2787162162162162, + "grad_norm": 0.5676165819168091, + "learning_rate": 1.1637e-05, + "loss": 0.0589, + "step": 3882 + }, + { + "epoch": 3.279560810810811, + "grad_norm": 0.573773205280304, + "learning_rate": 1.164e-05, + "loss": 0.0532, + "step": 3883 + }, + { + "epoch": 3.2804054054054053, + "grad_norm": 0.9980499744415283, + "learning_rate": 1.1643e-05, + "loss": 0.0785, + "step": 3884 + }, + { + "epoch": 3.28125, + "grad_norm": 0.8495071530342102, + "learning_rate": 1.1646e-05, + "loss": 0.0901, + "step": 3885 + }, + { + "epoch": 3.2820945945945947, + "grad_norm": 0.7087128758430481, + "learning_rate": 1.1649e-05, + "loss": 0.0744, + "step": 3886 + }, + { + "epoch": 3.282939189189189, + "grad_norm": 1.0572969913482666, + "learning_rate": 1.1652000000000001e-05, + "loss": 0.1363, + "step": 3887 + }, + { + "epoch": 3.2837837837837838, + "grad_norm": 1.4450515508651733, + "learning_rate": 1.1655000000000001e-05, + "loss": 0.0856, + "step": 3888 + }, + { + "epoch": 3.2846283783783785, + "grad_norm": 0.7715592980384827, + "learning_rate": 1.1658000000000001e-05, + "loss": 0.0533, + "step": 3889 + }, + { + "epoch": 3.285472972972973, + "grad_norm": 0.7316685318946838, + "learning_rate": 1.1661e-05, + "loss": 0.0551, + "step": 3890 + }, + { + "epoch": 3.2863175675675675, + "grad_norm": 0.9218887686729431, + "learning_rate": 1.1664e-05, + "loss": 0.0741, + "step": 3891 + }, + { + "epoch": 3.2871621621621623, + "grad_norm": 0.7153891921043396, + "learning_rate": 1.1667e-05, + "loss": 0.0877, + "step": 3892 + }, + { + "epoch": 3.2880067567567566, + "grad_norm": 0.6308774948120117, + "learning_rate": 1.167e-05, + "loss": 0.0711, + "step": 3893 + }, + { + "epoch": 3.2888513513513513, + "grad_norm": 0.7953395843505859, + "learning_rate": 1.1673e-05, + "loss": 0.0727, + "step": 3894 + }, + { + "epoch": 3.289695945945946, + "grad_norm": 0.7823639512062073, + "learning_rate": 1.1676e-05, + "loss": 0.0667, + "step": 3895 + }, + { + "epoch": 3.2905405405405403, + "grad_norm": 1.0176653861999512, + "learning_rate": 1.1679e-05, + "loss": 0.0907, + "step": 3896 + }, + { + "epoch": 3.291385135135135, + "grad_norm": 0.6995924115180969, + "learning_rate": 1.1682000000000001e-05, + "loss": 0.0469, + "step": 3897 + }, + { + "epoch": 3.29222972972973, + "grad_norm": 0.8960579633712769, + "learning_rate": 1.1685e-05, + "loss": 0.0772, + "step": 3898 + }, + { + "epoch": 3.2930743243243246, + "grad_norm": 1.0045534372329712, + "learning_rate": 1.1688e-05, + "loss": 0.1112, + "step": 3899 + }, + { + "epoch": 3.293918918918919, + "grad_norm": 0.9906949400901794, + "learning_rate": 1.1691e-05, + "loss": 0.0903, + "step": 3900 + }, + { + "epoch": 3.2947635135135136, + "grad_norm": 2.0007855892181396, + "learning_rate": 1.1694e-05, + "loss": 0.1254, + "step": 3901 + }, + { + "epoch": 3.295608108108108, + "grad_norm": 2.795485496520996, + "learning_rate": 1.1697000000000002e-05, + "loss": 0.1357, + "step": 3902 + }, + { + "epoch": 3.2964527027027026, + "grad_norm": 1.7441269159317017, + "learning_rate": 1.1700000000000001e-05, + "loss": 0.4938, + "step": 3903 + }, + { + "epoch": 3.2972972972972974, + "grad_norm": 1.0046314001083374, + "learning_rate": 1.1703e-05, + "loss": 0.2409, + "step": 3904 + }, + { + "epoch": 3.298141891891892, + "grad_norm": 0.7036058902740479, + "learning_rate": 1.1706e-05, + "loss": 0.1233, + "step": 3905 + }, + { + "epoch": 3.2989864864864864, + "grad_norm": 0.7122266888618469, + "learning_rate": 1.1709e-05, + "loss": 0.0854, + "step": 3906 + }, + { + "epoch": 3.299831081081081, + "grad_norm": 0.6676765084266663, + "learning_rate": 1.1712e-05, + "loss": 0.0801, + "step": 3907 + }, + { + "epoch": 3.3006756756756754, + "grad_norm": 0.8892258405685425, + "learning_rate": 1.1715e-05, + "loss": 0.0779, + "step": 3908 + }, + { + "epoch": 3.30152027027027, + "grad_norm": 0.7634930610656738, + "learning_rate": 1.1718e-05, + "loss": 0.0762, + "step": 3909 + }, + { + "epoch": 3.302364864864865, + "grad_norm": 0.5710618495941162, + "learning_rate": 1.1721e-05, + "loss": 0.045, + "step": 3910 + }, + { + "epoch": 3.3032094594594597, + "grad_norm": 0.63881516456604, + "learning_rate": 1.1724e-05, + "loss": 0.08, + "step": 3911 + }, + { + "epoch": 3.304054054054054, + "grad_norm": 0.49863728880882263, + "learning_rate": 1.1727000000000001e-05, + "loss": 0.045, + "step": 3912 + }, + { + "epoch": 3.3048986486486487, + "grad_norm": 0.7920162081718445, + "learning_rate": 1.1730000000000001e-05, + "loss": 0.0788, + "step": 3913 + }, + { + "epoch": 3.3057432432432434, + "grad_norm": 0.7156081795692444, + "learning_rate": 1.1733000000000001e-05, + "loss": 0.0696, + "step": 3914 + }, + { + "epoch": 3.3065878378378377, + "grad_norm": 0.6441711783409119, + "learning_rate": 1.1736e-05, + "loss": 0.0723, + "step": 3915 + }, + { + "epoch": 3.3074324324324325, + "grad_norm": 0.6569247245788574, + "learning_rate": 1.1738999999999999e-05, + "loss": 0.0547, + "step": 3916 + }, + { + "epoch": 3.308277027027027, + "grad_norm": 0.6304224133491516, + "learning_rate": 1.1742e-05, + "loss": 0.07, + "step": 3917 + }, + { + "epoch": 3.3091216216216215, + "grad_norm": 0.6787375211715698, + "learning_rate": 1.1745e-05, + "loss": 0.0608, + "step": 3918 + }, + { + "epoch": 3.3099662162162162, + "grad_norm": 0.5891062617301941, + "learning_rate": 1.1748e-05, + "loss": 0.0547, + "step": 3919 + }, + { + "epoch": 3.310810810810811, + "grad_norm": 0.6670762300491333, + "learning_rate": 1.1751e-05, + "loss": 0.0731, + "step": 3920 + }, + { + "epoch": 3.3116554054054053, + "grad_norm": 0.609984815120697, + "learning_rate": 1.1754e-05, + "loss": 0.066, + "step": 3921 + }, + { + "epoch": 3.3125, + "grad_norm": 0.6675212979316711, + "learning_rate": 1.1757000000000001e-05, + "loss": 0.0743, + "step": 3922 + }, + { + "epoch": 3.3133445945945947, + "grad_norm": 0.6948481202125549, + "learning_rate": 1.1760000000000001e-05, + "loss": 0.0891, + "step": 3923 + }, + { + "epoch": 3.314189189189189, + "grad_norm": 0.7549641132354736, + "learning_rate": 1.1763e-05, + "loss": 0.0672, + "step": 3924 + }, + { + "epoch": 3.3150337837837838, + "grad_norm": 0.7961869835853577, + "learning_rate": 1.1766e-05, + "loss": 0.0734, + "step": 3925 + }, + { + "epoch": 3.3158783783783785, + "grad_norm": 0.8594260811805725, + "learning_rate": 1.1769e-05, + "loss": 0.1189, + "step": 3926 + }, + { + "epoch": 3.316722972972973, + "grad_norm": 0.556521475315094, + "learning_rate": 1.1772000000000002e-05, + "loss": 0.0592, + "step": 3927 + }, + { + "epoch": 3.3175675675675675, + "grad_norm": 0.7132596373558044, + "learning_rate": 1.1775000000000002e-05, + "loss": 0.0593, + "step": 3928 + }, + { + "epoch": 3.3184121621621623, + "grad_norm": 0.6777545213699341, + "learning_rate": 1.1778e-05, + "loss": 0.0514, + "step": 3929 + }, + { + "epoch": 3.3192567567567566, + "grad_norm": 0.9635444283485413, + "learning_rate": 1.1781e-05, + "loss": 0.0825, + "step": 3930 + }, + { + "epoch": 3.3201013513513513, + "grad_norm": 0.6171340942382812, + "learning_rate": 1.1784e-05, + "loss": 0.0664, + "step": 3931 + }, + { + "epoch": 3.320945945945946, + "grad_norm": 0.860900342464447, + "learning_rate": 1.1787e-05, + "loss": 0.078, + "step": 3932 + }, + { + "epoch": 3.3217905405405403, + "grad_norm": 0.6942296028137207, + "learning_rate": 1.179e-05, + "loss": 0.0746, + "step": 3933 + }, + { + "epoch": 3.322635135135135, + "grad_norm": 0.597150981426239, + "learning_rate": 1.1793e-05, + "loss": 0.0575, + "step": 3934 + }, + { + "epoch": 3.32347972972973, + "grad_norm": 0.7416917085647583, + "learning_rate": 1.1796e-05, + "loss": 0.0781, + "step": 3935 + }, + { + "epoch": 3.3243243243243246, + "grad_norm": 2.237961530685425, + "learning_rate": 1.1799e-05, + "loss": 0.185, + "step": 3936 + }, + { + "epoch": 3.325168918918919, + "grad_norm": 0.5388156175613403, + "learning_rate": 1.1802000000000002e-05, + "loss": 0.0472, + "step": 3937 + }, + { + "epoch": 3.3260135135135136, + "grad_norm": 0.8778148889541626, + "learning_rate": 1.1805000000000001e-05, + "loss": 0.0756, + "step": 3938 + }, + { + "epoch": 3.326858108108108, + "grad_norm": 0.871222734451294, + "learning_rate": 1.1808000000000001e-05, + "loss": 0.1242, + "step": 3939 + }, + { + "epoch": 3.3277027027027026, + "grad_norm": 0.9781417846679688, + "learning_rate": 1.1811000000000001e-05, + "loss": 0.1009, + "step": 3940 + }, + { + "epoch": 3.3285472972972974, + "grad_norm": 0.8198784589767456, + "learning_rate": 1.1813999999999999e-05, + "loss": 0.0754, + "step": 3941 + }, + { + "epoch": 3.329391891891892, + "grad_norm": 1.1910786628723145, + "learning_rate": 1.1816999999999999e-05, + "loss": 0.0944, + "step": 3942 + }, + { + "epoch": 3.3302364864864864, + "grad_norm": 0.6850098967552185, + "learning_rate": 1.182e-05, + "loss": 0.0605, + "step": 3943 + }, + { + "epoch": 3.331081081081081, + "grad_norm": 1.125475287437439, + "learning_rate": 1.1823e-05, + "loss": 0.0846, + "step": 3944 + }, + { + "epoch": 3.3319256756756754, + "grad_norm": 0.6977891325950623, + "learning_rate": 1.1826e-05, + "loss": 0.058, + "step": 3945 + }, + { + "epoch": 3.33277027027027, + "grad_norm": 4.555857181549072, + "learning_rate": 1.1829e-05, + "loss": 0.0945, + "step": 3946 + }, + { + "epoch": 3.333614864864865, + "grad_norm": 0.9961831569671631, + "learning_rate": 1.1832e-05, + "loss": 0.0995, + "step": 3947 + }, + { + "epoch": 3.3344594594594597, + "grad_norm": 0.8152703046798706, + "learning_rate": 1.1835000000000001e-05, + "loss": 0.0778, + "step": 3948 + }, + { + "epoch": 3.335304054054054, + "grad_norm": 1.2011021375656128, + "learning_rate": 1.1838e-05, + "loss": 0.1012, + "step": 3949 + }, + { + "epoch": 3.3361486486486487, + "grad_norm": 1.1260334253311157, + "learning_rate": 1.1841e-05, + "loss": 0.1245, + "step": 3950 + }, + { + "epoch": 3.3369932432432434, + "grad_norm": 1.4148283004760742, + "learning_rate": 1.1844e-05, + "loss": 0.1466, + "step": 3951 + }, + { + "epoch": 3.3378378378378377, + "grad_norm": 1.859545350074768, + "learning_rate": 1.1847e-05, + "loss": 0.2154, + "step": 3952 + }, + { + "epoch": 3.3386824324324325, + "grad_norm": 2.1780481338500977, + "learning_rate": 1.185e-05, + "loss": 0.477, + "step": 3953 + }, + { + "epoch": 3.339527027027027, + "grad_norm": 2.3331525325775146, + "learning_rate": 1.1853e-05, + "loss": 0.4353, + "step": 3954 + }, + { + "epoch": 3.3403716216216215, + "grad_norm": 0.7704637050628662, + "learning_rate": 1.1856e-05, + "loss": 0.1342, + "step": 3955 + }, + { + "epoch": 3.3412162162162162, + "grad_norm": 0.6070114374160767, + "learning_rate": 1.1859e-05, + "loss": 0.0837, + "step": 3956 + }, + { + "epoch": 3.342060810810811, + "grad_norm": 0.5363158583641052, + "learning_rate": 1.1862e-05, + "loss": 0.0635, + "step": 3957 + }, + { + "epoch": 3.3429054054054053, + "grad_norm": 0.7309954166412354, + "learning_rate": 1.1865e-05, + "loss": 0.096, + "step": 3958 + }, + { + "epoch": 3.34375, + "grad_norm": 0.5602104067802429, + "learning_rate": 1.1868e-05, + "loss": 0.0668, + "step": 3959 + }, + { + "epoch": 3.3445945945945947, + "grad_norm": 0.598757803440094, + "learning_rate": 1.1871e-05, + "loss": 0.0645, + "step": 3960 + }, + { + "epoch": 3.345439189189189, + "grad_norm": 0.7280358076095581, + "learning_rate": 1.1874e-05, + "loss": 0.074, + "step": 3961 + }, + { + "epoch": 3.3462837837837838, + "grad_norm": 0.4402478337287903, + "learning_rate": 1.1877e-05, + "loss": 0.041, + "step": 3962 + }, + { + "epoch": 3.3471283783783785, + "grad_norm": 0.7299851775169373, + "learning_rate": 1.1880000000000001e-05, + "loss": 0.092, + "step": 3963 + }, + { + "epoch": 3.347972972972973, + "grad_norm": 0.5911464691162109, + "learning_rate": 1.1883000000000001e-05, + "loss": 0.0685, + "step": 3964 + }, + { + "epoch": 3.3488175675675675, + "grad_norm": 0.6729894280433655, + "learning_rate": 1.1886e-05, + "loss": 0.0822, + "step": 3965 + }, + { + "epoch": 3.3496621621621623, + "grad_norm": 0.5395510196685791, + "learning_rate": 1.1889e-05, + "loss": 0.0633, + "step": 3966 + }, + { + "epoch": 3.3505067567567566, + "grad_norm": 0.44593098759651184, + "learning_rate": 1.1891999999999999e-05, + "loss": 0.0552, + "step": 3967 + }, + { + "epoch": 3.3513513513513513, + "grad_norm": 0.48029038310050964, + "learning_rate": 1.1895e-05, + "loss": 0.0742, + "step": 3968 + }, + { + "epoch": 3.352195945945946, + "grad_norm": 0.6809232234954834, + "learning_rate": 1.1898e-05, + "loss": 0.0649, + "step": 3969 + }, + { + "epoch": 3.3530405405405403, + "grad_norm": 0.6150186061859131, + "learning_rate": 1.1901e-05, + "loss": 0.0434, + "step": 3970 + }, + { + "epoch": 3.353885135135135, + "grad_norm": 0.7431069016456604, + "learning_rate": 1.1904e-05, + "loss": 0.0868, + "step": 3971 + }, + { + "epoch": 3.35472972972973, + "grad_norm": 0.9875496625900269, + "learning_rate": 1.1907e-05, + "loss": 0.1135, + "step": 3972 + }, + { + "epoch": 3.3555743243243246, + "grad_norm": 0.6851582527160645, + "learning_rate": 1.1910000000000001e-05, + "loss": 0.0621, + "step": 3973 + }, + { + "epoch": 3.356418918918919, + "grad_norm": 0.7411687970161438, + "learning_rate": 1.1913000000000001e-05, + "loss": 0.0583, + "step": 3974 + }, + { + "epoch": 3.3572635135135136, + "grad_norm": 0.7120720148086548, + "learning_rate": 1.1916e-05, + "loss": 0.0775, + "step": 3975 + }, + { + "epoch": 3.358108108108108, + "grad_norm": 0.6394950747489929, + "learning_rate": 1.1919e-05, + "loss": 0.0651, + "step": 3976 + }, + { + "epoch": 3.3589527027027026, + "grad_norm": 0.7171327471733093, + "learning_rate": 1.1922e-05, + "loss": 0.0632, + "step": 3977 + }, + { + "epoch": 3.3597972972972974, + "grad_norm": 0.5487982034683228, + "learning_rate": 1.1925e-05, + "loss": 0.0717, + "step": 3978 + }, + { + "epoch": 3.360641891891892, + "grad_norm": 0.8264620900154114, + "learning_rate": 1.1928e-05, + "loss": 0.0873, + "step": 3979 + }, + { + "epoch": 3.3614864864864864, + "grad_norm": 1.1969830989837646, + "learning_rate": 1.1931e-05, + "loss": 0.0488, + "step": 3980 + }, + { + "epoch": 3.362331081081081, + "grad_norm": 0.598273515701294, + "learning_rate": 1.1934e-05, + "loss": 0.0521, + "step": 3981 + }, + { + "epoch": 3.3631756756756754, + "grad_norm": 0.6452941298484802, + "learning_rate": 1.1937e-05, + "loss": 0.0803, + "step": 3982 + }, + { + "epoch": 3.36402027027027, + "grad_norm": 1.2869802713394165, + "learning_rate": 1.1940000000000001e-05, + "loss": 0.0738, + "step": 3983 + }, + { + "epoch": 3.364864864864865, + "grad_norm": 0.6850610375404358, + "learning_rate": 1.1943e-05, + "loss": 0.0596, + "step": 3984 + }, + { + "epoch": 3.3657094594594597, + "grad_norm": 0.6550266742706299, + "learning_rate": 1.1946e-05, + "loss": 0.0595, + "step": 3985 + }, + { + "epoch": 3.366554054054054, + "grad_norm": 0.802826464176178, + "learning_rate": 1.1949e-05, + "loss": 0.0939, + "step": 3986 + }, + { + "epoch": 3.3673986486486487, + "grad_norm": 1.3286683559417725, + "learning_rate": 1.1952e-05, + "loss": 0.0788, + "step": 3987 + }, + { + "epoch": 3.3682432432432434, + "grad_norm": 0.8703361749649048, + "learning_rate": 1.1955000000000002e-05, + "loss": 0.0696, + "step": 3988 + }, + { + "epoch": 3.3690878378378377, + "grad_norm": 0.9478511810302734, + "learning_rate": 1.1958000000000001e-05, + "loss": 0.1194, + "step": 3989 + }, + { + "epoch": 3.3699324324324325, + "grad_norm": 0.7556315064430237, + "learning_rate": 1.1961e-05, + "loss": 0.0881, + "step": 3990 + }, + { + "epoch": 3.370777027027027, + "grad_norm": 0.6798964142799377, + "learning_rate": 1.1964e-05, + "loss": 0.0552, + "step": 3991 + }, + { + "epoch": 3.3716216216216215, + "grad_norm": 0.677491307258606, + "learning_rate": 1.1966999999999999e-05, + "loss": 0.0747, + "step": 3992 + }, + { + "epoch": 3.3724662162162162, + "grad_norm": 0.763908863067627, + "learning_rate": 1.197e-05, + "loss": 0.0785, + "step": 3993 + }, + { + "epoch": 3.373310810810811, + "grad_norm": 0.8131700754165649, + "learning_rate": 1.1973e-05, + "loss": 0.0657, + "step": 3994 + }, + { + "epoch": 3.3741554054054053, + "grad_norm": 0.9891523122787476, + "learning_rate": 1.1976e-05, + "loss": 0.0624, + "step": 3995 + }, + { + "epoch": 3.375, + "grad_norm": 0.7831671833992004, + "learning_rate": 1.1979e-05, + "loss": 0.0728, + "step": 3996 + }, + { + "epoch": 3.3758445945945947, + "grad_norm": 1.0534862279891968, + "learning_rate": 1.1982e-05, + "loss": 0.0957, + "step": 3997 + }, + { + "epoch": 3.376689189189189, + "grad_norm": 1.0769959688186646, + "learning_rate": 1.1985000000000001e-05, + "loss": 0.1031, + "step": 3998 + }, + { + "epoch": 3.3775337837837838, + "grad_norm": 0.8610873222351074, + "learning_rate": 1.1988000000000001e-05, + "loss": 0.0649, + "step": 3999 + }, + { + "epoch": 3.3783783783783785, + "grad_norm": 1.0583992004394531, + "learning_rate": 1.1991000000000001e-05, + "loss": 0.1017, + "step": 4000 + }, + { + "epoch": 3.3783783783783785, + "eval_cer": 0.11332423098546301, + "eval_loss": 0.3857017457485199, + "eval_runtime": 8.1555, + "eval_samples_per_second": 12.384, + "eval_steps_per_second": 0.49, + "eval_wer": 0.3531746031746032, + "step": 4000 + }, + { + "epoch": 3.379222972972973, + "grad_norm": 1.4095873832702637, + "learning_rate": 1.1994e-05, + "loss": 0.1251, + "step": 4001 + }, + { + "epoch": 3.3800675675675675, + "grad_norm": 1.5022720098495483, + "learning_rate": 1.1996999999999999e-05, + "loss": 0.1392, + "step": 4002 + }, + { + "epoch": 3.3809121621621623, + "grad_norm": 1.6833398342132568, + "learning_rate": 1.2e-05, + "loss": 0.4184, + "step": 4003 + }, + { + "epoch": 3.3817567567567566, + "grad_norm": 1.2138605117797852, + "learning_rate": 1.2003e-05, + "loss": 0.2921, + "step": 4004 + }, + { + "epoch": 3.3826013513513513, + "grad_norm": 0.8538126349449158, + "learning_rate": 1.2006e-05, + "loss": 0.1723, + "step": 4005 + }, + { + "epoch": 3.383445945945946, + "grad_norm": 1.0138370990753174, + "learning_rate": 1.2009e-05, + "loss": 0.1196, + "step": 4006 + }, + { + "epoch": 3.3842905405405403, + "grad_norm": 0.7465329170227051, + "learning_rate": 1.2012e-05, + "loss": 0.0973, + "step": 4007 + }, + { + "epoch": 3.385135135135135, + "grad_norm": 0.5002086162567139, + "learning_rate": 1.2015000000000001e-05, + "loss": 0.0658, + "step": 4008 + }, + { + "epoch": 3.38597972972973, + "grad_norm": 0.5266256928443909, + "learning_rate": 1.2018e-05, + "loss": 0.0578, + "step": 4009 + }, + { + "epoch": 3.3868243243243246, + "grad_norm": 0.693436324596405, + "learning_rate": 1.2021e-05, + "loss": 0.1131, + "step": 4010 + }, + { + "epoch": 3.387668918918919, + "grad_norm": 0.6322379112243652, + "learning_rate": 1.2024e-05, + "loss": 0.0749, + "step": 4011 + }, + { + "epoch": 3.3885135135135136, + "grad_norm": 0.8049827814102173, + "learning_rate": 1.2027e-05, + "loss": 0.1099, + "step": 4012 + }, + { + "epoch": 3.389358108108108, + "grad_norm": 0.5424456000328064, + "learning_rate": 1.2030000000000002e-05, + "loss": 0.0542, + "step": 4013 + }, + { + "epoch": 3.3902027027027026, + "grad_norm": 0.8745025396347046, + "learning_rate": 1.2033000000000002e-05, + "loss": 0.0537, + "step": 4014 + }, + { + "epoch": 3.3910472972972974, + "grad_norm": 0.5911552906036377, + "learning_rate": 1.2036e-05, + "loss": 0.0698, + "step": 4015 + }, + { + "epoch": 3.391891891891892, + "grad_norm": 0.6019206643104553, + "learning_rate": 1.2039e-05, + "loss": 0.0761, + "step": 4016 + }, + { + "epoch": 3.3927364864864864, + "grad_norm": 0.49862784147262573, + "learning_rate": 1.2042e-05, + "loss": 0.0687, + "step": 4017 + }, + { + "epoch": 3.393581081081081, + "grad_norm": 0.5550438165664673, + "learning_rate": 1.2045e-05, + "loss": 0.0674, + "step": 4018 + }, + { + "epoch": 3.3944256756756754, + "grad_norm": 0.7954584956169128, + "learning_rate": 1.2048e-05, + "loss": 0.0692, + "step": 4019 + }, + { + "epoch": 3.39527027027027, + "grad_norm": 0.8548818826675415, + "learning_rate": 1.2051e-05, + "loss": 0.0881, + "step": 4020 + }, + { + "epoch": 3.396114864864865, + "grad_norm": 0.5945700407028198, + "learning_rate": 1.2054e-05, + "loss": 0.0602, + "step": 4021 + }, + { + "epoch": 3.3969594594594597, + "grad_norm": 0.6880996227264404, + "learning_rate": 1.2057e-05, + "loss": 0.0695, + "step": 4022 + }, + { + "epoch": 3.397804054054054, + "grad_norm": 0.6182917356491089, + "learning_rate": 1.2060000000000001e-05, + "loss": 0.0593, + "step": 4023 + }, + { + "epoch": 3.3986486486486487, + "grad_norm": 0.4329436719417572, + "learning_rate": 1.2063000000000001e-05, + "loss": 0.0515, + "step": 4024 + }, + { + "epoch": 3.3994932432432434, + "grad_norm": 0.6675774455070496, + "learning_rate": 1.2066000000000001e-05, + "loss": 0.0685, + "step": 4025 + }, + { + "epoch": 3.4003378378378377, + "grad_norm": 0.6946617960929871, + "learning_rate": 1.2069e-05, + "loss": 0.0668, + "step": 4026 + }, + { + "epoch": 3.4011824324324325, + "grad_norm": 0.7148919701576233, + "learning_rate": 1.2071999999999999e-05, + "loss": 0.076, + "step": 4027 + }, + { + "epoch": 3.402027027027027, + "grad_norm": 0.7005245685577393, + "learning_rate": 1.2075e-05, + "loss": 0.0794, + "step": 4028 + }, + { + "epoch": 3.4028716216216215, + "grad_norm": 0.8174470663070679, + "learning_rate": 1.2078e-05, + "loss": 0.0788, + "step": 4029 + }, + { + "epoch": 3.4037162162162162, + "grad_norm": 0.8142012357711792, + "learning_rate": 1.2081e-05, + "loss": 0.0809, + "step": 4030 + }, + { + "epoch": 3.404560810810811, + "grad_norm": 0.729363739490509, + "learning_rate": 1.2084e-05, + "loss": 0.0703, + "step": 4031 + }, + { + "epoch": 3.4054054054054053, + "grad_norm": 0.6651347279548645, + "learning_rate": 1.2087e-05, + "loss": 0.0588, + "step": 4032 + }, + { + "epoch": 3.40625, + "grad_norm": 0.9170257449150085, + "learning_rate": 1.2090000000000001e-05, + "loss": 0.0962, + "step": 4033 + }, + { + "epoch": 3.4070945945945947, + "grad_norm": 0.7313915491104126, + "learning_rate": 1.2093000000000001e-05, + "loss": 0.0618, + "step": 4034 + }, + { + "epoch": 3.407939189189189, + "grad_norm": 0.8310635089874268, + "learning_rate": 1.2096e-05, + "loss": 0.0691, + "step": 4035 + }, + { + "epoch": 3.4087837837837838, + "grad_norm": 0.6817588806152344, + "learning_rate": 1.2099e-05, + "loss": 0.0746, + "step": 4036 + }, + { + "epoch": 3.4096283783783785, + "grad_norm": 0.7460519671440125, + "learning_rate": 1.2102e-05, + "loss": 0.0721, + "step": 4037 + }, + { + "epoch": 3.410472972972973, + "grad_norm": 1.5166233777999878, + "learning_rate": 1.2105000000000002e-05, + "loss": 0.0784, + "step": 4038 + }, + { + "epoch": 3.4113175675675675, + "grad_norm": 0.7825971841812134, + "learning_rate": 1.2108e-05, + "loss": 0.0734, + "step": 4039 + }, + { + "epoch": 3.4121621621621623, + "grad_norm": 0.666084885597229, + "learning_rate": 1.2111e-05, + "loss": 0.0582, + "step": 4040 + }, + { + "epoch": 3.4130067567567566, + "grad_norm": 0.6607237458229065, + "learning_rate": 1.2114e-05, + "loss": 0.0852, + "step": 4041 + }, + { + "epoch": 3.4138513513513513, + "grad_norm": 1.1115988492965698, + "learning_rate": 1.2117e-05, + "loss": 0.109, + "step": 4042 + }, + { + "epoch": 3.414695945945946, + "grad_norm": 0.6287840008735657, + "learning_rate": 1.2120000000000001e-05, + "loss": 0.0586, + "step": 4043 + }, + { + "epoch": 3.4155405405405403, + "grad_norm": 0.8936541676521301, + "learning_rate": 1.2123e-05, + "loss": 0.0635, + "step": 4044 + }, + { + "epoch": 3.416385135135135, + "grad_norm": 0.8310178518295288, + "learning_rate": 1.2126e-05, + "loss": 0.1057, + "step": 4045 + }, + { + "epoch": 3.41722972972973, + "grad_norm": 0.7518346905708313, + "learning_rate": 1.2129e-05, + "loss": 0.0489, + "step": 4046 + }, + { + "epoch": 3.4180743243243246, + "grad_norm": 0.881924569606781, + "learning_rate": 1.2132e-05, + "loss": 0.0641, + "step": 4047 + }, + { + "epoch": 3.418918918918919, + "grad_norm": 0.9821265339851379, + "learning_rate": 1.2135000000000002e-05, + "loss": 0.093, + "step": 4048 + }, + { + "epoch": 3.4197635135135136, + "grad_norm": 0.9290544986724854, + "learning_rate": 1.2138000000000001e-05, + "loss": 0.0797, + "step": 4049 + }, + { + "epoch": 3.420608108108108, + "grad_norm": 4.446503162384033, + "learning_rate": 1.2141000000000001e-05, + "loss": 0.2081, + "step": 4050 + }, + { + "epoch": 3.4214527027027026, + "grad_norm": 1.442778468132019, + "learning_rate": 1.2144e-05, + "loss": 0.1573, + "step": 4051 + }, + { + "epoch": 3.4222972972972974, + "grad_norm": 1.6602102518081665, + "learning_rate": 1.2146999999999999e-05, + "loss": 0.1081, + "step": 4052 + }, + { + "epoch": 3.423141891891892, + "grad_norm": 3.9706835746765137, + "learning_rate": 1.215e-05, + "loss": 0.4031, + "step": 4053 + }, + { + "epoch": 3.4239864864864864, + "grad_norm": 1.6365985870361328, + "learning_rate": 1.2153e-05, + "loss": 0.2714, + "step": 4054 + }, + { + "epoch": 3.424831081081081, + "grad_norm": 0.784578263759613, + "learning_rate": 1.2156e-05, + "loss": 0.1386, + "step": 4055 + }, + { + "epoch": 3.4256756756756754, + "grad_norm": 0.6822528839111328, + "learning_rate": 1.2159e-05, + "loss": 0.1248, + "step": 4056 + }, + { + "epoch": 3.42652027027027, + "grad_norm": 0.673846423625946, + "learning_rate": 1.2162e-05, + "loss": 0.0951, + "step": 4057 + }, + { + "epoch": 3.427364864864865, + "grad_norm": 0.6397392749786377, + "learning_rate": 1.2165000000000001e-05, + "loss": 0.099, + "step": 4058 + }, + { + "epoch": 3.4282094594594597, + "grad_norm": 0.78976970911026, + "learning_rate": 1.2168000000000001e-05, + "loss": 0.0668, + "step": 4059 + }, + { + "epoch": 3.429054054054054, + "grad_norm": 0.6037519574165344, + "learning_rate": 1.2171000000000001e-05, + "loss": 0.0695, + "step": 4060 + }, + { + "epoch": 3.4298986486486487, + "grad_norm": 0.9149604439735413, + "learning_rate": 1.2174e-05, + "loss": 0.0728, + "step": 4061 + }, + { + "epoch": 3.4307432432432434, + "grad_norm": 0.4959262013435364, + "learning_rate": 1.2177e-05, + "loss": 0.0642, + "step": 4062 + }, + { + "epoch": 3.4315878378378377, + "grad_norm": 0.4927270710468292, + "learning_rate": 1.2180000000000002e-05, + "loss": 0.0402, + "step": 4063 + }, + { + "epoch": 3.4324324324324325, + "grad_norm": 0.6432793140411377, + "learning_rate": 1.2183e-05, + "loss": 0.0739, + "step": 4064 + }, + { + "epoch": 3.433277027027027, + "grad_norm": 0.5762895941734314, + "learning_rate": 1.2186e-05, + "loss": 0.0428, + "step": 4065 + }, + { + "epoch": 3.4341216216216215, + "grad_norm": 0.8307366371154785, + "learning_rate": 1.2189e-05, + "loss": 0.0845, + "step": 4066 + }, + { + "epoch": 3.4349662162162162, + "grad_norm": 0.7485628128051758, + "learning_rate": 1.2192e-05, + "loss": 0.0843, + "step": 4067 + }, + { + "epoch": 3.435810810810811, + "grad_norm": 0.7366847991943359, + "learning_rate": 1.2195e-05, + "loss": 0.0687, + "step": 4068 + }, + { + "epoch": 3.4366554054054053, + "grad_norm": 1.6700825691223145, + "learning_rate": 1.2198e-05, + "loss": 0.0469, + "step": 4069 + }, + { + "epoch": 3.4375, + "grad_norm": 0.5955573916435242, + "learning_rate": 1.2201e-05, + "loss": 0.0478, + "step": 4070 + }, + { + "epoch": 3.4383445945945947, + "grad_norm": 0.5699429512023926, + "learning_rate": 1.2204e-05, + "loss": 0.0477, + "step": 4071 + }, + { + "epoch": 3.439189189189189, + "grad_norm": 0.6617381572723389, + "learning_rate": 1.2207e-05, + "loss": 0.1005, + "step": 4072 + }, + { + "epoch": 3.4400337837837838, + "grad_norm": 0.7031179070472717, + "learning_rate": 1.221e-05, + "loss": 0.0828, + "step": 4073 + }, + { + "epoch": 3.4408783783783785, + "grad_norm": 0.9543467164039612, + "learning_rate": 1.2213000000000001e-05, + "loss": 0.0663, + "step": 4074 + }, + { + "epoch": 3.441722972972973, + "grad_norm": 0.7065834403038025, + "learning_rate": 1.2216000000000001e-05, + "loss": 0.0846, + "step": 4075 + }, + { + "epoch": 3.4425675675675675, + "grad_norm": 1.3215631246566772, + "learning_rate": 1.2219e-05, + "loss": 0.1919, + "step": 4076 + }, + { + "epoch": 3.4434121621621623, + "grad_norm": 0.6607271432876587, + "learning_rate": 1.2222e-05, + "loss": 0.0757, + "step": 4077 + }, + { + "epoch": 3.4442567567567566, + "grad_norm": 0.5578023195266724, + "learning_rate": 1.2224999999999999e-05, + "loss": 0.0522, + "step": 4078 + }, + { + "epoch": 3.4451013513513513, + "grad_norm": 0.7703476548194885, + "learning_rate": 1.2228e-05, + "loss": 0.1044, + "step": 4079 + }, + { + "epoch": 3.445945945945946, + "grad_norm": 0.8342571258544922, + "learning_rate": 1.2231e-05, + "loss": 0.0734, + "step": 4080 + }, + { + "epoch": 3.4467905405405403, + "grad_norm": 0.6740601062774658, + "learning_rate": 1.2234e-05, + "loss": 0.0605, + "step": 4081 + }, + { + "epoch": 3.447635135135135, + "grad_norm": 0.7889488935470581, + "learning_rate": 1.2237e-05, + "loss": 0.085, + "step": 4082 + }, + { + "epoch": 3.44847972972973, + "grad_norm": 0.5411627292633057, + "learning_rate": 1.224e-05, + "loss": 0.0564, + "step": 4083 + }, + { + "epoch": 3.4493243243243246, + "grad_norm": 1.0016065835952759, + "learning_rate": 1.2243000000000001e-05, + "loss": 0.0878, + "step": 4084 + }, + { + "epoch": 3.450168918918919, + "grad_norm": 0.6957688927650452, + "learning_rate": 1.2246000000000001e-05, + "loss": 0.0532, + "step": 4085 + }, + { + "epoch": 3.4510135135135136, + "grad_norm": 0.6796482801437378, + "learning_rate": 1.2249e-05, + "loss": 0.0643, + "step": 4086 + }, + { + "epoch": 3.451858108108108, + "grad_norm": 0.7964410781860352, + "learning_rate": 1.2252e-05, + "loss": 0.0874, + "step": 4087 + }, + { + "epoch": 3.4527027027027026, + "grad_norm": 0.6555958390235901, + "learning_rate": 1.2254999999999999e-05, + "loss": 0.0769, + "step": 4088 + }, + { + "epoch": 3.4535472972972974, + "grad_norm": 0.9693285822868347, + "learning_rate": 1.2258e-05, + "loss": 0.0817, + "step": 4089 + }, + { + "epoch": 3.454391891891892, + "grad_norm": 0.639249324798584, + "learning_rate": 1.2261e-05, + "loss": 0.064, + "step": 4090 + }, + { + "epoch": 3.4552364864864864, + "grad_norm": 0.7004167437553406, + "learning_rate": 1.2264e-05, + "loss": 0.0615, + "step": 4091 + }, + { + "epoch": 3.456081081081081, + "grad_norm": 0.7403393387794495, + "learning_rate": 1.2267e-05, + "loss": 0.091, + "step": 4092 + }, + { + "epoch": 3.4569256756756754, + "grad_norm": 0.870163083076477, + "learning_rate": 1.227e-05, + "loss": 0.0682, + "step": 4093 + }, + { + "epoch": 3.45777027027027, + "grad_norm": 0.737073540687561, + "learning_rate": 1.2273000000000001e-05, + "loss": 0.0775, + "step": 4094 + }, + { + "epoch": 3.458614864864865, + "grad_norm": 0.7446344494819641, + "learning_rate": 1.2276e-05, + "loss": 0.0855, + "step": 4095 + }, + { + "epoch": 3.4594594594594597, + "grad_norm": 0.6928022503852844, + "learning_rate": 1.2279e-05, + "loss": 0.0808, + "step": 4096 + }, + { + "epoch": 3.460304054054054, + "grad_norm": 0.7195981740951538, + "learning_rate": 1.2282e-05, + "loss": 0.06, + "step": 4097 + }, + { + "epoch": 3.4611486486486487, + "grad_norm": 0.9217014312744141, + "learning_rate": 1.2285e-05, + "loss": 0.0826, + "step": 4098 + }, + { + "epoch": 3.4619932432432434, + "grad_norm": 1.0721522569656372, + "learning_rate": 1.2288000000000002e-05, + "loss": 0.0928, + "step": 4099 + }, + { + "epoch": 3.4628378378378377, + "grad_norm": 0.9630815386772156, + "learning_rate": 1.2291000000000001e-05, + "loss": 0.0881, + "step": 4100 + }, + { + "epoch": 3.4636824324324325, + "grad_norm": 1.8878014087677002, + "learning_rate": 1.2294e-05, + "loss": 0.1703, + "step": 4101 + }, + { + "epoch": 3.464527027027027, + "grad_norm": 1.6089435815811157, + "learning_rate": 1.2297e-05, + "loss": 0.1291, + "step": 4102 + }, + { + "epoch": 3.4653716216216215, + "grad_norm": 2.763936758041382, + "learning_rate": 1.2299999999999999e-05, + "loss": 0.5532, + "step": 4103 + }, + { + "epoch": 3.4662162162162162, + "grad_norm": 0.9310737252235413, + "learning_rate": 1.2303e-05, + "loss": 0.2678, + "step": 4104 + }, + { + "epoch": 3.467060810810811, + "grad_norm": 1.153751015663147, + "learning_rate": 1.2306e-05, + "loss": 0.1462, + "step": 4105 + }, + { + "epoch": 3.4679054054054053, + "grad_norm": 0.5843374729156494, + "learning_rate": 1.2309e-05, + "loss": 0.1, + "step": 4106 + }, + { + "epoch": 3.46875, + "grad_norm": 0.6983466744422913, + "learning_rate": 1.2312e-05, + "loss": 0.0846, + "step": 4107 + }, + { + "epoch": 3.4695945945945947, + "grad_norm": 0.6697255969047546, + "learning_rate": 1.2315e-05, + "loss": 0.0721, + "step": 4108 + }, + { + "epoch": 3.470439189189189, + "grad_norm": 0.7753097414970398, + "learning_rate": 1.2318000000000001e-05, + "loss": 0.0903, + "step": 4109 + }, + { + "epoch": 3.4712837837837838, + "grad_norm": 0.8670817017555237, + "learning_rate": 1.2321000000000001e-05, + "loss": 0.087, + "step": 4110 + }, + { + "epoch": 3.4721283783783785, + "grad_norm": 0.6172165870666504, + "learning_rate": 1.2324000000000001e-05, + "loss": 0.0653, + "step": 4111 + }, + { + "epoch": 3.472972972972973, + "grad_norm": 0.5330771803855896, + "learning_rate": 1.2327e-05, + "loss": 0.0632, + "step": 4112 + }, + { + "epoch": 3.4738175675675675, + "grad_norm": 1.1487345695495605, + "learning_rate": 1.2329999999999999e-05, + "loss": 0.0789, + "step": 4113 + }, + { + "epoch": 3.4746621621621623, + "grad_norm": 0.8288916945457458, + "learning_rate": 1.2333e-05, + "loss": 0.0729, + "step": 4114 + }, + { + "epoch": 3.4755067567567566, + "grad_norm": 0.5544078946113586, + "learning_rate": 1.2336e-05, + "loss": 0.0521, + "step": 4115 + }, + { + "epoch": 3.4763513513513513, + "grad_norm": 0.6867651343345642, + "learning_rate": 1.2339e-05, + "loss": 0.0681, + "step": 4116 + }, + { + "epoch": 3.477195945945946, + "grad_norm": 1.0527716875076294, + "learning_rate": 1.2342e-05, + "loss": 0.0759, + "step": 4117 + }, + { + "epoch": 3.4780405405405403, + "grad_norm": 0.6701065301895142, + "learning_rate": 1.2345e-05, + "loss": 0.0589, + "step": 4118 + }, + { + "epoch": 3.478885135135135, + "grad_norm": 0.6042844653129578, + "learning_rate": 1.2348000000000001e-05, + "loss": 0.0539, + "step": 4119 + }, + { + "epoch": 3.47972972972973, + "grad_norm": 0.7885158658027649, + "learning_rate": 1.2351e-05, + "loss": 0.0914, + "step": 4120 + }, + { + "epoch": 3.4805743243243246, + "grad_norm": 0.5141008496284485, + "learning_rate": 1.2354e-05, + "loss": 0.0426, + "step": 4121 + }, + { + "epoch": 3.481418918918919, + "grad_norm": 0.6331264972686768, + "learning_rate": 1.2357e-05, + "loss": 0.0681, + "step": 4122 + }, + { + "epoch": 3.4822635135135136, + "grad_norm": 0.5194094181060791, + "learning_rate": 1.236e-05, + "loss": 0.063, + "step": 4123 + }, + { + "epoch": 3.483108108108108, + "grad_norm": 0.6720242500305176, + "learning_rate": 1.2363000000000002e-05, + "loss": 0.0651, + "step": 4124 + }, + { + "epoch": 3.4839527027027026, + "grad_norm": 0.9294724464416504, + "learning_rate": 1.2366e-05, + "loss": 0.0723, + "step": 4125 + }, + { + "epoch": 3.4847972972972974, + "grad_norm": 0.6674201488494873, + "learning_rate": 1.2369e-05, + "loss": 0.0751, + "step": 4126 + }, + { + "epoch": 3.485641891891892, + "grad_norm": 0.6866880059242249, + "learning_rate": 1.2372e-05, + "loss": 0.0646, + "step": 4127 + }, + { + "epoch": 3.4864864864864864, + "grad_norm": 0.8540074229240417, + "learning_rate": 1.2375e-05, + "loss": 0.0682, + "step": 4128 + }, + { + "epoch": 3.487331081081081, + "grad_norm": 0.5334758162498474, + "learning_rate": 1.2378e-05, + "loss": 0.0517, + "step": 4129 + }, + { + "epoch": 3.4881756756756754, + "grad_norm": 0.6759699583053589, + "learning_rate": 1.2381e-05, + "loss": 0.0659, + "step": 4130 + }, + { + "epoch": 3.48902027027027, + "grad_norm": 0.6017577648162842, + "learning_rate": 1.2384e-05, + "loss": 0.0502, + "step": 4131 + }, + { + "epoch": 3.489864864864865, + "grad_norm": 0.6604937314987183, + "learning_rate": 1.2387e-05, + "loss": 0.0849, + "step": 4132 + }, + { + "epoch": 3.4907094594594597, + "grad_norm": 0.8820559978485107, + "learning_rate": 1.239e-05, + "loss": 0.0703, + "step": 4133 + }, + { + "epoch": 3.491554054054054, + "grad_norm": 0.6553615927696228, + "learning_rate": 1.2393000000000001e-05, + "loss": 0.0736, + "step": 4134 + }, + { + "epoch": 3.4923986486486487, + "grad_norm": 0.7434465885162354, + "learning_rate": 1.2396000000000001e-05, + "loss": 0.0695, + "step": 4135 + }, + { + "epoch": 3.4932432432432434, + "grad_norm": 0.7110425233840942, + "learning_rate": 1.2399000000000001e-05, + "loss": 0.0866, + "step": 4136 + }, + { + "epoch": 3.4940878378378377, + "grad_norm": 0.7158728241920471, + "learning_rate": 1.2402e-05, + "loss": 0.0576, + "step": 4137 + }, + { + "epoch": 3.4949324324324325, + "grad_norm": 0.5987915396690369, + "learning_rate": 1.2404999999999999e-05, + "loss": 0.0538, + "step": 4138 + }, + { + "epoch": 3.495777027027027, + "grad_norm": 1.1830060482025146, + "learning_rate": 1.2408e-05, + "loss": 0.1056, + "step": 4139 + }, + { + "epoch": 3.4966216216216215, + "grad_norm": 0.6556545495986938, + "learning_rate": 1.2411e-05, + "loss": 0.0616, + "step": 4140 + }, + { + "epoch": 3.4974662162162162, + "grad_norm": 0.6199811697006226, + "learning_rate": 1.2414e-05, + "loss": 0.0702, + "step": 4141 + }, + { + "epoch": 3.498310810810811, + "grad_norm": 0.6932005286216736, + "learning_rate": 1.2417e-05, + "loss": 0.0618, + "step": 4142 + }, + { + "epoch": 3.4991554054054053, + "grad_norm": 0.7317396998405457, + "learning_rate": 1.242e-05, + "loss": 0.0629, + "step": 4143 + }, + { + "epoch": 3.5, + "grad_norm": 0.8401498198509216, + "learning_rate": 1.2423000000000001e-05, + "loss": 0.0671, + "step": 4144 + }, + { + "epoch": 3.5008445945945947, + "grad_norm": 0.6792809367179871, + "learning_rate": 1.2426000000000001e-05, + "loss": 0.0721, + "step": 4145 + }, + { + "epoch": 3.501689189189189, + "grad_norm": 0.9104238152503967, + "learning_rate": 1.2429e-05, + "loss": 0.0542, + "step": 4146 + }, + { + "epoch": 3.5025337837837838, + "grad_norm": 1.3253798484802246, + "learning_rate": 1.2432e-05, + "loss": 0.0725, + "step": 4147 + }, + { + "epoch": 3.5033783783783785, + "grad_norm": 0.7717822194099426, + "learning_rate": 1.2435e-05, + "loss": 0.0708, + "step": 4148 + }, + { + "epoch": 3.504222972972973, + "grad_norm": 1.0167545080184937, + "learning_rate": 1.2438000000000002e-05, + "loss": 0.0764, + "step": 4149 + }, + { + "epoch": 3.5050675675675675, + "grad_norm": 1.013974666595459, + "learning_rate": 1.2441e-05, + "loss": 0.1145, + "step": 4150 + }, + { + "epoch": 3.5059121621621623, + "grad_norm": 1.0452824831008911, + "learning_rate": 1.2444e-05, + "loss": 0.1143, + "step": 4151 + }, + { + "epoch": 3.506756756756757, + "grad_norm": 1.8111730813980103, + "learning_rate": 1.2447e-05, + "loss": 0.1916, + "step": 4152 + }, + { + "epoch": 3.5076013513513513, + "grad_norm": 1.9601924419403076, + "learning_rate": 1.245e-05, + "loss": 0.4502, + "step": 4153 + }, + { + "epoch": 3.508445945945946, + "grad_norm": 1.2430309057235718, + "learning_rate": 1.2453000000000001e-05, + "loss": 0.243, + "step": 4154 + }, + { + "epoch": 3.5092905405405403, + "grad_norm": 0.6473004817962646, + "learning_rate": 1.2456e-05, + "loss": 0.1157, + "step": 4155 + }, + { + "epoch": 3.510135135135135, + "grad_norm": 0.6514524221420288, + "learning_rate": 1.2459e-05, + "loss": 0.1253, + "step": 4156 + }, + { + "epoch": 3.51097972972973, + "grad_norm": 0.7022819519042969, + "learning_rate": 1.2462e-05, + "loss": 0.1077, + "step": 4157 + }, + { + "epoch": 3.5118243243243246, + "grad_norm": 1.023717999458313, + "learning_rate": 1.2465e-05, + "loss": 0.1182, + "step": 4158 + }, + { + "epoch": 3.512668918918919, + "grad_norm": 0.6838351488113403, + "learning_rate": 1.2468000000000002e-05, + "loss": 0.0897, + "step": 4159 + }, + { + "epoch": 3.5135135135135136, + "grad_norm": 0.9117250442504883, + "learning_rate": 1.2471000000000001e-05, + "loss": 0.0745, + "step": 4160 + }, + { + "epoch": 3.514358108108108, + "grad_norm": 0.6636767387390137, + "learning_rate": 1.2474000000000001e-05, + "loss": 0.0736, + "step": 4161 + }, + { + "epoch": 3.5152027027027026, + "grad_norm": 0.6015921235084534, + "learning_rate": 1.2477e-05, + "loss": 0.0712, + "step": 4162 + }, + { + "epoch": 3.5160472972972974, + "grad_norm": 0.6836204528808594, + "learning_rate": 1.2479999999999999e-05, + "loss": 0.0644, + "step": 4163 + }, + { + "epoch": 3.516891891891892, + "grad_norm": 0.519741415977478, + "learning_rate": 1.2483e-05, + "loss": 0.062, + "step": 4164 + }, + { + "epoch": 3.5177364864864864, + "grad_norm": 0.7087646126747131, + "learning_rate": 1.2486e-05, + "loss": 0.0804, + "step": 4165 + }, + { + "epoch": 3.518581081081081, + "grad_norm": 0.5089424848556519, + "learning_rate": 1.2489e-05, + "loss": 0.0551, + "step": 4166 + }, + { + "epoch": 3.5194256756756754, + "grad_norm": 0.4815307855606079, + "learning_rate": 1.2492e-05, + "loss": 0.0615, + "step": 4167 + }, + { + "epoch": 3.52027027027027, + "grad_norm": 0.5566580891609192, + "learning_rate": 1.2495e-05, + "loss": 0.0675, + "step": 4168 + }, + { + "epoch": 3.521114864864865, + "grad_norm": 0.7243868708610535, + "learning_rate": 1.2498000000000001e-05, + "loss": 0.0696, + "step": 4169 + }, + { + "epoch": 3.5219594594594597, + "grad_norm": 0.9186239838600159, + "learning_rate": 1.2501000000000001e-05, + "loss": 0.0949, + "step": 4170 + }, + { + "epoch": 3.522804054054054, + "grad_norm": 0.4936967194080353, + "learning_rate": 1.2504000000000001e-05, + "loss": 0.0475, + "step": 4171 + }, + { + "epoch": 3.5236486486486487, + "grad_norm": 0.5168708562850952, + "learning_rate": 1.2507e-05, + "loss": 0.0499, + "step": 4172 + }, + { + "epoch": 3.524493243243243, + "grad_norm": 0.6472837328910828, + "learning_rate": 1.251e-05, + "loss": 0.0725, + "step": 4173 + }, + { + "epoch": 3.5253378378378377, + "grad_norm": 0.7847802639007568, + "learning_rate": 1.2513e-05, + "loss": 0.0755, + "step": 4174 + }, + { + "epoch": 3.5261824324324325, + "grad_norm": 0.6649460196495056, + "learning_rate": 1.2516e-05, + "loss": 0.0626, + "step": 4175 + }, + { + "epoch": 3.527027027027027, + "grad_norm": 0.6137911677360535, + "learning_rate": 1.2519e-05, + "loss": 0.0895, + "step": 4176 + }, + { + "epoch": 3.5278716216216215, + "grad_norm": 0.6711657643318176, + "learning_rate": 1.2522e-05, + "loss": 0.0784, + "step": 4177 + }, + { + "epoch": 3.5287162162162162, + "grad_norm": 0.6890873312950134, + "learning_rate": 1.2525e-05, + "loss": 0.0632, + "step": 4178 + }, + { + "epoch": 3.529560810810811, + "grad_norm": 0.7976025938987732, + "learning_rate": 1.2528000000000001e-05, + "loss": 0.0788, + "step": 4179 + }, + { + "epoch": 3.5304054054054053, + "grad_norm": 0.7949998378753662, + "learning_rate": 1.2531e-05, + "loss": 0.0723, + "step": 4180 + }, + { + "epoch": 3.53125, + "grad_norm": 0.7735546827316284, + "learning_rate": 1.2534e-05, + "loss": 0.0774, + "step": 4181 + }, + { + "epoch": 3.5320945945945947, + "grad_norm": 0.6867329478263855, + "learning_rate": 1.2537e-05, + "loss": 0.0794, + "step": 4182 + }, + { + "epoch": 3.532939189189189, + "grad_norm": 0.5837734937667847, + "learning_rate": 1.254e-05, + "loss": 0.0567, + "step": 4183 + }, + { + "epoch": 3.5337837837837838, + "grad_norm": 0.8084859848022461, + "learning_rate": 1.2543000000000002e-05, + "loss": 0.0726, + "step": 4184 + }, + { + "epoch": 3.5346283783783785, + "grad_norm": 0.43312305212020874, + "learning_rate": 1.2546000000000002e-05, + "loss": 0.0407, + "step": 4185 + }, + { + "epoch": 3.535472972972973, + "grad_norm": 0.8188486099243164, + "learning_rate": 1.2549000000000001e-05, + "loss": 0.0889, + "step": 4186 + }, + { + "epoch": 3.5363175675675675, + "grad_norm": 1.5989651679992676, + "learning_rate": 1.2552e-05, + "loss": 0.1203, + "step": 4187 + }, + { + "epoch": 3.5371621621621623, + "grad_norm": 0.7412142157554626, + "learning_rate": 1.2555e-05, + "loss": 0.0716, + "step": 4188 + }, + { + "epoch": 3.538006756756757, + "grad_norm": 0.7444847822189331, + "learning_rate": 1.2558e-05, + "loss": 0.0683, + "step": 4189 + }, + { + "epoch": 3.5388513513513513, + "grad_norm": 0.7937759160995483, + "learning_rate": 1.2561e-05, + "loss": 0.0662, + "step": 4190 + }, + { + "epoch": 3.539695945945946, + "grad_norm": 0.70552659034729, + "learning_rate": 1.2564e-05, + "loss": 0.0446, + "step": 4191 + }, + { + "epoch": 3.5405405405405403, + "grad_norm": 0.6090805530548096, + "learning_rate": 1.2567e-05, + "loss": 0.0689, + "step": 4192 + }, + { + "epoch": 3.541385135135135, + "grad_norm": 1.3529006242752075, + "learning_rate": 1.257e-05, + "loss": 0.1155, + "step": 4193 + }, + { + "epoch": 3.54222972972973, + "grad_norm": 0.8867320418357849, + "learning_rate": 1.2573e-05, + "loss": 0.0687, + "step": 4194 + }, + { + "epoch": 3.5430743243243246, + "grad_norm": 0.6109957098960876, + "learning_rate": 1.2576000000000001e-05, + "loss": 0.0649, + "step": 4195 + }, + { + "epoch": 3.543918918918919, + "grad_norm": 1.0778135061264038, + "learning_rate": 1.2579000000000001e-05, + "loss": 0.0941, + "step": 4196 + }, + { + "epoch": 3.5447635135135136, + "grad_norm": 0.715121865272522, + "learning_rate": 1.2582e-05, + "loss": 0.0784, + "step": 4197 + }, + { + "epoch": 3.545608108108108, + "grad_norm": 1.1501975059509277, + "learning_rate": 1.2585e-05, + "loss": 0.1221, + "step": 4198 + }, + { + "epoch": 3.5464527027027026, + "grad_norm": 1.2413458824157715, + "learning_rate": 1.2587999999999999e-05, + "loss": 0.0815, + "step": 4199 + }, + { + "epoch": 3.5472972972972974, + "grad_norm": 1.26583731174469, + "learning_rate": 1.2591e-05, + "loss": 0.1397, + "step": 4200 + }, + { + "epoch": 3.548141891891892, + "grad_norm": 1.4141852855682373, + "learning_rate": 1.2594e-05, + "loss": 0.1351, + "step": 4201 + }, + { + "epoch": 3.5489864864864864, + "grad_norm": 1.3965345621109009, + "learning_rate": 1.2597e-05, + "loss": 0.1308, + "step": 4202 + }, + { + "epoch": 3.549831081081081, + "grad_norm": 1.2049617767333984, + "learning_rate": 1.26e-05, + "loss": 0.4144, + "step": 4203 + }, + { + "epoch": 3.5506756756756754, + "grad_norm": 1.1272478103637695, + "learning_rate": 1.2603e-05, + "loss": 0.2679, + "step": 4204 + }, + { + "epoch": 3.55152027027027, + "grad_norm": 0.697226881980896, + "learning_rate": 1.2606000000000001e-05, + "loss": 0.1313, + "step": 4205 + }, + { + "epoch": 3.552364864864865, + "grad_norm": 0.8271681070327759, + "learning_rate": 1.2609e-05, + "loss": 0.1161, + "step": 4206 + }, + { + "epoch": 3.5532094594594597, + "grad_norm": 0.644217848777771, + "learning_rate": 1.2612e-05, + "loss": 0.0744, + "step": 4207 + }, + { + "epoch": 3.554054054054054, + "grad_norm": 0.6571239233016968, + "learning_rate": 1.2615e-05, + "loss": 0.0915, + "step": 4208 + }, + { + "epoch": 3.5548986486486487, + "grad_norm": 0.5156275629997253, + "learning_rate": 1.2618e-05, + "loss": 0.0811, + "step": 4209 + }, + { + "epoch": 3.555743243243243, + "grad_norm": 0.5217009782791138, + "learning_rate": 1.2621000000000002e-05, + "loss": 0.0629, + "step": 4210 + }, + { + "epoch": 3.5565878378378377, + "grad_norm": 0.5058425664901733, + "learning_rate": 1.2624e-05, + "loss": 0.0515, + "step": 4211 + }, + { + "epoch": 3.5574324324324325, + "grad_norm": 0.6139473915100098, + "learning_rate": 1.2627e-05, + "loss": 0.0907, + "step": 4212 + }, + { + "epoch": 3.558277027027027, + "grad_norm": 0.5460073947906494, + "learning_rate": 1.263e-05, + "loss": 0.0518, + "step": 4213 + }, + { + "epoch": 3.5591216216216215, + "grad_norm": 0.5547436475753784, + "learning_rate": 1.2633e-05, + "loss": 0.0656, + "step": 4214 + }, + { + "epoch": 3.5599662162162162, + "grad_norm": 0.649427592754364, + "learning_rate": 1.2636e-05, + "loss": 0.0593, + "step": 4215 + }, + { + "epoch": 3.560810810810811, + "grad_norm": 0.7396875023841858, + "learning_rate": 1.2639e-05, + "loss": 0.0905, + "step": 4216 + }, + { + "epoch": 3.5616554054054053, + "grad_norm": 0.9999665021896362, + "learning_rate": 1.2642e-05, + "loss": 0.0609, + "step": 4217 + }, + { + "epoch": 3.5625, + "grad_norm": 0.8858944773674011, + "learning_rate": 1.2645e-05, + "loss": 0.07, + "step": 4218 + }, + { + "epoch": 3.5633445945945947, + "grad_norm": 0.6326929330825806, + "learning_rate": 1.2648e-05, + "loss": 0.0685, + "step": 4219 + }, + { + "epoch": 3.564189189189189, + "grad_norm": 0.6866007447242737, + "learning_rate": 1.2651000000000001e-05, + "loss": 0.06, + "step": 4220 + }, + { + "epoch": 3.5650337837837838, + "grad_norm": 0.6123542785644531, + "learning_rate": 1.2654000000000001e-05, + "loss": 0.0525, + "step": 4221 + }, + { + "epoch": 3.5658783783783785, + "grad_norm": 0.7190589308738708, + "learning_rate": 1.2657000000000001e-05, + "loss": 0.0642, + "step": 4222 + }, + { + "epoch": 3.566722972972973, + "grad_norm": 0.5598136782646179, + "learning_rate": 1.2659999999999999e-05, + "loss": 0.0633, + "step": 4223 + }, + { + "epoch": 3.5675675675675675, + "grad_norm": 0.6840665936470032, + "learning_rate": 1.2662999999999999e-05, + "loss": 0.0671, + "step": 4224 + }, + { + "epoch": 3.5684121621621623, + "grad_norm": 0.4921876788139343, + "learning_rate": 1.2666e-05, + "loss": 0.0571, + "step": 4225 + }, + { + "epoch": 3.569256756756757, + "grad_norm": 0.6722495555877686, + "learning_rate": 1.2669e-05, + "loss": 0.0684, + "step": 4226 + }, + { + "epoch": 3.5701013513513513, + "grad_norm": 0.5780634880065918, + "learning_rate": 1.2672e-05, + "loss": 0.0616, + "step": 4227 + }, + { + "epoch": 3.570945945945946, + "grad_norm": 0.7230973243713379, + "learning_rate": 1.2675e-05, + "loss": 0.0717, + "step": 4228 + }, + { + "epoch": 3.5717905405405403, + "grad_norm": 0.5525695085525513, + "learning_rate": 1.2678e-05, + "loss": 0.0603, + "step": 4229 + }, + { + "epoch": 3.572635135135135, + "grad_norm": 0.5807981491088867, + "learning_rate": 1.2681000000000001e-05, + "loss": 0.044, + "step": 4230 + }, + { + "epoch": 3.57347972972973, + "grad_norm": 0.6807660460472107, + "learning_rate": 1.2684000000000001e-05, + "loss": 0.0628, + "step": 4231 + }, + { + "epoch": 3.5743243243243246, + "grad_norm": 0.6794142723083496, + "learning_rate": 1.2687e-05, + "loss": 0.0637, + "step": 4232 + }, + { + "epoch": 3.575168918918919, + "grad_norm": 0.6101224422454834, + "learning_rate": 1.269e-05, + "loss": 0.0701, + "step": 4233 + }, + { + "epoch": 3.5760135135135136, + "grad_norm": 0.6392484903335571, + "learning_rate": 1.2693e-05, + "loss": 0.0804, + "step": 4234 + }, + { + "epoch": 3.576858108108108, + "grad_norm": 0.8762181997299194, + "learning_rate": 1.2696000000000002e-05, + "loss": 0.0888, + "step": 4235 + }, + { + "epoch": 3.5777027027027026, + "grad_norm": 0.9101364016532898, + "learning_rate": 1.2699e-05, + "loss": 0.0579, + "step": 4236 + }, + { + "epoch": 3.5785472972972974, + "grad_norm": 0.7903633117675781, + "learning_rate": 1.2702e-05, + "loss": 0.062, + "step": 4237 + }, + { + "epoch": 3.579391891891892, + "grad_norm": 0.7026593685150146, + "learning_rate": 1.2705e-05, + "loss": 0.0558, + "step": 4238 + }, + { + "epoch": 3.5802364864864864, + "grad_norm": 0.6675370931625366, + "learning_rate": 1.2708e-05, + "loss": 0.0553, + "step": 4239 + }, + { + "epoch": 3.581081081081081, + "grad_norm": 0.8722484707832336, + "learning_rate": 1.2711e-05, + "loss": 0.0862, + "step": 4240 + }, + { + "epoch": 3.5819256756756754, + "grad_norm": 0.7811734080314636, + "learning_rate": 1.2714e-05, + "loss": 0.0943, + "step": 4241 + }, + { + "epoch": 3.58277027027027, + "grad_norm": 0.8651053309440613, + "learning_rate": 1.2717e-05, + "loss": 0.0858, + "step": 4242 + }, + { + "epoch": 3.583614864864865, + "grad_norm": 0.828088641166687, + "learning_rate": 1.272e-05, + "loss": 0.0725, + "step": 4243 + }, + { + "epoch": 3.5844594594594597, + "grad_norm": 0.6206476092338562, + "learning_rate": 1.2723e-05, + "loss": 0.0559, + "step": 4244 + }, + { + "epoch": 3.585304054054054, + "grad_norm": 0.732593834400177, + "learning_rate": 1.2726000000000001e-05, + "loss": 0.0812, + "step": 4245 + }, + { + "epoch": 3.5861486486486487, + "grad_norm": 0.8152467608451843, + "learning_rate": 1.2729000000000001e-05, + "loss": 0.0817, + "step": 4246 + }, + { + "epoch": 3.586993243243243, + "grad_norm": 1.0163419246673584, + "learning_rate": 1.2732000000000001e-05, + "loss": 0.0935, + "step": 4247 + }, + { + "epoch": 3.5878378378378377, + "grad_norm": 0.813235878944397, + "learning_rate": 1.2735e-05, + "loss": 0.1062, + "step": 4248 + }, + { + "epoch": 3.5886824324324325, + "grad_norm": 0.7851195335388184, + "learning_rate": 1.2737999999999999e-05, + "loss": 0.0753, + "step": 4249 + }, + { + "epoch": 3.589527027027027, + "grad_norm": 0.8691843152046204, + "learning_rate": 1.2741e-05, + "loss": 0.0766, + "step": 4250 + }, + { + "epoch": 3.5903716216216215, + "grad_norm": 1.2395846843719482, + "learning_rate": 1.2744e-05, + "loss": 0.1145, + "step": 4251 + }, + { + "epoch": 3.5912162162162162, + "grad_norm": 1.8590368032455444, + "learning_rate": 1.2747e-05, + "loss": 0.2575, + "step": 4252 + }, + { + "epoch": 3.592060810810811, + "grad_norm": 3.2572574615478516, + "learning_rate": 1.275e-05, + "loss": 0.4526, + "step": 4253 + }, + { + "epoch": 3.5929054054054053, + "grad_norm": 1.1184954643249512, + "learning_rate": 1.2753e-05, + "loss": 0.3102, + "step": 4254 + }, + { + "epoch": 3.59375, + "grad_norm": 0.7752662897109985, + "learning_rate": 1.2756000000000001e-05, + "loss": 0.1601, + "step": 4255 + }, + { + "epoch": 3.5945945945945947, + "grad_norm": 0.7107540965080261, + "learning_rate": 1.2759000000000001e-05, + "loss": 0.1075, + "step": 4256 + }, + { + "epoch": 3.595439189189189, + "grad_norm": 0.6793136596679688, + "learning_rate": 1.2762e-05, + "loss": 0.1072, + "step": 4257 + }, + { + "epoch": 3.5962837837837838, + "grad_norm": 0.492581844329834, + "learning_rate": 1.2765e-05, + "loss": 0.0819, + "step": 4258 + }, + { + "epoch": 3.5971283783783785, + "grad_norm": 0.47294485569000244, + "learning_rate": 1.2768e-05, + "loss": 0.0598, + "step": 4259 + }, + { + "epoch": 3.597972972972973, + "grad_norm": 0.625684380531311, + "learning_rate": 1.2771e-05, + "loss": 0.0595, + "step": 4260 + }, + { + "epoch": 3.5988175675675675, + "grad_norm": 0.5889713168144226, + "learning_rate": 1.2774e-05, + "loss": 0.0807, + "step": 4261 + }, + { + "epoch": 3.5996621621621623, + "grad_norm": 0.4488428235054016, + "learning_rate": 1.2777e-05, + "loss": 0.0714, + "step": 4262 + }, + { + "epoch": 3.600506756756757, + "grad_norm": 0.5318130254745483, + "learning_rate": 1.278e-05, + "loss": 0.0677, + "step": 4263 + }, + { + "epoch": 3.6013513513513513, + "grad_norm": 0.651491105556488, + "learning_rate": 1.2783e-05, + "loss": 0.0939, + "step": 4264 + }, + { + "epoch": 3.602195945945946, + "grad_norm": 0.7882003784179688, + "learning_rate": 1.2786000000000001e-05, + "loss": 0.0827, + "step": 4265 + }, + { + "epoch": 3.6030405405405403, + "grad_norm": 0.5249614715576172, + "learning_rate": 1.2789e-05, + "loss": 0.0673, + "step": 4266 + }, + { + "epoch": 3.603885135135135, + "grad_norm": 0.5094960331916809, + "learning_rate": 1.2792e-05, + "loss": 0.0567, + "step": 4267 + }, + { + "epoch": 3.60472972972973, + "grad_norm": 0.6038164496421814, + "learning_rate": 1.2795e-05, + "loss": 0.0804, + "step": 4268 + }, + { + "epoch": 3.6055743243243246, + "grad_norm": 0.6036378741264343, + "learning_rate": 1.2798e-05, + "loss": 0.0635, + "step": 4269 + }, + { + "epoch": 3.606418918918919, + "grad_norm": 0.6052785515785217, + "learning_rate": 1.2801000000000002e-05, + "loss": 0.0563, + "step": 4270 + }, + { + "epoch": 3.6072635135135136, + "grad_norm": 0.5918703079223633, + "learning_rate": 1.2804000000000001e-05, + "loss": 0.0776, + "step": 4271 + }, + { + "epoch": 3.608108108108108, + "grad_norm": 0.6624329090118408, + "learning_rate": 1.2807000000000001e-05, + "loss": 0.0581, + "step": 4272 + }, + { + "epoch": 3.6089527027027026, + "grad_norm": 0.48747044801712036, + "learning_rate": 1.281e-05, + "loss": 0.0478, + "step": 4273 + }, + { + "epoch": 3.6097972972972974, + "grad_norm": 0.6875655651092529, + "learning_rate": 1.2812999999999999e-05, + "loss": 0.074, + "step": 4274 + }, + { + "epoch": 3.610641891891892, + "grad_norm": 0.5757899880409241, + "learning_rate": 1.2816e-05, + "loss": 0.0672, + "step": 4275 + }, + { + "epoch": 3.6114864864864864, + "grad_norm": 0.47954994440078735, + "learning_rate": 1.2819e-05, + "loss": 0.0578, + "step": 4276 + }, + { + "epoch": 3.612331081081081, + "grad_norm": 0.6608002185821533, + "learning_rate": 1.2822e-05, + "loss": 0.0673, + "step": 4277 + }, + { + "epoch": 3.6131756756756754, + "grad_norm": 0.6318681240081787, + "learning_rate": 1.2825e-05, + "loss": 0.0672, + "step": 4278 + }, + { + "epoch": 3.61402027027027, + "grad_norm": 0.7673001289367676, + "learning_rate": 1.2828e-05, + "loss": 0.0812, + "step": 4279 + }, + { + "epoch": 3.614864864864865, + "grad_norm": 0.6322424411773682, + "learning_rate": 1.2831000000000001e-05, + "loss": 0.0582, + "step": 4280 + }, + { + "epoch": 3.6157094594594597, + "grad_norm": 0.6126437783241272, + "learning_rate": 1.2834000000000001e-05, + "loss": 0.0709, + "step": 4281 + }, + { + "epoch": 3.616554054054054, + "grad_norm": 0.8626749515533447, + "learning_rate": 1.2837000000000001e-05, + "loss": 0.0669, + "step": 4282 + }, + { + "epoch": 3.6173986486486487, + "grad_norm": 0.6661975383758545, + "learning_rate": 1.284e-05, + "loss": 0.0579, + "step": 4283 + }, + { + "epoch": 3.618243243243243, + "grad_norm": 0.6756268739700317, + "learning_rate": 1.2843e-05, + "loss": 0.0712, + "step": 4284 + }, + { + "epoch": 3.6190878378378377, + "grad_norm": 0.5665481090545654, + "learning_rate": 1.2846e-05, + "loss": 0.0634, + "step": 4285 + }, + { + "epoch": 3.6199324324324325, + "grad_norm": 0.6217775940895081, + "learning_rate": 1.2849e-05, + "loss": 0.0584, + "step": 4286 + }, + { + "epoch": 3.620777027027027, + "grad_norm": 0.7709856033325195, + "learning_rate": 1.2852e-05, + "loss": 0.0739, + "step": 4287 + }, + { + "epoch": 3.6216216216216215, + "grad_norm": 1.099442720413208, + "learning_rate": 1.2855e-05, + "loss": 0.0756, + "step": 4288 + }, + { + "epoch": 3.6224662162162162, + "grad_norm": 0.7725868821144104, + "learning_rate": 1.2858e-05, + "loss": 0.0683, + "step": 4289 + }, + { + "epoch": 3.623310810810811, + "grad_norm": 1.079200029373169, + "learning_rate": 1.2861000000000001e-05, + "loss": 0.0978, + "step": 4290 + }, + { + "epoch": 3.6241554054054053, + "grad_norm": 0.6733262538909912, + "learning_rate": 1.2864e-05, + "loss": 0.0718, + "step": 4291 + }, + { + "epoch": 3.625, + "grad_norm": 0.6306110620498657, + "learning_rate": 1.2867e-05, + "loss": 0.0523, + "step": 4292 + }, + { + "epoch": 3.6258445945945947, + "grad_norm": 0.7164384126663208, + "learning_rate": 1.287e-05, + "loss": 0.0722, + "step": 4293 + }, + { + "epoch": 3.626689189189189, + "grad_norm": 0.8333314061164856, + "learning_rate": 1.2873e-05, + "loss": 0.0712, + "step": 4294 + }, + { + "epoch": 3.6275337837837838, + "grad_norm": 0.9042384028434753, + "learning_rate": 1.2876000000000002e-05, + "loss": 0.0671, + "step": 4295 + }, + { + "epoch": 3.6283783783783785, + "grad_norm": 0.7228586077690125, + "learning_rate": 1.2879000000000002e-05, + "loss": 0.0638, + "step": 4296 + }, + { + "epoch": 3.629222972972973, + "grad_norm": 0.7387025356292725, + "learning_rate": 1.2882e-05, + "loss": 0.0621, + "step": 4297 + }, + { + "epoch": 3.6300675675675675, + "grad_norm": 0.7240515947341919, + "learning_rate": 1.2885e-05, + "loss": 0.0663, + "step": 4298 + }, + { + "epoch": 3.6309121621621623, + "grad_norm": 0.9068356156349182, + "learning_rate": 1.2888e-05, + "loss": 0.0922, + "step": 4299 + }, + { + "epoch": 3.631756756756757, + "grad_norm": 1.078941822052002, + "learning_rate": 1.2891e-05, + "loss": 0.112, + "step": 4300 + }, + { + "epoch": 3.6326013513513513, + "grad_norm": 1.891622543334961, + "learning_rate": 1.2894e-05, + "loss": 0.1214, + "step": 4301 + }, + { + "epoch": 3.633445945945946, + "grad_norm": 1.3556742668151855, + "learning_rate": 1.2897e-05, + "loss": 0.1116, + "step": 4302 + }, + { + "epoch": 3.6342905405405403, + "grad_norm": 2.4117178916931152, + "learning_rate": 1.29e-05, + "loss": 0.5213, + "step": 4303 + }, + { + "epoch": 3.635135135135135, + "grad_norm": 1.1419763565063477, + "learning_rate": 1.2903e-05, + "loss": 0.2456, + "step": 4304 + }, + { + "epoch": 3.63597972972973, + "grad_norm": 0.8766488432884216, + "learning_rate": 1.2906000000000001e-05, + "loss": 0.1477, + "step": 4305 + }, + { + "epoch": 3.6368243243243246, + "grad_norm": 0.6396113038063049, + "learning_rate": 1.2909000000000001e-05, + "loss": 0.0829, + "step": 4306 + }, + { + "epoch": 3.637668918918919, + "grad_norm": 0.580727219581604, + "learning_rate": 1.2912000000000001e-05, + "loss": 0.0989, + "step": 4307 + }, + { + "epoch": 3.6385135135135136, + "grad_norm": 0.522655725479126, + "learning_rate": 1.2915000000000001e-05, + "loss": 0.0647, + "step": 4308 + }, + { + "epoch": 3.639358108108108, + "grad_norm": 0.7009950280189514, + "learning_rate": 1.2917999999999999e-05, + "loss": 0.085, + "step": 4309 + }, + { + "epoch": 3.6402027027027026, + "grad_norm": 0.6010565757751465, + "learning_rate": 1.2921e-05, + "loss": 0.0646, + "step": 4310 + }, + { + "epoch": 3.6410472972972974, + "grad_norm": 0.6940047144889832, + "learning_rate": 1.2924e-05, + "loss": 0.0653, + "step": 4311 + }, + { + "epoch": 3.641891891891892, + "grad_norm": 0.718172013759613, + "learning_rate": 1.2927e-05, + "loss": 0.0655, + "step": 4312 + }, + { + "epoch": 3.6427364864864864, + "grad_norm": 0.6161797642707825, + "learning_rate": 1.293e-05, + "loss": 0.0803, + "step": 4313 + }, + { + "epoch": 3.643581081081081, + "grad_norm": 0.4671526551246643, + "learning_rate": 1.2933e-05, + "loss": 0.0452, + "step": 4314 + }, + { + "epoch": 3.6444256756756754, + "grad_norm": 0.6454151272773743, + "learning_rate": 1.2936000000000001e-05, + "loss": 0.0521, + "step": 4315 + }, + { + "epoch": 3.64527027027027, + "grad_norm": 0.5097796320915222, + "learning_rate": 1.2939000000000001e-05, + "loss": 0.0663, + "step": 4316 + }, + { + "epoch": 3.646114864864865, + "grad_norm": 0.6301630139350891, + "learning_rate": 1.2942e-05, + "loss": 0.0719, + "step": 4317 + }, + { + "epoch": 3.6469594594594597, + "grad_norm": 0.5742582082748413, + "learning_rate": 1.2945e-05, + "loss": 0.0798, + "step": 4318 + }, + { + "epoch": 3.647804054054054, + "grad_norm": 0.5007941126823425, + "learning_rate": 1.2948e-05, + "loss": 0.0437, + "step": 4319 + }, + { + "epoch": 3.6486486486486487, + "grad_norm": 0.613652765750885, + "learning_rate": 1.2951e-05, + "loss": 0.0622, + "step": 4320 + }, + { + "epoch": 3.649493243243243, + "grad_norm": 0.6072162389755249, + "learning_rate": 1.2954000000000002e-05, + "loss": 0.0553, + "step": 4321 + }, + { + "epoch": 3.6503378378378377, + "grad_norm": 0.5117518305778503, + "learning_rate": 1.2957e-05, + "loss": 0.0527, + "step": 4322 + }, + { + "epoch": 3.6511824324324325, + "grad_norm": 0.9214260578155518, + "learning_rate": 1.296e-05, + "loss": 0.0738, + "step": 4323 + }, + { + "epoch": 3.652027027027027, + "grad_norm": 0.43640783429145813, + "learning_rate": 1.2963e-05, + "loss": 0.0505, + "step": 4324 + }, + { + "epoch": 3.6528716216216215, + "grad_norm": 0.6208381652832031, + "learning_rate": 1.2966e-05, + "loss": 0.0565, + "step": 4325 + }, + { + "epoch": 3.6537162162162162, + "grad_norm": 0.5836608409881592, + "learning_rate": 1.2969e-05, + "loss": 0.0591, + "step": 4326 + }, + { + "epoch": 3.654560810810811, + "grad_norm": 0.4908498227596283, + "learning_rate": 1.2972e-05, + "loss": 0.0443, + "step": 4327 + }, + { + "epoch": 3.6554054054054053, + "grad_norm": 0.7367815375328064, + "learning_rate": 1.2975e-05, + "loss": 0.0871, + "step": 4328 + }, + { + "epoch": 3.65625, + "grad_norm": 0.6605685949325562, + "learning_rate": 1.2978e-05, + "loss": 0.0693, + "step": 4329 + }, + { + "epoch": 3.6570945945945947, + "grad_norm": 0.6546284556388855, + "learning_rate": 1.2981e-05, + "loss": 0.067, + "step": 4330 + }, + { + "epoch": 3.657939189189189, + "grad_norm": 0.5647033452987671, + "learning_rate": 1.2984000000000001e-05, + "loss": 0.0524, + "step": 4331 + }, + { + "epoch": 3.6587837837837838, + "grad_norm": 0.8602354526519775, + "learning_rate": 1.2987000000000001e-05, + "loss": 0.0849, + "step": 4332 + }, + { + "epoch": 3.6596283783783785, + "grad_norm": 0.6272715926170349, + "learning_rate": 1.2990000000000001e-05, + "loss": 0.0636, + "step": 4333 + }, + { + "epoch": 3.660472972972973, + "grad_norm": 0.6988127827644348, + "learning_rate": 1.2992999999999999e-05, + "loss": 0.062, + "step": 4334 + }, + { + "epoch": 3.6613175675675675, + "grad_norm": 1.1277846097946167, + "learning_rate": 1.2995999999999999e-05, + "loss": 0.0743, + "step": 4335 + }, + { + "epoch": 3.6621621621621623, + "grad_norm": 0.7625616788864136, + "learning_rate": 1.2999e-05, + "loss": 0.0768, + "step": 4336 + }, + { + "epoch": 3.663006756756757, + "grad_norm": 0.5632727742195129, + "learning_rate": 1.3002e-05, + "loss": 0.0616, + "step": 4337 + }, + { + "epoch": 3.6638513513513513, + "grad_norm": 0.5745845437049866, + "learning_rate": 1.3005e-05, + "loss": 0.0451, + "step": 4338 + }, + { + "epoch": 3.664695945945946, + "grad_norm": 0.5143485069274902, + "learning_rate": 1.3008e-05, + "loss": 0.0587, + "step": 4339 + }, + { + "epoch": 3.6655405405405403, + "grad_norm": 2.1483514308929443, + "learning_rate": 1.3011e-05, + "loss": 0.1821, + "step": 4340 + }, + { + "epoch": 3.666385135135135, + "grad_norm": 0.9208971858024597, + "learning_rate": 1.3014000000000001e-05, + "loss": 0.0591, + "step": 4341 + }, + { + "epoch": 3.66722972972973, + "grad_norm": 0.6284272074699402, + "learning_rate": 1.3017000000000001e-05, + "loss": 0.0599, + "step": 4342 + }, + { + "epoch": 3.6680743243243246, + "grad_norm": 0.713905394077301, + "learning_rate": 1.302e-05, + "loss": 0.0697, + "step": 4343 + }, + { + "epoch": 3.668918918918919, + "grad_norm": 0.7609804272651672, + "learning_rate": 1.3023e-05, + "loss": 0.064, + "step": 4344 + }, + { + "epoch": 3.6697635135135136, + "grad_norm": 0.7663131952285767, + "learning_rate": 1.3026e-05, + "loss": 0.076, + "step": 4345 + }, + { + "epoch": 3.670608108108108, + "grad_norm": 0.7385449409484863, + "learning_rate": 1.3029e-05, + "loss": 0.0771, + "step": 4346 + }, + { + "epoch": 3.6714527027027026, + "grad_norm": 0.8755495548248291, + "learning_rate": 1.3032e-05, + "loss": 0.1062, + "step": 4347 + }, + { + "epoch": 3.6722972972972974, + "grad_norm": 0.8840491771697998, + "learning_rate": 1.3035e-05, + "loss": 0.0915, + "step": 4348 + }, + { + "epoch": 3.673141891891892, + "grad_norm": 1.0501446723937988, + "learning_rate": 1.3038e-05, + "loss": 0.0883, + "step": 4349 + }, + { + "epoch": 3.6739864864864864, + "grad_norm": 0.8247696757316589, + "learning_rate": 1.3041e-05, + "loss": 0.084, + "step": 4350 + }, + { + "epoch": 3.674831081081081, + "grad_norm": 1.5452852249145508, + "learning_rate": 1.3044e-05, + "loss": 0.1098, + "step": 4351 + }, + { + "epoch": 3.6756756756756754, + "grad_norm": 1.730754017829895, + "learning_rate": 1.3047e-05, + "loss": 0.1586, + "step": 4352 + }, + { + "epoch": 3.67652027027027, + "grad_norm": 1.3615034818649292, + "learning_rate": 1.305e-05, + "loss": 0.3868, + "step": 4353 + }, + { + "epoch": 3.677364864864865, + "grad_norm": 0.7913135290145874, + "learning_rate": 1.3053e-05, + "loss": 0.2636, + "step": 4354 + }, + { + "epoch": 3.6782094594594597, + "grad_norm": 1.3400092124938965, + "learning_rate": 1.3056e-05, + "loss": 0.2098, + "step": 4355 + }, + { + "epoch": 3.679054054054054, + "grad_norm": 0.6760542392730713, + "learning_rate": 1.3059000000000002e-05, + "loss": 0.081, + "step": 4356 + }, + { + "epoch": 3.6798986486486487, + "grad_norm": 0.6332755088806152, + "learning_rate": 1.3062000000000001e-05, + "loss": 0.0809, + "step": 4357 + }, + { + "epoch": 3.680743243243243, + "grad_norm": 0.543788731098175, + "learning_rate": 1.3065000000000001e-05, + "loss": 0.0827, + "step": 4358 + }, + { + "epoch": 3.6815878378378377, + "grad_norm": 0.6470718383789062, + "learning_rate": 1.3068e-05, + "loss": 0.0717, + "step": 4359 + }, + { + "epoch": 3.6824324324324325, + "grad_norm": 2.591989755630493, + "learning_rate": 1.3070999999999999e-05, + "loss": 0.0701, + "step": 4360 + }, + { + "epoch": 3.683277027027027, + "grad_norm": 0.6707907915115356, + "learning_rate": 1.3074e-05, + "loss": 0.0663, + "step": 4361 + }, + { + "epoch": 3.6841216216216215, + "grad_norm": 0.7200556993484497, + "learning_rate": 1.3077e-05, + "loss": 0.0942, + "step": 4362 + }, + { + "epoch": 3.6849662162162162, + "grad_norm": 0.5748695135116577, + "learning_rate": 1.308e-05, + "loss": 0.0576, + "step": 4363 + }, + { + "epoch": 3.685810810810811, + "grad_norm": 0.7403507232666016, + "learning_rate": 1.3083e-05, + "loss": 0.0653, + "step": 4364 + }, + { + "epoch": 3.6866554054054053, + "grad_norm": 0.5613417625427246, + "learning_rate": 1.3086e-05, + "loss": 0.0543, + "step": 4365 + }, + { + "epoch": 3.6875, + "grad_norm": 0.6270257234573364, + "learning_rate": 1.3089000000000001e-05, + "loss": 0.0694, + "step": 4366 + }, + { + "epoch": 3.6883445945945947, + "grad_norm": 0.7461497783660889, + "learning_rate": 1.3092000000000001e-05, + "loss": 0.0912, + "step": 4367 + }, + { + "epoch": 3.689189189189189, + "grad_norm": 0.46702495217323303, + "learning_rate": 1.3095e-05, + "loss": 0.0563, + "step": 4368 + }, + { + "epoch": 3.6900337837837838, + "grad_norm": 0.7108515501022339, + "learning_rate": 1.3098e-05, + "loss": 0.1014, + "step": 4369 + }, + { + "epoch": 3.6908783783783785, + "grad_norm": 0.3925575613975525, + "learning_rate": 1.3101e-05, + "loss": 0.0413, + "step": 4370 + }, + { + "epoch": 3.691722972972973, + "grad_norm": 0.7093313932418823, + "learning_rate": 1.3104e-05, + "loss": 0.0662, + "step": 4371 + }, + { + "epoch": 3.6925675675675675, + "grad_norm": 1.8759934902191162, + "learning_rate": 1.3107e-05, + "loss": 0.1009, + "step": 4372 + }, + { + "epoch": 3.6934121621621623, + "grad_norm": 0.884941816329956, + "learning_rate": 1.311e-05, + "loss": 0.1965, + "step": 4373 + }, + { + "epoch": 3.694256756756757, + "grad_norm": 0.7409775257110596, + "learning_rate": 1.3113e-05, + "loss": 0.0778, + "step": 4374 + }, + { + "epoch": 3.6951013513513513, + "grad_norm": 0.7192912101745605, + "learning_rate": 1.3116e-05, + "loss": 0.076, + "step": 4375 + }, + { + "epoch": 3.695945945945946, + "grad_norm": 0.5903024077415466, + "learning_rate": 1.3119000000000001e-05, + "loss": 0.0865, + "step": 4376 + }, + { + "epoch": 3.6967905405405403, + "grad_norm": 0.5983407497406006, + "learning_rate": 1.3122e-05, + "loss": 0.0644, + "step": 4377 + }, + { + "epoch": 3.697635135135135, + "grad_norm": 0.5615967512130737, + "learning_rate": 1.3125e-05, + "loss": 0.0661, + "step": 4378 + }, + { + "epoch": 3.69847972972973, + "grad_norm": 1.0442308187484741, + "learning_rate": 1.3128e-05, + "loss": 0.0769, + "step": 4379 + }, + { + "epoch": 3.6993243243243246, + "grad_norm": 0.631382942199707, + "learning_rate": 1.3131e-05, + "loss": 0.0602, + "step": 4380 + }, + { + "epoch": 3.700168918918919, + "grad_norm": 0.5787543654441833, + "learning_rate": 1.3134000000000002e-05, + "loss": 0.0567, + "step": 4381 + }, + { + "epoch": 3.7010135135135136, + "grad_norm": 0.461303174495697, + "learning_rate": 1.3137000000000001e-05, + "loss": 0.048, + "step": 4382 + }, + { + "epoch": 3.701858108108108, + "grad_norm": 0.6615902781486511, + "learning_rate": 1.314e-05, + "loss": 0.0668, + "step": 4383 + }, + { + "epoch": 3.7027027027027026, + "grad_norm": 1.2878953218460083, + "learning_rate": 1.3143e-05, + "loss": 0.0769, + "step": 4384 + }, + { + "epoch": 3.7035472972972974, + "grad_norm": 0.7566766738891602, + "learning_rate": 1.3146e-05, + "loss": 0.0559, + "step": 4385 + }, + { + "epoch": 3.704391891891892, + "grad_norm": 1.0127397775650024, + "learning_rate": 1.3149e-05, + "loss": 0.0705, + "step": 4386 + }, + { + "epoch": 3.7052364864864864, + "grad_norm": 0.7674389481544495, + "learning_rate": 1.3152e-05, + "loss": 0.0587, + "step": 4387 + }, + { + "epoch": 3.706081081081081, + "grad_norm": 0.6819568872451782, + "learning_rate": 1.3155e-05, + "loss": 0.0926, + "step": 4388 + }, + { + "epoch": 3.7069256756756754, + "grad_norm": 0.7436239123344421, + "learning_rate": 1.3158e-05, + "loss": 0.065, + "step": 4389 + }, + { + "epoch": 3.70777027027027, + "grad_norm": 0.7562564611434937, + "learning_rate": 1.3161e-05, + "loss": 0.078, + "step": 4390 + }, + { + "epoch": 3.708614864864865, + "grad_norm": 0.8192824721336365, + "learning_rate": 1.3164000000000001e-05, + "loss": 0.1063, + "step": 4391 + }, + { + "epoch": 3.7094594594594597, + "grad_norm": 0.8452253341674805, + "learning_rate": 1.3167000000000001e-05, + "loss": 0.0795, + "step": 4392 + }, + { + "epoch": 3.710304054054054, + "grad_norm": 0.7303003668785095, + "learning_rate": 1.3170000000000001e-05, + "loss": 0.0843, + "step": 4393 + }, + { + "epoch": 3.7111486486486487, + "grad_norm": 0.9640312790870667, + "learning_rate": 1.3173e-05, + "loss": 0.0784, + "step": 4394 + }, + { + "epoch": 3.711993243243243, + "grad_norm": 0.6513113379478455, + "learning_rate": 1.3175999999999999e-05, + "loss": 0.0951, + "step": 4395 + }, + { + "epoch": 3.7128378378378377, + "grad_norm": 0.7431321144104004, + "learning_rate": 1.3179e-05, + "loss": 0.0844, + "step": 4396 + }, + { + "epoch": 3.7136824324324325, + "grad_norm": 0.5613250732421875, + "learning_rate": 1.3182e-05, + "loss": 0.0776, + "step": 4397 + }, + { + "epoch": 3.714527027027027, + "grad_norm": 0.724529504776001, + "learning_rate": 1.3185e-05, + "loss": 0.0841, + "step": 4398 + }, + { + "epoch": 3.7153716216216215, + "grad_norm": 0.7589484453201294, + "learning_rate": 1.3188e-05, + "loss": 0.0696, + "step": 4399 + }, + { + "epoch": 3.7162162162162162, + "grad_norm": 1.053653359413147, + "learning_rate": 1.3191e-05, + "loss": 0.1088, + "step": 4400 + }, + { + "epoch": 3.717060810810811, + "grad_norm": 1.4887278079986572, + "learning_rate": 1.3194000000000001e-05, + "loss": 0.1181, + "step": 4401 + }, + { + "epoch": 3.7179054054054053, + "grad_norm": 1.4127178192138672, + "learning_rate": 1.3197000000000001e-05, + "loss": 0.127, + "step": 4402 + }, + { + "epoch": 3.71875, + "grad_norm": 1.3693395853042603, + "learning_rate": 1.32e-05, + "loss": 0.4537, + "step": 4403 + }, + { + "epoch": 3.7195945945945947, + "grad_norm": 0.919558584690094, + "learning_rate": 1.3203e-05, + "loss": 0.2439, + "step": 4404 + }, + { + "epoch": 3.720439189189189, + "grad_norm": 0.7475987076759338, + "learning_rate": 1.3206e-05, + "loss": 0.1461, + "step": 4405 + }, + { + "epoch": 3.7212837837837838, + "grad_norm": 0.8330742716789246, + "learning_rate": 1.3209000000000002e-05, + "loss": 0.1208, + "step": 4406 + }, + { + "epoch": 3.7221283783783785, + "grad_norm": 0.6560646891593933, + "learning_rate": 1.3212000000000002e-05, + "loss": 0.0911, + "step": 4407 + }, + { + "epoch": 3.722972972972973, + "grad_norm": 0.5482650995254517, + "learning_rate": 1.3215e-05, + "loss": 0.063, + "step": 4408 + }, + { + "epoch": 3.7238175675675675, + "grad_norm": 0.4863010346889496, + "learning_rate": 1.3218e-05, + "loss": 0.0479, + "step": 4409 + }, + { + "epoch": 3.7246621621621623, + "grad_norm": 0.5560134649276733, + "learning_rate": 1.3221e-05, + "loss": 0.0623, + "step": 4410 + }, + { + "epoch": 3.725506756756757, + "grad_norm": 0.6217612028121948, + "learning_rate": 1.3224e-05, + "loss": 0.0754, + "step": 4411 + }, + { + "epoch": 3.7263513513513513, + "grad_norm": 0.5285851359367371, + "learning_rate": 1.3227e-05, + "loss": 0.0655, + "step": 4412 + }, + { + "epoch": 3.727195945945946, + "grad_norm": 0.5393053293228149, + "learning_rate": 1.323e-05, + "loss": 0.0735, + "step": 4413 + }, + { + "epoch": 3.7280405405405403, + "grad_norm": 0.5874719619750977, + "learning_rate": 1.3233e-05, + "loss": 0.0693, + "step": 4414 + }, + { + "epoch": 3.728885135135135, + "grad_norm": 0.7691531777381897, + "learning_rate": 1.3236e-05, + "loss": 0.0909, + "step": 4415 + }, + { + "epoch": 3.72972972972973, + "grad_norm": 0.6784599423408508, + "learning_rate": 1.3239000000000001e-05, + "loss": 0.086, + "step": 4416 + }, + { + "epoch": 3.7305743243243246, + "grad_norm": 0.4960681200027466, + "learning_rate": 1.3242000000000001e-05, + "loss": 0.0427, + "step": 4417 + }, + { + "epoch": 3.731418918918919, + "grad_norm": 0.505656361579895, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0485, + "step": 4418 + }, + { + "epoch": 3.7322635135135136, + "grad_norm": 0.6606845259666443, + "learning_rate": 1.3248000000000001e-05, + "loss": 0.0797, + "step": 4419 + }, + { + "epoch": 3.733108108108108, + "grad_norm": 0.5157758593559265, + "learning_rate": 1.3250999999999999e-05, + "loss": 0.05, + "step": 4420 + }, + { + "epoch": 3.7339527027027026, + "grad_norm": 0.6626462936401367, + "learning_rate": 1.3254e-05, + "loss": 0.0739, + "step": 4421 + }, + { + "epoch": 3.7347972972972974, + "grad_norm": 0.4822562038898468, + "learning_rate": 1.3257e-05, + "loss": 0.0447, + "step": 4422 + }, + { + "epoch": 3.735641891891892, + "grad_norm": 0.6129989624023438, + "learning_rate": 1.326e-05, + "loss": 0.0576, + "step": 4423 + }, + { + "epoch": 3.7364864864864864, + "grad_norm": 0.5348591804504395, + "learning_rate": 1.3263e-05, + "loss": 0.0595, + "step": 4424 + }, + { + "epoch": 3.737331081081081, + "grad_norm": 1.0880248546600342, + "learning_rate": 1.3266e-05, + "loss": 0.0567, + "step": 4425 + }, + { + "epoch": 3.7381756756756754, + "grad_norm": 0.5882156491279602, + "learning_rate": 1.3269000000000001e-05, + "loss": 0.0563, + "step": 4426 + }, + { + "epoch": 3.73902027027027, + "grad_norm": 0.5287323594093323, + "learning_rate": 1.3272000000000001e-05, + "loss": 0.0457, + "step": 4427 + }, + { + "epoch": 3.739864864864865, + "grad_norm": 0.847838819026947, + "learning_rate": 1.3275e-05, + "loss": 0.0928, + "step": 4428 + }, + { + "epoch": 3.7407094594594597, + "grad_norm": 0.6048473715782166, + "learning_rate": 1.3278e-05, + "loss": 0.0649, + "step": 4429 + }, + { + "epoch": 3.741554054054054, + "grad_norm": 0.782224178314209, + "learning_rate": 1.3281e-05, + "loss": 0.071, + "step": 4430 + }, + { + "epoch": 3.7423986486486487, + "grad_norm": 0.50554358959198, + "learning_rate": 1.3284000000000002e-05, + "loss": 0.0687, + "step": 4431 + }, + { + "epoch": 3.743243243243243, + "grad_norm": 0.8395126461982727, + "learning_rate": 1.3287e-05, + "loss": 0.0728, + "step": 4432 + }, + { + "epoch": 3.7440878378378377, + "grad_norm": 0.6896973252296448, + "learning_rate": 1.329e-05, + "loss": 0.0726, + "step": 4433 + }, + { + "epoch": 3.7449324324324325, + "grad_norm": 0.8726404905319214, + "learning_rate": 1.3293e-05, + "loss": 0.0893, + "step": 4434 + }, + { + "epoch": 3.745777027027027, + "grad_norm": 0.674228310585022, + "learning_rate": 1.3296e-05, + "loss": 0.0678, + "step": 4435 + }, + { + "epoch": 3.7466216216216215, + "grad_norm": 0.4686411917209625, + "learning_rate": 1.3299000000000001e-05, + "loss": 0.042, + "step": 4436 + }, + { + "epoch": 3.7474662162162162, + "grad_norm": 0.4663027822971344, + "learning_rate": 1.3302e-05, + "loss": 0.0523, + "step": 4437 + }, + { + "epoch": 3.748310810810811, + "grad_norm": 0.6230629086494446, + "learning_rate": 1.3305e-05, + "loss": 0.0685, + "step": 4438 + }, + { + "epoch": 3.7491554054054053, + "grad_norm": 0.5711079835891724, + "learning_rate": 1.3308e-05, + "loss": 0.0704, + "step": 4439 + }, + { + "epoch": 3.75, + "grad_norm": 0.9799067974090576, + "learning_rate": 1.3311e-05, + "loss": 0.089, + "step": 4440 + }, + { + "epoch": 3.7508445945945947, + "grad_norm": 0.5920965075492859, + "learning_rate": 1.3314e-05, + "loss": 0.057, + "step": 4441 + }, + { + "epoch": 3.751689189189189, + "grad_norm": 1.6515170335769653, + "learning_rate": 1.3317000000000001e-05, + "loss": 0.1052, + "step": 4442 + }, + { + "epoch": 3.7525337837837838, + "grad_norm": 0.5442373752593994, + "learning_rate": 1.3320000000000001e-05, + "loss": 0.0473, + "step": 4443 + }, + { + "epoch": 3.7533783783783785, + "grad_norm": 1.1958943605422974, + "learning_rate": 1.3323000000000001e-05, + "loss": 0.0691, + "step": 4444 + }, + { + "epoch": 3.754222972972973, + "grad_norm": 0.9085906744003296, + "learning_rate": 1.3325999999999999e-05, + "loss": 0.0811, + "step": 4445 + }, + { + "epoch": 3.7550675675675675, + "grad_norm": 0.8286060690879822, + "learning_rate": 1.3328999999999999e-05, + "loss": 0.0754, + "step": 4446 + }, + { + "epoch": 3.7559121621621623, + "grad_norm": 0.829378604888916, + "learning_rate": 1.3332e-05, + "loss": 0.0632, + "step": 4447 + }, + { + "epoch": 3.756756756756757, + "grad_norm": 1.1104214191436768, + "learning_rate": 1.3335e-05, + "loss": 0.1047, + "step": 4448 + }, + { + "epoch": 3.7576013513513513, + "grad_norm": 1.0992392301559448, + "learning_rate": 1.3338e-05, + "loss": 0.0934, + "step": 4449 + }, + { + "epoch": 3.758445945945946, + "grad_norm": 1.2259654998779297, + "learning_rate": 1.3341e-05, + "loss": 0.1217, + "step": 4450 + }, + { + "epoch": 3.7592905405405403, + "grad_norm": 1.2061785459518433, + "learning_rate": 1.3344e-05, + "loss": 0.0829, + "step": 4451 + }, + { + "epoch": 3.760135135135135, + "grad_norm": 2.9617972373962402, + "learning_rate": 1.3347000000000001e-05, + "loss": 0.1353, + "step": 4452 + }, + { + "epoch": 3.76097972972973, + "grad_norm": 2.5007293224334717, + "learning_rate": 1.3350000000000001e-05, + "loss": 0.4203, + "step": 4453 + }, + { + "epoch": 3.7618243243243246, + "grad_norm": 1.5905303955078125, + "learning_rate": 1.3353e-05, + "loss": 0.3294, + "step": 4454 + }, + { + "epoch": 3.762668918918919, + "grad_norm": 0.9540732502937317, + "learning_rate": 1.3356e-05, + "loss": 0.1653, + "step": 4455 + }, + { + "epoch": 3.7635135135135136, + "grad_norm": 0.7590221166610718, + "learning_rate": 1.3359e-05, + "loss": 0.1084, + "step": 4456 + }, + { + "epoch": 3.764358108108108, + "grad_norm": 0.47845515608787537, + "learning_rate": 1.3362e-05, + "loss": 0.0809, + "step": 4457 + }, + { + "epoch": 3.7652027027027026, + "grad_norm": 0.7109180688858032, + "learning_rate": 1.3365e-05, + "loss": 0.076, + "step": 4458 + }, + { + "epoch": 3.7660472972972974, + "grad_norm": 0.4651121497154236, + "learning_rate": 1.3368e-05, + "loss": 0.0563, + "step": 4459 + }, + { + "epoch": 3.766891891891892, + "grad_norm": 0.6596838235855103, + "learning_rate": 1.3371e-05, + "loss": 0.1052, + "step": 4460 + }, + { + "epoch": 3.7677364864864864, + "grad_norm": 0.42002230882644653, + "learning_rate": 1.3374e-05, + "loss": 0.0512, + "step": 4461 + }, + { + "epoch": 3.768581081081081, + "grad_norm": 0.6946735978126526, + "learning_rate": 1.3377e-05, + "loss": 0.0766, + "step": 4462 + }, + { + "epoch": 3.7694256756756754, + "grad_norm": 0.5518532395362854, + "learning_rate": 1.338e-05, + "loss": 0.0539, + "step": 4463 + }, + { + "epoch": 3.77027027027027, + "grad_norm": 0.8161900639533997, + "learning_rate": 1.3383e-05, + "loss": 0.0751, + "step": 4464 + }, + { + "epoch": 3.771114864864865, + "grad_norm": 0.45778846740722656, + "learning_rate": 1.3386e-05, + "loss": 0.0546, + "step": 4465 + }, + { + "epoch": 3.7719594594594597, + "grad_norm": 0.5143083333969116, + "learning_rate": 1.3389e-05, + "loss": 0.0556, + "step": 4466 + }, + { + "epoch": 3.772804054054054, + "grad_norm": 0.5575762391090393, + "learning_rate": 1.3392000000000002e-05, + "loss": 0.0626, + "step": 4467 + }, + { + "epoch": 3.7736486486486487, + "grad_norm": 0.521317183971405, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0655, + "step": 4468 + }, + { + "epoch": 3.774493243243243, + "grad_norm": 0.6761794090270996, + "learning_rate": 1.3398e-05, + "loss": 0.0482, + "step": 4469 + }, + { + "epoch": 3.7753378378378377, + "grad_norm": 0.6191100478172302, + "learning_rate": 1.3401e-05, + "loss": 0.0761, + "step": 4470 + }, + { + "epoch": 3.7761824324324325, + "grad_norm": 0.8260782361030579, + "learning_rate": 1.3403999999999999e-05, + "loss": 0.063, + "step": 4471 + }, + { + "epoch": 3.777027027027027, + "grad_norm": 0.46141088008880615, + "learning_rate": 1.3407e-05, + "loss": 0.0433, + "step": 4472 + }, + { + "epoch": 3.7778716216216215, + "grad_norm": 0.593956470489502, + "learning_rate": 1.341e-05, + "loss": 0.074, + "step": 4473 + }, + { + "epoch": 3.7787162162162162, + "grad_norm": 0.5106512904167175, + "learning_rate": 1.3413e-05, + "loss": 0.0573, + "step": 4474 + }, + { + "epoch": 3.779560810810811, + "grad_norm": 0.5885075926780701, + "learning_rate": 1.3416e-05, + "loss": 0.0581, + "step": 4475 + }, + { + "epoch": 3.7804054054054053, + "grad_norm": 0.6294794082641602, + "learning_rate": 1.3419e-05, + "loss": 0.0695, + "step": 4476 + }, + { + "epoch": 3.78125, + "grad_norm": 0.5750512480735779, + "learning_rate": 1.3422000000000001e-05, + "loss": 0.0561, + "step": 4477 + }, + { + "epoch": 3.7820945945945947, + "grad_norm": 0.6767889261245728, + "learning_rate": 1.3425000000000001e-05, + "loss": 0.0664, + "step": 4478 + }, + { + "epoch": 3.782939189189189, + "grad_norm": 0.7107561230659485, + "learning_rate": 1.3428000000000001e-05, + "loss": 0.0808, + "step": 4479 + }, + { + "epoch": 3.7837837837837838, + "grad_norm": 0.5781220197677612, + "learning_rate": 1.3431e-05, + "loss": 0.0452, + "step": 4480 + }, + { + "epoch": 3.7846283783783785, + "grad_norm": 0.7309671640396118, + "learning_rate": 1.3433999999999999e-05, + "loss": 0.0861, + "step": 4481 + }, + { + "epoch": 3.785472972972973, + "grad_norm": 0.6771832704544067, + "learning_rate": 1.3437e-05, + "loss": 0.074, + "step": 4482 + }, + { + "epoch": 3.7863175675675675, + "grad_norm": 0.5853342413902283, + "learning_rate": 1.344e-05, + "loss": 0.0644, + "step": 4483 + }, + { + "epoch": 3.7871621621621623, + "grad_norm": 0.7525617480278015, + "learning_rate": 1.3443e-05, + "loss": 0.0786, + "step": 4484 + }, + { + "epoch": 3.788006756756757, + "grad_norm": 0.8899219632148743, + "learning_rate": 1.3446e-05, + "loss": 0.0816, + "step": 4485 + }, + { + "epoch": 3.7888513513513513, + "grad_norm": 0.7791750431060791, + "learning_rate": 1.3449e-05, + "loss": 0.0716, + "step": 4486 + }, + { + "epoch": 3.789695945945946, + "grad_norm": 0.44735822081565857, + "learning_rate": 1.3452000000000001e-05, + "loss": 0.0339, + "step": 4487 + }, + { + "epoch": 3.7905405405405403, + "grad_norm": 0.7500439286231995, + "learning_rate": 1.3455e-05, + "loss": 0.085, + "step": 4488 + }, + { + "epoch": 3.791385135135135, + "grad_norm": 0.4834289848804474, + "learning_rate": 1.3458e-05, + "loss": 0.0423, + "step": 4489 + }, + { + "epoch": 3.79222972972973, + "grad_norm": 0.819252610206604, + "learning_rate": 1.3461e-05, + "loss": 0.0754, + "step": 4490 + }, + { + "epoch": 3.7930743243243246, + "grad_norm": 0.6881804466247559, + "learning_rate": 1.3464e-05, + "loss": 0.0718, + "step": 4491 + }, + { + "epoch": 3.793918918918919, + "grad_norm": 0.6739464998245239, + "learning_rate": 1.3467000000000002e-05, + "loss": 0.0807, + "step": 4492 + }, + { + "epoch": 3.7947635135135136, + "grad_norm": 0.8705262541770935, + "learning_rate": 1.3470000000000001e-05, + "loss": 0.0992, + "step": 4493 + }, + { + "epoch": 3.795608108108108, + "grad_norm": 0.8157890439033508, + "learning_rate": 1.3473e-05, + "loss": 0.0791, + "step": 4494 + }, + { + "epoch": 3.7964527027027026, + "grad_norm": 0.7081852555274963, + "learning_rate": 1.3476e-05, + "loss": 0.0568, + "step": 4495 + }, + { + "epoch": 3.7972972972972974, + "grad_norm": 0.8999429941177368, + "learning_rate": 1.3479e-05, + "loss": 0.0789, + "step": 4496 + }, + { + "epoch": 3.798141891891892, + "grad_norm": 0.8801524639129639, + "learning_rate": 1.3482e-05, + "loss": 0.096, + "step": 4497 + }, + { + "epoch": 3.7989864864864864, + "grad_norm": 1.2276580333709717, + "learning_rate": 1.3485e-05, + "loss": 0.101, + "step": 4498 + }, + { + "epoch": 3.799831081081081, + "grad_norm": 1.0851414203643799, + "learning_rate": 1.3488e-05, + "loss": 0.0963, + "step": 4499 + }, + { + "epoch": 3.8006756756756754, + "grad_norm": 0.8623608350753784, + "learning_rate": 1.3491e-05, + "loss": 0.0742, + "step": 4500 + }, + { + "epoch": 3.80152027027027, + "grad_norm": 1.885053277015686, + "learning_rate": 1.3494e-05, + "loss": 0.1235, + "step": 4501 + }, + { + "epoch": 3.802364864864865, + "grad_norm": 1.4717981815338135, + "learning_rate": 1.3497000000000001e-05, + "loss": 0.1386, + "step": 4502 + }, + { + "epoch": 3.8032094594594597, + "grad_norm": 3.1443285942077637, + "learning_rate": 1.3500000000000001e-05, + "loss": 0.4827, + "step": 4503 + }, + { + "epoch": 3.804054054054054, + "grad_norm": 1.1415032148361206, + "learning_rate": 1.3503000000000001e-05, + "loss": 0.3014, + "step": 4504 + }, + { + "epoch": 3.8048986486486487, + "grad_norm": 1.6906769275665283, + "learning_rate": 1.3506e-05, + "loss": 0.1806, + "step": 4505 + }, + { + "epoch": 3.805743243243243, + "grad_norm": 1.171152949333191, + "learning_rate": 1.3508999999999999e-05, + "loss": 0.1208, + "step": 4506 + }, + { + "epoch": 3.8065878378378377, + "grad_norm": 0.700550377368927, + "learning_rate": 1.3512e-05, + "loss": 0.1069, + "step": 4507 + }, + { + "epoch": 3.8074324324324325, + "grad_norm": 0.49418574571609497, + "learning_rate": 1.3515e-05, + "loss": 0.0765, + "step": 4508 + }, + { + "epoch": 3.808277027027027, + "grad_norm": 0.8745545744895935, + "learning_rate": 1.3518e-05, + "loss": 0.0746, + "step": 4509 + }, + { + "epoch": 3.8091216216216215, + "grad_norm": 0.7120918035507202, + "learning_rate": 1.3521e-05, + "loss": 0.0783, + "step": 4510 + }, + { + "epoch": 3.8099662162162162, + "grad_norm": 0.6318997144699097, + "learning_rate": 1.3524e-05, + "loss": 0.0599, + "step": 4511 + }, + { + "epoch": 3.810810810810811, + "grad_norm": 0.8581671714782715, + "learning_rate": 1.3527000000000001e-05, + "loss": 0.1009, + "step": 4512 + }, + { + "epoch": 3.8116554054054053, + "grad_norm": 0.6312020421028137, + "learning_rate": 1.3530000000000001e-05, + "loss": 0.0733, + "step": 4513 + }, + { + "epoch": 3.8125, + "grad_norm": 0.6843045949935913, + "learning_rate": 1.3533e-05, + "loss": 0.077, + "step": 4514 + }, + { + "epoch": 3.8133445945945947, + "grad_norm": 0.45423629879951477, + "learning_rate": 1.3536e-05, + "loss": 0.0586, + "step": 4515 + }, + { + "epoch": 3.814189189189189, + "grad_norm": 0.5964615345001221, + "learning_rate": 1.3539e-05, + "loss": 0.0497, + "step": 4516 + }, + { + "epoch": 3.8150337837837838, + "grad_norm": 0.8893651366233826, + "learning_rate": 1.3542000000000002e-05, + "loss": 0.0695, + "step": 4517 + }, + { + "epoch": 3.8158783783783785, + "grad_norm": 0.5105565190315247, + "learning_rate": 1.3545e-05, + "loss": 0.0659, + "step": 4518 + }, + { + "epoch": 3.816722972972973, + "grad_norm": 0.7465155124664307, + "learning_rate": 1.3548e-05, + "loss": 0.0804, + "step": 4519 + }, + { + "epoch": 3.8175675675675675, + "grad_norm": 0.5519887208938599, + "learning_rate": 1.3551e-05, + "loss": 0.0575, + "step": 4520 + }, + { + "epoch": 3.8184121621621623, + "grad_norm": 0.5325664281845093, + "learning_rate": 1.3554e-05, + "loss": 0.0703, + "step": 4521 + }, + { + "epoch": 3.819256756756757, + "grad_norm": 0.6138718724250793, + "learning_rate": 1.3557e-05, + "loss": 0.0729, + "step": 4522 + }, + { + "epoch": 3.8201013513513513, + "grad_norm": 0.7799266576766968, + "learning_rate": 1.356e-05, + "loss": 0.0888, + "step": 4523 + }, + { + "epoch": 3.820945945945946, + "grad_norm": 0.609050989151001, + "learning_rate": 1.3563e-05, + "loss": 0.059, + "step": 4524 + }, + { + "epoch": 3.8217905405405403, + "grad_norm": 0.859428346157074, + "learning_rate": 1.3566e-05, + "loss": 0.0776, + "step": 4525 + }, + { + "epoch": 3.822635135135135, + "grad_norm": 0.46672865748405457, + "learning_rate": 1.3569e-05, + "loss": 0.0502, + "step": 4526 + }, + { + "epoch": 3.82347972972973, + "grad_norm": 0.5739372372627258, + "learning_rate": 1.3572000000000002e-05, + "loss": 0.053, + "step": 4527 + }, + { + "epoch": 3.8243243243243246, + "grad_norm": 0.6419994235038757, + "learning_rate": 1.3575000000000001e-05, + "loss": 0.0437, + "step": 4528 + }, + { + "epoch": 3.825168918918919, + "grad_norm": 0.5082858800888062, + "learning_rate": 1.3578000000000001e-05, + "loss": 0.055, + "step": 4529 + }, + { + "epoch": 3.8260135135135136, + "grad_norm": 0.9427955746650696, + "learning_rate": 1.3581000000000001e-05, + "loss": 0.0916, + "step": 4530 + }, + { + "epoch": 3.826858108108108, + "grad_norm": 0.6221125721931458, + "learning_rate": 1.3583999999999999e-05, + "loss": 0.0571, + "step": 4531 + }, + { + "epoch": 3.8277027027027026, + "grad_norm": 0.7643008828163147, + "learning_rate": 1.3587e-05, + "loss": 0.0615, + "step": 4532 + }, + { + "epoch": 3.8285472972972974, + "grad_norm": 0.7007522583007812, + "learning_rate": 1.359e-05, + "loss": 0.0559, + "step": 4533 + }, + { + "epoch": 3.829391891891892, + "grad_norm": 0.4841384291648865, + "learning_rate": 1.3593e-05, + "loss": 0.0479, + "step": 4534 + }, + { + "epoch": 3.8302364864864864, + "grad_norm": 0.850412905216217, + "learning_rate": 1.3596e-05, + "loss": 0.0586, + "step": 4535 + }, + { + "epoch": 3.831081081081081, + "grad_norm": 0.743853747844696, + "learning_rate": 1.3599e-05, + "loss": 0.0945, + "step": 4536 + }, + { + "epoch": 3.8319256756756754, + "grad_norm": 0.6375693678855896, + "learning_rate": 1.3602000000000001e-05, + "loss": 0.0548, + "step": 4537 + }, + { + "epoch": 3.83277027027027, + "grad_norm": 0.7488872408866882, + "learning_rate": 1.3605000000000001e-05, + "loss": 0.0578, + "step": 4538 + }, + { + "epoch": 3.833614864864865, + "grad_norm": 0.6489161252975464, + "learning_rate": 1.3608e-05, + "loss": 0.0742, + "step": 4539 + }, + { + "epoch": 3.8344594594594597, + "grad_norm": 1.061689853668213, + "learning_rate": 1.3611e-05, + "loss": 0.0877, + "step": 4540 + }, + { + "epoch": 3.835304054054054, + "grad_norm": 0.8411803245544434, + "learning_rate": 1.3614e-05, + "loss": 0.097, + "step": 4541 + }, + { + "epoch": 3.8361486486486487, + "grad_norm": 0.6637545824050903, + "learning_rate": 1.3617000000000002e-05, + "loss": 0.0691, + "step": 4542 + }, + { + "epoch": 3.836993243243243, + "grad_norm": 0.9128286242485046, + "learning_rate": 1.362e-05, + "loss": 0.0695, + "step": 4543 + }, + { + "epoch": 3.8378378378378377, + "grad_norm": 0.7596311569213867, + "learning_rate": 1.3623e-05, + "loss": 0.0597, + "step": 4544 + }, + { + "epoch": 3.8386824324324325, + "grad_norm": 0.9908422231674194, + "learning_rate": 1.3626e-05, + "loss": 0.0876, + "step": 4545 + }, + { + "epoch": 3.839527027027027, + "grad_norm": 0.8905428051948547, + "learning_rate": 1.3629e-05, + "loss": 0.0791, + "step": 4546 + }, + { + "epoch": 3.8403716216216215, + "grad_norm": 1.2866846323013306, + "learning_rate": 1.3632000000000001e-05, + "loss": 0.1409, + "step": 4547 + }, + { + "epoch": 3.8412162162162162, + "grad_norm": 0.9755632877349854, + "learning_rate": 1.3635e-05, + "loss": 0.0655, + "step": 4548 + }, + { + "epoch": 3.842060810810811, + "grad_norm": 2.446683168411255, + "learning_rate": 1.3638e-05, + "loss": 0.0971, + "step": 4549 + }, + { + "epoch": 3.8429054054054053, + "grad_norm": 0.7075211405754089, + "learning_rate": 1.3641e-05, + "loss": 0.0698, + "step": 4550 + }, + { + "epoch": 3.84375, + "grad_norm": 0.9702792167663574, + "learning_rate": 1.3644e-05, + "loss": 0.0682, + "step": 4551 + }, + { + "epoch": 3.8445945945945947, + "grad_norm": 16.57729721069336, + "learning_rate": 1.3647000000000002e-05, + "loss": 0.1422, + "step": 4552 + }, + { + "epoch": 3.845439189189189, + "grad_norm": 4.694633960723877, + "learning_rate": 1.3650000000000001e-05, + "loss": 0.6506, + "step": 4553 + }, + { + "epoch": 3.8462837837837838, + "grad_norm": 1.1975760459899902, + "learning_rate": 1.3653000000000001e-05, + "loss": 0.3175, + "step": 4554 + }, + { + "epoch": 3.8471283783783785, + "grad_norm": 1.0960354804992676, + "learning_rate": 1.3656e-05, + "loss": 0.147, + "step": 4555 + }, + { + "epoch": 3.847972972972973, + "grad_norm": 0.8602277636528015, + "learning_rate": 1.3659e-05, + "loss": 0.0988, + "step": 4556 + }, + { + "epoch": 3.8488175675675675, + "grad_norm": 0.9277072548866272, + "learning_rate": 1.3662e-05, + "loss": 0.098, + "step": 4557 + }, + { + "epoch": 3.8496621621621623, + "grad_norm": 0.49671196937561035, + "learning_rate": 1.3665e-05, + "loss": 0.07, + "step": 4558 + }, + { + "epoch": 3.850506756756757, + "grad_norm": 0.7039651870727539, + "learning_rate": 1.3668e-05, + "loss": 0.1077, + "step": 4559 + }, + { + "epoch": 3.8513513513513513, + "grad_norm": 0.6993746757507324, + "learning_rate": 1.3671e-05, + "loss": 0.0852, + "step": 4560 + }, + { + "epoch": 3.852195945945946, + "grad_norm": 0.6626499891281128, + "learning_rate": 1.3674e-05, + "loss": 0.0808, + "step": 4561 + }, + { + "epoch": 3.8530405405405403, + "grad_norm": 0.7265749573707581, + "learning_rate": 1.3677000000000001e-05, + "loss": 0.0713, + "step": 4562 + }, + { + "epoch": 3.853885135135135, + "grad_norm": 0.5782411098480225, + "learning_rate": 1.3680000000000001e-05, + "loss": 0.0516, + "step": 4563 + }, + { + "epoch": 3.85472972972973, + "grad_norm": 0.684842586517334, + "learning_rate": 1.3683000000000001e-05, + "loss": 0.0576, + "step": 4564 + }, + { + "epoch": 3.8555743243243246, + "grad_norm": 0.6994024515151978, + "learning_rate": 1.3686e-05, + "loss": 0.1082, + "step": 4565 + }, + { + "epoch": 3.856418918918919, + "grad_norm": 0.6535413265228271, + "learning_rate": 1.3689e-05, + "loss": 0.0782, + "step": 4566 + }, + { + "epoch": 3.8572635135135136, + "grad_norm": 0.5733880400657654, + "learning_rate": 1.3691999999999999e-05, + "loss": 0.0457, + "step": 4567 + }, + { + "epoch": 3.858108108108108, + "grad_norm": 0.7388923168182373, + "learning_rate": 1.3695e-05, + "loss": 0.0719, + "step": 4568 + }, + { + "epoch": 3.8589527027027026, + "grad_norm": 0.7467928528785706, + "learning_rate": 1.3698e-05, + "loss": 0.0698, + "step": 4569 + }, + { + "epoch": 3.8597972972972974, + "grad_norm": 0.648360013961792, + "learning_rate": 1.3701e-05, + "loss": 0.0533, + "step": 4570 + }, + { + "epoch": 3.860641891891892, + "grad_norm": 0.5393189191818237, + "learning_rate": 1.3704e-05, + "loss": 0.0697, + "step": 4571 + }, + { + "epoch": 3.8614864864864864, + "grad_norm": 0.596689760684967, + "learning_rate": 1.3707e-05, + "loss": 0.0491, + "step": 4572 + }, + { + "epoch": 3.862331081081081, + "grad_norm": 0.6219120025634766, + "learning_rate": 1.3710000000000001e-05, + "loss": 0.0743, + "step": 4573 + }, + { + "epoch": 3.8631756756756754, + "grad_norm": 0.4795258045196533, + "learning_rate": 1.3713e-05, + "loss": 0.0542, + "step": 4574 + }, + { + "epoch": 3.86402027027027, + "grad_norm": 0.8023406863212585, + "learning_rate": 1.3716e-05, + "loss": 0.0781, + "step": 4575 + }, + { + "epoch": 3.864864864864865, + "grad_norm": 0.62820965051651, + "learning_rate": 1.3719e-05, + "loss": 0.0638, + "step": 4576 + }, + { + "epoch": 3.8657094594594597, + "grad_norm": 0.8141347765922546, + "learning_rate": 1.3722e-05, + "loss": 0.0732, + "step": 4577 + }, + { + "epoch": 3.866554054054054, + "grad_norm": 0.5773369073867798, + "learning_rate": 1.3725000000000002e-05, + "loss": 0.0565, + "step": 4578 + }, + { + "epoch": 3.8673986486486487, + "grad_norm": 0.6483588814735413, + "learning_rate": 1.3728000000000001e-05, + "loss": 0.0649, + "step": 4579 + }, + { + "epoch": 3.868243243243243, + "grad_norm": 0.5263703465461731, + "learning_rate": 1.3731e-05, + "loss": 0.0496, + "step": 4580 + }, + { + "epoch": 3.8690878378378377, + "grad_norm": 0.6590248942375183, + "learning_rate": 1.3734e-05, + "loss": 0.0568, + "step": 4581 + }, + { + "epoch": 3.8699324324324325, + "grad_norm": 0.9083133935928345, + "learning_rate": 1.3736999999999999e-05, + "loss": 0.0979, + "step": 4582 + }, + { + "epoch": 3.870777027027027, + "grad_norm": 0.5906864404678345, + "learning_rate": 1.374e-05, + "loss": 0.0516, + "step": 4583 + }, + { + "epoch": 3.8716216216216215, + "grad_norm": 0.6740853786468506, + "learning_rate": 1.3743e-05, + "loss": 0.0656, + "step": 4584 + }, + { + "epoch": 3.8724662162162162, + "grad_norm": 0.5260892510414124, + "learning_rate": 1.3746e-05, + "loss": 0.0688, + "step": 4585 + }, + { + "epoch": 3.873310810810811, + "grad_norm": 0.6097966432571411, + "learning_rate": 1.3749e-05, + "loss": 0.0466, + "step": 4586 + }, + { + "epoch": 3.8741554054054053, + "grad_norm": 0.5384581685066223, + "learning_rate": 1.3752e-05, + "loss": 0.0501, + "step": 4587 + }, + { + "epoch": 3.875, + "grad_norm": 0.5342409014701843, + "learning_rate": 1.3755000000000001e-05, + "loss": 0.0464, + "step": 4588 + }, + { + "epoch": 3.8758445945945947, + "grad_norm": 0.971831738948822, + "learning_rate": 1.3758000000000001e-05, + "loss": 0.0786, + "step": 4589 + }, + { + "epoch": 3.876689189189189, + "grad_norm": 0.5523898005485535, + "learning_rate": 1.3761000000000001e-05, + "loss": 0.0578, + "step": 4590 + }, + { + "epoch": 3.8775337837837838, + "grad_norm": 0.7846912145614624, + "learning_rate": 1.3764e-05, + "loss": 0.0664, + "step": 4591 + }, + { + "epoch": 3.8783783783783785, + "grad_norm": 0.6297100782394409, + "learning_rate": 1.3766999999999999e-05, + "loss": 0.0646, + "step": 4592 + }, + { + "epoch": 3.879222972972973, + "grad_norm": 0.6539124250411987, + "learning_rate": 1.377e-05, + "loss": 0.0647, + "step": 4593 + }, + { + "epoch": 3.8800675675675675, + "grad_norm": 0.7522905468940735, + "learning_rate": 1.3773e-05, + "loss": 0.0678, + "step": 4594 + }, + { + "epoch": 3.8809121621621623, + "grad_norm": 0.826954185962677, + "learning_rate": 1.3776e-05, + "loss": 0.0691, + "step": 4595 + }, + { + "epoch": 3.881756756756757, + "grad_norm": 1.1948769092559814, + "learning_rate": 1.3779e-05, + "loss": 0.1251, + "step": 4596 + }, + { + "epoch": 3.8826013513513513, + "grad_norm": 1.1093205213546753, + "learning_rate": 1.3782e-05, + "loss": 0.0834, + "step": 4597 + }, + { + "epoch": 3.883445945945946, + "grad_norm": 1.2598150968551636, + "learning_rate": 1.3785000000000001e-05, + "loss": 0.0787, + "step": 4598 + }, + { + "epoch": 3.8842905405405403, + "grad_norm": 0.930587112903595, + "learning_rate": 1.3788e-05, + "loss": 0.078, + "step": 4599 + }, + { + "epoch": 3.885135135135135, + "grad_norm": 0.7629300951957703, + "learning_rate": 1.3791e-05, + "loss": 0.0948, + "step": 4600 + }, + { + "epoch": 3.88597972972973, + "grad_norm": 1.1999558210372925, + "learning_rate": 1.3794e-05, + "loss": 0.11, + "step": 4601 + }, + { + "epoch": 3.8868243243243246, + "grad_norm": 1.634070634841919, + "learning_rate": 1.3797e-05, + "loss": 0.1146, + "step": 4602 + }, + { + "epoch": 3.887668918918919, + "grad_norm": 1.605832576751709, + "learning_rate": 1.3800000000000002e-05, + "loss": 0.4369, + "step": 4603 + }, + { + "epoch": 3.8885135135135136, + "grad_norm": 1.9803825616836548, + "learning_rate": 1.3803e-05, + "loss": 0.3007, + "step": 4604 + }, + { + "epoch": 3.889358108108108, + "grad_norm": 0.6499809622764587, + "learning_rate": 1.3806e-05, + "loss": 0.1553, + "step": 4605 + }, + { + "epoch": 3.8902027027027026, + "grad_norm": 0.6270037293434143, + "learning_rate": 1.3809e-05, + "loss": 0.0735, + "step": 4606 + }, + { + "epoch": 3.8910472972972974, + "grad_norm": 0.8585547804832458, + "learning_rate": 1.3812e-05, + "loss": 0.0978, + "step": 4607 + }, + { + "epoch": 3.891891891891892, + "grad_norm": 0.5838525295257568, + "learning_rate": 1.3815e-05, + "loss": 0.0749, + "step": 4608 + }, + { + "epoch": 3.8927364864864864, + "grad_norm": 0.588310718536377, + "learning_rate": 1.3818e-05, + "loss": 0.0762, + "step": 4609 + }, + { + "epoch": 3.893581081081081, + "grad_norm": 0.466155081987381, + "learning_rate": 1.3821e-05, + "loss": 0.0501, + "step": 4610 + }, + { + "epoch": 3.8944256756756754, + "grad_norm": 0.4126177728176117, + "learning_rate": 1.3824e-05, + "loss": 0.0538, + "step": 4611 + }, + { + "epoch": 3.89527027027027, + "grad_norm": 0.5249996781349182, + "learning_rate": 1.3827e-05, + "loss": 0.0644, + "step": 4612 + }, + { + "epoch": 3.896114864864865, + "grad_norm": 0.5934330224990845, + "learning_rate": 1.3830000000000001e-05, + "loss": 0.0653, + "step": 4613 + }, + { + "epoch": 3.8969594594594597, + "grad_norm": 0.6003490686416626, + "learning_rate": 1.3833000000000001e-05, + "loss": 0.0717, + "step": 4614 + }, + { + "epoch": 3.897804054054054, + "grad_norm": 0.8847383260726929, + "learning_rate": 1.3836000000000001e-05, + "loss": 0.0769, + "step": 4615 + }, + { + "epoch": 3.8986486486486487, + "grad_norm": 0.7587101459503174, + "learning_rate": 1.3839e-05, + "loss": 0.0525, + "step": 4616 + }, + { + "epoch": 3.899493243243243, + "grad_norm": 0.9544668197631836, + "learning_rate": 1.3841999999999999e-05, + "loss": 0.0931, + "step": 4617 + }, + { + "epoch": 3.9003378378378377, + "grad_norm": 0.5334459543228149, + "learning_rate": 1.3845e-05, + "loss": 0.068, + "step": 4618 + }, + { + "epoch": 3.9011824324324325, + "grad_norm": 0.635109543800354, + "learning_rate": 1.3848e-05, + "loss": 0.0629, + "step": 4619 + }, + { + "epoch": 3.902027027027027, + "grad_norm": 0.4126764237880707, + "learning_rate": 1.3851e-05, + "loss": 0.0427, + "step": 4620 + }, + { + "epoch": 3.9028716216216215, + "grad_norm": 0.5145339965820312, + "learning_rate": 1.3854e-05, + "loss": 0.0581, + "step": 4621 + }, + { + "epoch": 3.9037162162162162, + "grad_norm": 1.1512970924377441, + "learning_rate": 1.3857e-05, + "loss": 0.0401, + "step": 4622 + }, + { + "epoch": 3.904560810810811, + "grad_norm": 0.5689579844474792, + "learning_rate": 1.3860000000000001e-05, + "loss": 0.0613, + "step": 4623 + }, + { + "epoch": 3.9054054054054053, + "grad_norm": 0.5652652978897095, + "learning_rate": 1.3863000000000001e-05, + "loss": 0.0554, + "step": 4624 + }, + { + "epoch": 3.90625, + "grad_norm": 0.4627223312854767, + "learning_rate": 1.3866e-05, + "loss": 0.0388, + "step": 4625 + }, + { + "epoch": 3.9070945945945947, + "grad_norm": 0.6252832412719727, + "learning_rate": 1.3869e-05, + "loss": 0.0646, + "step": 4626 + }, + { + "epoch": 3.907939189189189, + "grad_norm": 0.5491419434547424, + "learning_rate": 1.3872e-05, + "loss": 0.0584, + "step": 4627 + }, + { + "epoch": 3.9087837837837838, + "grad_norm": 0.7211734652519226, + "learning_rate": 1.3875000000000002e-05, + "loss": 0.0544, + "step": 4628 + }, + { + "epoch": 3.9096283783783785, + "grad_norm": 0.9797558784484863, + "learning_rate": 1.3878e-05, + "loss": 0.0934, + "step": 4629 + }, + { + "epoch": 3.910472972972973, + "grad_norm": 0.6591535210609436, + "learning_rate": 1.3881e-05, + "loss": 0.0584, + "step": 4630 + }, + { + "epoch": 3.9113175675675675, + "grad_norm": 0.6008901000022888, + "learning_rate": 1.3884e-05, + "loss": 0.061, + "step": 4631 + }, + { + "epoch": 3.9121621621621623, + "grad_norm": 0.6215082406997681, + "learning_rate": 1.3887e-05, + "loss": 0.0459, + "step": 4632 + }, + { + "epoch": 3.913006756756757, + "grad_norm": 0.6810535192489624, + "learning_rate": 1.389e-05, + "loss": 0.0722, + "step": 4633 + }, + { + "epoch": 3.9138513513513513, + "grad_norm": 0.9543363451957703, + "learning_rate": 1.3893e-05, + "loss": 0.0676, + "step": 4634 + }, + { + "epoch": 3.914695945945946, + "grad_norm": 1.0094856023788452, + "learning_rate": 1.3896e-05, + "loss": 0.056, + "step": 4635 + }, + { + "epoch": 3.9155405405405403, + "grad_norm": 0.6195409893989563, + "learning_rate": 1.3899e-05, + "loss": 0.0724, + "step": 4636 + }, + { + "epoch": 3.916385135135135, + "grad_norm": 0.7741706371307373, + "learning_rate": 1.3902e-05, + "loss": 0.0606, + "step": 4637 + }, + { + "epoch": 3.91722972972973, + "grad_norm": 0.7414426207542419, + "learning_rate": 1.3905000000000002e-05, + "loss": 0.0581, + "step": 4638 + }, + { + "epoch": 3.9180743243243246, + "grad_norm": 0.8224761486053467, + "learning_rate": 1.3908000000000001e-05, + "loss": 0.0708, + "step": 4639 + }, + { + "epoch": 3.918918918918919, + "grad_norm": 0.8103757500648499, + "learning_rate": 1.3911000000000001e-05, + "loss": 0.0971, + "step": 4640 + }, + { + "epoch": 3.9197635135135136, + "grad_norm": 0.6145468950271606, + "learning_rate": 1.3914e-05, + "loss": 0.0565, + "step": 4641 + }, + { + "epoch": 3.920608108108108, + "grad_norm": 0.7376102209091187, + "learning_rate": 1.3916999999999999e-05, + "loss": 0.074, + "step": 4642 + }, + { + "epoch": 3.9214527027027026, + "grad_norm": 0.7010684609413147, + "learning_rate": 1.392e-05, + "loss": 0.0596, + "step": 4643 + }, + { + "epoch": 3.9222972972972974, + "grad_norm": 0.9380566477775574, + "learning_rate": 1.3923e-05, + "loss": 0.1096, + "step": 4644 + }, + { + "epoch": 3.923141891891892, + "grad_norm": 0.6298916935920715, + "learning_rate": 1.3926e-05, + "loss": 0.0577, + "step": 4645 + }, + { + "epoch": 3.9239864864864864, + "grad_norm": 0.714770495891571, + "learning_rate": 1.3929e-05, + "loss": 0.0849, + "step": 4646 + }, + { + "epoch": 3.924831081081081, + "grad_norm": 0.9976733922958374, + "learning_rate": 1.3932e-05, + "loss": 0.0978, + "step": 4647 + }, + { + "epoch": 3.9256756756756754, + "grad_norm": 0.9113912582397461, + "learning_rate": 1.3935000000000001e-05, + "loss": 0.0787, + "step": 4648 + }, + { + "epoch": 3.92652027027027, + "grad_norm": 1.1044806241989136, + "learning_rate": 1.3938000000000001e-05, + "loss": 0.1072, + "step": 4649 + }, + { + "epoch": 3.927364864864865, + "grad_norm": 1.1070258617401123, + "learning_rate": 1.3941000000000001e-05, + "loss": 0.0908, + "step": 4650 + }, + { + "epoch": 3.9282094594594597, + "grad_norm": 1.1042001247406006, + "learning_rate": 1.3944e-05, + "loss": 0.1139, + "step": 4651 + }, + { + "epoch": 3.929054054054054, + "grad_norm": 1.5824265480041504, + "learning_rate": 1.3947e-05, + "loss": 0.1758, + "step": 4652 + }, + { + "epoch": 3.9298986486486487, + "grad_norm": 2.1188583374023438, + "learning_rate": 1.395e-05, + "loss": 0.4248, + "step": 4653 + }, + { + "epoch": 3.930743243243243, + "grad_norm": 1.2013071775436401, + "learning_rate": 1.3953e-05, + "loss": 0.2581, + "step": 4654 + }, + { + "epoch": 3.9315878378378377, + "grad_norm": 1.0341354608535767, + "learning_rate": 1.3956e-05, + "loss": 0.1752, + "step": 4655 + }, + { + "epoch": 3.9324324324324325, + "grad_norm": 0.5329644680023193, + "learning_rate": 1.3959e-05, + "loss": 0.0697, + "step": 4656 + }, + { + "epoch": 3.933277027027027, + "grad_norm": 0.773332953453064, + "learning_rate": 1.3962e-05, + "loss": 0.0941, + "step": 4657 + }, + { + "epoch": 3.9341216216216215, + "grad_norm": 0.4608597159385681, + "learning_rate": 1.3965000000000001e-05, + "loss": 0.0571, + "step": 4658 + }, + { + "epoch": 3.9349662162162162, + "grad_norm": 0.6185877919197083, + "learning_rate": 1.3968e-05, + "loss": 0.0748, + "step": 4659 + }, + { + "epoch": 3.935810810810811, + "grad_norm": 0.5080185532569885, + "learning_rate": 1.3971e-05, + "loss": 0.0708, + "step": 4660 + }, + { + "epoch": 3.9366554054054053, + "grad_norm": 0.5305293798446655, + "learning_rate": 1.3974e-05, + "loss": 0.079, + "step": 4661 + }, + { + "epoch": 3.9375, + "grad_norm": 0.6832944750785828, + "learning_rate": 1.3977e-05, + "loss": 0.0891, + "step": 4662 + }, + { + "epoch": 3.9383445945945947, + "grad_norm": 0.6231105327606201, + "learning_rate": 1.3980000000000002e-05, + "loss": 0.0764, + "step": 4663 + }, + { + "epoch": 3.939189189189189, + "grad_norm": 0.47597700357437134, + "learning_rate": 1.3983000000000001e-05, + "loss": 0.0675, + "step": 4664 + }, + { + "epoch": 3.9400337837837838, + "grad_norm": 0.5736660957336426, + "learning_rate": 1.3986000000000001e-05, + "loss": 0.0558, + "step": 4665 + }, + { + "epoch": 3.9408783783783785, + "grad_norm": 0.6043276190757751, + "learning_rate": 1.3989e-05, + "loss": 0.0742, + "step": 4666 + }, + { + "epoch": 3.941722972972973, + "grad_norm": 0.5044993758201599, + "learning_rate": 1.3992e-05, + "loss": 0.048, + "step": 4667 + }, + { + "epoch": 3.9425675675675675, + "grad_norm": 0.5769449472427368, + "learning_rate": 1.3995e-05, + "loss": 0.0742, + "step": 4668 + }, + { + "epoch": 3.9434121621621623, + "grad_norm": 0.6376593112945557, + "learning_rate": 1.3998e-05, + "loss": 0.0564, + "step": 4669 + }, + { + "epoch": 3.944256756756757, + "grad_norm": 0.4952482581138611, + "learning_rate": 1.4001e-05, + "loss": 0.0618, + "step": 4670 + }, + { + "epoch": 3.9451013513513513, + "grad_norm": 0.6211197376251221, + "learning_rate": 1.4004e-05, + "loss": 0.092, + "step": 4671 + }, + { + "epoch": 3.945945945945946, + "grad_norm": 0.6414012908935547, + "learning_rate": 1.4007e-05, + "loss": 0.0922, + "step": 4672 + }, + { + "epoch": 3.9467905405405403, + "grad_norm": 0.6948162913322449, + "learning_rate": 1.4010000000000001e-05, + "loss": 0.0605, + "step": 4673 + }, + { + "epoch": 3.947635135135135, + "grad_norm": 0.6076042056083679, + "learning_rate": 1.4013000000000001e-05, + "loss": 0.0769, + "step": 4674 + }, + { + "epoch": 3.94847972972973, + "grad_norm": 0.7625834941864014, + "learning_rate": 1.4016000000000001e-05, + "loss": 0.0681, + "step": 4675 + }, + { + "epoch": 3.9493243243243246, + "grad_norm": 0.5347191691398621, + "learning_rate": 1.4019e-05, + "loss": 0.0672, + "step": 4676 + }, + { + "epoch": 3.950168918918919, + "grad_norm": 0.4505428075790405, + "learning_rate": 1.4022e-05, + "loss": 0.0528, + "step": 4677 + }, + { + "epoch": 3.9510135135135136, + "grad_norm": 0.5097553133964539, + "learning_rate": 1.4025e-05, + "loss": 0.0518, + "step": 4678 + }, + { + "epoch": 3.951858108108108, + "grad_norm": 0.7792648077011108, + "learning_rate": 1.4028e-05, + "loss": 0.0821, + "step": 4679 + }, + { + "epoch": 3.9527027027027026, + "grad_norm": 0.8837346434593201, + "learning_rate": 1.4031e-05, + "loss": 0.0821, + "step": 4680 + }, + { + "epoch": 3.9535472972972974, + "grad_norm": 0.5203768610954285, + "learning_rate": 1.4034e-05, + "loss": 0.0551, + "step": 4681 + }, + { + "epoch": 3.954391891891892, + "grad_norm": 0.6322460174560547, + "learning_rate": 1.4037e-05, + "loss": 0.0914, + "step": 4682 + }, + { + "epoch": 3.9552364864864864, + "grad_norm": 0.5834423899650574, + "learning_rate": 1.4040000000000001e-05, + "loss": 0.0642, + "step": 4683 + }, + { + "epoch": 3.956081081081081, + "grad_norm": 0.5401440858840942, + "learning_rate": 1.4043000000000001e-05, + "loss": 0.0623, + "step": 4684 + }, + { + "epoch": 3.9569256756756754, + "grad_norm": 0.6657136678695679, + "learning_rate": 1.4046e-05, + "loss": 0.0488, + "step": 4685 + }, + { + "epoch": 3.95777027027027, + "grad_norm": 0.6978393793106079, + "learning_rate": 1.4049e-05, + "loss": 0.0813, + "step": 4686 + }, + { + "epoch": 3.958614864864865, + "grad_norm": 0.9655654430389404, + "learning_rate": 1.4052e-05, + "loss": 0.0803, + "step": 4687 + }, + { + "epoch": 3.9594594594594597, + "grad_norm": 0.5556730031967163, + "learning_rate": 1.4055000000000002e-05, + "loss": 0.0569, + "step": 4688 + }, + { + "epoch": 3.960304054054054, + "grad_norm": 0.8332982063293457, + "learning_rate": 1.4058000000000002e-05, + "loss": 0.0652, + "step": 4689 + }, + { + "epoch": 3.9611486486486487, + "grad_norm": 0.6229501962661743, + "learning_rate": 1.4061e-05, + "loss": 0.0728, + "step": 4690 + }, + { + "epoch": 3.961993243243243, + "grad_norm": 0.5085963010787964, + "learning_rate": 1.4064e-05, + "loss": 0.0582, + "step": 4691 + }, + { + "epoch": 3.9628378378378377, + "grad_norm": 0.8646627068519592, + "learning_rate": 1.4067e-05, + "loss": 0.0873, + "step": 4692 + }, + { + "epoch": 3.9636824324324325, + "grad_norm": 0.8730455636978149, + "learning_rate": 1.4069999999999999e-05, + "loss": 0.0677, + "step": 4693 + }, + { + "epoch": 3.964527027027027, + "grad_norm": 0.7451529502868652, + "learning_rate": 1.4073e-05, + "loss": 0.0643, + "step": 4694 + }, + { + "epoch": 3.9653716216216215, + "grad_norm": 0.6635097861289978, + "learning_rate": 1.4076e-05, + "loss": 0.0673, + "step": 4695 + }, + { + "epoch": 3.9662162162162162, + "grad_norm": 1.0096977949142456, + "learning_rate": 1.4079e-05, + "loss": 0.1176, + "step": 4696 + }, + { + "epoch": 3.967060810810811, + "grad_norm": 0.6931459903717041, + "learning_rate": 1.4082e-05, + "loss": 0.0733, + "step": 4697 + }, + { + "epoch": 3.9679054054054053, + "grad_norm": 0.6994540691375732, + "learning_rate": 1.4085e-05, + "loss": 0.0808, + "step": 4698 + }, + { + "epoch": 3.96875, + "grad_norm": 0.8970212340354919, + "learning_rate": 1.4088000000000001e-05, + "loss": 0.1116, + "step": 4699 + }, + { + "epoch": 3.9695945945945947, + "grad_norm": 1.0381790399551392, + "learning_rate": 1.4091000000000001e-05, + "loss": 0.0967, + "step": 4700 + }, + { + "epoch": 3.970439189189189, + "grad_norm": 2.0756633281707764, + "learning_rate": 1.4094000000000001e-05, + "loss": 0.2229, + "step": 4701 + }, + { + "epoch": 3.9712837837837838, + "grad_norm": 1.4619795083999634, + "learning_rate": 1.4097e-05, + "loss": 0.1166, + "step": 4702 + }, + { + "epoch": 3.9721283783783785, + "grad_norm": 3.68719220161438, + "learning_rate": 1.4099999999999999e-05, + "loss": 0.5221, + "step": 4703 + }, + { + "epoch": 3.972972972972973, + "grad_norm": 0.8525757789611816, + "learning_rate": 1.4103e-05, + "loss": 0.1653, + "step": 4704 + }, + { + "epoch": 3.9738175675675675, + "grad_norm": 0.6685129404067993, + "learning_rate": 1.4106e-05, + "loss": 0.1345, + "step": 4705 + }, + { + "epoch": 3.9746621621621623, + "grad_norm": 0.614292323589325, + "learning_rate": 1.4109e-05, + "loss": 0.0841, + "step": 4706 + }, + { + "epoch": 3.975506756756757, + "grad_norm": 0.4470475912094116, + "learning_rate": 1.4112e-05, + "loss": 0.059, + "step": 4707 + }, + { + "epoch": 3.9763513513513513, + "grad_norm": 0.5616071224212646, + "learning_rate": 1.4115e-05, + "loss": 0.082, + "step": 4708 + }, + { + "epoch": 3.977195945945946, + "grad_norm": 0.5062105059623718, + "learning_rate": 1.4118000000000001e-05, + "loss": 0.0502, + "step": 4709 + }, + { + "epoch": 3.9780405405405403, + "grad_norm": 0.4427416920661926, + "learning_rate": 1.4121e-05, + "loss": 0.0459, + "step": 4710 + }, + { + "epoch": 3.978885135135135, + "grad_norm": 0.6375781893730164, + "learning_rate": 1.4124e-05, + "loss": 0.067, + "step": 4711 + }, + { + "epoch": 3.97972972972973, + "grad_norm": 0.711895227432251, + "learning_rate": 1.4127e-05, + "loss": 0.0719, + "step": 4712 + }, + { + "epoch": 3.9805743243243246, + "grad_norm": 0.49138113856315613, + "learning_rate": 1.413e-05, + "loss": 0.0437, + "step": 4713 + }, + { + "epoch": 3.981418918918919, + "grad_norm": 0.6944342851638794, + "learning_rate": 1.4133000000000002e-05, + "loss": 0.0666, + "step": 4714 + }, + { + "epoch": 3.9822635135135136, + "grad_norm": 0.6298285722732544, + "learning_rate": 1.4136e-05, + "loss": 0.0648, + "step": 4715 + }, + { + "epoch": 3.983108108108108, + "grad_norm": 0.6491815447807312, + "learning_rate": 1.4139e-05, + "loss": 0.0728, + "step": 4716 + }, + { + "epoch": 3.9839527027027026, + "grad_norm": 1.7449971437454224, + "learning_rate": 1.4142e-05, + "loss": 0.0607, + "step": 4717 + }, + { + "epoch": 3.9847972972972974, + "grad_norm": 0.5967217683792114, + "learning_rate": 1.4145e-05, + "loss": 0.0514, + "step": 4718 + }, + { + "epoch": 3.985641891891892, + "grad_norm": 0.55485999584198, + "learning_rate": 1.4148e-05, + "loss": 0.0587, + "step": 4719 + }, + { + "epoch": 3.9864864864864864, + "grad_norm": 0.447922945022583, + "learning_rate": 1.4151e-05, + "loss": 0.0274, + "step": 4720 + }, + { + "epoch": 3.987331081081081, + "grad_norm": 0.5048686265945435, + "learning_rate": 1.4154e-05, + "loss": 0.0537, + "step": 4721 + }, + { + "epoch": 3.9881756756756754, + "grad_norm": 0.6335474848747253, + "learning_rate": 1.4157e-05, + "loss": 0.0641, + "step": 4722 + }, + { + "epoch": 3.98902027027027, + "grad_norm": 0.7054678201675415, + "learning_rate": 1.416e-05, + "loss": 0.0675, + "step": 4723 + }, + { + "epoch": 3.989864864864865, + "grad_norm": 0.6109403371810913, + "learning_rate": 1.4163000000000001e-05, + "loss": 0.0468, + "step": 4724 + }, + { + "epoch": 3.9907094594594597, + "grad_norm": 0.5923206806182861, + "learning_rate": 1.4166000000000001e-05, + "loss": 0.0793, + "step": 4725 + }, + { + "epoch": 3.991554054054054, + "grad_norm": 0.7271683216094971, + "learning_rate": 1.4169000000000001e-05, + "loss": 0.0828, + "step": 4726 + }, + { + "epoch": 3.9923986486486487, + "grad_norm": 0.7700018286705017, + "learning_rate": 1.4172e-05, + "loss": 0.0709, + "step": 4727 + }, + { + "epoch": 3.993243243243243, + "grad_norm": 0.8280394673347473, + "learning_rate": 1.4174999999999999e-05, + "loss": 0.0846, + "step": 4728 + }, + { + "epoch": 3.9940878378378377, + "grad_norm": 0.7896605730056763, + "learning_rate": 1.4178e-05, + "loss": 0.0847, + "step": 4729 + }, + { + "epoch": 3.9949324324324325, + "grad_norm": 0.6547603607177734, + "learning_rate": 1.4181e-05, + "loss": 0.0837, + "step": 4730 + }, + { + "epoch": 3.995777027027027, + "grad_norm": 1.0659641027450562, + "learning_rate": 1.4184e-05, + "loss": 0.0678, + "step": 4731 + }, + { + "epoch": 3.9966216216216215, + "grad_norm": 0.6148770451545715, + "learning_rate": 1.4187e-05, + "loss": 0.0563, + "step": 4732 + }, + { + "epoch": 3.9974662162162162, + "grad_norm": 3.795145034790039, + "learning_rate": 1.419e-05, + "loss": 0.1046, + "step": 4733 + }, + { + "epoch": 3.998310810810811, + "grad_norm": 1.2310853004455566, + "learning_rate": 1.4193000000000001e-05, + "loss": 0.1268, + "step": 4734 + }, + { + "epoch": 3.9991554054054053, + "grad_norm": 0.6967158913612366, + "learning_rate": 1.4196000000000001e-05, + "loss": 0.081, + "step": 4735 + }, + { + "epoch": 4.0, + "grad_norm": 1.8275134563446045, + "learning_rate": 1.4199e-05, + "loss": 0.1926, + "step": 4736 + }, + { + "epoch": 4.000844594594595, + "grad_norm": 2.281329393386841, + "learning_rate": 1.4202e-05, + "loss": 0.4542, + "step": 4737 + }, + { + "epoch": 4.0016891891891895, + "grad_norm": 1.5186901092529297, + "learning_rate": 1.4205e-05, + "loss": 0.2628, + "step": 4738 + }, + { + "epoch": 4.002533783783784, + "grad_norm": 0.571540892124176, + "learning_rate": 1.4208e-05, + "loss": 0.1217, + "step": 4739 + }, + { + "epoch": 4.003378378378378, + "grad_norm": 0.49620187282562256, + "learning_rate": 1.4211e-05, + "loss": 0.0632, + "step": 4740 + }, + { + "epoch": 4.004222972972973, + "grad_norm": 0.4659908413887024, + "learning_rate": 1.4214e-05, + "loss": 0.0866, + "step": 4741 + }, + { + "epoch": 4.0050675675675675, + "grad_norm": 0.6372514367103577, + "learning_rate": 1.4217e-05, + "loss": 0.089, + "step": 4742 + }, + { + "epoch": 4.005912162162162, + "grad_norm": 0.4808290898799896, + "learning_rate": 1.422e-05, + "loss": 0.0596, + "step": 4743 + }, + { + "epoch": 4.006756756756757, + "grad_norm": 0.4735528826713562, + "learning_rate": 1.4223000000000001e-05, + "loss": 0.0546, + "step": 4744 + }, + { + "epoch": 4.007601351351352, + "grad_norm": 0.6463225483894348, + "learning_rate": 1.4226e-05, + "loss": 0.0913, + "step": 4745 + }, + { + "epoch": 4.008445945945946, + "grad_norm": 0.48817357420921326, + "learning_rate": 1.4229e-05, + "loss": 0.0644, + "step": 4746 + }, + { + "epoch": 4.00929054054054, + "grad_norm": 0.586111843585968, + "learning_rate": 1.4232e-05, + "loss": 0.0612, + "step": 4747 + }, + { + "epoch": 4.010135135135135, + "grad_norm": 0.6292346119880676, + "learning_rate": 1.4235e-05, + "loss": 0.0611, + "step": 4748 + }, + { + "epoch": 4.01097972972973, + "grad_norm": 0.5416604280471802, + "learning_rate": 1.4238000000000002e-05, + "loss": 0.0604, + "step": 4749 + }, + { + "epoch": 4.011824324324325, + "grad_norm": 0.44902825355529785, + "learning_rate": 1.4241000000000001e-05, + "loss": 0.0551, + "step": 4750 + }, + { + "epoch": 4.012668918918919, + "grad_norm": 0.7712932825088501, + "learning_rate": 1.4244000000000001e-05, + "loss": 0.0513, + "step": 4751 + }, + { + "epoch": 4.013513513513513, + "grad_norm": 0.5038570165634155, + "learning_rate": 1.4247e-05, + "loss": 0.0761, + "step": 4752 + }, + { + "epoch": 4.014358108108108, + "grad_norm": 0.45758509635925293, + "learning_rate": 1.4249999999999999e-05, + "loss": 0.0601, + "step": 4753 + }, + { + "epoch": 4.015202702702703, + "grad_norm": 0.4037126898765564, + "learning_rate": 1.4253e-05, + "loss": 0.033, + "step": 4754 + }, + { + "epoch": 4.016047297297297, + "grad_norm": 0.5764806270599365, + "learning_rate": 1.4256e-05, + "loss": 0.0669, + "step": 4755 + }, + { + "epoch": 4.016891891891892, + "grad_norm": 0.49403107166290283, + "learning_rate": 1.4259e-05, + "loss": 0.0689, + "step": 4756 + }, + { + "epoch": 4.017736486486487, + "grad_norm": 0.5201539993286133, + "learning_rate": 1.4262e-05, + "loss": 0.0539, + "step": 4757 + }, + { + "epoch": 4.018581081081081, + "grad_norm": 0.40986135601997375, + "learning_rate": 1.4265e-05, + "loss": 0.0397, + "step": 4758 + }, + { + "epoch": 4.019425675675675, + "grad_norm": 0.7142712473869324, + "learning_rate": 1.4268000000000001e-05, + "loss": 0.0657, + "step": 4759 + }, + { + "epoch": 4.02027027027027, + "grad_norm": 0.49589914083480835, + "learning_rate": 1.4271000000000001e-05, + "loss": 0.0644, + "step": 4760 + }, + { + "epoch": 4.021114864864865, + "grad_norm": 0.5580615401268005, + "learning_rate": 1.4274000000000001e-05, + "loss": 0.0448, + "step": 4761 + }, + { + "epoch": 4.02195945945946, + "grad_norm": 0.6536552309989929, + "learning_rate": 1.4277e-05, + "loss": 0.0673, + "step": 4762 + }, + { + "epoch": 4.022804054054054, + "grad_norm": 0.5092374086380005, + "learning_rate": 1.428e-05, + "loss": 0.047, + "step": 4763 + }, + { + "epoch": 4.023648648648648, + "grad_norm": 0.33516958355903625, + "learning_rate": 1.4283e-05, + "loss": 0.0334, + "step": 4764 + }, + { + "epoch": 4.024493243243243, + "grad_norm": 0.5772778391838074, + "learning_rate": 1.4286e-05, + "loss": 0.0576, + "step": 4765 + }, + { + "epoch": 4.025337837837838, + "grad_norm": 1.0482308864593506, + "learning_rate": 1.4289e-05, + "loss": 0.0578, + "step": 4766 + }, + { + "epoch": 4.0261824324324325, + "grad_norm": 0.8246139287948608, + "learning_rate": 1.4292e-05, + "loss": 0.0725, + "step": 4767 + }, + { + "epoch": 4.027027027027027, + "grad_norm": 0.596606969833374, + "learning_rate": 1.4295e-05, + "loss": 0.0498, + "step": 4768 + }, + { + "epoch": 4.027871621621622, + "grad_norm": 0.6632739305496216, + "learning_rate": 1.4298000000000001e-05, + "loss": 0.048, + "step": 4769 + }, + { + "epoch": 4.028716216216216, + "grad_norm": 0.4626615345478058, + "learning_rate": 1.4301e-05, + "loss": 0.0574, + "step": 4770 + }, + { + "epoch": 4.0295608108108105, + "grad_norm": 0.5249585509300232, + "learning_rate": 1.4304e-05, + "loss": 0.055, + "step": 4771 + }, + { + "epoch": 4.030405405405405, + "grad_norm": 0.5842897891998291, + "learning_rate": 1.4307e-05, + "loss": 0.0516, + "step": 4772 + }, + { + "epoch": 4.03125, + "grad_norm": 0.8643218874931335, + "learning_rate": 1.431e-05, + "loss": 0.0777, + "step": 4773 + }, + { + "epoch": 4.032094594594595, + "grad_norm": 0.6301710605621338, + "learning_rate": 1.4313000000000002e-05, + "loss": 0.0585, + "step": 4774 + }, + { + "epoch": 4.0329391891891895, + "grad_norm": 0.6994414329528809, + "learning_rate": 1.4316000000000002e-05, + "loss": 0.0693, + "step": 4775 + }, + { + "epoch": 4.033783783783784, + "grad_norm": 0.5855177044868469, + "learning_rate": 1.4319e-05, + "loss": 0.0582, + "step": 4776 + }, + { + "epoch": 4.034628378378378, + "grad_norm": 0.744791567325592, + "learning_rate": 1.4322e-05, + "loss": 0.0754, + "step": 4777 + }, + { + "epoch": 4.035472972972973, + "grad_norm": 0.7024312019348145, + "learning_rate": 1.4325e-05, + "loss": 0.0526, + "step": 4778 + }, + { + "epoch": 4.0363175675675675, + "grad_norm": 0.4448970854282379, + "learning_rate": 1.4328e-05, + "loss": 0.0481, + "step": 4779 + }, + { + "epoch": 4.037162162162162, + "grad_norm": 0.6519774198532104, + "learning_rate": 1.4331e-05, + "loss": 0.0958, + "step": 4780 + }, + { + "epoch": 4.038006756756757, + "grad_norm": 0.7855120301246643, + "learning_rate": 1.4334e-05, + "loss": 0.0651, + "step": 4781 + }, + { + "epoch": 4.038851351351352, + "grad_norm": 0.7072463631629944, + "learning_rate": 1.4337e-05, + "loss": 0.0824, + "step": 4782 + }, + { + "epoch": 4.039695945945946, + "grad_norm": 0.6264456510543823, + "learning_rate": 1.434e-05, + "loss": 0.0649, + "step": 4783 + }, + { + "epoch": 4.04054054054054, + "grad_norm": 0.8227925896644592, + "learning_rate": 1.4343000000000001e-05, + "loss": 0.0862, + "step": 4784 + }, + { + "epoch": 4.041385135135135, + "grad_norm": 0.9634701013565063, + "learning_rate": 1.4346000000000001e-05, + "loss": 0.0876, + "step": 4785 + }, + { + "epoch": 4.04222972972973, + "grad_norm": 1.299111247062683, + "learning_rate": 1.4349000000000001e-05, + "loss": 0.1262, + "step": 4786 + }, + { + "epoch": 4.043074324324325, + "grad_norm": 1.4934135675430298, + "learning_rate": 1.4352e-05, + "loss": 0.3969, + "step": 4787 + }, + { + "epoch": 4.043918918918919, + "grad_norm": 1.3134921789169312, + "learning_rate": 1.4355e-05, + "loss": 0.2704, + "step": 4788 + }, + { + "epoch": 4.044763513513513, + "grad_norm": 0.5573056936264038, + "learning_rate": 1.4358e-05, + "loss": 0.0961, + "step": 4789 + }, + { + "epoch": 4.045608108108108, + "grad_norm": 0.6817238926887512, + "learning_rate": 1.4361e-05, + "loss": 0.0767, + "step": 4790 + }, + { + "epoch": 4.046452702702703, + "grad_norm": 0.5463603734970093, + "learning_rate": 1.4364e-05, + "loss": 0.0624, + "step": 4791 + }, + { + "epoch": 4.047297297297297, + "grad_norm": 0.6186942458152771, + "learning_rate": 1.4367e-05, + "loss": 0.0487, + "step": 4792 + }, + { + "epoch": 4.048141891891892, + "grad_norm": 0.48300275206565857, + "learning_rate": 1.437e-05, + "loss": 0.0658, + "step": 4793 + }, + { + "epoch": 4.048986486486487, + "grad_norm": 0.5022290945053101, + "learning_rate": 1.4373000000000001e-05, + "loss": 0.0581, + "step": 4794 + }, + { + "epoch": 4.049831081081081, + "grad_norm": 0.4919896721839905, + "learning_rate": 1.4376000000000001e-05, + "loss": 0.0615, + "step": 4795 + }, + { + "epoch": 4.050675675675675, + "grad_norm": 0.5936094522476196, + "learning_rate": 1.4379e-05, + "loss": 0.051, + "step": 4796 + }, + { + "epoch": 4.05152027027027, + "grad_norm": 0.44979843497276306, + "learning_rate": 1.4382e-05, + "loss": 0.0619, + "step": 4797 + }, + { + "epoch": 4.052364864864865, + "grad_norm": 0.4704039394855499, + "learning_rate": 1.4385e-05, + "loss": 0.0507, + "step": 4798 + }, + { + "epoch": 4.05320945945946, + "grad_norm": 0.4576042890548706, + "learning_rate": 1.4388000000000002e-05, + "loss": 0.0421, + "step": 4799 + }, + { + "epoch": 4.054054054054054, + "grad_norm": 0.49256542325019836, + "learning_rate": 1.4391000000000002e-05, + "loss": 0.0477, + "step": 4800 + }, + { + "epoch": 4.054898648648648, + "grad_norm": 0.6136898398399353, + "learning_rate": 1.4394e-05, + "loss": 0.0417, + "step": 4801 + }, + { + "epoch": 4.055743243243243, + "grad_norm": 0.5576212406158447, + "learning_rate": 1.4397e-05, + "loss": 0.051, + "step": 4802 + }, + { + "epoch": 4.056587837837838, + "grad_norm": 0.5657324194908142, + "learning_rate": 1.44e-05, + "loss": 0.0442, + "step": 4803 + }, + { + "epoch": 4.0574324324324325, + "grad_norm": 1.0310015678405762, + "learning_rate": 1.4403e-05, + "loss": 0.0563, + "step": 4804 + }, + { + "epoch": 4.058277027027027, + "grad_norm": 0.5960704684257507, + "learning_rate": 1.4406e-05, + "loss": 0.0622, + "step": 4805 + }, + { + "epoch": 4.059121621621622, + "grad_norm": 0.6393682956695557, + "learning_rate": 1.4409e-05, + "loss": 0.0848, + "step": 4806 + }, + { + "epoch": 4.059966216216216, + "grad_norm": 0.39675596356391907, + "learning_rate": 1.4412e-05, + "loss": 0.0442, + "step": 4807 + }, + { + "epoch": 4.0608108108108105, + "grad_norm": 0.7990335822105408, + "learning_rate": 1.4415e-05, + "loss": 0.0571, + "step": 4808 + }, + { + "epoch": 4.061655405405405, + "grad_norm": 0.6425875425338745, + "learning_rate": 1.4418000000000002e-05, + "loss": 0.0672, + "step": 4809 + }, + { + "epoch": 4.0625, + "grad_norm": 0.5819877982139587, + "learning_rate": 1.4421000000000001e-05, + "loss": 0.0706, + "step": 4810 + }, + { + "epoch": 4.063344594594595, + "grad_norm": 0.42553743720054626, + "learning_rate": 1.4424000000000001e-05, + "loss": 0.0317, + "step": 4811 + }, + { + "epoch": 4.0641891891891895, + "grad_norm": 1.0005099773406982, + "learning_rate": 1.4427000000000001e-05, + "loss": 0.055, + "step": 4812 + }, + { + "epoch": 4.065033783783784, + "grad_norm": 0.5327931046485901, + "learning_rate": 1.4429999999999999e-05, + "loss": 0.0467, + "step": 4813 + }, + { + "epoch": 4.065878378378378, + "grad_norm": 0.7563168406486511, + "learning_rate": 1.4433e-05, + "loss": 0.0554, + "step": 4814 + }, + { + "epoch": 4.066722972972973, + "grad_norm": 0.5439898371696472, + "learning_rate": 1.4436e-05, + "loss": 0.0508, + "step": 4815 + }, + { + "epoch": 4.0675675675675675, + "grad_norm": 0.663450300693512, + "learning_rate": 1.4439e-05, + "loss": 0.0718, + "step": 4816 + }, + { + "epoch": 4.068412162162162, + "grad_norm": 0.7261356115341187, + "learning_rate": 1.4442e-05, + "loss": 0.0505, + "step": 4817 + }, + { + "epoch": 4.069256756756757, + "grad_norm": 0.48752689361572266, + "learning_rate": 1.4445e-05, + "loss": 0.0447, + "step": 4818 + }, + { + "epoch": 4.070101351351352, + "grad_norm": 0.5111473798751831, + "learning_rate": 1.4448e-05, + "loss": 0.0477, + "step": 4819 + }, + { + "epoch": 4.070945945945946, + "grad_norm": 1.00303053855896, + "learning_rate": 1.4451000000000001e-05, + "loss": 0.0637, + "step": 4820 + }, + { + "epoch": 4.07179054054054, + "grad_norm": 0.694969892501831, + "learning_rate": 1.4454000000000001e-05, + "loss": 0.0743, + "step": 4821 + }, + { + "epoch": 4.072635135135135, + "grad_norm": 0.4509445130825043, + "learning_rate": 1.4457e-05, + "loss": 0.0397, + "step": 4822 + }, + { + "epoch": 4.07347972972973, + "grad_norm": 0.8516058325767517, + "learning_rate": 1.446e-05, + "loss": 0.0637, + "step": 4823 + }, + { + "epoch": 4.074324324324325, + "grad_norm": 0.992294192314148, + "learning_rate": 1.4463e-05, + "loss": 0.0814, + "step": 4824 + }, + { + "epoch": 4.075168918918919, + "grad_norm": 0.5853064060211182, + "learning_rate": 1.4466e-05, + "loss": 0.0541, + "step": 4825 + }, + { + "epoch": 4.076013513513513, + "grad_norm": 0.623371958732605, + "learning_rate": 1.4469e-05, + "loss": 0.0538, + "step": 4826 + }, + { + "epoch": 4.076858108108108, + "grad_norm": 0.7239195704460144, + "learning_rate": 1.4472e-05, + "loss": 0.0687, + "step": 4827 + }, + { + "epoch": 4.077702702702703, + "grad_norm": 0.45801517367362976, + "learning_rate": 1.4475e-05, + "loss": 0.0387, + "step": 4828 + }, + { + "epoch": 4.078547297297297, + "grad_norm": 0.8891844153404236, + "learning_rate": 1.4478e-05, + "loss": 0.0559, + "step": 4829 + }, + { + "epoch": 4.079391891891892, + "grad_norm": 0.9538540244102478, + "learning_rate": 1.4481e-05, + "loss": 0.0565, + "step": 4830 + }, + { + "epoch": 4.080236486486487, + "grad_norm": 0.9160658121109009, + "learning_rate": 1.4484e-05, + "loss": 0.0791, + "step": 4831 + }, + { + "epoch": 4.081081081081081, + "grad_norm": 0.8498881459236145, + "learning_rate": 1.4487e-05, + "loss": 0.0712, + "step": 4832 + }, + { + "epoch": 4.081925675675675, + "grad_norm": 0.7427356839179993, + "learning_rate": 1.449e-05, + "loss": 0.0532, + "step": 4833 + }, + { + "epoch": 4.08277027027027, + "grad_norm": 0.9204626679420471, + "learning_rate": 1.4493e-05, + "loss": 0.0736, + "step": 4834 + }, + { + "epoch": 4.083614864864865, + "grad_norm": 0.9350427985191345, + "learning_rate": 1.4496000000000001e-05, + "loss": 0.0905, + "step": 4835 + }, + { + "epoch": 4.08445945945946, + "grad_norm": 1.2916715145111084, + "learning_rate": 1.4499000000000001e-05, + "loss": 0.0854, + "step": 4836 + }, + { + "epoch": 4.085304054054054, + "grad_norm": 2.0289950370788574, + "learning_rate": 1.4502000000000001e-05, + "loss": 0.4003, + "step": 4837 + }, + { + "epoch": 4.086148648648648, + "grad_norm": 0.9566875696182251, + "learning_rate": 1.4505e-05, + "loss": 0.2565, + "step": 4838 + }, + { + "epoch": 4.086993243243243, + "grad_norm": 1.4430283308029175, + "learning_rate": 1.4507999999999999e-05, + "loss": 0.1285, + "step": 4839 + }, + { + "epoch": 4.087837837837838, + "grad_norm": 1.0399959087371826, + "learning_rate": 1.4511e-05, + "loss": 0.1123, + "step": 4840 + }, + { + "epoch": 4.0886824324324325, + "grad_norm": 0.5435648560523987, + "learning_rate": 1.4514e-05, + "loss": 0.0606, + "step": 4841 + }, + { + "epoch": 4.089527027027027, + "grad_norm": 0.6230497360229492, + "learning_rate": 1.4517e-05, + "loss": 0.0908, + "step": 4842 + }, + { + "epoch": 4.090371621621622, + "grad_norm": 0.5741007328033447, + "learning_rate": 1.452e-05, + "loss": 0.0845, + "step": 4843 + }, + { + "epoch": 4.091216216216216, + "grad_norm": 0.5218200087547302, + "learning_rate": 1.4523e-05, + "loss": 0.0658, + "step": 4844 + }, + { + "epoch": 4.0920608108108105, + "grad_norm": 0.5230473875999451, + "learning_rate": 1.4526000000000001e-05, + "loss": 0.0454, + "step": 4845 + }, + { + "epoch": 4.092905405405405, + "grad_norm": 0.468885600566864, + "learning_rate": 1.4529000000000001e-05, + "loss": 0.051, + "step": 4846 + }, + { + "epoch": 4.09375, + "grad_norm": 0.49080193042755127, + "learning_rate": 1.4532e-05, + "loss": 0.0478, + "step": 4847 + }, + { + "epoch": 4.094594594594595, + "grad_norm": 0.6727127432823181, + "learning_rate": 1.4535e-05, + "loss": 0.0693, + "step": 4848 + }, + { + "epoch": 4.0954391891891895, + "grad_norm": 0.5081314444541931, + "learning_rate": 1.4538e-05, + "loss": 0.0538, + "step": 4849 + }, + { + "epoch": 4.096283783783784, + "grad_norm": 0.5993751287460327, + "learning_rate": 1.4541e-05, + "loss": 0.0584, + "step": 4850 + }, + { + "epoch": 4.097128378378378, + "grad_norm": 0.4310128092765808, + "learning_rate": 1.4544e-05, + "loss": 0.046, + "step": 4851 + }, + { + "epoch": 4.097972972972973, + "grad_norm": 0.7385318875312805, + "learning_rate": 1.4547e-05, + "loss": 0.0594, + "step": 4852 + }, + { + "epoch": 4.0988175675675675, + "grad_norm": 0.5979647040367126, + "learning_rate": 1.455e-05, + "loss": 0.0471, + "step": 4853 + }, + { + "epoch": 4.099662162162162, + "grad_norm": 0.5388291478157043, + "learning_rate": 1.4553e-05, + "loss": 0.0692, + "step": 4854 + }, + { + "epoch": 4.100506756756757, + "grad_norm": 0.7341147661209106, + "learning_rate": 1.4556000000000001e-05, + "loss": 0.0583, + "step": 4855 + }, + { + "epoch": 4.101351351351352, + "grad_norm": 0.663439929485321, + "learning_rate": 1.4559e-05, + "loss": 0.0813, + "step": 4856 + }, + { + "epoch": 4.102195945945946, + "grad_norm": 0.9023957252502441, + "learning_rate": 1.4562e-05, + "loss": 0.0561, + "step": 4857 + }, + { + "epoch": 4.10304054054054, + "grad_norm": 0.6621078848838806, + "learning_rate": 1.4565e-05, + "loss": 0.0568, + "step": 4858 + }, + { + "epoch": 4.103885135135135, + "grad_norm": 0.6865058541297913, + "learning_rate": 1.4568e-05, + "loss": 0.0591, + "step": 4859 + }, + { + "epoch": 4.10472972972973, + "grad_norm": 0.5354077816009521, + "learning_rate": 1.4571000000000002e-05, + "loss": 0.055, + "step": 4860 + }, + { + "epoch": 4.105574324324325, + "grad_norm": 0.5561379194259644, + "learning_rate": 1.4574000000000001e-05, + "loss": 0.0663, + "step": 4861 + }, + { + "epoch": 4.106418918918919, + "grad_norm": 0.6218866109848022, + "learning_rate": 1.4577e-05, + "loss": 0.0681, + "step": 4862 + }, + { + "epoch": 4.107263513513513, + "grad_norm": 0.6575968861579895, + "learning_rate": 1.458e-05, + "loss": 0.0658, + "step": 4863 + }, + { + "epoch": 4.108108108108108, + "grad_norm": 0.5164707899093628, + "learning_rate": 1.4582999999999999e-05, + "loss": 0.0404, + "step": 4864 + }, + { + "epoch": 4.108952702702703, + "grad_norm": 0.8400449156761169, + "learning_rate": 1.4586e-05, + "loss": 0.0583, + "step": 4865 + }, + { + "epoch": 4.109797297297297, + "grad_norm": 0.5322386622428894, + "learning_rate": 1.4589e-05, + "loss": 0.0463, + "step": 4866 + }, + { + "epoch": 4.110641891891892, + "grad_norm": 0.8856278657913208, + "learning_rate": 1.4592e-05, + "loss": 0.0512, + "step": 4867 + }, + { + "epoch": 4.111486486486487, + "grad_norm": 0.5282100439071655, + "learning_rate": 1.4595e-05, + "loss": 0.0484, + "step": 4868 + }, + { + "epoch": 4.112331081081081, + "grad_norm": 0.47450384497642517, + "learning_rate": 1.4598e-05, + "loss": 0.0315, + "step": 4869 + }, + { + "epoch": 4.113175675675675, + "grad_norm": 0.8285987973213196, + "learning_rate": 1.4601000000000001e-05, + "loss": 0.0734, + "step": 4870 + }, + { + "epoch": 4.11402027027027, + "grad_norm": 0.7544177770614624, + "learning_rate": 1.4604000000000001e-05, + "loss": 0.0541, + "step": 4871 + }, + { + "epoch": 4.114864864864865, + "grad_norm": 0.5830057263374329, + "learning_rate": 1.4607000000000001e-05, + "loss": 0.0477, + "step": 4872 + }, + { + "epoch": 4.11570945945946, + "grad_norm": 0.9400501251220703, + "learning_rate": 1.461e-05, + "loss": 0.1103, + "step": 4873 + }, + { + "epoch": 4.116554054054054, + "grad_norm": 0.545468270778656, + "learning_rate": 1.4613e-05, + "loss": 0.0413, + "step": 4874 + }, + { + "epoch": 4.117398648648648, + "grad_norm": 0.889522910118103, + "learning_rate": 1.4616e-05, + "loss": 0.0756, + "step": 4875 + }, + { + "epoch": 4.118243243243243, + "grad_norm": 0.8450178503990173, + "learning_rate": 1.4619e-05, + "loss": 0.0648, + "step": 4876 + }, + { + "epoch": 4.119087837837838, + "grad_norm": 0.8452561497688293, + "learning_rate": 1.4622e-05, + "loss": 0.0732, + "step": 4877 + }, + { + "epoch": 4.1199324324324325, + "grad_norm": 0.6517895460128784, + "learning_rate": 1.4625e-05, + "loss": 0.0483, + "step": 4878 + }, + { + "epoch": 4.120777027027027, + "grad_norm": 0.7837688326835632, + "learning_rate": 1.4628e-05, + "loss": 0.053, + "step": 4879 + }, + { + "epoch": 4.121621621621622, + "grad_norm": 1.2283865213394165, + "learning_rate": 1.4631000000000001e-05, + "loss": 0.0859, + "step": 4880 + }, + { + "epoch": 4.122466216216216, + "grad_norm": 0.7784660458564758, + "learning_rate": 1.4634e-05, + "loss": 0.0683, + "step": 4881 + }, + { + "epoch": 4.1233108108108105, + "grad_norm": 1.2542133331298828, + "learning_rate": 1.4637e-05, + "loss": 0.1121, + "step": 4882 + }, + { + "epoch": 4.124155405405405, + "grad_norm": 0.6868534684181213, + "learning_rate": 1.464e-05, + "loss": 0.0609, + "step": 4883 + }, + { + "epoch": 4.125, + "grad_norm": 1.1894160509109497, + "learning_rate": 1.4643e-05, + "loss": 0.0835, + "step": 4884 + }, + { + "epoch": 4.125844594594595, + "grad_norm": 2.4752016067504883, + "learning_rate": 1.4646000000000002e-05, + "loss": 0.1231, + "step": 4885 + }, + { + "epoch": 4.1266891891891895, + "grad_norm": 1.1996315717697144, + "learning_rate": 1.4649000000000002e-05, + "loss": 0.0983, + "step": 4886 + }, + { + "epoch": 4.127533783783784, + "grad_norm": 1.8493499755859375, + "learning_rate": 1.4652e-05, + "loss": 0.3968, + "step": 4887 + }, + { + "epoch": 4.128378378378378, + "grad_norm": 0.8934518098831177, + "learning_rate": 1.4655e-05, + "loss": 0.2099, + "step": 4888 + }, + { + "epoch": 4.129222972972973, + "grad_norm": 0.6042224168777466, + "learning_rate": 1.4658e-05, + "loss": 0.1173, + "step": 4889 + }, + { + "epoch": 4.1300675675675675, + "grad_norm": 0.7328402400016785, + "learning_rate": 1.4661e-05, + "loss": 0.1089, + "step": 4890 + }, + { + "epoch": 4.130912162162162, + "grad_norm": 0.6784640550613403, + "learning_rate": 1.4664e-05, + "loss": 0.0853, + "step": 4891 + }, + { + "epoch": 4.131756756756757, + "grad_norm": 0.4118046164512634, + "learning_rate": 1.4667e-05, + "loss": 0.0517, + "step": 4892 + }, + { + "epoch": 4.132601351351352, + "grad_norm": 0.8033551573753357, + "learning_rate": 1.467e-05, + "loss": 0.0752, + "step": 4893 + }, + { + "epoch": 4.133445945945946, + "grad_norm": 0.4541693329811096, + "learning_rate": 1.4673e-05, + "loss": 0.0472, + "step": 4894 + }, + { + "epoch": 4.13429054054054, + "grad_norm": 0.6021104454994202, + "learning_rate": 1.4676000000000001e-05, + "loss": 0.0816, + "step": 4895 + }, + { + "epoch": 4.135135135135135, + "grad_norm": 0.6223790645599365, + "learning_rate": 1.4679000000000001e-05, + "loss": 0.0711, + "step": 4896 + }, + { + "epoch": 4.13597972972973, + "grad_norm": 0.5320374965667725, + "learning_rate": 1.4682000000000001e-05, + "loss": 0.0603, + "step": 4897 + }, + { + "epoch": 4.136824324324325, + "grad_norm": 0.41336214542388916, + "learning_rate": 1.4685000000000001e-05, + "loss": 0.062, + "step": 4898 + }, + { + "epoch": 4.137668918918919, + "grad_norm": 0.5913623571395874, + "learning_rate": 1.4687999999999999e-05, + "loss": 0.066, + "step": 4899 + }, + { + "epoch": 4.138513513513513, + "grad_norm": 0.5456079244613647, + "learning_rate": 1.4691e-05, + "loss": 0.053, + "step": 4900 + }, + { + "epoch": 4.139358108108108, + "grad_norm": 0.6346492767333984, + "learning_rate": 1.4694e-05, + "loss": 0.0855, + "step": 4901 + }, + { + "epoch": 4.140202702702703, + "grad_norm": 0.5462587475776672, + "learning_rate": 1.4697e-05, + "loss": 0.0552, + "step": 4902 + }, + { + "epoch": 4.141047297297297, + "grad_norm": 0.6149448752403259, + "learning_rate": 1.47e-05, + "loss": 0.0708, + "step": 4903 + }, + { + "epoch": 4.141891891891892, + "grad_norm": 0.7794519066810608, + "learning_rate": 1.4703e-05, + "loss": 0.0844, + "step": 4904 + }, + { + "epoch": 4.142736486486487, + "grad_norm": 0.505948543548584, + "learning_rate": 1.4706000000000001e-05, + "loss": 0.0559, + "step": 4905 + }, + { + "epoch": 4.143581081081081, + "grad_norm": 0.4051482081413269, + "learning_rate": 1.4709000000000001e-05, + "loss": 0.0468, + "step": 4906 + }, + { + "epoch": 4.144425675675675, + "grad_norm": 0.7706531286239624, + "learning_rate": 1.4712e-05, + "loss": 0.0847, + "step": 4907 + }, + { + "epoch": 4.14527027027027, + "grad_norm": 0.5022884607315063, + "learning_rate": 1.4715e-05, + "loss": 0.0495, + "step": 4908 + }, + { + "epoch": 4.146114864864865, + "grad_norm": 0.5237668752670288, + "learning_rate": 1.4718e-05, + "loss": 0.0633, + "step": 4909 + }, + { + "epoch": 4.14695945945946, + "grad_norm": 0.5789461731910706, + "learning_rate": 1.4721000000000002e-05, + "loss": 0.0514, + "step": 4910 + }, + { + "epoch": 4.147804054054054, + "grad_norm": 0.5383216142654419, + "learning_rate": 1.4724e-05, + "loss": 0.0592, + "step": 4911 + }, + { + "epoch": 4.148648648648648, + "grad_norm": 0.5594640970230103, + "learning_rate": 1.4727e-05, + "loss": 0.0504, + "step": 4912 + }, + { + "epoch": 4.149493243243243, + "grad_norm": 0.6977416276931763, + "learning_rate": 1.473e-05, + "loss": 0.0881, + "step": 4913 + }, + { + "epoch": 4.150337837837838, + "grad_norm": 0.663324236869812, + "learning_rate": 1.4733e-05, + "loss": 0.0623, + "step": 4914 + }, + { + "epoch": 4.1511824324324325, + "grad_norm": 0.4850424826145172, + "learning_rate": 1.4736000000000001e-05, + "loss": 0.0386, + "step": 4915 + }, + { + "epoch": 4.152027027027027, + "grad_norm": 0.7780945897102356, + "learning_rate": 1.4739e-05, + "loss": 0.0601, + "step": 4916 + }, + { + "epoch": 4.152871621621622, + "grad_norm": 0.7947589755058289, + "learning_rate": 1.4742e-05, + "loss": 0.0629, + "step": 4917 + }, + { + "epoch": 4.153716216216216, + "grad_norm": 0.7742847800254822, + "learning_rate": 1.4745e-05, + "loss": 0.0816, + "step": 4918 + }, + { + "epoch": 4.1545608108108105, + "grad_norm": 0.6817547082901001, + "learning_rate": 1.4748e-05, + "loss": 0.0631, + "step": 4919 + }, + { + "epoch": 4.155405405405405, + "grad_norm": 0.9340263605117798, + "learning_rate": 1.4751000000000002e-05, + "loss": 0.0606, + "step": 4920 + }, + { + "epoch": 4.15625, + "grad_norm": 1.1250633001327515, + "learning_rate": 1.4754000000000001e-05, + "loss": 0.0771, + "step": 4921 + }, + { + "epoch": 4.157094594594595, + "grad_norm": 0.8580924272537231, + "learning_rate": 1.4757000000000001e-05, + "loss": 0.0649, + "step": 4922 + }, + { + "epoch": 4.1579391891891895, + "grad_norm": 0.758780837059021, + "learning_rate": 1.4760000000000001e-05, + "loss": 0.0536, + "step": 4923 + }, + { + "epoch": 4.158783783783784, + "grad_norm": 0.5670645236968994, + "learning_rate": 1.4762999999999999e-05, + "loss": 0.0425, + "step": 4924 + }, + { + "epoch": 4.159628378378378, + "grad_norm": 0.5339645743370056, + "learning_rate": 1.4766e-05, + "loss": 0.0471, + "step": 4925 + }, + { + "epoch": 4.160472972972973, + "grad_norm": 0.5180341601371765, + "learning_rate": 1.4769e-05, + "loss": 0.0523, + "step": 4926 + }, + { + "epoch": 4.1613175675675675, + "grad_norm": 0.9787354469299316, + "learning_rate": 1.4772e-05, + "loss": 0.0685, + "step": 4927 + }, + { + "epoch": 4.162162162162162, + "grad_norm": 0.6979410648345947, + "learning_rate": 1.4775e-05, + "loss": 0.0691, + "step": 4928 + }, + { + "epoch": 4.163006756756757, + "grad_norm": 0.6033633351325989, + "learning_rate": 1.4778e-05, + "loss": 0.051, + "step": 4929 + }, + { + "epoch": 4.163851351351352, + "grad_norm": 0.55157470703125, + "learning_rate": 1.4781000000000001e-05, + "loss": 0.0537, + "step": 4930 + }, + { + "epoch": 4.164695945945946, + "grad_norm": 0.5504067540168762, + "learning_rate": 1.4784000000000001e-05, + "loss": 0.0615, + "step": 4931 + }, + { + "epoch": 4.16554054054054, + "grad_norm": 0.7136526107788086, + "learning_rate": 1.4787000000000001e-05, + "loss": 0.0778, + "step": 4932 + }, + { + "epoch": 4.166385135135135, + "grad_norm": 0.8927635550498962, + "learning_rate": 1.479e-05, + "loss": 0.0622, + "step": 4933 + }, + { + "epoch": 4.16722972972973, + "grad_norm": 1.337172269821167, + "learning_rate": 1.4793e-05, + "loss": 0.0713, + "step": 4934 + }, + { + "epoch": 4.168074324324325, + "grad_norm": 0.9616535305976868, + "learning_rate": 1.4796000000000002e-05, + "loss": 0.0887, + "step": 4935 + }, + { + "epoch": 4.168918918918919, + "grad_norm": 1.5765537023544312, + "learning_rate": 1.4799e-05, + "loss": 0.1396, + "step": 4936 + }, + { + "epoch": 4.169763513513513, + "grad_norm": 1.786867380142212, + "learning_rate": 1.4802e-05, + "loss": 0.4942, + "step": 4937 + }, + { + "epoch": 4.170608108108108, + "grad_norm": 1.3523914813995361, + "learning_rate": 1.4805e-05, + "loss": 0.2903, + "step": 4938 + }, + { + "epoch": 4.171452702702703, + "grad_norm": 0.8626089096069336, + "learning_rate": 1.4808e-05, + "loss": 0.1514, + "step": 4939 + }, + { + "epoch": 4.172297297297297, + "grad_norm": 0.9175999760627747, + "learning_rate": 1.4811000000000001e-05, + "loss": 0.1211, + "step": 4940 + }, + { + "epoch": 4.173141891891892, + "grad_norm": 1.0036052465438843, + "learning_rate": 1.4814e-05, + "loss": 0.081, + "step": 4941 + }, + { + "epoch": 4.173986486486487, + "grad_norm": 0.6865316033363342, + "learning_rate": 1.4817e-05, + "loss": 0.0656, + "step": 4942 + }, + { + "epoch": 4.174831081081081, + "grad_norm": 0.7496147155761719, + "learning_rate": 1.482e-05, + "loss": 0.0642, + "step": 4943 + }, + { + "epoch": 4.175675675675675, + "grad_norm": 0.4832127094268799, + "learning_rate": 1.4823e-05, + "loss": 0.0521, + "step": 4944 + }, + { + "epoch": 4.17652027027027, + "grad_norm": 0.48372599482536316, + "learning_rate": 1.4826e-05, + "loss": 0.0619, + "step": 4945 + }, + { + "epoch": 4.177364864864865, + "grad_norm": 0.5148786306381226, + "learning_rate": 1.4829000000000002e-05, + "loss": 0.0515, + "step": 4946 + }, + { + "epoch": 4.17820945945946, + "grad_norm": 0.891856849193573, + "learning_rate": 1.4832000000000001e-05, + "loss": 0.0683, + "step": 4947 + }, + { + "epoch": 4.179054054054054, + "grad_norm": 0.4710070788860321, + "learning_rate": 1.4835e-05, + "loss": 0.0459, + "step": 4948 + }, + { + "epoch": 4.179898648648648, + "grad_norm": 0.6386625170707703, + "learning_rate": 1.4838e-05, + "loss": 0.0504, + "step": 4949 + }, + { + "epoch": 4.180743243243243, + "grad_norm": 0.6237409114837646, + "learning_rate": 1.4840999999999999e-05, + "loss": 0.0587, + "step": 4950 + }, + { + "epoch": 4.181587837837838, + "grad_norm": 0.9041457176208496, + "learning_rate": 1.4844e-05, + "loss": 0.0524, + "step": 4951 + }, + { + "epoch": 4.1824324324324325, + "grad_norm": 0.7410102486610413, + "learning_rate": 1.4847e-05, + "loss": 0.0578, + "step": 4952 + }, + { + "epoch": 4.183277027027027, + "grad_norm": 0.5459177494049072, + "learning_rate": 1.485e-05, + "loss": 0.0593, + "step": 4953 + }, + { + "epoch": 4.184121621621622, + "grad_norm": 0.7869409322738647, + "learning_rate": 1.4853e-05, + "loss": 0.0668, + "step": 4954 + }, + { + "epoch": 4.184966216216216, + "grad_norm": 0.5658491849899292, + "learning_rate": 1.4856e-05, + "loss": 0.0542, + "step": 4955 + }, + { + "epoch": 4.1858108108108105, + "grad_norm": 0.9238007068634033, + "learning_rate": 1.4859000000000001e-05, + "loss": 0.0689, + "step": 4956 + }, + { + "epoch": 4.186655405405405, + "grad_norm": 0.6523988842964172, + "learning_rate": 1.4862000000000001e-05, + "loss": 0.0656, + "step": 4957 + }, + { + "epoch": 4.1875, + "grad_norm": 0.895550549030304, + "learning_rate": 1.4865e-05, + "loss": 0.0681, + "step": 4958 + }, + { + "epoch": 4.188344594594595, + "grad_norm": 0.7100493907928467, + "learning_rate": 1.4868e-05, + "loss": 0.0718, + "step": 4959 + }, + { + "epoch": 4.1891891891891895, + "grad_norm": 0.5400139093399048, + "learning_rate": 1.4871e-05, + "loss": 0.0396, + "step": 4960 + }, + { + "epoch": 4.190033783783784, + "grad_norm": 0.5174754858016968, + "learning_rate": 1.4874e-05, + "loss": 0.0455, + "step": 4961 + }, + { + "epoch": 4.190878378378378, + "grad_norm": 0.5339234471321106, + "learning_rate": 1.4877e-05, + "loss": 0.0511, + "step": 4962 + }, + { + "epoch": 4.191722972972973, + "grad_norm": 0.6445474028587341, + "learning_rate": 1.488e-05, + "loss": 0.062, + "step": 4963 + }, + { + "epoch": 4.1925675675675675, + "grad_norm": 0.7873409390449524, + "learning_rate": 1.4883e-05, + "loss": 0.0702, + "step": 4964 + }, + { + "epoch": 4.193412162162162, + "grad_norm": 0.5140800476074219, + "learning_rate": 1.4886e-05, + "loss": 0.0422, + "step": 4965 + }, + { + "epoch": 4.194256756756757, + "grad_norm": 0.9375565648078918, + "learning_rate": 1.4889000000000001e-05, + "loss": 0.068, + "step": 4966 + }, + { + "epoch": 4.195101351351352, + "grad_norm": 0.6081218123435974, + "learning_rate": 1.4892e-05, + "loss": 0.0687, + "step": 4967 + }, + { + "epoch": 4.195945945945946, + "grad_norm": 0.7428207993507385, + "learning_rate": 1.4895e-05, + "loss": 0.0636, + "step": 4968 + }, + { + "epoch": 4.19679054054054, + "grad_norm": 0.47065961360931396, + "learning_rate": 1.4898e-05, + "loss": 0.0362, + "step": 4969 + }, + { + "epoch": 4.197635135135135, + "grad_norm": 1.5028847455978394, + "learning_rate": 1.4901e-05, + "loss": 0.0659, + "step": 4970 + }, + { + "epoch": 4.19847972972973, + "grad_norm": 0.5672389268875122, + "learning_rate": 1.4904000000000002e-05, + "loss": 0.0525, + "step": 4971 + }, + { + "epoch": 4.199324324324325, + "grad_norm": 0.5397771000862122, + "learning_rate": 1.4907000000000001e-05, + "loss": 0.0292, + "step": 4972 + }, + { + "epoch": 4.200168918918919, + "grad_norm": 0.7091238498687744, + "learning_rate": 1.491e-05, + "loss": 0.0586, + "step": 4973 + }, + { + "epoch": 4.201013513513513, + "grad_norm": 0.9190642833709717, + "learning_rate": 1.4913e-05, + "loss": 0.0688, + "step": 4974 + }, + { + "epoch": 4.201858108108108, + "grad_norm": 0.4657061994075775, + "learning_rate": 1.4915999999999999e-05, + "loss": 0.0294, + "step": 4975 + }, + { + "epoch": 4.202702702702703, + "grad_norm": 0.7958877682685852, + "learning_rate": 1.4919e-05, + "loss": 0.0551, + "step": 4976 + }, + { + "epoch": 4.203547297297297, + "grad_norm": 0.762326180934906, + "learning_rate": 1.4922e-05, + "loss": 0.0615, + "step": 4977 + }, + { + "epoch": 4.204391891891892, + "grad_norm": 0.678066611289978, + "learning_rate": 1.4925e-05, + "loss": 0.0421, + "step": 4978 + }, + { + "epoch": 4.205236486486487, + "grad_norm": 0.8839058876037598, + "learning_rate": 1.4928e-05, + "loss": 0.0646, + "step": 4979 + }, + { + "epoch": 4.206081081081081, + "grad_norm": 0.6959027647972107, + "learning_rate": 1.4931e-05, + "loss": 0.0665, + "step": 4980 + }, + { + "epoch": 4.206925675675675, + "grad_norm": 1.2680926322937012, + "learning_rate": 1.4934000000000001e-05, + "loss": 0.0848, + "step": 4981 + }, + { + "epoch": 4.20777027027027, + "grad_norm": 1.3186136484146118, + "learning_rate": 1.4937000000000001e-05, + "loss": 0.1025, + "step": 4982 + }, + { + "epoch": 4.208614864864865, + "grad_norm": 0.7894458770751953, + "learning_rate": 1.4940000000000001e-05, + "loss": 0.0748, + "step": 4983 + }, + { + "epoch": 4.20945945945946, + "grad_norm": 0.8671680688858032, + "learning_rate": 1.4943e-05, + "loss": 0.0845, + "step": 4984 + }, + { + "epoch": 4.210304054054054, + "grad_norm": 0.9941083192825317, + "learning_rate": 1.4945999999999999e-05, + "loss": 0.079, + "step": 4985 + }, + { + "epoch": 4.211148648648648, + "grad_norm": 1.377968430519104, + "learning_rate": 1.4949e-05, + "loss": 0.131, + "step": 4986 + }, + { + "epoch": 4.211993243243243, + "grad_norm": 1.2127043008804321, + "learning_rate": 1.4952e-05, + "loss": 0.3152, + "step": 4987 + }, + { + "epoch": 4.212837837837838, + "grad_norm": 1.40300714969635, + "learning_rate": 1.4955e-05, + "loss": 0.3596, + "step": 4988 + }, + { + "epoch": 4.2136824324324325, + "grad_norm": 0.8045386075973511, + "learning_rate": 1.4958e-05, + "loss": 0.1751, + "step": 4989 + }, + { + "epoch": 4.214527027027027, + "grad_norm": 0.7932912707328796, + "learning_rate": 1.4961e-05, + "loss": 0.1078, + "step": 4990 + }, + { + "epoch": 4.215371621621622, + "grad_norm": 0.4319070279598236, + "learning_rate": 1.4964000000000001e-05, + "loss": 0.0543, + "step": 4991 + }, + { + "epoch": 4.216216216216216, + "grad_norm": 0.6449800133705139, + "learning_rate": 1.4967000000000001e-05, + "loss": 0.0887, + "step": 4992 + }, + { + "epoch": 4.2170608108108105, + "grad_norm": 0.47273075580596924, + "learning_rate": 1.497e-05, + "loss": 0.0703, + "step": 4993 + }, + { + "epoch": 4.217905405405405, + "grad_norm": 0.6631123423576355, + "learning_rate": 1.4973e-05, + "loss": 0.0548, + "step": 4994 + }, + { + "epoch": 4.21875, + "grad_norm": 0.5409554839134216, + "learning_rate": 1.4976e-05, + "loss": 0.0669, + "step": 4995 + }, + { + "epoch": 4.219594594594595, + "grad_norm": 0.5482361912727356, + "learning_rate": 1.4979000000000002e-05, + "loss": 0.0546, + "step": 4996 + }, + { + "epoch": 4.2204391891891895, + "grad_norm": 0.42345497012138367, + "learning_rate": 1.4982e-05, + "loss": 0.0461, + "step": 4997 + }, + { + "epoch": 4.221283783783784, + "grad_norm": 0.6804132461547852, + "learning_rate": 1.4985e-05, + "loss": 0.0678, + "step": 4998 + }, + { + "epoch": 4.222128378378378, + "grad_norm": 0.43267789483070374, + "learning_rate": 1.4988e-05, + "loss": 0.0377, + "step": 4999 + }, + { + "epoch": 4.222972972972973, + "grad_norm": 0.49326592683792114, + "learning_rate": 1.4991e-05, + "loss": 0.036, + "step": 5000 + }, + { + "epoch": 4.222972972972973, + "eval_cer": 0.11605493534655852, + "eval_loss": 0.3765541613101959, + "eval_runtime": 8.4928, + "eval_samples_per_second": 11.892, + "eval_steps_per_second": 0.471, + "eval_wer": 0.345679012345679, + "step": 5000 + }, + { + "epoch": 4.2238175675675675, + "grad_norm": 0.5448799133300781, + "learning_rate": 1.4994e-05, + "loss": 0.0387, + "step": 5001 + }, + { + "epoch": 4.224662162162162, + "grad_norm": 0.5545806884765625, + "learning_rate": 1.4997e-05, + "loss": 0.0536, + "step": 5002 + }, + { + "epoch": 4.225506756756757, + "grad_norm": 0.42051300406455994, + "learning_rate": 1.5e-05, + "loss": 0.0406, + "step": 5003 + }, + { + "epoch": 4.226351351351352, + "grad_norm": 0.5478152632713318, + "learning_rate": 1.5003e-05, + "loss": 0.0332, + "step": 5004 + }, + { + "epoch": 4.227195945945946, + "grad_norm": 0.7195851802825928, + "learning_rate": 1.5006e-05, + "loss": 0.0456, + "step": 5005 + }, + { + "epoch": 4.22804054054054, + "grad_norm": 0.5017030239105225, + "learning_rate": 1.5009e-05, + "loss": 0.0386, + "step": 5006 + }, + { + "epoch": 4.228885135135135, + "grad_norm": 1.0524559020996094, + "learning_rate": 1.5012e-05, + "loss": 0.0656, + "step": 5007 + }, + { + "epoch": 4.22972972972973, + "grad_norm": 0.4443204998970032, + "learning_rate": 1.5015e-05, + "loss": 0.0411, + "step": 5008 + }, + { + "epoch": 4.230574324324325, + "grad_norm": 0.3886638581752777, + "learning_rate": 1.5018000000000001e-05, + "loss": 0.0362, + "step": 5009 + }, + { + "epoch": 4.231418918918919, + "grad_norm": 0.5084425210952759, + "learning_rate": 1.5021e-05, + "loss": 0.043, + "step": 5010 + }, + { + "epoch": 4.232263513513513, + "grad_norm": 0.5433184504508972, + "learning_rate": 1.5024e-05, + "loss": 0.0539, + "step": 5011 + }, + { + "epoch": 4.233108108108108, + "grad_norm": 0.7179712653160095, + "learning_rate": 1.5027e-05, + "loss": 0.0666, + "step": 5012 + }, + { + "epoch": 4.233952702702703, + "grad_norm": 0.8574289083480835, + "learning_rate": 1.503e-05, + "loss": 0.0464, + "step": 5013 + }, + { + "epoch": 4.234797297297297, + "grad_norm": 0.8178863525390625, + "learning_rate": 1.5033e-05, + "loss": 0.05, + "step": 5014 + }, + { + "epoch": 4.235641891891892, + "grad_norm": 0.6429744362831116, + "learning_rate": 1.5036e-05, + "loss": 0.0497, + "step": 5015 + }, + { + "epoch": 4.236486486486487, + "grad_norm": 0.6776488423347473, + "learning_rate": 1.5039e-05, + "loss": 0.0436, + "step": 5016 + }, + { + "epoch": 4.237331081081081, + "grad_norm": 0.6507218480110168, + "learning_rate": 1.5042e-05, + "loss": 0.0687, + "step": 5017 + }, + { + "epoch": 4.238175675675675, + "grad_norm": 0.584656298160553, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.065, + "step": 5018 + }, + { + "epoch": 4.23902027027027, + "grad_norm": 0.7508190870285034, + "learning_rate": 1.5048000000000002e-05, + "loss": 0.0513, + "step": 5019 + }, + { + "epoch": 4.239864864864865, + "grad_norm": 0.7809785604476929, + "learning_rate": 1.5051000000000002e-05, + "loss": 0.0652, + "step": 5020 + }, + { + "epoch": 4.24070945945946, + "grad_norm": 0.5846214294433594, + "learning_rate": 1.5054000000000002e-05, + "loss": 0.0595, + "step": 5021 + }, + { + "epoch": 4.241554054054054, + "grad_norm": 0.49419909715652466, + "learning_rate": 1.5057e-05, + "loss": 0.0539, + "step": 5022 + }, + { + "epoch": 4.242398648648648, + "grad_norm": 0.4016555845737457, + "learning_rate": 1.506e-05, + "loss": 0.0449, + "step": 5023 + }, + { + "epoch": 4.243243243243243, + "grad_norm": 0.713300883769989, + "learning_rate": 1.5063e-05, + "loss": 0.069, + "step": 5024 + }, + { + "epoch": 4.244087837837838, + "grad_norm": 0.589903712272644, + "learning_rate": 1.5066e-05, + "loss": 0.0547, + "step": 5025 + }, + { + "epoch": 4.2449324324324325, + "grad_norm": 0.6505820751190186, + "learning_rate": 1.5069e-05, + "loss": 0.0638, + "step": 5026 + }, + { + "epoch": 4.245777027027027, + "grad_norm": 0.9928831458091736, + "learning_rate": 1.5071999999999999e-05, + "loss": 0.0586, + "step": 5027 + }, + { + "epoch": 4.246621621621622, + "grad_norm": 0.8662407994270325, + "learning_rate": 1.5074999999999999e-05, + "loss": 0.0715, + "step": 5028 + }, + { + "epoch": 4.247466216216216, + "grad_norm": 0.9342048168182373, + "learning_rate": 1.5078000000000002e-05, + "loss": 0.0953, + "step": 5029 + }, + { + "epoch": 4.2483108108108105, + "grad_norm": 0.5597781538963318, + "learning_rate": 1.5081000000000002e-05, + "loss": 0.051, + "step": 5030 + }, + { + "epoch": 4.249155405405405, + "grad_norm": 1.2447429895401, + "learning_rate": 1.5084000000000002e-05, + "loss": 0.0616, + "step": 5031 + }, + { + "epoch": 4.25, + "grad_norm": 1.4213801622390747, + "learning_rate": 1.5087000000000001e-05, + "loss": 0.1033, + "step": 5032 + }, + { + "epoch": 4.250844594594595, + "grad_norm": 0.6617791056632996, + "learning_rate": 1.5090000000000001e-05, + "loss": 0.0534, + "step": 5033 + }, + { + "epoch": 4.2516891891891895, + "grad_norm": 1.0475573539733887, + "learning_rate": 1.5093e-05, + "loss": 0.1285, + "step": 5034 + }, + { + "epoch": 4.252533783783784, + "grad_norm": 1.013972520828247, + "learning_rate": 1.5095999999999999e-05, + "loss": 0.0963, + "step": 5035 + }, + { + "epoch": 4.253378378378378, + "grad_norm": 1.5534616708755493, + "learning_rate": 1.5098999999999999e-05, + "loss": 0.1156, + "step": 5036 + }, + { + "epoch": 4.254222972972973, + "grad_norm": 3.887407064437866, + "learning_rate": 1.5101999999999999e-05, + "loss": 0.4834, + "step": 5037 + }, + { + "epoch": 4.2550675675675675, + "grad_norm": 1.02928626537323, + "learning_rate": 1.5104999999999999e-05, + "loss": 0.2411, + "step": 5038 + }, + { + "epoch": 4.255912162162162, + "grad_norm": 1.3928227424621582, + "learning_rate": 1.5108000000000002e-05, + "loss": 0.1425, + "step": 5039 + }, + { + "epoch": 4.256756756756757, + "grad_norm": 0.7450975775718689, + "learning_rate": 1.5111000000000002e-05, + "loss": 0.0919, + "step": 5040 + }, + { + "epoch": 4.257601351351352, + "grad_norm": 0.7682592272758484, + "learning_rate": 1.5114000000000001e-05, + "loss": 0.0907, + "step": 5041 + }, + { + "epoch": 4.258445945945946, + "grad_norm": 0.541253387928009, + "learning_rate": 1.5117000000000001e-05, + "loss": 0.0624, + "step": 5042 + }, + { + "epoch": 4.25929054054054, + "grad_norm": 1.0815351009368896, + "learning_rate": 1.5120000000000001e-05, + "loss": 0.078, + "step": 5043 + }, + { + "epoch": 4.260135135135135, + "grad_norm": 0.6072978377342224, + "learning_rate": 1.5123e-05, + "loss": 0.078, + "step": 5044 + }, + { + "epoch": 4.26097972972973, + "grad_norm": 0.6183551549911499, + "learning_rate": 1.5126e-05, + "loss": 0.0725, + "step": 5045 + }, + { + "epoch": 4.261824324324325, + "grad_norm": 0.4095716178417206, + "learning_rate": 1.5129e-05, + "loss": 0.0513, + "step": 5046 + }, + { + "epoch": 4.262668918918919, + "grad_norm": 0.6089488863945007, + "learning_rate": 1.5131999999999998e-05, + "loss": 0.0572, + "step": 5047 + }, + { + "epoch": 4.263513513513513, + "grad_norm": 0.6959583759307861, + "learning_rate": 1.5134999999999998e-05, + "loss": 0.0655, + "step": 5048 + }, + { + "epoch": 4.264358108108108, + "grad_norm": 0.5745880603790283, + "learning_rate": 1.5138000000000001e-05, + "loss": 0.0618, + "step": 5049 + }, + { + "epoch": 4.265202702702703, + "grad_norm": 1.2552298307418823, + "learning_rate": 1.5141000000000001e-05, + "loss": 0.0576, + "step": 5050 + }, + { + "epoch": 4.266047297297297, + "grad_norm": 0.6201310157775879, + "learning_rate": 1.5144000000000001e-05, + "loss": 0.0837, + "step": 5051 + }, + { + "epoch": 4.266891891891892, + "grad_norm": 0.4221488833427429, + "learning_rate": 1.5147e-05, + "loss": 0.0416, + "step": 5052 + }, + { + "epoch": 4.267736486486487, + "grad_norm": 0.6285022497177124, + "learning_rate": 1.515e-05, + "loss": 0.0624, + "step": 5053 + }, + { + "epoch": 4.268581081081081, + "grad_norm": 0.6923889517784119, + "learning_rate": 1.5153e-05, + "loss": 0.0646, + "step": 5054 + }, + { + "epoch": 4.269425675675675, + "grad_norm": 0.623680055141449, + "learning_rate": 1.5156e-05, + "loss": 0.0355, + "step": 5055 + }, + { + "epoch": 4.27027027027027, + "grad_norm": 1.5258187055587769, + "learning_rate": 1.5159e-05, + "loss": 0.0546, + "step": 5056 + }, + { + "epoch": 4.271114864864865, + "grad_norm": 0.48807743191719055, + "learning_rate": 1.5162e-05, + "loss": 0.0425, + "step": 5057 + }, + { + "epoch": 4.27195945945946, + "grad_norm": 0.49416613578796387, + "learning_rate": 1.5165e-05, + "loss": 0.0567, + "step": 5058 + }, + { + "epoch": 4.272804054054054, + "grad_norm": 0.690803050994873, + "learning_rate": 1.5168000000000001e-05, + "loss": 0.0581, + "step": 5059 + }, + { + "epoch": 4.273648648648648, + "grad_norm": 0.35638177394866943, + "learning_rate": 1.5171000000000001e-05, + "loss": 0.0314, + "step": 5060 + }, + { + "epoch": 4.274493243243243, + "grad_norm": 0.6498668193817139, + "learning_rate": 1.5174e-05, + "loss": 0.0581, + "step": 5061 + }, + { + "epoch": 4.275337837837838, + "grad_norm": 0.6622974276542664, + "learning_rate": 1.5177e-05, + "loss": 0.0463, + "step": 5062 + }, + { + "epoch": 4.2761824324324325, + "grad_norm": 0.5681681632995605, + "learning_rate": 1.518e-05, + "loss": 0.048, + "step": 5063 + }, + { + "epoch": 4.277027027027027, + "grad_norm": 0.6665590405464172, + "learning_rate": 1.5183e-05, + "loss": 0.055, + "step": 5064 + }, + { + "epoch": 4.277871621621622, + "grad_norm": 0.7586027979850769, + "learning_rate": 1.5186e-05, + "loss": 0.062, + "step": 5065 + }, + { + "epoch": 4.278716216216216, + "grad_norm": 0.842563271522522, + "learning_rate": 1.5189e-05, + "loss": 0.0564, + "step": 5066 + }, + { + "epoch": 4.2795608108108105, + "grad_norm": 0.5060229301452637, + "learning_rate": 1.5192e-05, + "loss": 0.058, + "step": 5067 + }, + { + "epoch": 4.280405405405405, + "grad_norm": 0.6362931132316589, + "learning_rate": 1.5195e-05, + "loss": 0.0625, + "step": 5068 + }, + { + "epoch": 4.28125, + "grad_norm": 0.7375402450561523, + "learning_rate": 1.5198000000000003e-05, + "loss": 0.0632, + "step": 5069 + }, + { + "epoch": 4.282094594594595, + "grad_norm": 0.5706396102905273, + "learning_rate": 1.5201000000000002e-05, + "loss": 0.0489, + "step": 5070 + }, + { + "epoch": 4.2829391891891895, + "grad_norm": 0.8014612197875977, + "learning_rate": 1.5204e-05, + "loss": 0.0882, + "step": 5071 + }, + { + "epoch": 4.283783783783784, + "grad_norm": 0.7179405689239502, + "learning_rate": 1.5207e-05, + "loss": 0.0441, + "step": 5072 + }, + { + "epoch": 4.284628378378378, + "grad_norm": 0.4238022267818451, + "learning_rate": 1.521e-05, + "loss": 0.039, + "step": 5073 + }, + { + "epoch": 4.285472972972973, + "grad_norm": 0.614352285861969, + "learning_rate": 1.5213e-05, + "loss": 0.0697, + "step": 5074 + }, + { + "epoch": 4.2863175675675675, + "grad_norm": 0.7804334759712219, + "learning_rate": 1.5216e-05, + "loss": 0.0584, + "step": 5075 + }, + { + "epoch": 4.287162162162162, + "grad_norm": 0.5636054277420044, + "learning_rate": 1.5219e-05, + "loss": 0.0574, + "step": 5076 + }, + { + "epoch": 4.288006756756757, + "grad_norm": 0.7467213869094849, + "learning_rate": 1.5222e-05, + "loss": 0.0577, + "step": 5077 + }, + { + "epoch": 4.288851351351352, + "grad_norm": 0.5637481212615967, + "learning_rate": 1.5224999999999999e-05, + "loss": 0.0595, + "step": 5078 + }, + { + "epoch": 4.289695945945946, + "grad_norm": 1.1654994487762451, + "learning_rate": 1.5228000000000002e-05, + "loss": 0.0653, + "step": 5079 + }, + { + "epoch": 4.29054054054054, + "grad_norm": 0.7749419212341309, + "learning_rate": 1.5231000000000002e-05, + "loss": 0.0741, + "step": 5080 + }, + { + "epoch": 4.291385135135135, + "grad_norm": 0.579532265663147, + "learning_rate": 1.5234000000000002e-05, + "loss": 0.0638, + "step": 5081 + }, + { + "epoch": 4.29222972972973, + "grad_norm": 0.7007715702056885, + "learning_rate": 1.5237000000000002e-05, + "loss": 0.0713, + "step": 5082 + }, + { + "epoch": 4.293074324324325, + "grad_norm": 1.2230037450790405, + "learning_rate": 1.524e-05, + "loss": 0.1391, + "step": 5083 + }, + { + "epoch": 4.293918918918919, + "grad_norm": 1.6910736560821533, + "learning_rate": 1.5243e-05, + "loss": 0.1195, + "step": 5084 + }, + { + "epoch": 4.294763513513513, + "grad_norm": 0.8681906461715698, + "learning_rate": 1.5246e-05, + "loss": 0.0849, + "step": 5085 + }, + { + "epoch": 4.295608108108108, + "grad_norm": 1.3028020858764648, + "learning_rate": 1.5249e-05, + "loss": 0.1473, + "step": 5086 + }, + { + "epoch": 4.296452702702703, + "grad_norm": 5.482728004455566, + "learning_rate": 1.5251999999999999e-05, + "loss": 0.6176, + "step": 5087 + }, + { + "epoch": 4.297297297297297, + "grad_norm": 2.9318723678588867, + "learning_rate": 1.5254999999999999e-05, + "loss": 0.4184, + "step": 5088 + }, + { + "epoch": 4.298141891891892, + "grad_norm": 0.9533354640007019, + "learning_rate": 1.5258000000000002e-05, + "loss": 0.2151, + "step": 5089 + }, + { + "epoch": 4.298986486486487, + "grad_norm": 1.0481446981430054, + "learning_rate": 1.5261000000000002e-05, + "loss": 0.1046, + "step": 5090 + }, + { + "epoch": 4.299831081081081, + "grad_norm": 1.015673041343689, + "learning_rate": 1.5264e-05, + "loss": 0.0943, + "step": 5091 + }, + { + "epoch": 4.300675675675675, + "grad_norm": 0.7817702293395996, + "learning_rate": 1.5267e-05, + "loss": 0.0687, + "step": 5092 + }, + { + "epoch": 4.30152027027027, + "grad_norm": 0.46297314763069153, + "learning_rate": 1.527e-05, + "loss": 0.0585, + "step": 5093 + }, + { + "epoch": 4.302364864864865, + "grad_norm": 0.45118045806884766, + "learning_rate": 1.5273e-05, + "loss": 0.0569, + "step": 5094 + }, + { + "epoch": 4.30320945945946, + "grad_norm": 0.7296449542045593, + "learning_rate": 1.5276e-05, + "loss": 0.0589, + "step": 5095 + }, + { + "epoch": 4.304054054054054, + "grad_norm": 0.5867288112640381, + "learning_rate": 1.5279e-05, + "loss": 0.0586, + "step": 5096 + }, + { + "epoch": 4.304898648648648, + "grad_norm": 0.7942385077476501, + "learning_rate": 1.5282e-05, + "loss": 0.0702, + "step": 5097 + }, + { + "epoch": 4.305743243243243, + "grad_norm": 0.49260082840919495, + "learning_rate": 1.5285e-05, + "loss": 0.0519, + "step": 5098 + }, + { + "epoch": 4.306587837837838, + "grad_norm": 0.562162458896637, + "learning_rate": 1.5288000000000003e-05, + "loss": 0.0533, + "step": 5099 + }, + { + "epoch": 4.3074324324324325, + "grad_norm": 0.46436503529548645, + "learning_rate": 1.5291000000000003e-05, + "loss": 0.0353, + "step": 5100 + }, + { + "epoch": 4.308277027027027, + "grad_norm": 0.7071244120597839, + "learning_rate": 1.5294000000000003e-05, + "loss": 0.1032, + "step": 5101 + }, + { + "epoch": 4.309121621621622, + "grad_norm": 0.42579856514930725, + "learning_rate": 1.5297e-05, + "loss": 0.046, + "step": 5102 + }, + { + "epoch": 4.309966216216216, + "grad_norm": 0.6186003684997559, + "learning_rate": 1.53e-05, + "loss": 0.0584, + "step": 5103 + }, + { + "epoch": 4.3108108108108105, + "grad_norm": 0.5025733113288879, + "learning_rate": 1.5303e-05, + "loss": 0.0483, + "step": 5104 + }, + { + "epoch": 4.311655405405405, + "grad_norm": 0.5630097985267639, + "learning_rate": 1.5306e-05, + "loss": 0.039, + "step": 5105 + }, + { + "epoch": 4.3125, + "grad_norm": 0.6066762804985046, + "learning_rate": 1.5309e-05, + "loss": 0.0448, + "step": 5106 + }, + { + "epoch": 4.313344594594595, + "grad_norm": 0.5667372345924377, + "learning_rate": 1.5312e-05, + "loss": 0.0509, + "step": 5107 + }, + { + "epoch": 4.3141891891891895, + "grad_norm": 0.5682874917984009, + "learning_rate": 1.5314999999999998e-05, + "loss": 0.0643, + "step": 5108 + }, + { + "epoch": 4.315033783783784, + "grad_norm": 0.6950831413269043, + "learning_rate": 1.5318e-05, + "loss": 0.0569, + "step": 5109 + }, + { + "epoch": 4.315878378378378, + "grad_norm": 0.45864295959472656, + "learning_rate": 1.5321e-05, + "loss": 0.0403, + "step": 5110 + }, + { + "epoch": 4.316722972972973, + "grad_norm": 0.7386707663536072, + "learning_rate": 1.5324e-05, + "loss": 0.1068, + "step": 5111 + }, + { + "epoch": 4.3175675675675675, + "grad_norm": 0.5829067230224609, + "learning_rate": 1.5327e-05, + "loss": 0.0506, + "step": 5112 + }, + { + "epoch": 4.318412162162162, + "grad_norm": 0.5513954162597656, + "learning_rate": 1.533e-05, + "loss": 0.052, + "step": 5113 + }, + { + "epoch": 4.319256756756757, + "grad_norm": 0.6124868392944336, + "learning_rate": 1.5333e-05, + "loss": 0.0513, + "step": 5114 + }, + { + "epoch": 4.320101351351352, + "grad_norm": 0.5787503123283386, + "learning_rate": 1.5336e-05, + "loss": 0.0492, + "step": 5115 + }, + { + "epoch": 4.320945945945946, + "grad_norm": 0.47505301237106323, + "learning_rate": 1.5339e-05, + "loss": 0.06, + "step": 5116 + }, + { + "epoch": 4.32179054054054, + "grad_norm": 0.45581820607185364, + "learning_rate": 1.5342e-05, + "loss": 0.0473, + "step": 5117 + }, + { + "epoch": 4.322635135135135, + "grad_norm": 0.6880620121955872, + "learning_rate": 1.5345e-05, + "loss": 0.0621, + "step": 5118 + }, + { + "epoch": 4.32347972972973, + "grad_norm": 0.6745285987854004, + "learning_rate": 1.5348000000000003e-05, + "loss": 0.0687, + "step": 5119 + }, + { + "epoch": 4.324324324324325, + "grad_norm": 0.5815983414649963, + "learning_rate": 1.5351000000000003e-05, + "loss": 0.0475, + "step": 5120 + }, + { + "epoch": 4.325168918918919, + "grad_norm": 0.7142229080200195, + "learning_rate": 1.5354000000000002e-05, + "loss": 0.0714, + "step": 5121 + }, + { + "epoch": 4.326013513513513, + "grad_norm": 0.5627449154853821, + "learning_rate": 1.5357000000000002e-05, + "loss": 0.0536, + "step": 5122 + }, + { + "epoch": 4.326858108108108, + "grad_norm": 0.7778940200805664, + "learning_rate": 1.5360000000000002e-05, + "loss": 0.0821, + "step": 5123 + }, + { + "epoch": 4.327702702702703, + "grad_norm": 0.5213956236839294, + "learning_rate": 1.5363000000000002e-05, + "loss": 0.0365, + "step": 5124 + }, + { + "epoch": 4.328547297297297, + "grad_norm": 0.6927341222763062, + "learning_rate": 1.5366e-05, + "loss": 0.0557, + "step": 5125 + }, + { + "epoch": 4.329391891891892, + "grad_norm": 0.9065855145454407, + "learning_rate": 1.5368999999999998e-05, + "loss": 0.0607, + "step": 5126 + }, + { + "epoch": 4.330236486486487, + "grad_norm": 0.7593888640403748, + "learning_rate": 1.5371999999999998e-05, + "loss": 0.0685, + "step": 5127 + }, + { + "epoch": 4.331081081081081, + "grad_norm": 0.6997899413108826, + "learning_rate": 1.5374999999999998e-05, + "loss": 0.0715, + "step": 5128 + }, + { + "epoch": 4.331925675675675, + "grad_norm": 0.5945335626602173, + "learning_rate": 1.5377999999999997e-05, + "loss": 0.0623, + "step": 5129 + }, + { + "epoch": 4.33277027027027, + "grad_norm": 0.5941228866577148, + "learning_rate": 1.5381e-05, + "loss": 0.0414, + "step": 5130 + }, + { + "epoch": 4.333614864864865, + "grad_norm": 1.7456544637680054, + "learning_rate": 1.5384e-05, + "loss": 0.1133, + "step": 5131 + }, + { + "epoch": 4.33445945945946, + "grad_norm": 0.7735204696655273, + "learning_rate": 1.5387e-05, + "loss": 0.0691, + "step": 5132 + }, + { + "epoch": 4.335304054054054, + "grad_norm": 0.6033904552459717, + "learning_rate": 1.539e-05, + "loss": 0.059, + "step": 5133 + }, + { + "epoch": 4.336148648648648, + "grad_norm": 0.7872196435928345, + "learning_rate": 1.5393e-05, + "loss": 0.0534, + "step": 5134 + }, + { + "epoch": 4.336993243243243, + "grad_norm": 0.8442234992980957, + "learning_rate": 1.5396e-05, + "loss": 0.0674, + "step": 5135 + }, + { + "epoch": 4.337837837837838, + "grad_norm": 2.0348997116088867, + "learning_rate": 1.5399e-05, + "loss": 0.1327, + "step": 5136 + }, + { + "epoch": 4.3386824324324325, + "grad_norm": 0.9536567330360413, + "learning_rate": 1.5402e-05, + "loss": 0.3626, + "step": 5137 + }, + { + "epoch": 4.339527027027027, + "grad_norm": 1.2645407915115356, + "learning_rate": 1.5405e-05, + "loss": 0.2999, + "step": 5138 + }, + { + "epoch": 4.340371621621622, + "grad_norm": 0.5262454748153687, + "learning_rate": 1.5408e-05, + "loss": 0.1171, + "step": 5139 + }, + { + "epoch": 4.341216216216216, + "grad_norm": 0.44447875022888184, + "learning_rate": 1.5411000000000002e-05, + "loss": 0.0745, + "step": 5140 + }, + { + "epoch": 4.3420608108108105, + "grad_norm": 0.651574969291687, + "learning_rate": 1.5414000000000002e-05, + "loss": 0.0636, + "step": 5141 + }, + { + "epoch": 4.342905405405405, + "grad_norm": 0.7981892824172974, + "learning_rate": 1.5417e-05, + "loss": 0.0913, + "step": 5142 + }, + { + "epoch": 4.34375, + "grad_norm": 0.6075350642204285, + "learning_rate": 1.542e-05, + "loss": 0.0769, + "step": 5143 + }, + { + "epoch": 4.344594594594595, + "grad_norm": 0.49109330773353577, + "learning_rate": 1.5423e-05, + "loss": 0.0615, + "step": 5144 + }, + { + "epoch": 4.3454391891891895, + "grad_norm": 0.6578825116157532, + "learning_rate": 1.5426e-05, + "loss": 0.0569, + "step": 5145 + }, + { + "epoch": 4.346283783783784, + "grad_norm": 0.508225679397583, + "learning_rate": 1.5429e-05, + "loss": 0.0585, + "step": 5146 + }, + { + "epoch": 4.347128378378378, + "grad_norm": 0.9738094210624695, + "learning_rate": 1.5432e-05, + "loss": 0.1108, + "step": 5147 + }, + { + "epoch": 4.347972972972973, + "grad_norm": 0.5897241830825806, + "learning_rate": 1.5435e-05, + "loss": 0.0403, + "step": 5148 + }, + { + "epoch": 4.3488175675675675, + "grad_norm": 0.5650955438613892, + "learning_rate": 1.5438e-05, + "loss": 0.0915, + "step": 5149 + }, + { + "epoch": 4.349662162162162, + "grad_norm": 0.4639889597892761, + "learning_rate": 1.5441000000000003e-05, + "loss": 0.0643, + "step": 5150 + }, + { + "epoch": 4.350506756756757, + "grad_norm": 0.35230883955955505, + "learning_rate": 1.5444e-05, + "loss": 0.0492, + "step": 5151 + }, + { + "epoch": 4.351351351351352, + "grad_norm": 0.4308423697948456, + "learning_rate": 1.5447e-05, + "loss": 0.0424, + "step": 5152 + }, + { + "epoch": 4.352195945945946, + "grad_norm": 0.9767122268676758, + "learning_rate": 1.545e-05, + "loss": 0.0527, + "step": 5153 + }, + { + "epoch": 4.35304054054054, + "grad_norm": 0.39967888593673706, + "learning_rate": 1.5453e-05, + "loss": 0.0417, + "step": 5154 + }, + { + "epoch": 4.353885135135135, + "grad_norm": 0.48781639337539673, + "learning_rate": 1.5456e-05, + "loss": 0.058, + "step": 5155 + }, + { + "epoch": 4.35472972972973, + "grad_norm": 0.5070879459381104, + "learning_rate": 1.5459e-05, + "loss": 0.0599, + "step": 5156 + }, + { + "epoch": 4.355574324324325, + "grad_norm": 0.4980398714542389, + "learning_rate": 1.5462e-05, + "loss": 0.0601, + "step": 5157 + }, + { + "epoch": 4.356418918918919, + "grad_norm": 0.6004948616027832, + "learning_rate": 1.5465e-05, + "loss": 0.0686, + "step": 5158 + }, + { + "epoch": 4.357263513513513, + "grad_norm": 0.45606404542922974, + "learning_rate": 1.5467999999999998e-05, + "loss": 0.0402, + "step": 5159 + }, + { + "epoch": 4.358108108108108, + "grad_norm": 0.8320892453193665, + "learning_rate": 1.5471e-05, + "loss": 0.0863, + "step": 5160 + }, + { + "epoch": 4.358952702702703, + "grad_norm": 0.41763558983802795, + "learning_rate": 1.5474e-05, + "loss": 0.0508, + "step": 5161 + }, + { + "epoch": 4.359797297297297, + "grad_norm": 0.38736575841903687, + "learning_rate": 1.5477e-05, + "loss": 0.0446, + "step": 5162 + }, + { + "epoch": 4.360641891891892, + "grad_norm": 0.5330506563186646, + "learning_rate": 1.548e-05, + "loss": 0.0698, + "step": 5163 + }, + { + "epoch": 4.361486486486487, + "grad_norm": 0.5247194766998291, + "learning_rate": 1.5483e-05, + "loss": 0.0538, + "step": 5164 + }, + { + "epoch": 4.362331081081081, + "grad_norm": 0.6237626075744629, + "learning_rate": 1.5486e-05, + "loss": 0.0454, + "step": 5165 + }, + { + "epoch": 4.363175675675675, + "grad_norm": 0.6426681876182556, + "learning_rate": 1.5489e-05, + "loss": 0.0492, + "step": 5166 + }, + { + "epoch": 4.36402027027027, + "grad_norm": 0.7702094912528992, + "learning_rate": 1.5492e-05, + "loss": 0.0583, + "step": 5167 + }, + { + "epoch": 4.364864864864865, + "grad_norm": 0.9119678735733032, + "learning_rate": 1.5495e-05, + "loss": 0.0785, + "step": 5168 + }, + { + "epoch": 4.36570945945946, + "grad_norm": 0.4792095124721527, + "learning_rate": 1.5498e-05, + "loss": 0.0464, + "step": 5169 + }, + { + "epoch": 4.366554054054054, + "grad_norm": 0.5972126722335815, + "learning_rate": 1.5501000000000003e-05, + "loss": 0.0455, + "step": 5170 + }, + { + "epoch": 4.367398648648648, + "grad_norm": 0.9491291046142578, + "learning_rate": 1.5504000000000003e-05, + "loss": 0.1098, + "step": 5171 + }, + { + "epoch": 4.368243243243243, + "grad_norm": 0.5871490240097046, + "learning_rate": 1.5507000000000002e-05, + "loss": 0.0526, + "step": 5172 + }, + { + "epoch": 4.369087837837838, + "grad_norm": 0.6887626051902771, + "learning_rate": 1.5510000000000002e-05, + "loss": 0.0487, + "step": 5173 + }, + { + "epoch": 4.3699324324324325, + "grad_norm": 0.7938011288642883, + "learning_rate": 1.5513000000000002e-05, + "loss": 0.0471, + "step": 5174 + }, + { + "epoch": 4.370777027027027, + "grad_norm": 0.40066060423851013, + "learning_rate": 1.5516000000000002e-05, + "loss": 0.0265, + "step": 5175 + }, + { + "epoch": 4.371621621621622, + "grad_norm": 0.6764049530029297, + "learning_rate": 1.5518999999999998e-05, + "loss": 0.0569, + "step": 5176 + }, + { + "epoch": 4.372466216216216, + "grad_norm": 1.1267480850219727, + "learning_rate": 1.5521999999999998e-05, + "loss": 0.0798, + "step": 5177 + }, + { + "epoch": 4.3733108108108105, + "grad_norm": 1.179287314414978, + "learning_rate": 1.5524999999999998e-05, + "loss": 0.0633, + "step": 5178 + }, + { + "epoch": 4.374155405405405, + "grad_norm": 1.00077486038208, + "learning_rate": 1.5527999999999998e-05, + "loss": 0.0704, + "step": 5179 + }, + { + "epoch": 4.375, + "grad_norm": 0.9771939516067505, + "learning_rate": 1.5531e-05, + "loss": 0.0729, + "step": 5180 + }, + { + "epoch": 4.375844594594595, + "grad_norm": 0.6370094418525696, + "learning_rate": 1.5534e-05, + "loss": 0.0564, + "step": 5181 + }, + { + "epoch": 4.3766891891891895, + "grad_norm": 0.5832936763763428, + "learning_rate": 1.5537e-05, + "loss": 0.0644, + "step": 5182 + }, + { + "epoch": 4.377533783783784, + "grad_norm": 1.4792169332504272, + "learning_rate": 1.554e-05, + "loss": 0.0779, + "step": 5183 + }, + { + "epoch": 4.378378378378378, + "grad_norm": 0.6959331035614014, + "learning_rate": 1.5543e-05, + "loss": 0.0627, + "step": 5184 + }, + { + "epoch": 4.379222972972973, + "grad_norm": 1.1883541345596313, + "learning_rate": 1.5546e-05, + "loss": 0.0975, + "step": 5185 + }, + { + "epoch": 4.3800675675675675, + "grad_norm": 2.1050174236297607, + "learning_rate": 1.5549e-05, + "loss": 0.1682, + "step": 5186 + }, + { + "epoch": 4.380912162162162, + "grad_norm": 1.9826523065567017, + "learning_rate": 1.5552e-05, + "loss": 0.4321, + "step": 5187 + }, + { + "epoch": 4.381756756756757, + "grad_norm": 0.9781120419502258, + "learning_rate": 1.5555e-05, + "loss": 0.2283, + "step": 5188 + }, + { + "epoch": 4.382601351351352, + "grad_norm": 0.8462255597114563, + "learning_rate": 1.5558e-05, + "loss": 0.1541, + "step": 5189 + }, + { + "epoch": 4.383445945945946, + "grad_norm": 0.8265193104743958, + "learning_rate": 1.5561000000000002e-05, + "loss": 0.0918, + "step": 5190 + }, + { + "epoch": 4.38429054054054, + "grad_norm": 0.6480406522750854, + "learning_rate": 1.5564000000000002e-05, + "loss": 0.0731, + "step": 5191 + }, + { + "epoch": 4.385135135135135, + "grad_norm": 0.54775470495224, + "learning_rate": 1.5567000000000002e-05, + "loss": 0.0788, + "step": 5192 + }, + { + "epoch": 4.38597972972973, + "grad_norm": 0.6640938520431519, + "learning_rate": 1.5570000000000002e-05, + "loss": 0.062, + "step": 5193 + }, + { + "epoch": 4.386824324324325, + "grad_norm": 0.5124265551567078, + "learning_rate": 1.5573e-05, + "loss": 0.0499, + "step": 5194 + }, + { + "epoch": 4.387668918918919, + "grad_norm": 0.6968497037887573, + "learning_rate": 1.5576e-05, + "loss": 0.0773, + "step": 5195 + }, + { + "epoch": 4.388513513513513, + "grad_norm": 0.4090768098831177, + "learning_rate": 1.5579e-05, + "loss": 0.048, + "step": 5196 + }, + { + "epoch": 4.389358108108108, + "grad_norm": 0.45388132333755493, + "learning_rate": 1.5582e-05, + "loss": 0.0525, + "step": 5197 + }, + { + "epoch": 4.390202702702703, + "grad_norm": 0.3541025221347809, + "learning_rate": 1.5585e-05, + "loss": 0.0403, + "step": 5198 + }, + { + "epoch": 4.391047297297297, + "grad_norm": 0.6059007048606873, + "learning_rate": 1.5588e-05, + "loss": 0.0544, + "step": 5199 + }, + { + "epoch": 4.391891891891892, + "grad_norm": 0.6453835964202881, + "learning_rate": 1.5591e-05, + "loss": 0.0659, + "step": 5200 + }, + { + "epoch": 4.392736486486487, + "grad_norm": 0.5190640687942505, + "learning_rate": 1.5594e-05, + "loss": 0.0564, + "step": 5201 + }, + { + "epoch": 4.393581081081081, + "grad_norm": 0.5628987550735474, + "learning_rate": 1.5597e-05, + "loss": 0.0357, + "step": 5202 + }, + { + "epoch": 4.394425675675675, + "grad_norm": 0.4878053665161133, + "learning_rate": 1.56e-05, + "loss": 0.0501, + "step": 5203 + }, + { + "epoch": 4.39527027027027, + "grad_norm": 0.4100075662136078, + "learning_rate": 1.5603e-05, + "loss": 0.0389, + "step": 5204 + }, + { + "epoch": 4.396114864864865, + "grad_norm": 0.5934538841247559, + "learning_rate": 1.5606e-05, + "loss": 0.0541, + "step": 5205 + }, + { + "epoch": 4.39695945945946, + "grad_norm": 0.4739060401916504, + "learning_rate": 1.5609e-05, + "loss": 0.0558, + "step": 5206 + }, + { + "epoch": 4.397804054054054, + "grad_norm": 0.6083192229270935, + "learning_rate": 1.5612e-05, + "loss": 0.059, + "step": 5207 + }, + { + "epoch": 4.398648648648648, + "grad_norm": 0.5564382076263428, + "learning_rate": 1.5615e-05, + "loss": 0.0554, + "step": 5208 + }, + { + "epoch": 4.399493243243243, + "grad_norm": 0.6483569741249084, + "learning_rate": 1.5618e-05, + "loss": 0.0643, + "step": 5209 + }, + { + "epoch": 4.400337837837838, + "grad_norm": 0.671732485294342, + "learning_rate": 1.5621000000000002e-05, + "loss": 0.0516, + "step": 5210 + }, + { + "epoch": 4.4011824324324325, + "grad_norm": 0.4515582323074341, + "learning_rate": 1.5624e-05, + "loss": 0.0384, + "step": 5211 + }, + { + "epoch": 4.402027027027027, + "grad_norm": 0.940356969833374, + "learning_rate": 1.5627e-05, + "loss": 0.0988, + "step": 5212 + }, + { + "epoch": 4.402871621621622, + "grad_norm": 0.7349386215209961, + "learning_rate": 1.563e-05, + "loss": 0.0589, + "step": 5213 + }, + { + "epoch": 4.403716216216216, + "grad_norm": 0.5887504816055298, + "learning_rate": 1.5633e-05, + "loss": 0.0566, + "step": 5214 + }, + { + "epoch": 4.4045608108108105, + "grad_norm": 0.5768280625343323, + "learning_rate": 1.5636e-05, + "loss": 0.0365, + "step": 5215 + }, + { + "epoch": 4.405405405405405, + "grad_norm": 0.7014636397361755, + "learning_rate": 1.5639e-05, + "loss": 0.05, + "step": 5216 + }, + { + "epoch": 4.40625, + "grad_norm": 0.9009767770767212, + "learning_rate": 1.5642e-05, + "loss": 0.0608, + "step": 5217 + }, + { + "epoch": 4.407094594594595, + "grad_norm": 0.665556788444519, + "learning_rate": 1.5645e-05, + "loss": 0.0617, + "step": 5218 + }, + { + "epoch": 4.4079391891891895, + "grad_norm": 0.44501322507858276, + "learning_rate": 1.5648e-05, + "loss": 0.0495, + "step": 5219 + }, + { + "epoch": 4.408783783783784, + "grad_norm": 0.8238680958747864, + "learning_rate": 1.5651000000000003e-05, + "loss": 0.0599, + "step": 5220 + }, + { + "epoch": 4.409628378378378, + "grad_norm": 0.6142075657844543, + "learning_rate": 1.5654000000000003e-05, + "loss": 0.0451, + "step": 5221 + }, + { + "epoch": 4.410472972972973, + "grad_norm": 0.9337335228919983, + "learning_rate": 1.5657000000000003e-05, + "loss": 0.0609, + "step": 5222 + }, + { + "epoch": 4.4113175675675675, + "grad_norm": 1.1949255466461182, + "learning_rate": 1.5660000000000003e-05, + "loss": 0.0783, + "step": 5223 + }, + { + "epoch": 4.412162162162162, + "grad_norm": 0.6634476780891418, + "learning_rate": 1.5663000000000002e-05, + "loss": 0.081, + "step": 5224 + }, + { + "epoch": 4.413006756756757, + "grad_norm": 0.6562100052833557, + "learning_rate": 1.5666e-05, + "loss": 0.0493, + "step": 5225 + }, + { + "epoch": 4.413851351351352, + "grad_norm": 0.7232336401939392, + "learning_rate": 1.5669e-05, + "loss": 0.0798, + "step": 5226 + }, + { + "epoch": 4.414695945945946, + "grad_norm": 0.5519424080848694, + "learning_rate": 1.5672e-05, + "loss": 0.0583, + "step": 5227 + }, + { + "epoch": 4.41554054054054, + "grad_norm": 0.7350763082504272, + "learning_rate": 1.5674999999999998e-05, + "loss": 0.0663, + "step": 5228 + }, + { + "epoch": 4.416385135135135, + "grad_norm": 0.6287031173706055, + "learning_rate": 1.5677999999999998e-05, + "loss": 0.045, + "step": 5229 + }, + { + "epoch": 4.41722972972973, + "grad_norm": 0.6348652243614197, + "learning_rate": 1.5681e-05, + "loss": 0.0441, + "step": 5230 + }, + { + "epoch": 4.418074324324325, + "grad_norm": 0.7635819911956787, + "learning_rate": 1.5684e-05, + "loss": 0.0625, + "step": 5231 + }, + { + "epoch": 4.418918918918919, + "grad_norm": 0.8495739698410034, + "learning_rate": 1.5687e-05, + "loss": 0.0552, + "step": 5232 + }, + { + "epoch": 4.419763513513513, + "grad_norm": 1.0102564096450806, + "learning_rate": 1.569e-05, + "loss": 0.0936, + "step": 5233 + }, + { + "epoch": 4.420608108108108, + "grad_norm": 1.2050734758377075, + "learning_rate": 1.5693e-05, + "loss": 0.0917, + "step": 5234 + }, + { + "epoch": 4.421452702702703, + "grad_norm": 1.215422511100769, + "learning_rate": 1.5696e-05, + "loss": 0.1027, + "step": 5235 + }, + { + "epoch": 4.422297297297297, + "grad_norm": 1.6870367527008057, + "learning_rate": 1.5699e-05, + "loss": 0.1284, + "step": 5236 + }, + { + "epoch": 4.423141891891892, + "grad_norm": 1.5120272636413574, + "learning_rate": 1.5702e-05, + "loss": 0.4015, + "step": 5237 + }, + { + "epoch": 4.423986486486487, + "grad_norm": 0.8607744574546814, + "learning_rate": 1.5705e-05, + "loss": 0.2593, + "step": 5238 + }, + { + "epoch": 4.424831081081081, + "grad_norm": 0.6325886249542236, + "learning_rate": 1.5708e-05, + "loss": 0.1038, + "step": 5239 + }, + { + "epoch": 4.425675675675675, + "grad_norm": 0.6127646565437317, + "learning_rate": 1.5711000000000003e-05, + "loss": 0.0791, + "step": 5240 + }, + { + "epoch": 4.42652027027027, + "grad_norm": 0.43941113352775574, + "learning_rate": 1.5714000000000002e-05, + "loss": 0.0687, + "step": 5241 + }, + { + "epoch": 4.427364864864865, + "grad_norm": 0.505367636680603, + "learning_rate": 1.5717000000000002e-05, + "loss": 0.0766, + "step": 5242 + }, + { + "epoch": 4.42820945945946, + "grad_norm": 0.3945299983024597, + "learning_rate": 1.5720000000000002e-05, + "loss": 0.0469, + "step": 5243 + }, + { + "epoch": 4.429054054054054, + "grad_norm": 0.49745479226112366, + "learning_rate": 1.5723000000000002e-05, + "loss": 0.0732, + "step": 5244 + }, + { + "epoch": 4.429898648648648, + "grad_norm": 0.7761968374252319, + "learning_rate": 1.5726e-05, + "loss": 0.0673, + "step": 5245 + }, + { + "epoch": 4.430743243243243, + "grad_norm": 0.613798201084137, + "learning_rate": 1.5729e-05, + "loss": 0.0655, + "step": 5246 + }, + { + "epoch": 4.431587837837838, + "grad_norm": 0.6386746764183044, + "learning_rate": 1.5732e-05, + "loss": 0.0606, + "step": 5247 + }, + { + "epoch": 4.4324324324324325, + "grad_norm": 0.625376284122467, + "learning_rate": 1.5735e-05, + "loss": 0.0641, + "step": 5248 + }, + { + "epoch": 4.433277027027027, + "grad_norm": 0.6329139471054077, + "learning_rate": 1.5737999999999997e-05, + "loss": 0.0637, + "step": 5249 + }, + { + "epoch": 4.434121621621622, + "grad_norm": 0.4096701443195343, + "learning_rate": 1.5741e-05, + "loss": 0.0536, + "step": 5250 + }, + { + "epoch": 4.434966216216216, + "grad_norm": 0.4017358720302582, + "learning_rate": 1.5744e-05, + "loss": 0.0372, + "step": 5251 + }, + { + "epoch": 4.4358108108108105, + "grad_norm": 0.5494780540466309, + "learning_rate": 1.5747e-05, + "loss": 0.0552, + "step": 5252 + }, + { + "epoch": 4.436655405405405, + "grad_norm": 0.502872884273529, + "learning_rate": 1.575e-05, + "loss": 0.0462, + "step": 5253 + }, + { + "epoch": 4.4375, + "grad_norm": 0.6036773920059204, + "learning_rate": 1.5753e-05, + "loss": 0.0581, + "step": 5254 + }, + { + "epoch": 4.438344594594595, + "grad_norm": 0.45385798811912537, + "learning_rate": 1.5756e-05, + "loss": 0.0501, + "step": 5255 + }, + { + "epoch": 4.4391891891891895, + "grad_norm": 0.5820817947387695, + "learning_rate": 1.5759e-05, + "loss": 0.0563, + "step": 5256 + }, + { + "epoch": 4.440033783783784, + "grad_norm": 0.5085216164588928, + "learning_rate": 1.5762e-05, + "loss": 0.0366, + "step": 5257 + }, + { + "epoch": 4.440878378378378, + "grad_norm": 0.38069504499435425, + "learning_rate": 1.5765e-05, + "loss": 0.0453, + "step": 5258 + }, + { + "epoch": 4.441722972972973, + "grad_norm": 1.7778654098510742, + "learning_rate": 1.5768e-05, + "loss": 0.1878, + "step": 5259 + }, + { + "epoch": 4.4425675675675675, + "grad_norm": 0.8826878666877747, + "learning_rate": 1.5771e-05, + "loss": 0.0637, + "step": 5260 + }, + { + "epoch": 4.443412162162162, + "grad_norm": 0.5284957885742188, + "learning_rate": 1.5774000000000002e-05, + "loss": 0.0585, + "step": 5261 + }, + { + "epoch": 4.444256756756757, + "grad_norm": 0.6221334338188171, + "learning_rate": 1.5777e-05, + "loss": 0.072, + "step": 5262 + }, + { + "epoch": 4.445101351351352, + "grad_norm": 0.5259972214698792, + "learning_rate": 1.578e-05, + "loss": 0.069, + "step": 5263 + }, + { + "epoch": 4.445945945945946, + "grad_norm": 0.5917832851409912, + "learning_rate": 1.5783e-05, + "loss": 0.0617, + "step": 5264 + }, + { + "epoch": 4.44679054054054, + "grad_norm": 0.5930662751197815, + "learning_rate": 1.5786e-05, + "loss": 0.0575, + "step": 5265 + }, + { + "epoch": 4.447635135135135, + "grad_norm": 0.5758991241455078, + "learning_rate": 1.5789e-05, + "loss": 0.0829, + "step": 5266 + }, + { + "epoch": 4.44847972972973, + "grad_norm": 0.37629732489585876, + "learning_rate": 1.5792e-05, + "loss": 0.0503, + "step": 5267 + }, + { + "epoch": 4.449324324324325, + "grad_norm": 0.6552395820617676, + "learning_rate": 1.5795e-05, + "loss": 0.0776, + "step": 5268 + }, + { + "epoch": 4.450168918918919, + "grad_norm": 0.49852192401885986, + "learning_rate": 1.5798e-05, + "loss": 0.0477, + "step": 5269 + }, + { + "epoch": 4.451013513513513, + "grad_norm": 0.5580707788467407, + "learning_rate": 1.5801e-05, + "loss": 0.0447, + "step": 5270 + }, + { + "epoch": 4.451858108108108, + "grad_norm": 0.5863687992095947, + "learning_rate": 1.5804000000000003e-05, + "loss": 0.0554, + "step": 5271 + }, + { + "epoch": 4.452702702702703, + "grad_norm": 0.6541157960891724, + "learning_rate": 1.5807000000000003e-05, + "loss": 0.0464, + "step": 5272 + }, + { + "epoch": 4.453547297297297, + "grad_norm": 0.5734925866127014, + "learning_rate": 1.5810000000000003e-05, + "loss": 0.0573, + "step": 5273 + }, + { + "epoch": 4.454391891891892, + "grad_norm": 0.620335578918457, + "learning_rate": 1.5813e-05, + "loss": 0.0523, + "step": 5274 + }, + { + "epoch": 4.455236486486487, + "grad_norm": 1.4815568923950195, + "learning_rate": 1.5816e-05, + "loss": 0.0866, + "step": 5275 + }, + { + "epoch": 4.456081081081081, + "grad_norm": 0.6700912714004517, + "learning_rate": 1.5819e-05, + "loss": 0.082, + "step": 5276 + }, + { + "epoch": 4.456925675675675, + "grad_norm": 0.47280755639076233, + "learning_rate": 1.5822e-05, + "loss": 0.0453, + "step": 5277 + }, + { + "epoch": 4.45777027027027, + "grad_norm": 0.6107532978057861, + "learning_rate": 1.5825e-05, + "loss": 0.077, + "step": 5278 + }, + { + "epoch": 4.458614864864865, + "grad_norm": 0.4701656699180603, + "learning_rate": 1.5827999999999998e-05, + "loss": 0.0454, + "step": 5279 + }, + { + "epoch": 4.45945945945946, + "grad_norm": 0.6697069406509399, + "learning_rate": 1.5830999999999998e-05, + "loss": 0.0602, + "step": 5280 + }, + { + "epoch": 4.460304054054054, + "grad_norm": 0.9507044553756714, + "learning_rate": 1.5834e-05, + "loss": 0.0955, + "step": 5281 + }, + { + "epoch": 4.461148648648648, + "grad_norm": 0.621407687664032, + "learning_rate": 1.5837e-05, + "loss": 0.0859, + "step": 5282 + }, + { + "epoch": 4.461993243243243, + "grad_norm": 0.6351586580276489, + "learning_rate": 1.584e-05, + "loss": 0.0554, + "step": 5283 + }, + { + "epoch": 4.462837837837838, + "grad_norm": 0.8363577127456665, + "learning_rate": 1.5843e-05, + "loss": 0.0788, + "step": 5284 + }, + { + "epoch": 4.4636824324324325, + "grad_norm": 12.866697311401367, + "learning_rate": 1.5846e-05, + "loss": 0.2877, + "step": 5285 + }, + { + "epoch": 4.464527027027027, + "grad_norm": 1.4411394596099854, + "learning_rate": 1.5849e-05, + "loss": 0.118, + "step": 5286 + }, + { + "epoch": 4.465371621621622, + "grad_norm": 3.274339199066162, + "learning_rate": 1.5852e-05, + "loss": 0.5186, + "step": 5287 + }, + { + "epoch": 4.466216216216216, + "grad_norm": 1.0391407012939453, + "learning_rate": 1.5855e-05, + "loss": 0.2111, + "step": 5288 + }, + { + "epoch": 4.4670608108108105, + "grad_norm": 0.9478889107704163, + "learning_rate": 1.5858e-05, + "loss": 0.1476, + "step": 5289 + }, + { + "epoch": 4.467905405405405, + "grad_norm": 0.5420424342155457, + "learning_rate": 1.5861e-05, + "loss": 0.0761, + "step": 5290 + }, + { + "epoch": 4.46875, + "grad_norm": 0.3764718174934387, + "learning_rate": 1.5864000000000003e-05, + "loss": 0.0437, + "step": 5291 + }, + { + "epoch": 4.469594594594595, + "grad_norm": 0.9313369989395142, + "learning_rate": 1.5867000000000002e-05, + "loss": 0.0836, + "step": 5292 + }, + { + "epoch": 4.4704391891891895, + "grad_norm": 0.4318457841873169, + "learning_rate": 1.5870000000000002e-05, + "loss": 0.0509, + "step": 5293 + }, + { + "epoch": 4.471283783783784, + "grad_norm": 0.6355682611465454, + "learning_rate": 1.5873000000000002e-05, + "loss": 0.0515, + "step": 5294 + }, + { + "epoch": 4.472128378378378, + "grad_norm": 0.5292975306510925, + "learning_rate": 1.5876000000000002e-05, + "loss": 0.057, + "step": 5295 + }, + { + "epoch": 4.472972972972973, + "grad_norm": 0.7071337103843689, + "learning_rate": 1.5879e-05, + "loss": 0.0416, + "step": 5296 + }, + { + "epoch": 4.4738175675675675, + "grad_norm": 0.5093319416046143, + "learning_rate": 1.5882e-05, + "loss": 0.0533, + "step": 5297 + }, + { + "epoch": 4.474662162162162, + "grad_norm": 0.3776058852672577, + "learning_rate": 1.5884999999999998e-05, + "loss": 0.043, + "step": 5298 + }, + { + "epoch": 4.475506756756757, + "grad_norm": 0.7790098786354065, + "learning_rate": 1.5887999999999998e-05, + "loss": 0.0661, + "step": 5299 + }, + { + "epoch": 4.476351351351352, + "grad_norm": 0.5735142230987549, + "learning_rate": 1.5890999999999997e-05, + "loss": 0.0476, + "step": 5300 + }, + { + "epoch": 4.477195945945946, + "grad_norm": 0.3798805773258209, + "learning_rate": 1.5894e-05, + "loss": 0.0375, + "step": 5301 + }, + { + "epoch": 4.47804054054054, + "grad_norm": 0.825415313243866, + "learning_rate": 1.5897e-05, + "loss": 0.0515, + "step": 5302 + }, + { + "epoch": 4.478885135135135, + "grad_norm": 0.5304741263389587, + "learning_rate": 1.59e-05, + "loss": 0.0515, + "step": 5303 + }, + { + "epoch": 4.47972972972973, + "grad_norm": 0.5278835296630859, + "learning_rate": 1.5903e-05, + "loss": 0.0491, + "step": 5304 + }, + { + "epoch": 4.480574324324325, + "grad_norm": 0.5562613606452942, + "learning_rate": 1.5906e-05, + "loss": 0.0502, + "step": 5305 + }, + { + "epoch": 4.481418918918919, + "grad_norm": 1.0637096166610718, + "learning_rate": 1.5909e-05, + "loss": 0.1459, + "step": 5306 + }, + { + "epoch": 4.482263513513513, + "grad_norm": 0.43723824620246887, + "learning_rate": 1.5912e-05, + "loss": 0.0449, + "step": 5307 + }, + { + "epoch": 4.483108108108108, + "grad_norm": 0.5493561029434204, + "learning_rate": 1.5915e-05, + "loss": 0.0496, + "step": 5308 + }, + { + "epoch": 4.483952702702703, + "grad_norm": 0.6143279671669006, + "learning_rate": 1.5918e-05, + "loss": 0.0626, + "step": 5309 + }, + { + "epoch": 4.484797297297297, + "grad_norm": 0.47034910321235657, + "learning_rate": 1.5921e-05, + "loss": 0.0489, + "step": 5310 + }, + { + "epoch": 4.485641891891892, + "grad_norm": 0.3727784752845764, + "learning_rate": 1.5924000000000002e-05, + "loss": 0.0373, + "step": 5311 + }, + { + "epoch": 4.486486486486487, + "grad_norm": 0.6231664419174194, + "learning_rate": 1.5927000000000002e-05, + "loss": 0.0633, + "step": 5312 + }, + { + "epoch": 4.487331081081081, + "grad_norm": 0.4632599353790283, + "learning_rate": 1.593e-05, + "loss": 0.0358, + "step": 5313 + }, + { + "epoch": 4.488175675675675, + "grad_norm": 1.1620701551437378, + "learning_rate": 1.5933e-05, + "loss": 0.0537, + "step": 5314 + }, + { + "epoch": 4.48902027027027, + "grad_norm": 0.464005708694458, + "learning_rate": 1.5936e-05, + "loss": 0.0452, + "step": 5315 + }, + { + "epoch": 4.489864864864865, + "grad_norm": 0.6537627577781677, + "learning_rate": 1.5939e-05, + "loss": 0.0627, + "step": 5316 + }, + { + "epoch": 4.49070945945946, + "grad_norm": 0.7327545285224915, + "learning_rate": 1.5942e-05, + "loss": 0.0523, + "step": 5317 + }, + { + "epoch": 4.491554054054054, + "grad_norm": 0.7932940125465393, + "learning_rate": 1.5945e-05, + "loss": 0.0846, + "step": 5318 + }, + { + "epoch": 4.492398648648648, + "grad_norm": 0.8526169657707214, + "learning_rate": 1.5948e-05, + "loss": 0.0672, + "step": 5319 + }, + { + "epoch": 4.493243243243243, + "grad_norm": 0.6355641484260559, + "learning_rate": 1.5951e-05, + "loss": 0.0753, + "step": 5320 + }, + { + "epoch": 4.494087837837838, + "grad_norm": 0.9835295677185059, + "learning_rate": 1.5954000000000003e-05, + "loss": 0.0949, + "step": 5321 + }, + { + "epoch": 4.4949324324324325, + "grad_norm": 0.6733484268188477, + "learning_rate": 1.5957000000000003e-05, + "loss": 0.0485, + "step": 5322 + }, + { + "epoch": 4.495777027027027, + "grad_norm": 0.6083396673202515, + "learning_rate": 1.596e-05, + "loss": 0.0524, + "step": 5323 + }, + { + "epoch": 4.496621621621622, + "grad_norm": 0.935813844203949, + "learning_rate": 1.5963e-05, + "loss": 0.0862, + "step": 5324 + }, + { + "epoch": 4.497466216216216, + "grad_norm": 0.6256188154220581, + "learning_rate": 1.5966e-05, + "loss": 0.053, + "step": 5325 + }, + { + "epoch": 4.4983108108108105, + "grad_norm": 0.6301610469818115, + "learning_rate": 1.5969e-05, + "loss": 0.068, + "step": 5326 + }, + { + "epoch": 4.499155405405405, + "grad_norm": 0.624972939491272, + "learning_rate": 1.5972e-05, + "loss": 0.0516, + "step": 5327 + }, + { + "epoch": 4.5, + "grad_norm": 0.8221753239631653, + "learning_rate": 1.5975e-05, + "loss": 0.0545, + "step": 5328 + }, + { + "epoch": 4.500844594594595, + "grad_norm": 0.9089149832725525, + "learning_rate": 1.5978e-05, + "loss": 0.0813, + "step": 5329 + }, + { + "epoch": 4.5016891891891895, + "grad_norm": 0.6328731179237366, + "learning_rate": 1.5980999999999998e-05, + "loss": 0.0535, + "step": 5330 + }, + { + "epoch": 4.502533783783784, + "grad_norm": 1.3084322214126587, + "learning_rate": 1.5984e-05, + "loss": 0.0803, + "step": 5331 + }, + { + "epoch": 4.503378378378378, + "grad_norm": 0.808246374130249, + "learning_rate": 1.5987e-05, + "loss": 0.0672, + "step": 5332 + }, + { + "epoch": 4.504222972972973, + "grad_norm": 0.7401777505874634, + "learning_rate": 1.599e-05, + "loss": 0.0429, + "step": 5333 + }, + { + "epoch": 4.5050675675675675, + "grad_norm": 1.2076795101165771, + "learning_rate": 1.5993e-05, + "loss": 0.0834, + "step": 5334 + }, + { + "epoch": 4.505912162162162, + "grad_norm": 1.42909836769104, + "learning_rate": 1.5996e-05, + "loss": 0.1034, + "step": 5335 + }, + { + "epoch": 4.506756756756757, + "grad_norm": 1.3943357467651367, + "learning_rate": 1.5999e-05, + "loss": 0.1097, + "step": 5336 + }, + { + "epoch": 4.507601351351351, + "grad_norm": 1.3018845319747925, + "learning_rate": 1.6002e-05, + "loss": 0.4389, + "step": 5337 + }, + { + "epoch": 4.508445945945946, + "grad_norm": 0.8437857627868652, + "learning_rate": 1.6005e-05, + "loss": 0.1964, + "step": 5338 + }, + { + "epoch": 4.50929054054054, + "grad_norm": 0.5815865397453308, + "learning_rate": 1.6008e-05, + "loss": 0.1428, + "step": 5339 + }, + { + "epoch": 4.510135135135135, + "grad_norm": 0.7429203391075134, + "learning_rate": 1.6011e-05, + "loss": 0.0778, + "step": 5340 + }, + { + "epoch": 4.51097972972973, + "grad_norm": 0.7239660620689392, + "learning_rate": 1.6014000000000003e-05, + "loss": 0.0976, + "step": 5341 + }, + { + "epoch": 4.511824324324325, + "grad_norm": 0.4305734932422638, + "learning_rate": 1.6017000000000003e-05, + "loss": 0.0629, + "step": 5342 + }, + { + "epoch": 4.512668918918919, + "grad_norm": 0.4426572918891907, + "learning_rate": 1.6020000000000002e-05, + "loss": 0.0677, + "step": 5343 + }, + { + "epoch": 4.513513513513513, + "grad_norm": 0.5053314566612244, + "learning_rate": 1.6023000000000002e-05, + "loss": 0.0434, + "step": 5344 + }, + { + "epoch": 4.514358108108108, + "grad_norm": 0.3959544003009796, + "learning_rate": 1.6026000000000002e-05, + "loss": 0.0521, + "step": 5345 + }, + { + "epoch": 4.515202702702703, + "grad_norm": 0.5415417551994324, + "learning_rate": 1.6029000000000002e-05, + "loss": 0.077, + "step": 5346 + }, + { + "epoch": 4.516047297297297, + "grad_norm": 0.5541843175888062, + "learning_rate": 1.6032e-05, + "loss": 0.0441, + "step": 5347 + }, + { + "epoch": 4.516891891891892, + "grad_norm": 0.5933812856674194, + "learning_rate": 1.6034999999999998e-05, + "loss": 0.056, + "step": 5348 + }, + { + "epoch": 4.517736486486487, + "grad_norm": 0.6260378956794739, + "learning_rate": 1.6037999999999998e-05, + "loss": 0.0768, + "step": 5349 + }, + { + "epoch": 4.518581081081081, + "grad_norm": 0.5637577176094055, + "learning_rate": 1.6040999999999998e-05, + "loss": 0.0511, + "step": 5350 + }, + { + "epoch": 4.519425675675675, + "grad_norm": 0.5659944415092468, + "learning_rate": 1.6044e-05, + "loss": 0.0571, + "step": 5351 + }, + { + "epoch": 4.52027027027027, + "grad_norm": 0.5037603974342346, + "learning_rate": 1.6047e-05, + "loss": 0.063, + "step": 5352 + }, + { + "epoch": 4.521114864864865, + "grad_norm": 0.5560306310653687, + "learning_rate": 1.605e-05, + "loss": 0.0671, + "step": 5353 + }, + { + "epoch": 4.52195945945946, + "grad_norm": 0.5305623412132263, + "learning_rate": 1.6053e-05, + "loss": 0.0508, + "step": 5354 + }, + { + "epoch": 4.522804054054054, + "grad_norm": 0.7153782844543457, + "learning_rate": 1.6056e-05, + "loss": 0.066, + "step": 5355 + }, + { + "epoch": 4.523648648648649, + "grad_norm": 0.5356406569480896, + "learning_rate": 1.6059e-05, + "loss": 0.0425, + "step": 5356 + }, + { + "epoch": 4.524493243243243, + "grad_norm": 0.6721763610839844, + "learning_rate": 1.6062e-05, + "loss": 0.0414, + "step": 5357 + }, + { + "epoch": 4.525337837837838, + "grad_norm": 0.6138339638710022, + "learning_rate": 1.6065e-05, + "loss": 0.0553, + "step": 5358 + }, + { + "epoch": 4.5261824324324325, + "grad_norm": 0.6481594443321228, + "learning_rate": 1.6068e-05, + "loss": 0.0574, + "step": 5359 + }, + { + "epoch": 4.527027027027027, + "grad_norm": 0.6725938320159912, + "learning_rate": 1.6071e-05, + "loss": 0.0595, + "step": 5360 + }, + { + "epoch": 4.527871621621622, + "grad_norm": 0.5353283286094666, + "learning_rate": 1.6074000000000002e-05, + "loss": 0.0662, + "step": 5361 + }, + { + "epoch": 4.528716216216216, + "grad_norm": 0.5040176510810852, + "learning_rate": 1.6077000000000002e-05, + "loss": 0.041, + "step": 5362 + }, + { + "epoch": 4.5295608108108105, + "grad_norm": 0.7655412554740906, + "learning_rate": 1.6080000000000002e-05, + "loss": 0.0648, + "step": 5363 + }, + { + "epoch": 4.530405405405405, + "grad_norm": 0.55818772315979, + "learning_rate": 1.6083000000000002e-05, + "loss": 0.0634, + "step": 5364 + }, + { + "epoch": 4.53125, + "grad_norm": 0.5545558929443359, + "learning_rate": 1.6086e-05, + "loss": 0.0456, + "step": 5365 + }, + { + "epoch": 4.532094594594595, + "grad_norm": 0.7824040651321411, + "learning_rate": 1.6089e-05, + "loss": 0.0662, + "step": 5366 + }, + { + "epoch": 4.5329391891891895, + "grad_norm": 0.8503099083900452, + "learning_rate": 1.6092e-05, + "loss": 0.0573, + "step": 5367 + }, + { + "epoch": 4.533783783783784, + "grad_norm": 0.5391951203346252, + "learning_rate": 1.6095e-05, + "loss": 0.0472, + "step": 5368 + }, + { + "epoch": 4.534628378378378, + "grad_norm": 0.7189595103263855, + "learning_rate": 1.6098e-05, + "loss": 0.0551, + "step": 5369 + }, + { + "epoch": 4.535472972972973, + "grad_norm": 0.7475507259368896, + "learning_rate": 1.6101e-05, + "loss": 0.0521, + "step": 5370 + }, + { + "epoch": 4.5363175675675675, + "grad_norm": 0.5766203999519348, + "learning_rate": 1.6104000000000004e-05, + "loss": 0.0642, + "step": 5371 + }, + { + "epoch": 4.537162162162162, + "grad_norm": 0.6805794835090637, + "learning_rate": 1.6107e-05, + "loss": 0.0687, + "step": 5372 + }, + { + "epoch": 4.538006756756757, + "grad_norm": 0.7535654306411743, + "learning_rate": 1.611e-05, + "loss": 0.0487, + "step": 5373 + }, + { + "epoch": 4.538851351351351, + "grad_norm": 0.6450232863426208, + "learning_rate": 1.6113e-05, + "loss": 0.0586, + "step": 5374 + }, + { + "epoch": 4.539695945945946, + "grad_norm": 0.6489740610122681, + "learning_rate": 1.6116e-05, + "loss": 0.0609, + "step": 5375 + }, + { + "epoch": 4.54054054054054, + "grad_norm": 0.6373089551925659, + "learning_rate": 1.6119e-05, + "loss": 0.0458, + "step": 5376 + }, + { + "epoch": 4.541385135135135, + "grad_norm": 0.6753019690513611, + "learning_rate": 1.6122e-05, + "loss": 0.0783, + "step": 5377 + }, + { + "epoch": 4.54222972972973, + "grad_norm": 0.6403706073760986, + "learning_rate": 1.6125e-05, + "loss": 0.0597, + "step": 5378 + }, + { + "epoch": 4.543074324324325, + "grad_norm": 0.7513359189033508, + "learning_rate": 1.6128e-05, + "loss": 0.0845, + "step": 5379 + }, + { + "epoch": 4.543918918918919, + "grad_norm": 0.524751603603363, + "learning_rate": 1.6131e-05, + "loss": 0.0535, + "step": 5380 + }, + { + "epoch": 4.544763513513513, + "grad_norm": 0.7826579809188843, + "learning_rate": 1.6134e-05, + "loss": 0.0578, + "step": 5381 + }, + { + "epoch": 4.545608108108108, + "grad_norm": 0.8104684352874756, + "learning_rate": 1.6137e-05, + "loss": 0.0575, + "step": 5382 + }, + { + "epoch": 4.546452702702703, + "grad_norm": 0.7464123368263245, + "learning_rate": 1.614e-05, + "loss": 0.063, + "step": 5383 + }, + { + "epoch": 4.547297297297297, + "grad_norm": 0.7550793290138245, + "learning_rate": 1.6143e-05, + "loss": 0.0683, + "step": 5384 + }, + { + "epoch": 4.548141891891892, + "grad_norm": 1.1352920532226562, + "learning_rate": 1.6146e-05, + "loss": 0.1062, + "step": 5385 + }, + { + "epoch": 4.548986486486487, + "grad_norm": 2.0754995346069336, + "learning_rate": 1.6149e-05, + "loss": 0.1388, + "step": 5386 + }, + { + "epoch": 4.549831081081081, + "grad_norm": 1.1141464710235596, + "learning_rate": 1.6152e-05, + "loss": 0.3461, + "step": 5387 + }, + { + "epoch": 4.550675675675675, + "grad_norm": 1.21857750415802, + "learning_rate": 1.6155e-05, + "loss": 0.3198, + "step": 5388 + }, + { + "epoch": 4.55152027027027, + "grad_norm": 0.6944521069526672, + "learning_rate": 1.6158e-05, + "loss": 0.1129, + "step": 5389 + }, + { + "epoch": 4.552364864864865, + "grad_norm": 0.5013948082923889, + "learning_rate": 1.6161e-05, + "loss": 0.0692, + "step": 5390 + }, + { + "epoch": 4.55320945945946, + "grad_norm": 0.47754034399986267, + "learning_rate": 1.6164e-05, + "loss": 0.0579, + "step": 5391 + }, + { + "epoch": 4.554054054054054, + "grad_norm": 0.46976733207702637, + "learning_rate": 1.6167000000000003e-05, + "loss": 0.0641, + "step": 5392 + }, + { + "epoch": 4.554898648648649, + "grad_norm": 0.44793960452079773, + "learning_rate": 1.6170000000000003e-05, + "loss": 0.0564, + "step": 5393 + }, + { + "epoch": 4.555743243243243, + "grad_norm": 0.4524334967136383, + "learning_rate": 1.6173000000000003e-05, + "loss": 0.0532, + "step": 5394 + }, + { + "epoch": 4.556587837837838, + "grad_norm": 0.4304547607898712, + "learning_rate": 1.6176000000000002e-05, + "loss": 0.0525, + "step": 5395 + }, + { + "epoch": 4.5574324324324325, + "grad_norm": 0.5377526879310608, + "learning_rate": 1.6179000000000002e-05, + "loss": 0.0621, + "step": 5396 + }, + { + "epoch": 4.558277027027027, + "grad_norm": 0.38059037923812866, + "learning_rate": 1.6182e-05, + "loss": 0.0364, + "step": 5397 + }, + { + "epoch": 4.559121621621622, + "grad_norm": 0.6563615798950195, + "learning_rate": 1.6185e-05, + "loss": 0.0427, + "step": 5398 + }, + { + "epoch": 4.559966216216216, + "grad_norm": 0.5824326276779175, + "learning_rate": 1.6187999999999998e-05, + "loss": 0.045, + "step": 5399 + }, + { + "epoch": 4.5608108108108105, + "grad_norm": 0.5176152586936951, + "learning_rate": 1.6190999999999998e-05, + "loss": 0.0746, + "step": 5400 + }, + { + "epoch": 4.561655405405405, + "grad_norm": 0.5918850898742676, + "learning_rate": 1.6193999999999998e-05, + "loss": 0.0502, + "step": 5401 + }, + { + "epoch": 4.5625, + "grad_norm": 0.5432378649711609, + "learning_rate": 1.6197e-05, + "loss": 0.056, + "step": 5402 + }, + { + "epoch": 4.563344594594595, + "grad_norm": 0.5529295206069946, + "learning_rate": 1.62e-05, + "loss": 0.0408, + "step": 5403 + }, + { + "epoch": 4.5641891891891895, + "grad_norm": 0.468763142824173, + "learning_rate": 1.6203e-05, + "loss": 0.0434, + "step": 5404 + }, + { + "epoch": 4.565033783783784, + "grad_norm": 0.4834555685520172, + "learning_rate": 1.6206e-05, + "loss": 0.0572, + "step": 5405 + }, + { + "epoch": 4.565878378378378, + "grad_norm": 0.5159100294113159, + "learning_rate": 1.6209e-05, + "loss": 0.0483, + "step": 5406 + }, + { + "epoch": 4.566722972972973, + "grad_norm": 0.5545820593833923, + "learning_rate": 1.6212e-05, + "loss": 0.0424, + "step": 5407 + }, + { + "epoch": 4.5675675675675675, + "grad_norm": 0.6570683717727661, + "learning_rate": 1.6215e-05, + "loss": 0.0669, + "step": 5408 + }, + { + "epoch": 4.568412162162162, + "grad_norm": 0.5338613986968994, + "learning_rate": 1.6218e-05, + "loss": 0.0445, + "step": 5409 + }, + { + "epoch": 4.569256756756757, + "grad_norm": 0.6133916974067688, + "learning_rate": 1.6221e-05, + "loss": 0.0651, + "step": 5410 + }, + { + "epoch": 4.570101351351351, + "grad_norm": 0.44685718417167664, + "learning_rate": 1.6224e-05, + "loss": 0.0408, + "step": 5411 + }, + { + "epoch": 4.570945945945946, + "grad_norm": 0.6291643977165222, + "learning_rate": 1.6227000000000002e-05, + "loss": 0.0891, + "step": 5412 + }, + { + "epoch": 4.57179054054054, + "grad_norm": 0.6077958941459656, + "learning_rate": 1.6230000000000002e-05, + "loss": 0.0484, + "step": 5413 + }, + { + "epoch": 4.572635135135135, + "grad_norm": 0.5444808602333069, + "learning_rate": 1.6233000000000002e-05, + "loss": 0.0541, + "step": 5414 + }, + { + "epoch": 4.57347972972973, + "grad_norm": 0.5074161291122437, + "learning_rate": 1.6236000000000002e-05, + "loss": 0.0586, + "step": 5415 + }, + { + "epoch": 4.574324324324325, + "grad_norm": 0.552762508392334, + "learning_rate": 1.6239e-05, + "loss": 0.0471, + "step": 5416 + }, + { + "epoch": 4.575168918918919, + "grad_norm": 0.4765016436576843, + "learning_rate": 1.6242e-05, + "loss": 0.0524, + "step": 5417 + }, + { + "epoch": 4.576013513513513, + "grad_norm": 0.40983694791793823, + "learning_rate": 1.6245e-05, + "loss": 0.0456, + "step": 5418 + }, + { + "epoch": 4.576858108108108, + "grad_norm": 0.5186775922775269, + "learning_rate": 1.6248e-05, + "loss": 0.0493, + "step": 5419 + }, + { + "epoch": 4.577702702702703, + "grad_norm": 0.8250090479850769, + "learning_rate": 1.6251e-05, + "loss": 0.0645, + "step": 5420 + }, + { + "epoch": 4.578547297297297, + "grad_norm": 0.7789498567581177, + "learning_rate": 1.6253999999999997e-05, + "loss": 0.0633, + "step": 5421 + }, + { + "epoch": 4.579391891891892, + "grad_norm": 0.43418246507644653, + "learning_rate": 1.6257e-05, + "loss": 0.039, + "step": 5422 + }, + { + "epoch": 4.580236486486487, + "grad_norm": 0.4592551290988922, + "learning_rate": 1.626e-05, + "loss": 0.0515, + "step": 5423 + }, + { + "epoch": 4.581081081081081, + "grad_norm": 1.0573437213897705, + "learning_rate": 1.6263e-05, + "loss": 0.0701, + "step": 5424 + }, + { + "epoch": 4.581925675675675, + "grad_norm": 0.6678667068481445, + "learning_rate": 1.6266e-05, + "loss": 0.0645, + "step": 5425 + }, + { + "epoch": 4.58277027027027, + "grad_norm": 0.7506866455078125, + "learning_rate": 1.6269e-05, + "loss": 0.0687, + "step": 5426 + }, + { + "epoch": 4.583614864864865, + "grad_norm": 0.80996173620224, + "learning_rate": 1.6272e-05, + "loss": 0.0656, + "step": 5427 + }, + { + "epoch": 4.58445945945946, + "grad_norm": 0.7084985971450806, + "learning_rate": 1.6275e-05, + "loss": 0.0631, + "step": 5428 + }, + { + "epoch": 4.585304054054054, + "grad_norm": 0.7053719162940979, + "learning_rate": 1.6278e-05, + "loss": 0.0649, + "step": 5429 + }, + { + "epoch": 4.586148648648649, + "grad_norm": 0.5596732497215271, + "learning_rate": 1.6281e-05, + "loss": 0.0475, + "step": 5430 + }, + { + "epoch": 4.586993243243243, + "grad_norm": 0.7429049015045166, + "learning_rate": 1.6284e-05, + "loss": 0.0982, + "step": 5431 + }, + { + "epoch": 4.587837837837838, + "grad_norm": 0.5740141868591309, + "learning_rate": 1.6287000000000002e-05, + "loss": 0.0596, + "step": 5432 + }, + { + "epoch": 4.5886824324324325, + "grad_norm": 0.7364721894264221, + "learning_rate": 1.629e-05, + "loss": 0.055, + "step": 5433 + }, + { + "epoch": 4.589527027027027, + "grad_norm": 1.2055492401123047, + "learning_rate": 1.6293e-05, + "loss": 0.0952, + "step": 5434 + }, + { + "epoch": 4.590371621621622, + "grad_norm": 0.9361098408699036, + "learning_rate": 1.6296e-05, + "loss": 0.1163, + "step": 5435 + }, + { + "epoch": 4.591216216216216, + "grad_norm": 1.6204766035079956, + "learning_rate": 1.6299e-05, + "loss": 0.155, + "step": 5436 + }, + { + "epoch": 4.5920608108108105, + "grad_norm": 2.167940378189087, + "learning_rate": 1.6302e-05, + "loss": 0.4509, + "step": 5437 + }, + { + "epoch": 4.592905405405405, + "grad_norm": 1.0365160703659058, + "learning_rate": 1.6305e-05, + "loss": 0.2802, + "step": 5438 + }, + { + "epoch": 4.59375, + "grad_norm": 0.8541480302810669, + "learning_rate": 1.6308e-05, + "loss": 0.1431, + "step": 5439 + }, + { + "epoch": 4.594594594594595, + "grad_norm": 0.8081040978431702, + "learning_rate": 1.6311e-05, + "loss": 0.101, + "step": 5440 + }, + { + "epoch": 4.5954391891891895, + "grad_norm": 0.6348864436149597, + "learning_rate": 1.6314e-05, + "loss": 0.0766, + "step": 5441 + }, + { + "epoch": 4.596283783783784, + "grad_norm": 0.45235341787338257, + "learning_rate": 1.6317000000000003e-05, + "loss": 0.0608, + "step": 5442 + }, + { + "epoch": 4.597128378378378, + "grad_norm": 0.3616316318511963, + "learning_rate": 1.6320000000000003e-05, + "loss": 0.0313, + "step": 5443 + }, + { + "epoch": 4.597972972972973, + "grad_norm": 0.6195970177650452, + "learning_rate": 1.6323000000000003e-05, + "loss": 0.0533, + "step": 5444 + }, + { + "epoch": 4.5988175675675675, + "grad_norm": 0.5284826755523682, + "learning_rate": 1.6326000000000003e-05, + "loss": 0.0727, + "step": 5445 + }, + { + "epoch": 4.599662162162162, + "grad_norm": 0.5334276556968689, + "learning_rate": 1.6329e-05, + "loss": 0.0587, + "step": 5446 + }, + { + "epoch": 4.600506756756757, + "grad_norm": 0.561461329460144, + "learning_rate": 1.6332e-05, + "loss": 0.0678, + "step": 5447 + }, + { + "epoch": 4.601351351351351, + "grad_norm": 0.6040999889373779, + "learning_rate": 1.6335e-05, + "loss": 0.0432, + "step": 5448 + }, + { + "epoch": 4.602195945945946, + "grad_norm": 0.5098923444747925, + "learning_rate": 1.6338e-05, + "loss": 0.0414, + "step": 5449 + }, + { + "epoch": 4.60304054054054, + "grad_norm": 0.628916323184967, + "learning_rate": 1.6340999999999998e-05, + "loss": 0.0624, + "step": 5450 + }, + { + "epoch": 4.603885135135135, + "grad_norm": 0.5270816087722778, + "learning_rate": 1.6343999999999998e-05, + "loss": 0.0538, + "step": 5451 + }, + { + "epoch": 4.60472972972973, + "grad_norm": 0.5248250961303711, + "learning_rate": 1.6347e-05, + "loss": 0.0597, + "step": 5452 + }, + { + "epoch": 4.605574324324325, + "grad_norm": 0.5423628091812134, + "learning_rate": 1.635e-05, + "loss": 0.0703, + "step": 5453 + }, + { + "epoch": 4.606418918918919, + "grad_norm": 0.7432615160942078, + "learning_rate": 1.6353e-05, + "loss": 0.0475, + "step": 5454 + }, + { + "epoch": 4.607263513513513, + "grad_norm": 0.7227054238319397, + "learning_rate": 1.6356e-05, + "loss": 0.0587, + "step": 5455 + }, + { + "epoch": 4.608108108108108, + "grad_norm": 0.5577089190483093, + "learning_rate": 1.6359e-05, + "loss": 0.0421, + "step": 5456 + }, + { + "epoch": 4.608952702702703, + "grad_norm": 0.5722436308860779, + "learning_rate": 1.6362e-05, + "loss": 0.0538, + "step": 5457 + }, + { + "epoch": 4.609797297297297, + "grad_norm": 0.34566327929496765, + "learning_rate": 1.6365e-05, + "loss": 0.0476, + "step": 5458 + }, + { + "epoch": 4.610641891891892, + "grad_norm": 0.6068854331970215, + "learning_rate": 1.6368e-05, + "loss": 0.0516, + "step": 5459 + }, + { + "epoch": 4.611486486486487, + "grad_norm": 0.605590283870697, + "learning_rate": 1.6371e-05, + "loss": 0.0535, + "step": 5460 + }, + { + "epoch": 4.612331081081081, + "grad_norm": 0.8784629702568054, + "learning_rate": 1.6374e-05, + "loss": 0.0557, + "step": 5461 + }, + { + "epoch": 4.613175675675675, + "grad_norm": 0.6169149875640869, + "learning_rate": 1.6377000000000003e-05, + "loss": 0.0489, + "step": 5462 + }, + { + "epoch": 4.61402027027027, + "grad_norm": 0.47094881534576416, + "learning_rate": 1.6380000000000002e-05, + "loss": 0.0383, + "step": 5463 + }, + { + "epoch": 4.614864864864865, + "grad_norm": 0.5335904955863953, + "learning_rate": 1.6383000000000002e-05, + "loss": 0.047, + "step": 5464 + }, + { + "epoch": 4.61570945945946, + "grad_norm": 0.43236953020095825, + "learning_rate": 1.6386000000000002e-05, + "loss": 0.0501, + "step": 5465 + }, + { + "epoch": 4.616554054054054, + "grad_norm": 0.5294243693351746, + "learning_rate": 1.6389000000000002e-05, + "loss": 0.0518, + "step": 5466 + }, + { + "epoch": 4.617398648648649, + "grad_norm": 0.7957121133804321, + "learning_rate": 1.6392e-05, + "loss": 0.0567, + "step": 5467 + }, + { + "epoch": 4.618243243243243, + "grad_norm": 0.566878080368042, + "learning_rate": 1.6395e-05, + "loss": 0.0548, + "step": 5468 + }, + { + "epoch": 4.619087837837838, + "grad_norm": 0.550688624382019, + "learning_rate": 1.6398e-05, + "loss": 0.0595, + "step": 5469 + }, + { + "epoch": 4.6199324324324325, + "grad_norm": 0.6553354263305664, + "learning_rate": 1.6400999999999998e-05, + "loss": 0.0441, + "step": 5470 + }, + { + "epoch": 4.620777027027027, + "grad_norm": 0.3417656123638153, + "learning_rate": 1.6403999999999997e-05, + "loss": 0.0328, + "step": 5471 + }, + { + "epoch": 4.621621621621622, + "grad_norm": 0.5982502102851868, + "learning_rate": 1.6407e-05, + "loss": 0.0578, + "step": 5472 + }, + { + "epoch": 4.622466216216216, + "grad_norm": 0.5211771130561829, + "learning_rate": 1.641e-05, + "loss": 0.0557, + "step": 5473 + }, + { + "epoch": 4.6233108108108105, + "grad_norm": 0.7739983797073364, + "learning_rate": 1.6413e-05, + "loss": 0.0663, + "step": 5474 + }, + { + "epoch": 4.624155405405405, + "grad_norm": 0.3638344407081604, + "learning_rate": 1.6416e-05, + "loss": 0.045, + "step": 5475 + }, + { + "epoch": 4.625, + "grad_norm": 1.0330721139907837, + "learning_rate": 1.6419e-05, + "loss": 0.0701, + "step": 5476 + }, + { + "epoch": 4.625844594594595, + "grad_norm": 0.8450912833213806, + "learning_rate": 1.6422e-05, + "loss": 0.0646, + "step": 5477 + }, + { + "epoch": 4.6266891891891895, + "grad_norm": 0.5245097279548645, + "learning_rate": 1.6425e-05, + "loss": 0.0428, + "step": 5478 + }, + { + "epoch": 4.627533783783784, + "grad_norm": 0.799034059047699, + "learning_rate": 1.6428e-05, + "loss": 0.0624, + "step": 5479 + }, + { + "epoch": 4.628378378378378, + "grad_norm": 0.7811905741691589, + "learning_rate": 1.6431e-05, + "loss": 0.054, + "step": 5480 + }, + { + "epoch": 4.629222972972973, + "grad_norm": 1.5127700567245483, + "learning_rate": 1.6434e-05, + "loss": 0.1693, + "step": 5481 + }, + { + "epoch": 4.6300675675675675, + "grad_norm": 0.677727222442627, + "learning_rate": 1.6437000000000002e-05, + "loss": 0.049, + "step": 5482 + }, + { + "epoch": 4.630912162162162, + "grad_norm": 0.9104031324386597, + "learning_rate": 1.6440000000000002e-05, + "loss": 0.0766, + "step": 5483 + }, + { + "epoch": 4.631756756756757, + "grad_norm": 0.6769290566444397, + "learning_rate": 1.6443e-05, + "loss": 0.0916, + "step": 5484 + }, + { + "epoch": 4.632601351351351, + "grad_norm": 0.894454836845398, + "learning_rate": 1.6446e-05, + "loss": 0.1153, + "step": 5485 + }, + { + "epoch": 4.633445945945946, + "grad_norm": 1.7054411172866821, + "learning_rate": 1.6449e-05, + "loss": 0.132, + "step": 5486 + }, + { + "epoch": 4.63429054054054, + "grad_norm": 3.166762351989746, + "learning_rate": 1.6452e-05, + "loss": 0.4255, + "step": 5487 + }, + { + "epoch": 4.635135135135135, + "grad_norm": 0.7585452795028687, + "learning_rate": 1.6455e-05, + "loss": 0.2107, + "step": 5488 + }, + { + "epoch": 4.63597972972973, + "grad_norm": 0.7580211162567139, + "learning_rate": 1.6458e-05, + "loss": 0.1624, + "step": 5489 + }, + { + "epoch": 4.636824324324325, + "grad_norm": 1.415305256843567, + "learning_rate": 1.6461e-05, + "loss": 0.0887, + "step": 5490 + }, + { + "epoch": 4.637668918918919, + "grad_norm": 0.4979839622974396, + "learning_rate": 1.6464e-05, + "loss": 0.0676, + "step": 5491 + }, + { + "epoch": 4.638513513513513, + "grad_norm": 0.49806928634643555, + "learning_rate": 1.6467000000000003e-05, + "loss": 0.0799, + "step": 5492 + }, + { + "epoch": 4.639358108108108, + "grad_norm": 0.6990793943405151, + "learning_rate": 1.6470000000000003e-05, + "loss": 0.0695, + "step": 5493 + }, + { + "epoch": 4.640202702702703, + "grad_norm": 0.6900742053985596, + "learning_rate": 1.6473000000000003e-05, + "loss": 0.0526, + "step": 5494 + }, + { + "epoch": 4.641047297297297, + "grad_norm": 0.42689403891563416, + "learning_rate": 1.6476e-05, + "loss": 0.0587, + "step": 5495 + }, + { + "epoch": 4.641891891891892, + "grad_norm": 0.5893619060516357, + "learning_rate": 1.6479e-05, + "loss": 0.0844, + "step": 5496 + }, + { + "epoch": 4.642736486486487, + "grad_norm": 0.5333355665206909, + "learning_rate": 1.6482e-05, + "loss": 0.0365, + "step": 5497 + }, + { + "epoch": 4.643581081081081, + "grad_norm": 0.5503154993057251, + "learning_rate": 1.6485e-05, + "loss": 0.0537, + "step": 5498 + }, + { + "epoch": 4.644425675675675, + "grad_norm": 0.4933074116706848, + "learning_rate": 1.6488e-05, + "loss": 0.0531, + "step": 5499 + }, + { + "epoch": 4.64527027027027, + "grad_norm": 0.5913088917732239, + "learning_rate": 1.6491e-05, + "loss": 0.075, + "step": 5500 + }, + { + "epoch": 4.646114864864865, + "grad_norm": 0.5199641585350037, + "learning_rate": 1.6493999999999998e-05, + "loss": 0.0564, + "step": 5501 + }, + { + "epoch": 4.64695945945946, + "grad_norm": 0.5967236757278442, + "learning_rate": 1.6497e-05, + "loss": 0.0603, + "step": 5502 + }, + { + "epoch": 4.647804054054054, + "grad_norm": 0.4646868109703064, + "learning_rate": 1.65e-05, + "loss": 0.0586, + "step": 5503 + }, + { + "epoch": 4.648648648648649, + "grad_norm": 0.45207104086875916, + "learning_rate": 1.6503e-05, + "loss": 0.0574, + "step": 5504 + }, + { + "epoch": 4.649493243243243, + "grad_norm": 0.5936165452003479, + "learning_rate": 1.6506e-05, + "loss": 0.0546, + "step": 5505 + }, + { + "epoch": 4.650337837837838, + "grad_norm": 0.5914079546928406, + "learning_rate": 1.6509e-05, + "loss": 0.0536, + "step": 5506 + }, + { + "epoch": 4.6511824324324325, + "grad_norm": 0.46666693687438965, + "learning_rate": 1.6512e-05, + "loss": 0.0461, + "step": 5507 + }, + { + "epoch": 4.652027027027027, + "grad_norm": 0.47415411472320557, + "learning_rate": 1.6515e-05, + "loss": 0.062, + "step": 5508 + }, + { + "epoch": 4.652871621621622, + "grad_norm": 0.48921501636505127, + "learning_rate": 1.6518e-05, + "loss": 0.0562, + "step": 5509 + }, + { + "epoch": 4.653716216216216, + "grad_norm": 0.40402477979660034, + "learning_rate": 1.6521e-05, + "loss": 0.0428, + "step": 5510 + }, + { + "epoch": 4.6545608108108105, + "grad_norm": 0.40634673833847046, + "learning_rate": 1.6524e-05, + "loss": 0.055, + "step": 5511 + }, + { + "epoch": 4.655405405405405, + "grad_norm": 0.49655967950820923, + "learning_rate": 1.6527e-05, + "loss": 0.0369, + "step": 5512 + }, + { + "epoch": 4.65625, + "grad_norm": 1.4466661214828491, + "learning_rate": 1.6530000000000003e-05, + "loss": 0.0674, + "step": 5513 + }, + { + "epoch": 4.657094594594595, + "grad_norm": 0.6513037085533142, + "learning_rate": 1.6533000000000002e-05, + "loss": 0.0707, + "step": 5514 + }, + { + "epoch": 4.6579391891891895, + "grad_norm": 0.6361933350563049, + "learning_rate": 1.6536000000000002e-05, + "loss": 0.0504, + "step": 5515 + }, + { + "epoch": 4.658783783783784, + "grad_norm": 0.5036759376525879, + "learning_rate": 1.6539000000000002e-05, + "loss": 0.04, + "step": 5516 + }, + { + "epoch": 4.659628378378378, + "grad_norm": 0.5309173464775085, + "learning_rate": 1.6542000000000002e-05, + "loss": 0.048, + "step": 5517 + }, + { + "epoch": 4.660472972972973, + "grad_norm": 1.1079068183898926, + "learning_rate": 1.6545e-05, + "loss": 0.0801, + "step": 5518 + }, + { + "epoch": 4.6613175675675675, + "grad_norm": 0.5201704502105713, + "learning_rate": 1.6548e-05, + "loss": 0.0694, + "step": 5519 + }, + { + "epoch": 4.662162162162162, + "grad_norm": 0.43828678131103516, + "learning_rate": 1.6550999999999998e-05, + "loss": 0.0368, + "step": 5520 + }, + { + "epoch": 4.663006756756757, + "grad_norm": 0.7237197756767273, + "learning_rate": 1.6553999999999998e-05, + "loss": 0.058, + "step": 5521 + }, + { + "epoch": 4.663851351351351, + "grad_norm": 0.5656716227531433, + "learning_rate": 1.6556999999999998e-05, + "loss": 0.0592, + "step": 5522 + }, + { + "epoch": 4.664695945945946, + "grad_norm": 0.5414818525314331, + "learning_rate": 1.656e-05, + "loss": 0.0577, + "step": 5523 + }, + { + "epoch": 4.66554054054054, + "grad_norm": 0.3809826076030731, + "learning_rate": 1.6563e-05, + "loss": 0.035, + "step": 5524 + }, + { + "epoch": 4.666385135135135, + "grad_norm": 0.7896703481674194, + "learning_rate": 1.6566e-05, + "loss": 0.0747, + "step": 5525 + }, + { + "epoch": 4.66722972972973, + "grad_norm": 0.6982415318489075, + "learning_rate": 1.6569e-05, + "loss": 0.0608, + "step": 5526 + }, + { + "epoch": 4.668074324324325, + "grad_norm": 0.6664999723434448, + "learning_rate": 1.6572e-05, + "loss": 0.0595, + "step": 5527 + }, + { + "epoch": 4.668918918918919, + "grad_norm": 0.6854673624038696, + "learning_rate": 1.6575e-05, + "loss": 0.052, + "step": 5528 + }, + { + "epoch": 4.669763513513513, + "grad_norm": 0.7859840393066406, + "learning_rate": 1.6578e-05, + "loss": 0.0822, + "step": 5529 + }, + { + "epoch": 4.670608108108108, + "grad_norm": 1.1380531787872314, + "learning_rate": 1.6581e-05, + "loss": 0.0634, + "step": 5530 + }, + { + "epoch": 4.671452702702703, + "grad_norm": 0.7174597382545471, + "learning_rate": 1.6584e-05, + "loss": 0.0653, + "step": 5531 + }, + { + "epoch": 4.672297297297297, + "grad_norm": 0.5764419436454773, + "learning_rate": 1.6587e-05, + "loss": 0.0452, + "step": 5532 + }, + { + "epoch": 4.673141891891892, + "grad_norm": 1.1555219888687134, + "learning_rate": 1.6590000000000002e-05, + "loss": 0.1066, + "step": 5533 + }, + { + "epoch": 4.673986486486487, + "grad_norm": 1.724372148513794, + "learning_rate": 1.6593000000000002e-05, + "loss": 0.0896, + "step": 5534 + }, + { + "epoch": 4.674831081081081, + "grad_norm": 0.9637579321861267, + "learning_rate": 1.6596000000000002e-05, + "loss": 0.0829, + "step": 5535 + }, + { + "epoch": 4.675675675675675, + "grad_norm": 1.398245096206665, + "learning_rate": 1.6599e-05, + "loss": 0.1058, + "step": 5536 + }, + { + "epoch": 4.67652027027027, + "grad_norm": 2.626197099685669, + "learning_rate": 1.6602e-05, + "loss": 0.4917, + "step": 5537 + }, + { + "epoch": 4.677364864864865, + "grad_norm": 0.7926445603370667, + "learning_rate": 1.6605e-05, + "loss": 0.1946, + "step": 5538 + }, + { + "epoch": 4.67820945945946, + "grad_norm": 0.7525971531867981, + "learning_rate": 1.6608e-05, + "loss": 0.1191, + "step": 5539 + }, + { + "epoch": 4.679054054054054, + "grad_norm": 0.5221357941627502, + "learning_rate": 1.6611e-05, + "loss": 0.0924, + "step": 5540 + }, + { + "epoch": 4.679898648648649, + "grad_norm": 0.6572176218032837, + "learning_rate": 1.6614e-05, + "loss": 0.0854, + "step": 5541 + }, + { + "epoch": 4.680743243243243, + "grad_norm": 0.4563084542751312, + "learning_rate": 1.6617e-05, + "loss": 0.0715, + "step": 5542 + }, + { + "epoch": 4.681587837837838, + "grad_norm": 0.4520714282989502, + "learning_rate": 1.6620000000000004e-05, + "loss": 0.0578, + "step": 5543 + }, + { + "epoch": 4.6824324324324325, + "grad_norm": 0.6761728525161743, + "learning_rate": 1.6623e-05, + "loss": 0.0735, + "step": 5544 + }, + { + "epoch": 4.683277027027027, + "grad_norm": 0.5593559741973877, + "learning_rate": 1.6626e-05, + "loss": 0.0672, + "step": 5545 + }, + { + "epoch": 4.684121621621622, + "grad_norm": 0.5228694677352905, + "learning_rate": 1.6629e-05, + "loss": 0.0513, + "step": 5546 + }, + { + "epoch": 4.684966216216216, + "grad_norm": 1.0017011165618896, + "learning_rate": 1.6632e-05, + "loss": 0.0691, + "step": 5547 + }, + { + "epoch": 4.6858108108108105, + "grad_norm": 0.6809964179992676, + "learning_rate": 1.6635e-05, + "loss": 0.0578, + "step": 5548 + }, + { + "epoch": 4.686655405405405, + "grad_norm": 0.5667756199836731, + "learning_rate": 1.6638e-05, + "loss": 0.0536, + "step": 5549 + }, + { + "epoch": 4.6875, + "grad_norm": 0.5753899812698364, + "learning_rate": 1.6641e-05, + "loss": 0.0846, + "step": 5550 + }, + { + "epoch": 4.688344594594595, + "grad_norm": 0.45798060297966003, + "learning_rate": 1.6644e-05, + "loss": 0.0545, + "step": 5551 + }, + { + "epoch": 4.6891891891891895, + "grad_norm": 0.5006407499313354, + "learning_rate": 1.6647e-05, + "loss": 0.0437, + "step": 5552 + }, + { + "epoch": 4.690033783783784, + "grad_norm": 0.5854042172431946, + "learning_rate": 1.665e-05, + "loss": 0.0525, + "step": 5553 + }, + { + "epoch": 4.690878378378378, + "grad_norm": 0.5363278388977051, + "learning_rate": 1.6653e-05, + "loss": 0.0545, + "step": 5554 + }, + { + "epoch": 4.691722972972973, + "grad_norm": 0.4628943204879761, + "learning_rate": 1.6656e-05, + "loss": 0.0543, + "step": 5555 + }, + { + "epoch": 4.6925675675675675, + "grad_norm": 0.6001430153846741, + "learning_rate": 1.6659e-05, + "loss": 0.0644, + "step": 5556 + }, + { + "epoch": 4.693412162162162, + "grad_norm": 0.47989729046821594, + "learning_rate": 1.6662e-05, + "loss": 0.0544, + "step": 5557 + }, + { + "epoch": 4.694256756756757, + "grad_norm": 0.5375245213508606, + "learning_rate": 1.6665e-05, + "loss": 0.0582, + "step": 5558 + }, + { + "epoch": 4.695101351351351, + "grad_norm": 0.8952156901359558, + "learning_rate": 1.6668e-05, + "loss": 0.1185, + "step": 5559 + }, + { + "epoch": 4.695945945945946, + "grad_norm": 0.4248574376106262, + "learning_rate": 1.6671e-05, + "loss": 0.0363, + "step": 5560 + }, + { + "epoch": 4.69679054054054, + "grad_norm": 0.5073045492172241, + "learning_rate": 1.6674e-05, + "loss": 0.0413, + "step": 5561 + }, + { + "epoch": 4.697635135135135, + "grad_norm": 0.6598958373069763, + "learning_rate": 1.6677e-05, + "loss": 0.0584, + "step": 5562 + }, + { + "epoch": 4.69847972972973, + "grad_norm": 0.4871501922607422, + "learning_rate": 1.6680000000000003e-05, + "loss": 0.0732, + "step": 5563 + }, + { + "epoch": 4.699324324324325, + "grad_norm": 0.3643069267272949, + "learning_rate": 1.6683000000000003e-05, + "loss": 0.0379, + "step": 5564 + }, + { + "epoch": 4.700168918918919, + "grad_norm": 0.47735193371772766, + "learning_rate": 1.6686000000000003e-05, + "loss": 0.067, + "step": 5565 + }, + { + "epoch": 4.701013513513513, + "grad_norm": 0.5434446930885315, + "learning_rate": 1.6689000000000002e-05, + "loss": 0.0512, + "step": 5566 + }, + { + "epoch": 4.701858108108108, + "grad_norm": 0.5685212016105652, + "learning_rate": 1.6692000000000002e-05, + "loss": 0.0582, + "step": 5567 + }, + { + "epoch": 4.702702702702703, + "grad_norm": 0.5000768899917603, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0631, + "step": 5568 + }, + { + "epoch": 4.703547297297297, + "grad_norm": 0.5299729704856873, + "learning_rate": 1.6698e-05, + "loss": 0.0587, + "step": 5569 + }, + { + "epoch": 4.704391891891892, + "grad_norm": 0.4458503723144531, + "learning_rate": 1.6700999999999998e-05, + "loss": 0.0508, + "step": 5570 + }, + { + "epoch": 4.705236486486487, + "grad_norm": 0.5022680163383484, + "learning_rate": 1.6703999999999998e-05, + "loss": 0.0493, + "step": 5571 + }, + { + "epoch": 4.706081081081081, + "grad_norm": 0.7613198161125183, + "learning_rate": 1.6706999999999998e-05, + "loss": 0.0691, + "step": 5572 + }, + { + "epoch": 4.706925675675675, + "grad_norm": 0.5584785342216492, + "learning_rate": 1.671e-05, + "loss": 0.0543, + "step": 5573 + }, + { + "epoch": 4.70777027027027, + "grad_norm": 0.7181403040885925, + "learning_rate": 1.6713e-05, + "loss": 0.1023, + "step": 5574 + }, + { + "epoch": 4.708614864864865, + "grad_norm": 0.6569494605064392, + "learning_rate": 1.6716e-05, + "loss": 0.0712, + "step": 5575 + }, + { + "epoch": 4.70945945945946, + "grad_norm": 0.7321613430976868, + "learning_rate": 1.6719e-05, + "loss": 0.0545, + "step": 5576 + }, + { + "epoch": 4.710304054054054, + "grad_norm": 0.6320468187332153, + "learning_rate": 1.6722e-05, + "loss": 0.0657, + "step": 5577 + }, + { + "epoch": 4.711148648648649, + "grad_norm": 0.57533198595047, + "learning_rate": 1.6725e-05, + "loss": 0.0427, + "step": 5578 + }, + { + "epoch": 4.711993243243243, + "grad_norm": 0.6790969967842102, + "learning_rate": 1.6728e-05, + "loss": 0.0701, + "step": 5579 + }, + { + "epoch": 4.712837837837838, + "grad_norm": 0.6434995532035828, + "learning_rate": 1.6731e-05, + "loss": 0.0703, + "step": 5580 + }, + { + "epoch": 4.7136824324324325, + "grad_norm": 0.9674848914146423, + "learning_rate": 1.6734e-05, + "loss": 0.0507, + "step": 5581 + }, + { + "epoch": 4.714527027027027, + "grad_norm": 0.7145622968673706, + "learning_rate": 1.6737e-05, + "loss": 0.0594, + "step": 5582 + }, + { + "epoch": 4.715371621621622, + "grad_norm": 1.4449117183685303, + "learning_rate": 1.6740000000000002e-05, + "loss": 0.0897, + "step": 5583 + }, + { + "epoch": 4.716216216216216, + "grad_norm": 0.7083641886711121, + "learning_rate": 1.6743000000000002e-05, + "loss": 0.0444, + "step": 5584 + }, + { + "epoch": 4.7170608108108105, + "grad_norm": 1.2270339727401733, + "learning_rate": 1.6746000000000002e-05, + "loss": 0.1179, + "step": 5585 + }, + { + "epoch": 4.717905405405405, + "grad_norm": 1.6381903886795044, + "learning_rate": 1.6749000000000002e-05, + "loss": 0.1824, + "step": 5586 + }, + { + "epoch": 4.71875, + "grad_norm": 1.205571174621582, + "learning_rate": 1.6752e-05, + "loss": 0.3958, + "step": 5587 + }, + { + "epoch": 4.719594594594595, + "grad_norm": 0.6692399382591248, + "learning_rate": 1.6755e-05, + "loss": 0.185, + "step": 5588 + }, + { + "epoch": 4.7204391891891895, + "grad_norm": 0.5169681906700134, + "learning_rate": 1.6758e-05, + "loss": 0.0753, + "step": 5589 + }, + { + "epoch": 4.721283783783784, + "grad_norm": 0.846386730670929, + "learning_rate": 1.6761e-05, + "loss": 0.0618, + "step": 5590 + }, + { + "epoch": 4.722128378378378, + "grad_norm": 0.5349334478378296, + "learning_rate": 1.6764e-05, + "loss": 0.0778, + "step": 5591 + }, + { + "epoch": 4.722972972972973, + "grad_norm": 0.5537841320037842, + "learning_rate": 1.6767e-05, + "loss": 0.0869, + "step": 5592 + }, + { + "epoch": 4.7238175675675675, + "grad_norm": 0.5434600710868835, + "learning_rate": 1.677e-05, + "loss": 0.076, + "step": 5593 + }, + { + "epoch": 4.724662162162162, + "grad_norm": 0.45717090368270874, + "learning_rate": 1.6773e-05, + "loss": 0.0561, + "step": 5594 + }, + { + "epoch": 4.725506756756757, + "grad_norm": 0.8013296127319336, + "learning_rate": 1.6776e-05, + "loss": 0.0545, + "step": 5595 + }, + { + "epoch": 4.726351351351351, + "grad_norm": 0.44061529636383057, + "learning_rate": 1.6779e-05, + "loss": 0.0542, + "step": 5596 + }, + { + "epoch": 4.727195945945946, + "grad_norm": 0.5887645483016968, + "learning_rate": 1.6782e-05, + "loss": 0.0461, + "step": 5597 + }, + { + "epoch": 4.72804054054054, + "grad_norm": 0.597136378288269, + "learning_rate": 1.6785e-05, + "loss": 0.075, + "step": 5598 + }, + { + "epoch": 4.728885135135135, + "grad_norm": 0.5341084599494934, + "learning_rate": 1.6788e-05, + "loss": 0.0495, + "step": 5599 + }, + { + "epoch": 4.72972972972973, + "grad_norm": 0.5551856756210327, + "learning_rate": 1.6791e-05, + "loss": 0.0471, + "step": 5600 + }, + { + "epoch": 4.730574324324325, + "grad_norm": 0.5708430409431458, + "learning_rate": 1.6794e-05, + "loss": 0.0472, + "step": 5601 + }, + { + "epoch": 4.731418918918919, + "grad_norm": 0.6631910800933838, + "learning_rate": 1.6797e-05, + "loss": 0.049, + "step": 5602 + }, + { + "epoch": 4.732263513513513, + "grad_norm": 0.46851450204849243, + "learning_rate": 1.6800000000000002e-05, + "loss": 0.0455, + "step": 5603 + }, + { + "epoch": 4.733108108108108, + "grad_norm": 0.5137978196144104, + "learning_rate": 1.6803e-05, + "loss": 0.0483, + "step": 5604 + }, + { + "epoch": 4.733952702702703, + "grad_norm": 0.6044975519180298, + "learning_rate": 1.6806e-05, + "loss": 0.058, + "step": 5605 + }, + { + "epoch": 4.734797297297297, + "grad_norm": 0.4825671911239624, + "learning_rate": 1.6809e-05, + "loss": 0.0599, + "step": 5606 + }, + { + "epoch": 4.735641891891892, + "grad_norm": 0.6840959787368774, + "learning_rate": 1.6812e-05, + "loss": 0.0655, + "step": 5607 + }, + { + "epoch": 4.736486486486487, + "grad_norm": 0.5607829689979553, + "learning_rate": 1.6815e-05, + "loss": 0.039, + "step": 5608 + }, + { + "epoch": 4.737331081081081, + "grad_norm": 0.49310898780822754, + "learning_rate": 1.6818e-05, + "loss": 0.0503, + "step": 5609 + }, + { + "epoch": 4.738175675675675, + "grad_norm": 0.5867869853973389, + "learning_rate": 1.6821e-05, + "loss": 0.0815, + "step": 5610 + }, + { + "epoch": 4.73902027027027, + "grad_norm": 0.4832451045513153, + "learning_rate": 1.6824e-05, + "loss": 0.0399, + "step": 5611 + }, + { + "epoch": 4.739864864864865, + "grad_norm": 0.6541415452957153, + "learning_rate": 1.6827e-05, + "loss": 0.0664, + "step": 5612 + }, + { + "epoch": 4.74070945945946, + "grad_norm": 0.5272122025489807, + "learning_rate": 1.6830000000000003e-05, + "loss": 0.0426, + "step": 5613 + }, + { + "epoch": 4.741554054054054, + "grad_norm": 0.6032171845436096, + "learning_rate": 1.6833000000000003e-05, + "loss": 0.0434, + "step": 5614 + }, + { + "epoch": 4.742398648648649, + "grad_norm": 0.9342778325080872, + "learning_rate": 1.6836000000000003e-05, + "loss": 0.0446, + "step": 5615 + }, + { + "epoch": 4.743243243243243, + "grad_norm": 0.7066278457641602, + "learning_rate": 1.6839000000000003e-05, + "loss": 0.0652, + "step": 5616 + }, + { + "epoch": 4.744087837837838, + "grad_norm": 0.4819711744785309, + "learning_rate": 1.6842000000000002e-05, + "loss": 0.045, + "step": 5617 + }, + { + "epoch": 4.7449324324324325, + "grad_norm": 0.5282353162765503, + "learning_rate": 1.6845e-05, + "loss": 0.0388, + "step": 5618 + }, + { + "epoch": 4.745777027027027, + "grad_norm": 0.9695351719856262, + "learning_rate": 1.6848e-05, + "loss": 0.0593, + "step": 5619 + }, + { + "epoch": 4.746621621621622, + "grad_norm": 0.9080561995506287, + "learning_rate": 1.6851e-05, + "loss": 0.0768, + "step": 5620 + }, + { + "epoch": 4.747466216216216, + "grad_norm": 0.6622799634933472, + "learning_rate": 1.6853999999999998e-05, + "loss": 0.0569, + "step": 5621 + }, + { + "epoch": 4.7483108108108105, + "grad_norm": 0.7592962384223938, + "learning_rate": 1.6856999999999998e-05, + "loss": 0.0637, + "step": 5622 + }, + { + "epoch": 4.749155405405405, + "grad_norm": 0.7489179372787476, + "learning_rate": 1.686e-05, + "loss": 0.0756, + "step": 5623 + }, + { + "epoch": 4.75, + "grad_norm": 0.8272799253463745, + "learning_rate": 1.6863e-05, + "loss": 0.0938, + "step": 5624 + }, + { + "epoch": 4.750844594594595, + "grad_norm": 0.7484639883041382, + "learning_rate": 1.6866e-05, + "loss": 0.0578, + "step": 5625 + }, + { + "epoch": 4.7516891891891895, + "grad_norm": 0.6539040803909302, + "learning_rate": 1.6869e-05, + "loss": 0.0661, + "step": 5626 + }, + { + "epoch": 4.752533783783784, + "grad_norm": 0.7144765853881836, + "learning_rate": 1.6872e-05, + "loss": 0.0761, + "step": 5627 + }, + { + "epoch": 4.753378378378378, + "grad_norm": 0.7024363875389099, + "learning_rate": 1.6875e-05, + "loss": 0.0663, + "step": 5628 + }, + { + "epoch": 4.754222972972973, + "grad_norm": 0.7529762387275696, + "learning_rate": 1.6878e-05, + "loss": 0.0689, + "step": 5629 + }, + { + "epoch": 4.7550675675675675, + "grad_norm": 0.6539063453674316, + "learning_rate": 1.6881e-05, + "loss": 0.0674, + "step": 5630 + }, + { + "epoch": 4.755912162162162, + "grad_norm": 0.4898500144481659, + "learning_rate": 1.6884e-05, + "loss": 0.0446, + "step": 5631 + }, + { + "epoch": 4.756756756756757, + "grad_norm": 1.0953607559204102, + "learning_rate": 1.6887e-05, + "loss": 0.1063, + "step": 5632 + }, + { + "epoch": 4.757601351351351, + "grad_norm": 0.5813139081001282, + "learning_rate": 1.689e-05, + "loss": 0.0763, + "step": 5633 + }, + { + "epoch": 4.758445945945946, + "grad_norm": 0.9793069362640381, + "learning_rate": 1.6893000000000002e-05, + "loss": 0.0716, + "step": 5634 + }, + { + "epoch": 4.75929054054054, + "grad_norm": 1.4879308938980103, + "learning_rate": 1.6896000000000002e-05, + "loss": 0.137, + "step": 5635 + }, + { + "epoch": 4.760135135135135, + "grad_norm": 1.51597261428833, + "learning_rate": 1.6899000000000002e-05, + "loss": 0.0831, + "step": 5636 + }, + { + "epoch": 4.76097972972973, + "grad_norm": 1.0722283124923706, + "learning_rate": 1.6902000000000002e-05, + "loss": 0.3376, + "step": 5637 + }, + { + "epoch": 4.761824324324325, + "grad_norm": 1.5171514749526978, + "learning_rate": 1.6905e-05, + "loss": 0.2772, + "step": 5638 + }, + { + "epoch": 4.762668918918919, + "grad_norm": 0.7186717987060547, + "learning_rate": 1.6908e-05, + "loss": 0.103, + "step": 5639 + }, + { + "epoch": 4.763513513513513, + "grad_norm": 0.7187658548355103, + "learning_rate": 1.6911e-05, + "loss": 0.1379, + "step": 5640 + }, + { + "epoch": 4.764358108108108, + "grad_norm": 0.6645406484603882, + "learning_rate": 1.6914e-05, + "loss": 0.0879, + "step": 5641 + }, + { + "epoch": 4.765202702702703, + "grad_norm": 0.5689976811408997, + "learning_rate": 1.6916999999999997e-05, + "loss": 0.0708, + "step": 5642 + }, + { + "epoch": 4.766047297297297, + "grad_norm": 0.9031285643577576, + "learning_rate": 1.6919999999999997e-05, + "loss": 0.0851, + "step": 5643 + }, + { + "epoch": 4.766891891891892, + "grad_norm": 0.4771413207054138, + "learning_rate": 1.6923e-05, + "loss": 0.0513, + "step": 5644 + }, + { + "epoch": 4.767736486486487, + "grad_norm": 0.5224307775497437, + "learning_rate": 1.6926e-05, + "loss": 0.0541, + "step": 5645 + }, + { + "epoch": 4.768581081081081, + "grad_norm": 0.9383451342582703, + "learning_rate": 1.6929e-05, + "loss": 0.0538, + "step": 5646 + }, + { + "epoch": 4.769425675675675, + "grad_norm": 0.5333069562911987, + "learning_rate": 1.6932e-05, + "loss": 0.0504, + "step": 5647 + }, + { + "epoch": 4.77027027027027, + "grad_norm": 1.8076860904693604, + "learning_rate": 1.6935e-05, + "loss": 0.0394, + "step": 5648 + }, + { + "epoch": 4.771114864864865, + "grad_norm": 0.6955750584602356, + "learning_rate": 1.6938e-05, + "loss": 0.067, + "step": 5649 + }, + { + "epoch": 4.77195945945946, + "grad_norm": 0.8738977909088135, + "learning_rate": 1.6941e-05, + "loss": 0.0656, + "step": 5650 + }, + { + "epoch": 4.772804054054054, + "grad_norm": 0.5699975490570068, + "learning_rate": 1.6944e-05, + "loss": 0.06, + "step": 5651 + }, + { + "epoch": 4.773648648648649, + "grad_norm": 0.45507752895355225, + "learning_rate": 1.6947e-05, + "loss": 0.0531, + "step": 5652 + }, + { + "epoch": 4.774493243243243, + "grad_norm": 0.6570853590965271, + "learning_rate": 1.695e-05, + "loss": 0.0557, + "step": 5653 + }, + { + "epoch": 4.775337837837838, + "grad_norm": 0.7291904091835022, + "learning_rate": 1.6953000000000002e-05, + "loss": 0.0822, + "step": 5654 + }, + { + "epoch": 4.7761824324324325, + "grad_norm": 0.6088243126869202, + "learning_rate": 1.6956e-05, + "loss": 0.0489, + "step": 5655 + }, + { + "epoch": 4.777027027027027, + "grad_norm": 0.45136597752571106, + "learning_rate": 1.6959e-05, + "loss": 0.0514, + "step": 5656 + }, + { + "epoch": 4.777871621621622, + "grad_norm": 0.7353313565254211, + "learning_rate": 1.6962e-05, + "loss": 0.068, + "step": 5657 + }, + { + "epoch": 4.778716216216216, + "grad_norm": 0.4624195098876953, + "learning_rate": 1.6965e-05, + "loss": 0.0417, + "step": 5658 + }, + { + "epoch": 4.7795608108108105, + "grad_norm": 0.556844174861908, + "learning_rate": 1.6968e-05, + "loss": 0.077, + "step": 5659 + }, + { + "epoch": 4.780405405405405, + "grad_norm": 0.504425048828125, + "learning_rate": 1.6971e-05, + "loss": 0.0416, + "step": 5660 + }, + { + "epoch": 4.78125, + "grad_norm": 0.6266406178474426, + "learning_rate": 1.6974e-05, + "loss": 0.0536, + "step": 5661 + }, + { + "epoch": 4.782094594594595, + "grad_norm": 0.41266071796417236, + "learning_rate": 1.6977e-05, + "loss": 0.0405, + "step": 5662 + }, + { + "epoch": 4.7829391891891895, + "grad_norm": 0.6183227896690369, + "learning_rate": 1.698e-05, + "loss": 0.0558, + "step": 5663 + }, + { + "epoch": 4.783783783783784, + "grad_norm": 0.5186048746109009, + "learning_rate": 1.6983000000000003e-05, + "loss": 0.041, + "step": 5664 + }, + { + "epoch": 4.784628378378378, + "grad_norm": 0.5043569207191467, + "learning_rate": 1.6986000000000003e-05, + "loss": 0.0471, + "step": 5665 + }, + { + "epoch": 4.785472972972973, + "grad_norm": 0.7437917590141296, + "learning_rate": 1.6989000000000003e-05, + "loss": 0.0555, + "step": 5666 + }, + { + "epoch": 4.7863175675675675, + "grad_norm": 0.555513322353363, + "learning_rate": 1.6992e-05, + "loss": 0.0565, + "step": 5667 + }, + { + "epoch": 4.787162162162162, + "grad_norm": 0.6397068500518799, + "learning_rate": 1.6995e-05, + "loss": 0.0695, + "step": 5668 + }, + { + "epoch": 4.788006756756757, + "grad_norm": 0.6040367484092712, + "learning_rate": 1.6998e-05, + "loss": 0.0571, + "step": 5669 + }, + { + "epoch": 4.788851351351351, + "grad_norm": 1.3658065795898438, + "learning_rate": 1.7001e-05, + "loss": 0.1461, + "step": 5670 + }, + { + "epoch": 4.789695945945946, + "grad_norm": 1.0259313583374023, + "learning_rate": 1.7004e-05, + "loss": 0.0904, + "step": 5671 + }, + { + "epoch": 4.79054054054054, + "grad_norm": 0.6376520395278931, + "learning_rate": 1.7006999999999998e-05, + "loss": 0.0631, + "step": 5672 + }, + { + "epoch": 4.791385135135135, + "grad_norm": 0.620550274848938, + "learning_rate": 1.7009999999999998e-05, + "loss": 0.0585, + "step": 5673 + }, + { + "epoch": 4.79222972972973, + "grad_norm": 0.6263018846511841, + "learning_rate": 1.7013e-05, + "loss": 0.0481, + "step": 5674 + }, + { + "epoch": 4.793074324324325, + "grad_norm": 0.5308980941772461, + "learning_rate": 1.7016e-05, + "loss": 0.0461, + "step": 5675 + }, + { + "epoch": 4.793918918918919, + "grad_norm": 0.8603367805480957, + "learning_rate": 1.7019e-05, + "loss": 0.0604, + "step": 5676 + }, + { + "epoch": 4.794763513513513, + "grad_norm": 0.663484513759613, + "learning_rate": 1.7022e-05, + "loss": 0.0531, + "step": 5677 + }, + { + "epoch": 4.795608108108108, + "grad_norm": 0.6153522729873657, + "learning_rate": 1.7025e-05, + "loss": 0.0634, + "step": 5678 + }, + { + "epoch": 4.796452702702703, + "grad_norm": 0.5674811601638794, + "learning_rate": 1.7028e-05, + "loss": 0.0552, + "step": 5679 + }, + { + "epoch": 4.797297297297297, + "grad_norm": 0.7993826270103455, + "learning_rate": 1.7031e-05, + "loss": 0.0622, + "step": 5680 + }, + { + "epoch": 4.798141891891892, + "grad_norm": 0.7709949016571045, + "learning_rate": 1.7034e-05, + "loss": 0.0618, + "step": 5681 + }, + { + "epoch": 4.798986486486487, + "grad_norm": 0.5477169156074524, + "learning_rate": 1.7037e-05, + "loss": 0.0444, + "step": 5682 + }, + { + "epoch": 4.799831081081081, + "grad_norm": 0.7139126658439636, + "learning_rate": 1.704e-05, + "loss": 0.0586, + "step": 5683 + }, + { + "epoch": 4.800675675675675, + "grad_norm": 0.9326269626617432, + "learning_rate": 1.7043000000000003e-05, + "loss": 0.0975, + "step": 5684 + }, + { + "epoch": 4.80152027027027, + "grad_norm": 1.2237995862960815, + "learning_rate": 1.7046000000000002e-05, + "loss": 0.097, + "step": 5685 + }, + { + "epoch": 4.802364864864865, + "grad_norm": 1.3948596715927124, + "learning_rate": 1.7049000000000002e-05, + "loss": 0.1325, + "step": 5686 + }, + { + "epoch": 4.80320945945946, + "grad_norm": 2.0935816764831543, + "learning_rate": 1.7052000000000002e-05, + "loss": 0.418, + "step": 5687 + }, + { + "epoch": 4.804054054054054, + "grad_norm": 4.869734287261963, + "learning_rate": 1.7055000000000002e-05, + "loss": 0.2476, + "step": 5688 + }, + { + "epoch": 4.804898648648649, + "grad_norm": 0.7868203520774841, + "learning_rate": 1.7058e-05, + "loss": 0.1447, + "step": 5689 + }, + { + "epoch": 4.805743243243243, + "grad_norm": 0.602412223815918, + "learning_rate": 1.7061e-05, + "loss": 0.0845, + "step": 5690 + }, + { + "epoch": 4.806587837837838, + "grad_norm": 0.621201753616333, + "learning_rate": 1.7064e-05, + "loss": 0.0841, + "step": 5691 + }, + { + "epoch": 4.8074324324324325, + "grad_norm": 0.7370939254760742, + "learning_rate": 1.7066999999999998e-05, + "loss": 0.082, + "step": 5692 + }, + { + "epoch": 4.808277027027027, + "grad_norm": 0.5819330811500549, + "learning_rate": 1.7069999999999998e-05, + "loss": 0.0677, + "step": 5693 + }, + { + "epoch": 4.809121621621622, + "grad_norm": 0.5243901610374451, + "learning_rate": 1.7073e-05, + "loss": 0.0514, + "step": 5694 + }, + { + "epoch": 4.809966216216216, + "grad_norm": 0.416206419467926, + "learning_rate": 1.7076e-05, + "loss": 0.0509, + "step": 5695 + }, + { + "epoch": 4.8108108108108105, + "grad_norm": 0.4408889412879944, + "learning_rate": 1.7079e-05, + "loss": 0.0487, + "step": 5696 + }, + { + "epoch": 4.811655405405405, + "grad_norm": 0.5383387207984924, + "learning_rate": 1.7082e-05, + "loss": 0.0494, + "step": 5697 + }, + { + "epoch": 4.8125, + "grad_norm": 0.4357639253139496, + "learning_rate": 1.7085e-05, + "loss": 0.0471, + "step": 5698 + }, + { + "epoch": 4.813344594594595, + "grad_norm": 1.0315231084823608, + "learning_rate": 1.7088e-05, + "loss": 0.0666, + "step": 5699 + }, + { + "epoch": 4.8141891891891895, + "grad_norm": 0.4724467396736145, + "learning_rate": 1.7091e-05, + "loss": 0.0468, + "step": 5700 + }, + { + "epoch": 4.815033783783784, + "grad_norm": 0.40078288316726685, + "learning_rate": 1.7094e-05, + "loss": 0.0493, + "step": 5701 + }, + { + "epoch": 4.815878378378378, + "grad_norm": 0.520650327205658, + "learning_rate": 1.7097e-05, + "loss": 0.0591, + "step": 5702 + }, + { + "epoch": 4.816722972972973, + "grad_norm": 0.5333505868911743, + "learning_rate": 1.71e-05, + "loss": 0.0457, + "step": 5703 + }, + { + "epoch": 4.8175675675675675, + "grad_norm": 0.6845664978027344, + "learning_rate": 1.7103000000000002e-05, + "loss": 0.0746, + "step": 5704 + }, + { + "epoch": 4.818412162162162, + "grad_norm": 0.5156315565109253, + "learning_rate": 1.7106000000000002e-05, + "loss": 0.0493, + "step": 5705 + }, + { + "epoch": 4.819256756756757, + "grad_norm": 0.48184195160865784, + "learning_rate": 1.7109000000000002e-05, + "loss": 0.0649, + "step": 5706 + }, + { + "epoch": 4.820101351351351, + "grad_norm": 0.5714427828788757, + "learning_rate": 1.7112e-05, + "loss": 0.0549, + "step": 5707 + }, + { + "epoch": 4.820945945945946, + "grad_norm": 0.5593718886375427, + "learning_rate": 1.7115e-05, + "loss": 0.0547, + "step": 5708 + }, + { + "epoch": 4.82179054054054, + "grad_norm": 0.600506603717804, + "learning_rate": 1.7118e-05, + "loss": 0.0515, + "step": 5709 + }, + { + "epoch": 4.822635135135135, + "grad_norm": 0.4911515414714813, + "learning_rate": 1.7121e-05, + "loss": 0.0569, + "step": 5710 + }, + { + "epoch": 4.82347972972973, + "grad_norm": 1.077505111694336, + "learning_rate": 1.7124e-05, + "loss": 0.0987, + "step": 5711 + }, + { + "epoch": 4.824324324324325, + "grad_norm": 0.462476909160614, + "learning_rate": 1.7127e-05, + "loss": 0.0465, + "step": 5712 + }, + { + "epoch": 4.825168918918919, + "grad_norm": 0.9253546595573425, + "learning_rate": 1.713e-05, + "loss": 0.0802, + "step": 5713 + }, + { + "epoch": 4.826013513513513, + "grad_norm": 0.5968599319458008, + "learning_rate": 1.7133000000000004e-05, + "loss": 0.0632, + "step": 5714 + }, + { + "epoch": 4.826858108108108, + "grad_norm": 0.6212867498397827, + "learning_rate": 1.7136000000000003e-05, + "loss": 0.075, + "step": 5715 + }, + { + "epoch": 4.827702702702703, + "grad_norm": 0.4175722002983093, + "learning_rate": 1.7139e-05, + "loss": 0.0409, + "step": 5716 + }, + { + "epoch": 4.828547297297297, + "grad_norm": 0.6232370734214783, + "learning_rate": 1.7142e-05, + "loss": 0.0561, + "step": 5717 + }, + { + "epoch": 4.829391891891892, + "grad_norm": 0.577519416809082, + "learning_rate": 1.7145e-05, + "loss": 0.064, + "step": 5718 + }, + { + "epoch": 4.830236486486487, + "grad_norm": 0.9135692715644836, + "learning_rate": 1.7148e-05, + "loss": 0.0695, + "step": 5719 + }, + { + "epoch": 4.831081081081081, + "grad_norm": 0.44991689920425415, + "learning_rate": 1.7151e-05, + "loss": 0.0414, + "step": 5720 + }, + { + "epoch": 4.831925675675675, + "grad_norm": 0.6445221304893494, + "learning_rate": 1.7154e-05, + "loss": 0.0516, + "step": 5721 + }, + { + "epoch": 4.83277027027027, + "grad_norm": 0.7608097195625305, + "learning_rate": 1.7157e-05, + "loss": 0.0628, + "step": 5722 + }, + { + "epoch": 4.833614864864865, + "grad_norm": 0.8959970474243164, + "learning_rate": 1.716e-05, + "loss": 0.0783, + "step": 5723 + }, + { + "epoch": 4.83445945945946, + "grad_norm": 1.1727932691574097, + "learning_rate": 1.7163e-05, + "loss": 0.0827, + "step": 5724 + }, + { + "epoch": 4.835304054054054, + "grad_norm": 0.5813509821891785, + "learning_rate": 1.7166e-05, + "loss": 0.0503, + "step": 5725 + }, + { + "epoch": 4.836148648648649, + "grad_norm": 0.7383803725242615, + "learning_rate": 1.7169e-05, + "loss": 0.0474, + "step": 5726 + }, + { + "epoch": 4.836993243243243, + "grad_norm": 0.6227375864982605, + "learning_rate": 1.7172e-05, + "loss": 0.0485, + "step": 5727 + }, + { + "epoch": 4.837837837837838, + "grad_norm": 0.9409237504005432, + "learning_rate": 1.7175e-05, + "loss": 0.0658, + "step": 5728 + }, + { + "epoch": 4.8386824324324325, + "grad_norm": 0.8674283027648926, + "learning_rate": 1.7178e-05, + "loss": 0.0638, + "step": 5729 + }, + { + "epoch": 4.839527027027027, + "grad_norm": 0.6573564410209656, + "learning_rate": 1.7181e-05, + "loss": 0.0626, + "step": 5730 + }, + { + "epoch": 4.840371621621622, + "grad_norm": 1.084365725517273, + "learning_rate": 1.7184e-05, + "loss": 0.0884, + "step": 5731 + }, + { + "epoch": 4.841216216216216, + "grad_norm": 1.2341418266296387, + "learning_rate": 1.7187e-05, + "loss": 0.0763, + "step": 5732 + }, + { + "epoch": 4.8420608108108105, + "grad_norm": 0.9200505018234253, + "learning_rate": 1.719e-05, + "loss": 0.079, + "step": 5733 + }, + { + "epoch": 4.842905405405405, + "grad_norm": 0.9252824783325195, + "learning_rate": 1.7193000000000003e-05, + "loss": 0.0908, + "step": 5734 + }, + { + "epoch": 4.84375, + "grad_norm": 1.0349466800689697, + "learning_rate": 1.7196000000000003e-05, + "loss": 0.1018, + "step": 5735 + }, + { + "epoch": 4.844594594594595, + "grad_norm": 2.495802164077759, + "learning_rate": 1.7199000000000003e-05, + "loss": 0.1826, + "step": 5736 + }, + { + "epoch": 4.8454391891891895, + "grad_norm": 1.071743130683899, + "learning_rate": 1.7202000000000002e-05, + "loss": 0.3921, + "step": 5737 + }, + { + "epoch": 4.846283783783784, + "grad_norm": 0.9701576232910156, + "learning_rate": 1.7205000000000002e-05, + "loss": 0.2354, + "step": 5738 + }, + { + "epoch": 4.847128378378378, + "grad_norm": 0.6954513788223267, + "learning_rate": 1.7208000000000002e-05, + "loss": 0.1783, + "step": 5739 + }, + { + "epoch": 4.847972972972973, + "grad_norm": 0.6216431260108948, + "learning_rate": 1.7211000000000002e-05, + "loss": 0.1085, + "step": 5740 + }, + { + "epoch": 4.8488175675675675, + "grad_norm": 0.5416164398193359, + "learning_rate": 1.7213999999999998e-05, + "loss": 0.0728, + "step": 5741 + }, + { + "epoch": 4.849662162162162, + "grad_norm": 0.4883697032928467, + "learning_rate": 1.7216999999999998e-05, + "loss": 0.0761, + "step": 5742 + }, + { + "epoch": 4.850506756756757, + "grad_norm": 0.45925676822662354, + "learning_rate": 1.7219999999999998e-05, + "loss": 0.041, + "step": 5743 + }, + { + "epoch": 4.851351351351351, + "grad_norm": 0.47717002034187317, + "learning_rate": 1.7223e-05, + "loss": 0.062, + "step": 5744 + }, + { + "epoch": 4.852195945945946, + "grad_norm": 0.5277879238128662, + "learning_rate": 1.7226e-05, + "loss": 0.0598, + "step": 5745 + }, + { + "epoch": 4.85304054054054, + "grad_norm": 0.5216675400733948, + "learning_rate": 1.7229e-05, + "loss": 0.0614, + "step": 5746 + }, + { + "epoch": 4.853885135135135, + "grad_norm": 0.45407211780548096, + "learning_rate": 1.7232e-05, + "loss": 0.0498, + "step": 5747 + }, + { + "epoch": 4.85472972972973, + "grad_norm": 0.5336042046546936, + "learning_rate": 1.7235e-05, + "loss": 0.0559, + "step": 5748 + }, + { + "epoch": 4.855574324324325, + "grad_norm": 0.5143003463745117, + "learning_rate": 1.7238e-05, + "loss": 0.0502, + "step": 5749 + }, + { + "epoch": 4.856418918918919, + "grad_norm": 0.640812337398529, + "learning_rate": 1.7241e-05, + "loss": 0.0573, + "step": 5750 + }, + { + "epoch": 4.857263513513513, + "grad_norm": 0.4560999274253845, + "learning_rate": 1.7244e-05, + "loss": 0.0456, + "step": 5751 + }, + { + "epoch": 4.858108108108108, + "grad_norm": 0.6662977337837219, + "learning_rate": 1.7247e-05, + "loss": 0.0658, + "step": 5752 + }, + { + "epoch": 4.858952702702703, + "grad_norm": 0.6368479132652283, + "learning_rate": 1.725e-05, + "loss": 0.0801, + "step": 5753 + }, + { + "epoch": 4.859797297297297, + "grad_norm": 0.49520862102508545, + "learning_rate": 1.7253e-05, + "loss": 0.0401, + "step": 5754 + }, + { + "epoch": 4.860641891891892, + "grad_norm": 0.5635510683059692, + "learning_rate": 1.7256000000000002e-05, + "loss": 0.0697, + "step": 5755 + }, + { + "epoch": 4.861486486486487, + "grad_norm": 0.6306589245796204, + "learning_rate": 1.7259000000000002e-05, + "loss": 0.0365, + "step": 5756 + }, + { + "epoch": 4.862331081081081, + "grad_norm": 0.5287733674049377, + "learning_rate": 1.7262000000000002e-05, + "loss": 0.0556, + "step": 5757 + }, + { + "epoch": 4.863175675675675, + "grad_norm": 0.5919682383537292, + "learning_rate": 1.7265e-05, + "loss": 0.0567, + "step": 5758 + }, + { + "epoch": 4.86402027027027, + "grad_norm": 0.9311250448226929, + "learning_rate": 1.7268e-05, + "loss": 0.0838, + "step": 5759 + }, + { + "epoch": 4.864864864864865, + "grad_norm": 0.7700218558311462, + "learning_rate": 1.7271e-05, + "loss": 0.0656, + "step": 5760 + }, + { + "epoch": 4.86570945945946, + "grad_norm": 0.4456535577774048, + "learning_rate": 1.7274e-05, + "loss": 0.0474, + "step": 5761 + }, + { + "epoch": 4.866554054054054, + "grad_norm": 0.3674728572368622, + "learning_rate": 1.7277e-05, + "loss": 0.0296, + "step": 5762 + }, + { + "epoch": 4.867398648648649, + "grad_norm": 0.5663643479347229, + "learning_rate": 1.728e-05, + "loss": 0.0425, + "step": 5763 + }, + { + "epoch": 4.868243243243243, + "grad_norm": 0.5236560702323914, + "learning_rate": 1.7283e-05, + "loss": 0.0505, + "step": 5764 + }, + { + "epoch": 4.869087837837838, + "grad_norm": 0.5592062473297119, + "learning_rate": 1.7286e-05, + "loss": 0.0523, + "step": 5765 + }, + { + "epoch": 4.8699324324324325, + "grad_norm": 0.6177638173103333, + "learning_rate": 1.7289e-05, + "loss": 0.0551, + "step": 5766 + }, + { + "epoch": 4.870777027027027, + "grad_norm": 0.6175156831741333, + "learning_rate": 1.7292e-05, + "loss": 0.0533, + "step": 5767 + }, + { + "epoch": 4.871621621621622, + "grad_norm": 0.723200798034668, + "learning_rate": 1.7295e-05, + "loss": 0.0499, + "step": 5768 + }, + { + "epoch": 4.872466216216216, + "grad_norm": 0.41551586985588074, + "learning_rate": 1.7298e-05, + "loss": 0.0302, + "step": 5769 + }, + { + "epoch": 4.8733108108108105, + "grad_norm": 0.5683813691139221, + "learning_rate": 1.7301e-05, + "loss": 0.0372, + "step": 5770 + }, + { + "epoch": 4.874155405405405, + "grad_norm": 0.552814245223999, + "learning_rate": 1.7304e-05, + "loss": 0.0495, + "step": 5771 + }, + { + "epoch": 4.875, + "grad_norm": 0.5738871097564697, + "learning_rate": 1.7307e-05, + "loss": 0.0481, + "step": 5772 + }, + { + "epoch": 4.875844594594595, + "grad_norm": 0.674717128276825, + "learning_rate": 1.731e-05, + "loss": 0.0585, + "step": 5773 + }, + { + "epoch": 4.8766891891891895, + "grad_norm": 0.6745892763137817, + "learning_rate": 1.7313e-05, + "loss": 0.0543, + "step": 5774 + }, + { + "epoch": 4.877533783783784, + "grad_norm": 0.8548738360404968, + "learning_rate": 1.7316e-05, + "loss": 0.0639, + "step": 5775 + }, + { + "epoch": 4.878378378378378, + "grad_norm": 0.6957017183303833, + "learning_rate": 1.7319e-05, + "loss": 0.0433, + "step": 5776 + }, + { + "epoch": 4.879222972972973, + "grad_norm": 0.8398239016532898, + "learning_rate": 1.7322e-05, + "loss": 0.0764, + "step": 5777 + }, + { + "epoch": 4.8800675675675675, + "grad_norm": 0.736411988735199, + "learning_rate": 1.7325e-05, + "loss": 0.0786, + "step": 5778 + }, + { + "epoch": 4.880912162162162, + "grad_norm": 0.5808204412460327, + "learning_rate": 1.7328e-05, + "loss": 0.0763, + "step": 5779 + }, + { + "epoch": 4.881756756756757, + "grad_norm": 0.6140557527542114, + "learning_rate": 1.7331e-05, + "loss": 0.0725, + "step": 5780 + }, + { + "epoch": 4.882601351351351, + "grad_norm": 0.5965267419815063, + "learning_rate": 1.7334e-05, + "loss": 0.0416, + "step": 5781 + }, + { + "epoch": 4.883445945945946, + "grad_norm": 0.6203557848930359, + "learning_rate": 1.7337e-05, + "loss": 0.0461, + "step": 5782 + }, + { + "epoch": 4.88429054054054, + "grad_norm": 0.6783068776130676, + "learning_rate": 1.734e-05, + "loss": 0.0643, + "step": 5783 + }, + { + "epoch": 4.885135135135135, + "grad_norm": 0.629927933216095, + "learning_rate": 1.7343e-05, + "loss": 0.06, + "step": 5784 + }, + { + "epoch": 4.88597972972973, + "grad_norm": 0.9637052416801453, + "learning_rate": 1.7346000000000003e-05, + "loss": 0.0969, + "step": 5785 + }, + { + "epoch": 4.886824324324325, + "grad_norm": 1.3824925422668457, + "learning_rate": 1.7349000000000003e-05, + "loss": 0.1305, + "step": 5786 + }, + { + "epoch": 4.887668918918919, + "grad_norm": 1.436968445777893, + "learning_rate": 1.7352000000000003e-05, + "loss": 0.3692, + "step": 5787 + }, + { + "epoch": 4.888513513513513, + "grad_norm": 0.9501616358757019, + "learning_rate": 1.7355000000000002e-05, + "loss": 0.2189, + "step": 5788 + }, + { + "epoch": 4.889358108108108, + "grad_norm": 0.6712062954902649, + "learning_rate": 1.7358000000000002e-05, + "loss": 0.1068, + "step": 5789 + }, + { + "epoch": 4.890202702702703, + "grad_norm": 0.5525882244110107, + "learning_rate": 1.7361e-05, + "loss": 0.1073, + "step": 5790 + }, + { + "epoch": 4.891047297297297, + "grad_norm": 0.4816790223121643, + "learning_rate": 1.7364e-05, + "loss": 0.0488, + "step": 5791 + }, + { + "epoch": 4.891891891891892, + "grad_norm": 0.41385698318481445, + "learning_rate": 1.7366999999999998e-05, + "loss": 0.056, + "step": 5792 + }, + { + "epoch": 4.892736486486487, + "grad_norm": 0.429973840713501, + "learning_rate": 1.7369999999999998e-05, + "loss": 0.0514, + "step": 5793 + }, + { + "epoch": 4.893581081081081, + "grad_norm": 0.505784809589386, + "learning_rate": 1.7372999999999998e-05, + "loss": 0.0831, + "step": 5794 + }, + { + "epoch": 4.894425675675675, + "grad_norm": 0.5421444177627563, + "learning_rate": 1.7376e-05, + "loss": 0.0638, + "step": 5795 + }, + { + "epoch": 4.89527027027027, + "grad_norm": 0.5041525363922119, + "learning_rate": 1.7379e-05, + "loss": 0.0623, + "step": 5796 + }, + { + "epoch": 4.896114864864865, + "grad_norm": 0.39018404483795166, + "learning_rate": 1.7382e-05, + "loss": 0.051, + "step": 5797 + }, + { + "epoch": 4.89695945945946, + "grad_norm": 0.44696536660194397, + "learning_rate": 1.7385e-05, + "loss": 0.0485, + "step": 5798 + }, + { + "epoch": 4.897804054054054, + "grad_norm": 0.4276726245880127, + "learning_rate": 1.7388e-05, + "loss": 0.0507, + "step": 5799 + }, + { + "epoch": 4.898648648648649, + "grad_norm": 0.47953638434410095, + "learning_rate": 1.7391e-05, + "loss": 0.0558, + "step": 5800 + }, + { + "epoch": 4.899493243243243, + "grad_norm": 0.582108736038208, + "learning_rate": 1.7394e-05, + "loss": 0.0506, + "step": 5801 + }, + { + "epoch": 4.900337837837838, + "grad_norm": 0.3950868844985962, + "learning_rate": 1.7397e-05, + "loss": 0.0408, + "step": 5802 + }, + { + "epoch": 4.9011824324324325, + "grad_norm": 0.6265711784362793, + "learning_rate": 1.74e-05, + "loss": 0.0577, + "step": 5803 + }, + { + "epoch": 4.902027027027027, + "grad_norm": 0.4710981547832489, + "learning_rate": 1.7403e-05, + "loss": 0.0413, + "step": 5804 + }, + { + "epoch": 4.902871621621622, + "grad_norm": 0.5036196112632751, + "learning_rate": 1.7406000000000002e-05, + "loss": 0.0551, + "step": 5805 + }, + { + "epoch": 4.903716216216216, + "grad_norm": 0.4583391845226288, + "learning_rate": 1.7409000000000002e-05, + "loss": 0.0501, + "step": 5806 + }, + { + "epoch": 4.9045608108108105, + "grad_norm": 0.5910520553588867, + "learning_rate": 1.7412000000000002e-05, + "loss": 0.0468, + "step": 5807 + }, + { + "epoch": 4.905405405405405, + "grad_norm": 0.4638925790786743, + "learning_rate": 1.7415000000000002e-05, + "loss": 0.0324, + "step": 5808 + }, + { + "epoch": 4.90625, + "grad_norm": 0.62339186668396, + "learning_rate": 1.7418e-05, + "loss": 0.0563, + "step": 5809 + }, + { + "epoch": 4.907094594594595, + "grad_norm": 0.5006028413772583, + "learning_rate": 1.7421e-05, + "loss": 0.0477, + "step": 5810 + }, + { + "epoch": 4.9079391891891895, + "grad_norm": 0.5384613871574402, + "learning_rate": 1.7424e-05, + "loss": 0.0725, + "step": 5811 + }, + { + "epoch": 4.908783783783784, + "grad_norm": 0.6026734113693237, + "learning_rate": 1.7427e-05, + "loss": 0.0565, + "step": 5812 + }, + { + "epoch": 4.909628378378378, + "grad_norm": 0.4528965950012207, + "learning_rate": 1.743e-05, + "loss": 0.0425, + "step": 5813 + }, + { + "epoch": 4.910472972972973, + "grad_norm": 0.5403547883033752, + "learning_rate": 1.7432999999999997e-05, + "loss": 0.047, + "step": 5814 + }, + { + "epoch": 4.9113175675675675, + "grad_norm": 0.8809278011322021, + "learning_rate": 1.7436e-05, + "loss": 0.0534, + "step": 5815 + }, + { + "epoch": 4.912162162162162, + "grad_norm": 0.640774667263031, + "learning_rate": 1.7439e-05, + "loss": 0.0524, + "step": 5816 + }, + { + "epoch": 4.913006756756757, + "grad_norm": 0.4504512846469879, + "learning_rate": 1.7442e-05, + "loss": 0.0493, + "step": 5817 + }, + { + "epoch": 4.913851351351351, + "grad_norm": 0.7596952319145203, + "learning_rate": 1.7445e-05, + "loss": 0.0526, + "step": 5818 + }, + { + "epoch": 4.914695945945946, + "grad_norm": 0.45548537373542786, + "learning_rate": 1.7448e-05, + "loss": 0.0422, + "step": 5819 + }, + { + "epoch": 4.91554054054054, + "grad_norm": 0.6601645946502686, + "learning_rate": 1.7451e-05, + "loss": 0.0737, + "step": 5820 + }, + { + "epoch": 4.916385135135135, + "grad_norm": 0.7073277831077576, + "learning_rate": 1.7454e-05, + "loss": 0.0714, + "step": 5821 + }, + { + "epoch": 4.91722972972973, + "grad_norm": 0.5744006037712097, + "learning_rate": 1.7457e-05, + "loss": 0.0633, + "step": 5822 + }, + { + "epoch": 4.918074324324325, + "grad_norm": 0.6954755187034607, + "learning_rate": 1.746e-05, + "loss": 0.0829, + "step": 5823 + }, + { + "epoch": 4.918918918918919, + "grad_norm": 0.5397576093673706, + "learning_rate": 1.7463e-05, + "loss": 0.0407, + "step": 5824 + }, + { + "epoch": 4.919763513513513, + "grad_norm": 0.9713366031646729, + "learning_rate": 1.7466000000000002e-05, + "loss": 0.0658, + "step": 5825 + }, + { + "epoch": 4.920608108108108, + "grad_norm": 0.8651922941207886, + "learning_rate": 1.7469e-05, + "loss": 0.0685, + "step": 5826 + }, + { + "epoch": 4.921452702702703, + "grad_norm": 0.3779897391796112, + "learning_rate": 1.7472e-05, + "loss": 0.0338, + "step": 5827 + }, + { + "epoch": 4.922297297297297, + "grad_norm": 0.7523493766784668, + "learning_rate": 1.7475e-05, + "loss": 0.0828, + "step": 5828 + }, + { + "epoch": 4.923141891891892, + "grad_norm": 0.4782276749610901, + "learning_rate": 1.7478e-05, + "loss": 0.0531, + "step": 5829 + }, + { + "epoch": 4.923986486486487, + "grad_norm": 0.6803292632102966, + "learning_rate": 1.7481e-05, + "loss": 0.0609, + "step": 5830 + }, + { + "epoch": 4.924831081081081, + "grad_norm": 1.0319997072219849, + "learning_rate": 1.7484e-05, + "loss": 0.0586, + "step": 5831 + }, + { + "epoch": 4.925675675675675, + "grad_norm": 1.496357798576355, + "learning_rate": 1.7487e-05, + "loss": 0.0811, + "step": 5832 + }, + { + "epoch": 4.92652027027027, + "grad_norm": 0.6105113625526428, + "learning_rate": 1.749e-05, + "loss": 0.054, + "step": 5833 + }, + { + "epoch": 4.927364864864865, + "grad_norm": 1.1069386005401611, + "learning_rate": 1.7493e-05, + "loss": 0.0674, + "step": 5834 + }, + { + "epoch": 4.92820945945946, + "grad_norm": 0.8773531913757324, + "learning_rate": 1.7496000000000003e-05, + "loss": 0.0833, + "step": 5835 + }, + { + "epoch": 4.929054054054054, + "grad_norm": 0.9674264788627625, + "learning_rate": 1.7499000000000003e-05, + "loss": 0.0817, + "step": 5836 + }, + { + "epoch": 4.929898648648649, + "grad_norm": 4.488339900970459, + "learning_rate": 1.7502000000000003e-05, + "loss": 0.6123, + "step": 5837 + }, + { + "epoch": 4.930743243243243, + "grad_norm": 0.8728847503662109, + "learning_rate": 1.7505000000000003e-05, + "loss": 0.2381, + "step": 5838 + }, + { + "epoch": 4.931587837837838, + "grad_norm": 0.7662221789360046, + "learning_rate": 1.7508e-05, + "loss": 0.1188, + "step": 5839 + }, + { + "epoch": 4.9324324324324325, + "grad_norm": 0.5905394554138184, + "learning_rate": 1.7511e-05, + "loss": 0.0773, + "step": 5840 + }, + { + "epoch": 4.933277027027027, + "grad_norm": 0.7420361042022705, + "learning_rate": 1.7514e-05, + "loss": 0.0972, + "step": 5841 + }, + { + "epoch": 4.934121621621622, + "grad_norm": 0.4611552655696869, + "learning_rate": 1.7517e-05, + "loss": 0.0555, + "step": 5842 + }, + { + "epoch": 4.934966216216216, + "grad_norm": 0.4369223415851593, + "learning_rate": 1.7519999999999998e-05, + "loss": 0.0512, + "step": 5843 + }, + { + "epoch": 4.9358108108108105, + "grad_norm": 0.6843777298927307, + "learning_rate": 1.7522999999999998e-05, + "loss": 0.049, + "step": 5844 + }, + { + "epoch": 4.936655405405405, + "grad_norm": 0.3521738052368164, + "learning_rate": 1.7526e-05, + "loss": 0.0392, + "step": 5845 + }, + { + "epoch": 4.9375, + "grad_norm": 0.41082313656806946, + "learning_rate": 1.7529e-05, + "loss": 0.0522, + "step": 5846 + }, + { + "epoch": 4.938344594594595, + "grad_norm": 0.354959636926651, + "learning_rate": 1.7532e-05, + "loss": 0.0242, + "step": 5847 + }, + { + "epoch": 4.9391891891891895, + "grad_norm": 0.5153846144676208, + "learning_rate": 1.7535e-05, + "loss": 0.0599, + "step": 5848 + }, + { + "epoch": 4.940033783783784, + "grad_norm": 0.580745279788971, + "learning_rate": 1.7538e-05, + "loss": 0.0502, + "step": 5849 + }, + { + "epoch": 4.940878378378378, + "grad_norm": 0.5819529891014099, + "learning_rate": 1.7541e-05, + "loss": 0.0882, + "step": 5850 + }, + { + "epoch": 4.941722972972973, + "grad_norm": 0.48622605204582214, + "learning_rate": 1.7544e-05, + "loss": 0.0564, + "step": 5851 + }, + { + "epoch": 4.9425675675675675, + "grad_norm": 0.5241417288780212, + "learning_rate": 1.7547e-05, + "loss": 0.0818, + "step": 5852 + }, + { + "epoch": 4.943412162162162, + "grad_norm": 0.47372865676879883, + "learning_rate": 1.755e-05, + "loss": 0.0577, + "step": 5853 + }, + { + "epoch": 4.944256756756757, + "grad_norm": 0.5343509912490845, + "learning_rate": 1.7553e-05, + "loss": 0.0621, + "step": 5854 + }, + { + "epoch": 4.945101351351351, + "grad_norm": 0.577201247215271, + "learning_rate": 1.7556000000000003e-05, + "loss": 0.0488, + "step": 5855 + }, + { + "epoch": 4.945945945945946, + "grad_norm": 0.749046266078949, + "learning_rate": 1.7559000000000002e-05, + "loss": 0.0629, + "step": 5856 + }, + { + "epoch": 4.94679054054054, + "grad_norm": 0.5737521648406982, + "learning_rate": 1.7562000000000002e-05, + "loss": 0.0563, + "step": 5857 + }, + { + "epoch": 4.947635135135135, + "grad_norm": 0.414582222700119, + "learning_rate": 1.7565000000000002e-05, + "loss": 0.0421, + "step": 5858 + }, + { + "epoch": 4.94847972972973, + "grad_norm": 0.57423335313797, + "learning_rate": 1.7568000000000002e-05, + "loss": 0.0583, + "step": 5859 + }, + { + "epoch": 4.949324324324325, + "grad_norm": 0.37115445733070374, + "learning_rate": 1.7571e-05, + "loss": 0.0401, + "step": 5860 + }, + { + "epoch": 4.950168918918919, + "grad_norm": 0.5499172806739807, + "learning_rate": 1.7574e-05, + "loss": 0.0584, + "step": 5861 + }, + { + "epoch": 4.951013513513513, + "grad_norm": 0.7356054186820984, + "learning_rate": 1.7577e-05, + "loss": 0.0753, + "step": 5862 + }, + { + "epoch": 4.951858108108108, + "grad_norm": 0.6282176375389099, + "learning_rate": 1.758e-05, + "loss": 0.0471, + "step": 5863 + }, + { + "epoch": 4.952702702702703, + "grad_norm": 0.5010102391242981, + "learning_rate": 1.7582999999999998e-05, + "loss": 0.038, + "step": 5864 + }, + { + "epoch": 4.953547297297297, + "grad_norm": 0.600834310054779, + "learning_rate": 1.7586e-05, + "loss": 0.0484, + "step": 5865 + }, + { + "epoch": 4.954391891891892, + "grad_norm": 0.6536915302276611, + "learning_rate": 1.7589e-05, + "loss": 0.0439, + "step": 5866 + }, + { + "epoch": 4.955236486486487, + "grad_norm": 0.47289812564849854, + "learning_rate": 1.7592e-05, + "loss": 0.0396, + "step": 5867 + }, + { + "epoch": 4.956081081081081, + "grad_norm": 0.6964331269264221, + "learning_rate": 1.7595e-05, + "loss": 0.053, + "step": 5868 + }, + { + "epoch": 4.956925675675675, + "grad_norm": 0.4878780245780945, + "learning_rate": 1.7598e-05, + "loss": 0.0359, + "step": 5869 + }, + { + "epoch": 4.95777027027027, + "grad_norm": 0.7372329235076904, + "learning_rate": 1.7601e-05, + "loss": 0.0871, + "step": 5870 + }, + { + "epoch": 4.958614864864865, + "grad_norm": 0.7691588997840881, + "learning_rate": 1.7604e-05, + "loss": 0.0687, + "step": 5871 + }, + { + "epoch": 4.95945945945946, + "grad_norm": 0.5152127146720886, + "learning_rate": 1.7607e-05, + "loss": 0.0347, + "step": 5872 + }, + { + "epoch": 4.960304054054054, + "grad_norm": 0.6540361046791077, + "learning_rate": 1.761e-05, + "loss": 0.0662, + "step": 5873 + }, + { + "epoch": 4.961148648648649, + "grad_norm": 0.9209172129631042, + "learning_rate": 1.7613e-05, + "loss": 0.0519, + "step": 5874 + }, + { + "epoch": 4.961993243243243, + "grad_norm": 0.6733456254005432, + "learning_rate": 1.7616000000000002e-05, + "loss": 0.0739, + "step": 5875 + }, + { + "epoch": 4.962837837837838, + "grad_norm": 0.5594066381454468, + "learning_rate": 1.7619000000000002e-05, + "loss": 0.0538, + "step": 5876 + }, + { + "epoch": 4.9636824324324325, + "grad_norm": 0.8665258884429932, + "learning_rate": 1.7622000000000002e-05, + "loss": 0.0699, + "step": 5877 + }, + { + "epoch": 4.964527027027027, + "grad_norm": 0.5116548538208008, + "learning_rate": 1.7625e-05, + "loss": 0.0525, + "step": 5878 + }, + { + "epoch": 4.965371621621622, + "grad_norm": 0.5690612196922302, + "learning_rate": 1.7628e-05, + "loss": 0.0381, + "step": 5879 + }, + { + "epoch": 4.966216216216216, + "grad_norm": 0.8611814975738525, + "learning_rate": 1.7631e-05, + "loss": 0.0831, + "step": 5880 + }, + { + "epoch": 4.9670608108108105, + "grad_norm": 0.5499203205108643, + "learning_rate": 1.7634e-05, + "loss": 0.0579, + "step": 5881 + }, + { + "epoch": 4.967905405405405, + "grad_norm": 0.5737179517745972, + "learning_rate": 1.7637e-05, + "loss": 0.0767, + "step": 5882 + }, + { + "epoch": 4.96875, + "grad_norm": 0.7215269804000854, + "learning_rate": 1.764e-05, + "loss": 0.0603, + "step": 5883 + }, + { + "epoch": 4.969594594594595, + "grad_norm": 1.1962121725082397, + "learning_rate": 1.7643e-05, + "loss": 0.1126, + "step": 5884 + }, + { + "epoch": 4.9704391891891895, + "grad_norm": 0.8684900999069214, + "learning_rate": 1.7646e-05, + "loss": 0.0652, + "step": 5885 + }, + { + "epoch": 4.971283783783784, + "grad_norm": 2.095871925354004, + "learning_rate": 1.7649000000000003e-05, + "loss": 0.1424, + "step": 5886 + }, + { + "epoch": 4.972128378378378, + "grad_norm": 2.1127853393554688, + "learning_rate": 1.7652000000000003e-05, + "loss": 0.379, + "step": 5887 + }, + { + "epoch": 4.972972972972973, + "grad_norm": 0.8904876708984375, + "learning_rate": 1.7655e-05, + "loss": 0.1115, + "step": 5888 + }, + { + "epoch": 4.9738175675675675, + "grad_norm": 0.4335446357727051, + "learning_rate": 1.7658e-05, + "loss": 0.0572, + "step": 5889 + }, + { + "epoch": 4.974662162162162, + "grad_norm": 0.5819356441497803, + "learning_rate": 1.7661e-05, + "loss": 0.0732, + "step": 5890 + }, + { + "epoch": 4.975506756756757, + "grad_norm": 0.6207284331321716, + "learning_rate": 1.7664e-05, + "loss": 0.074, + "step": 5891 + }, + { + "epoch": 4.976351351351351, + "grad_norm": 0.6526414155960083, + "learning_rate": 1.7667e-05, + "loss": 0.0566, + "step": 5892 + }, + { + "epoch": 4.977195945945946, + "grad_norm": 0.6530730128288269, + "learning_rate": 1.767e-05, + "loss": 0.0519, + "step": 5893 + }, + { + "epoch": 4.97804054054054, + "grad_norm": 0.5613729357719421, + "learning_rate": 1.7673e-05, + "loss": 0.0698, + "step": 5894 + }, + { + "epoch": 4.978885135135135, + "grad_norm": 0.6228047609329224, + "learning_rate": 1.7675999999999998e-05, + "loss": 0.0692, + "step": 5895 + }, + { + "epoch": 4.97972972972973, + "grad_norm": 0.5190767645835876, + "learning_rate": 1.7679e-05, + "loss": 0.0591, + "step": 5896 + }, + { + "epoch": 4.980574324324325, + "grad_norm": 0.5448602437973022, + "learning_rate": 1.7682e-05, + "loss": 0.0447, + "step": 5897 + }, + { + "epoch": 4.981418918918919, + "grad_norm": 0.6992744207382202, + "learning_rate": 1.7685e-05, + "loss": 0.0666, + "step": 5898 + }, + { + "epoch": 4.982263513513513, + "grad_norm": 0.573494553565979, + "learning_rate": 1.7688e-05, + "loss": 0.0538, + "step": 5899 + }, + { + "epoch": 4.983108108108108, + "grad_norm": 0.7396647334098816, + "learning_rate": 1.7691e-05, + "loss": 0.0648, + "step": 5900 + }, + { + "epoch": 4.983952702702703, + "grad_norm": 0.48593658208847046, + "learning_rate": 1.7694e-05, + "loss": 0.0662, + "step": 5901 + }, + { + "epoch": 4.984797297297297, + "grad_norm": 0.5608000159263611, + "learning_rate": 1.7697e-05, + "loss": 0.0612, + "step": 5902 + }, + { + "epoch": 4.985641891891892, + "grad_norm": 0.5618636012077332, + "learning_rate": 1.77e-05, + "loss": 0.0783, + "step": 5903 + }, + { + "epoch": 4.986486486486487, + "grad_norm": 0.4292460083961487, + "learning_rate": 1.7703e-05, + "loss": 0.0386, + "step": 5904 + }, + { + "epoch": 4.987331081081081, + "grad_norm": 0.9261767268180847, + "learning_rate": 1.7706e-05, + "loss": 0.0675, + "step": 5905 + }, + { + "epoch": 4.988175675675675, + "grad_norm": 0.5260120034217834, + "learning_rate": 1.7709000000000003e-05, + "loss": 0.0561, + "step": 5906 + }, + { + "epoch": 4.98902027027027, + "grad_norm": 0.5197432637214661, + "learning_rate": 1.7712000000000003e-05, + "loss": 0.0673, + "step": 5907 + }, + { + "epoch": 4.989864864864865, + "grad_norm": 0.5980039834976196, + "learning_rate": 1.7715000000000002e-05, + "loss": 0.0502, + "step": 5908 + }, + { + "epoch": 4.99070945945946, + "grad_norm": 0.5634139180183411, + "learning_rate": 1.7718000000000002e-05, + "loss": 0.0454, + "step": 5909 + }, + { + "epoch": 4.991554054054054, + "grad_norm": 0.4084548056125641, + "learning_rate": 1.7721000000000002e-05, + "loss": 0.052, + "step": 5910 + }, + { + "epoch": 4.992398648648649, + "grad_norm": 0.48547232151031494, + "learning_rate": 1.7724000000000002e-05, + "loss": 0.0483, + "step": 5911 + }, + { + "epoch": 4.993243243243243, + "grad_norm": 0.5420435070991516, + "learning_rate": 1.7727e-05, + "loss": 0.0508, + "step": 5912 + }, + { + "epoch": 4.994087837837838, + "grad_norm": 0.7431482076644897, + "learning_rate": 1.7729999999999998e-05, + "loss": 0.06, + "step": 5913 + }, + { + "epoch": 4.9949324324324325, + "grad_norm": 0.7357265949249268, + "learning_rate": 1.7732999999999998e-05, + "loss": 0.0894, + "step": 5914 + }, + { + "epoch": 4.995777027027027, + "grad_norm": 0.6696967482566833, + "learning_rate": 1.7735999999999998e-05, + "loss": 0.062, + "step": 5915 + }, + { + "epoch": 4.996621621621622, + "grad_norm": 0.629500150680542, + "learning_rate": 1.7739e-05, + "loss": 0.0633, + "step": 5916 + }, + { + "epoch": 4.997466216216216, + "grad_norm": 0.48385781049728394, + "learning_rate": 1.7742e-05, + "loss": 0.044, + "step": 5917 + }, + { + "epoch": 4.9983108108108105, + "grad_norm": 0.6386424899101257, + "learning_rate": 1.7745e-05, + "loss": 0.0772, + "step": 5918 + }, + { + "epoch": 4.999155405405405, + "grad_norm": 1.170325517654419, + "learning_rate": 1.7748e-05, + "loss": 0.0676, + "step": 5919 + }, + { + "epoch": 5.0, + "grad_norm": 1.5523377656936646, + "learning_rate": 1.7751e-05, + "loss": 0.1021, + "step": 5920 + }, + { + "epoch": 5.000844594594595, + "grad_norm": 4.488789081573486, + "learning_rate": 1.7754e-05, + "loss": 0.573, + "step": 5921 + }, + { + "epoch": 5.0016891891891895, + "grad_norm": 1.4761546850204468, + "learning_rate": 1.7757e-05, + "loss": 0.2808, + "step": 5922 + }, + { + "epoch": 5.002533783783784, + "grad_norm": 0.6066982746124268, + "learning_rate": 1.776e-05, + "loss": 0.1228, + "step": 5923 + }, + { + "epoch": 5.003378378378378, + "grad_norm": 0.9462047219276428, + "learning_rate": 1.7763e-05, + "loss": 0.0975, + "step": 5924 + }, + { + "epoch": 5.004222972972973, + "grad_norm": 0.6668618321418762, + "learning_rate": 1.7766e-05, + "loss": 0.0518, + "step": 5925 + }, + { + "epoch": 5.0050675675675675, + "grad_norm": 0.6981812119483948, + "learning_rate": 1.7769000000000002e-05, + "loss": 0.0687, + "step": 5926 + }, + { + "epoch": 5.005912162162162, + "grad_norm": 0.5309522747993469, + "learning_rate": 1.7772000000000002e-05, + "loss": 0.0627, + "step": 5927 + }, + { + "epoch": 5.006756756756757, + "grad_norm": 0.44969046115875244, + "learning_rate": 1.7775000000000002e-05, + "loss": 0.0473, + "step": 5928 + }, + { + "epoch": 5.007601351351352, + "grad_norm": 0.603171706199646, + "learning_rate": 1.7778e-05, + "loss": 0.0761, + "step": 5929 + }, + { + "epoch": 5.008445945945946, + "grad_norm": 0.6309815049171448, + "learning_rate": 1.7781e-05, + "loss": 0.0438, + "step": 5930 + }, + { + "epoch": 5.00929054054054, + "grad_norm": 0.5808737874031067, + "learning_rate": 1.7784e-05, + "loss": 0.0505, + "step": 5931 + }, + { + "epoch": 5.010135135135135, + "grad_norm": 0.47982197999954224, + "learning_rate": 1.7787e-05, + "loss": 0.0472, + "step": 5932 + }, + { + "epoch": 5.01097972972973, + "grad_norm": 0.5337579846382141, + "learning_rate": 1.779e-05, + "loss": 0.0593, + "step": 5933 + }, + { + "epoch": 5.011824324324325, + "grad_norm": 0.5650960206985474, + "learning_rate": 1.7793e-05, + "loss": 0.0422, + "step": 5934 + }, + { + "epoch": 5.012668918918919, + "grad_norm": 0.40230223536491394, + "learning_rate": 1.7796e-05, + "loss": 0.0406, + "step": 5935 + }, + { + "epoch": 5.013513513513513, + "grad_norm": 0.40810808539390564, + "learning_rate": 1.7799000000000004e-05, + "loss": 0.0419, + "step": 5936 + }, + { + "epoch": 5.014358108108108, + "grad_norm": 0.9277350306510925, + "learning_rate": 1.7802e-05, + "loss": 0.0423, + "step": 5937 + }, + { + "epoch": 5.015202702702703, + "grad_norm": 0.4797435700893402, + "learning_rate": 1.7805e-05, + "loss": 0.0314, + "step": 5938 + }, + { + "epoch": 5.016047297297297, + "grad_norm": 0.4573366940021515, + "learning_rate": 1.7808e-05, + "loss": 0.048, + "step": 5939 + }, + { + "epoch": 5.016891891891892, + "grad_norm": 0.5688931941986084, + "learning_rate": 1.7811e-05, + "loss": 0.0457, + "step": 5940 + }, + { + "epoch": 5.017736486486487, + "grad_norm": 0.49845045804977417, + "learning_rate": 1.7814e-05, + "loss": 0.0487, + "step": 5941 + }, + { + "epoch": 5.018581081081081, + "grad_norm": 0.4352245330810547, + "learning_rate": 1.7817e-05, + "loss": 0.0437, + "step": 5942 + }, + { + "epoch": 5.019425675675675, + "grad_norm": 0.5305733680725098, + "learning_rate": 1.782e-05, + "loss": 0.0579, + "step": 5943 + }, + { + "epoch": 5.02027027027027, + "grad_norm": 0.689723014831543, + "learning_rate": 1.7823e-05, + "loss": 0.0462, + "step": 5944 + }, + { + "epoch": 5.021114864864865, + "grad_norm": 0.6228036880493164, + "learning_rate": 1.7826e-05, + "loss": 0.0496, + "step": 5945 + }, + { + "epoch": 5.02195945945946, + "grad_norm": 0.8175864815711975, + "learning_rate": 1.7829e-05, + "loss": 0.0397, + "step": 5946 + }, + { + "epoch": 5.022804054054054, + "grad_norm": 0.36321380734443665, + "learning_rate": 1.7832e-05, + "loss": 0.037, + "step": 5947 + }, + { + "epoch": 5.023648648648648, + "grad_norm": 0.43443021178245544, + "learning_rate": 1.7835e-05, + "loss": 0.0472, + "step": 5948 + }, + { + "epoch": 5.024493243243243, + "grad_norm": 0.3203311562538147, + "learning_rate": 1.7838e-05, + "loss": 0.0298, + "step": 5949 + }, + { + "epoch": 5.025337837837838, + "grad_norm": 0.4812799394130707, + "learning_rate": 1.7841e-05, + "loss": 0.043, + "step": 5950 + }, + { + "epoch": 5.0261824324324325, + "grad_norm": 0.6321377754211426, + "learning_rate": 1.7844e-05, + "loss": 0.045, + "step": 5951 + }, + { + "epoch": 5.027027027027027, + "grad_norm": 0.5063523650169373, + "learning_rate": 1.7847e-05, + "loss": 0.0446, + "step": 5952 + }, + { + "epoch": 5.027871621621622, + "grad_norm": 0.5208126902580261, + "learning_rate": 1.785e-05, + "loss": 0.0497, + "step": 5953 + }, + { + "epoch": 5.028716216216216, + "grad_norm": 0.46928995847702026, + "learning_rate": 1.7853e-05, + "loss": 0.0345, + "step": 5954 + }, + { + "epoch": 5.0295608108108105, + "grad_norm": 0.786331057548523, + "learning_rate": 1.7856e-05, + "loss": 0.0461, + "step": 5955 + }, + { + "epoch": 5.030405405405405, + "grad_norm": 0.6636422276496887, + "learning_rate": 1.7859000000000003e-05, + "loss": 0.0613, + "step": 5956 + }, + { + "epoch": 5.03125, + "grad_norm": 0.3651231527328491, + "learning_rate": 1.7862000000000003e-05, + "loss": 0.024, + "step": 5957 + }, + { + "epoch": 5.032094594594595, + "grad_norm": 0.5417736768722534, + "learning_rate": 1.7865000000000003e-05, + "loss": 0.0529, + "step": 5958 + }, + { + "epoch": 5.0329391891891895, + "grad_norm": 0.8374013304710388, + "learning_rate": 1.7868000000000002e-05, + "loss": 0.0552, + "step": 5959 + }, + { + "epoch": 5.033783783783784, + "grad_norm": 0.559799313545227, + "learning_rate": 1.7871000000000002e-05, + "loss": 0.0402, + "step": 5960 + }, + { + "epoch": 5.034628378378378, + "grad_norm": 0.6948674321174622, + "learning_rate": 1.7874000000000002e-05, + "loss": 0.0534, + "step": 5961 + }, + { + "epoch": 5.035472972972973, + "grad_norm": 0.4696570634841919, + "learning_rate": 1.7877e-05, + "loss": 0.0432, + "step": 5962 + }, + { + "epoch": 5.0363175675675675, + "grad_norm": 0.5668541193008423, + "learning_rate": 1.7879999999999998e-05, + "loss": 0.0458, + "step": 5963 + }, + { + "epoch": 5.037162162162162, + "grad_norm": 0.6182718276977539, + "learning_rate": 1.7882999999999998e-05, + "loss": 0.0577, + "step": 5964 + }, + { + "epoch": 5.038006756756757, + "grad_norm": 0.5837206840515137, + "learning_rate": 1.7885999999999998e-05, + "loss": 0.0374, + "step": 5965 + }, + { + "epoch": 5.038851351351352, + "grad_norm": 0.7688050866127014, + "learning_rate": 1.7889e-05, + "loss": 0.06, + "step": 5966 + }, + { + "epoch": 5.039695945945946, + "grad_norm": 0.5936143398284912, + "learning_rate": 1.7892e-05, + "loss": 0.0482, + "step": 5967 + }, + { + "epoch": 5.04054054054054, + "grad_norm": 0.8056132197380066, + "learning_rate": 1.7895e-05, + "loss": 0.0846, + "step": 5968 + }, + { + "epoch": 5.041385135135135, + "grad_norm": 0.629310667514801, + "learning_rate": 1.7898e-05, + "loss": 0.0605, + "step": 5969 + }, + { + "epoch": 5.04222972972973, + "grad_norm": 1.5959864854812622, + "learning_rate": 1.7901e-05, + "loss": 0.0873, + "step": 5970 + }, + { + "epoch": 5.043074324324325, + "grad_norm": 1.2125381231307983, + "learning_rate": 1.7904e-05, + "loss": 0.4068, + "step": 5971 + }, + { + "epoch": 5.043918918918919, + "grad_norm": 0.9157169461250305, + "learning_rate": 1.7907e-05, + "loss": 0.2381, + "step": 5972 + }, + { + "epoch": 5.044763513513513, + "grad_norm": 0.7591609358787537, + "learning_rate": 1.791e-05, + "loss": 0.1399, + "step": 5973 + }, + { + "epoch": 5.045608108108108, + "grad_norm": 0.5126935243606567, + "learning_rate": 1.7913e-05, + "loss": 0.0593, + "step": 5974 + }, + { + "epoch": 5.046452702702703, + "grad_norm": 0.6880307793617249, + "learning_rate": 1.7916e-05, + "loss": 0.0837, + "step": 5975 + }, + { + "epoch": 5.047297297297297, + "grad_norm": 0.5559912919998169, + "learning_rate": 1.7919000000000002e-05, + "loss": 0.058, + "step": 5976 + }, + { + "epoch": 5.048141891891892, + "grad_norm": 0.6804177165031433, + "learning_rate": 1.7922000000000002e-05, + "loss": 0.0788, + "step": 5977 + }, + { + "epoch": 5.048986486486487, + "grad_norm": 0.7075167298316956, + "learning_rate": 1.7925000000000002e-05, + "loss": 0.0631, + "step": 5978 + }, + { + "epoch": 5.049831081081081, + "grad_norm": 0.5644777417182922, + "learning_rate": 1.7928000000000002e-05, + "loss": 0.0693, + "step": 5979 + }, + { + "epoch": 5.050675675675675, + "grad_norm": 0.4475374221801758, + "learning_rate": 1.7931e-05, + "loss": 0.0407, + "step": 5980 + }, + { + "epoch": 5.05152027027027, + "grad_norm": 0.540752649307251, + "learning_rate": 1.7934e-05, + "loss": 0.0499, + "step": 5981 + }, + { + "epoch": 5.052364864864865, + "grad_norm": 0.6897169351577759, + "learning_rate": 1.7937e-05, + "loss": 0.0656, + "step": 5982 + }, + { + "epoch": 5.05320945945946, + "grad_norm": 0.6174749135971069, + "learning_rate": 1.794e-05, + "loss": 0.0437, + "step": 5983 + }, + { + "epoch": 5.054054054054054, + "grad_norm": 0.5276663303375244, + "learning_rate": 1.7943e-05, + "loss": 0.0618, + "step": 5984 + }, + { + "epoch": 5.054898648648648, + "grad_norm": 0.5178970098495483, + "learning_rate": 1.7946e-05, + "loss": 0.0431, + "step": 5985 + }, + { + "epoch": 5.055743243243243, + "grad_norm": 0.5739356875419617, + "learning_rate": 1.7949e-05, + "loss": 0.0671, + "step": 5986 + }, + { + "epoch": 5.056587837837838, + "grad_norm": 0.7834247946739197, + "learning_rate": 1.7952e-05, + "loss": 0.0479, + "step": 5987 + }, + { + "epoch": 5.0574324324324325, + "grad_norm": 0.5922560095787048, + "learning_rate": 1.7955e-05, + "loss": 0.0475, + "step": 5988 + }, + { + "epoch": 5.058277027027027, + "grad_norm": 0.7523877620697021, + "learning_rate": 1.7958e-05, + "loss": 0.0409, + "step": 5989 + }, + { + "epoch": 5.059121621621622, + "grad_norm": 0.5933565497398376, + "learning_rate": 1.7961e-05, + "loss": 0.0512, + "step": 5990 + }, + { + "epoch": 5.059966216216216, + "grad_norm": 0.5658991932868958, + "learning_rate": 1.7964e-05, + "loss": 0.0448, + "step": 5991 + }, + { + "epoch": 5.0608108108108105, + "grad_norm": 0.7173256874084473, + "learning_rate": 1.7967e-05, + "loss": 0.0656, + "step": 5992 + }, + { + "epoch": 5.061655405405405, + "grad_norm": 0.5471627116203308, + "learning_rate": 1.797e-05, + "loss": 0.0504, + "step": 5993 + }, + { + "epoch": 5.0625, + "grad_norm": 0.4931020140647888, + "learning_rate": 1.7973e-05, + "loss": 0.0402, + "step": 5994 + }, + { + "epoch": 5.063344594594595, + "grad_norm": 0.642970085144043, + "learning_rate": 1.7976e-05, + "loss": 0.0398, + "step": 5995 + }, + { + "epoch": 5.0641891891891895, + "grad_norm": 0.45459941029548645, + "learning_rate": 1.7979000000000002e-05, + "loss": 0.0437, + "step": 5996 + }, + { + "epoch": 5.065033783783784, + "grad_norm": 0.5177255272865295, + "learning_rate": 1.7982e-05, + "loss": 0.0457, + "step": 5997 + }, + { + "epoch": 5.065878378378378, + "grad_norm": 0.5046541690826416, + "learning_rate": 1.7985e-05, + "loss": 0.043, + "step": 5998 + }, + { + "epoch": 5.066722972972973, + "grad_norm": 0.41965803503990173, + "learning_rate": 1.7988e-05, + "loss": 0.0407, + "step": 5999 + }, + { + "epoch": 5.0675675675675675, + "grad_norm": 0.4317670166492462, + "learning_rate": 1.7991e-05, + "loss": 0.0509, + "step": 6000 + }, + { + "epoch": 5.0675675675675675, + "eval_cer": 0.11493052766846036, + "eval_loss": 0.3433251976966858, + "eval_runtime": 8.3626, + "eval_samples_per_second": 12.078, + "eval_steps_per_second": 0.478, + "eval_wer": 0.3408289241622575, + "step": 6000 + }, + { + "epoch": 5.068412162162162, + "grad_norm": 0.4680194556713104, + "learning_rate": 1.7994e-05, + "loss": 0.044, + "step": 6001 + }, + { + "epoch": 5.069256756756757, + "grad_norm": 0.44977134466171265, + "learning_rate": 1.7997e-05, + "loss": 0.0381, + "step": 6002 + }, + { + "epoch": 5.070101351351352, + "grad_norm": 0.33362433314323425, + "learning_rate": 1.8e-05, + "loss": 0.0398, + "step": 6003 + }, + { + "epoch": 5.070945945945946, + "grad_norm": 0.5821490287780762, + "learning_rate": 1.8003e-05, + "loss": 0.0336, + "step": 6004 + }, + { + "epoch": 5.07179054054054, + "grad_norm": 0.6062755584716797, + "learning_rate": 1.8006e-05, + "loss": 0.0565, + "step": 6005 + }, + { + "epoch": 5.072635135135135, + "grad_norm": 0.6532291173934937, + "learning_rate": 1.8009e-05, + "loss": 0.0475, + "step": 6006 + }, + { + "epoch": 5.07347972972973, + "grad_norm": 0.631094753742218, + "learning_rate": 1.8012000000000003e-05, + "loss": 0.0449, + "step": 6007 + }, + { + "epoch": 5.074324324324325, + "grad_norm": 0.45918598771095276, + "learning_rate": 1.8015000000000003e-05, + "loss": 0.0601, + "step": 6008 + }, + { + "epoch": 5.075168918918919, + "grad_norm": 0.6274428963661194, + "learning_rate": 1.8018000000000003e-05, + "loss": 0.0528, + "step": 6009 + }, + { + "epoch": 5.076013513513513, + "grad_norm": 0.4459342360496521, + "learning_rate": 1.8021000000000002e-05, + "loss": 0.0467, + "step": 6010 + }, + { + "epoch": 5.076858108108108, + "grad_norm": 0.6554032564163208, + "learning_rate": 1.8024e-05, + "loss": 0.0598, + "step": 6011 + }, + { + "epoch": 5.077702702702703, + "grad_norm": 0.4292629361152649, + "learning_rate": 1.8027e-05, + "loss": 0.0483, + "step": 6012 + }, + { + "epoch": 5.078547297297297, + "grad_norm": 1.1446447372436523, + "learning_rate": 1.803e-05, + "loss": 0.0538, + "step": 6013 + }, + { + "epoch": 5.079391891891892, + "grad_norm": 0.5953379273414612, + "learning_rate": 1.8032999999999998e-05, + "loss": 0.0451, + "step": 6014 + }, + { + "epoch": 5.080236486486487, + "grad_norm": 0.6362884640693665, + "learning_rate": 1.8035999999999998e-05, + "loss": 0.0461, + "step": 6015 + }, + { + "epoch": 5.081081081081081, + "grad_norm": 0.6051636934280396, + "learning_rate": 1.8038999999999998e-05, + "loss": 0.071, + "step": 6016 + }, + { + "epoch": 5.081925675675675, + "grad_norm": 0.6091210842132568, + "learning_rate": 1.8042e-05, + "loss": 0.0552, + "step": 6017 + }, + { + "epoch": 5.08277027027027, + "grad_norm": 0.585029125213623, + "learning_rate": 1.8045e-05, + "loss": 0.0482, + "step": 6018 + }, + { + "epoch": 5.083614864864865, + "grad_norm": 1.3531073331832886, + "learning_rate": 1.8048e-05, + "loss": 0.0902, + "step": 6019 + }, + { + "epoch": 5.08445945945946, + "grad_norm": 1.208301305770874, + "learning_rate": 1.8051e-05, + "loss": 0.1219, + "step": 6020 + }, + { + "epoch": 5.085304054054054, + "grad_norm": 1.9880722761154175, + "learning_rate": 1.8054e-05, + "loss": 0.4068, + "step": 6021 + }, + { + "epoch": 5.086148648648648, + "grad_norm": 0.7268407940864563, + "learning_rate": 1.8057e-05, + "loss": 0.1999, + "step": 6022 + }, + { + "epoch": 5.086993243243243, + "grad_norm": 0.6477835774421692, + "learning_rate": 1.806e-05, + "loss": 0.1062, + "step": 6023 + }, + { + "epoch": 5.087837837837838, + "grad_norm": 0.47461360692977905, + "learning_rate": 1.8063e-05, + "loss": 0.0594, + "step": 6024 + }, + { + "epoch": 5.0886824324324325, + "grad_norm": 0.44729965925216675, + "learning_rate": 1.8066e-05, + "loss": 0.0736, + "step": 6025 + }, + { + "epoch": 5.089527027027027, + "grad_norm": 0.5600784420967102, + "learning_rate": 1.8069e-05, + "loss": 0.0664, + "step": 6026 + }, + { + "epoch": 5.090371621621622, + "grad_norm": 0.44747933745384216, + "learning_rate": 1.8072000000000002e-05, + "loss": 0.0461, + "step": 6027 + }, + { + "epoch": 5.091216216216216, + "grad_norm": 0.29515042901039124, + "learning_rate": 1.8075000000000002e-05, + "loss": 0.0392, + "step": 6028 + }, + { + "epoch": 5.0920608108108105, + "grad_norm": 0.515205979347229, + "learning_rate": 1.8078000000000002e-05, + "loss": 0.0389, + "step": 6029 + }, + { + "epoch": 5.092905405405405, + "grad_norm": 0.38443708419799805, + "learning_rate": 1.8081000000000002e-05, + "loss": 0.0519, + "step": 6030 + }, + { + "epoch": 5.09375, + "grad_norm": 0.6289327144622803, + "learning_rate": 1.8084e-05, + "loss": 0.0826, + "step": 6031 + }, + { + "epoch": 5.094594594594595, + "grad_norm": 0.5552347302436829, + "learning_rate": 1.8087e-05, + "loss": 0.0363, + "step": 6032 + }, + { + "epoch": 5.0954391891891895, + "grad_norm": 0.4112178385257721, + "learning_rate": 1.809e-05, + "loss": 0.0617, + "step": 6033 + }, + { + "epoch": 5.096283783783784, + "grad_norm": 0.7474162578582764, + "learning_rate": 1.8093e-05, + "loss": 0.0447, + "step": 6034 + }, + { + "epoch": 5.097128378378378, + "grad_norm": 0.40639135241508484, + "learning_rate": 1.8096e-05, + "loss": 0.0364, + "step": 6035 + }, + { + "epoch": 5.097972972972973, + "grad_norm": 0.4348788857460022, + "learning_rate": 1.8098999999999997e-05, + "loss": 0.0394, + "step": 6036 + }, + { + "epoch": 5.0988175675675675, + "grad_norm": 0.6652030348777771, + "learning_rate": 1.8102e-05, + "loss": 0.0411, + "step": 6037 + }, + { + "epoch": 5.099662162162162, + "grad_norm": 0.521934449672699, + "learning_rate": 1.8105e-05, + "loss": 0.044, + "step": 6038 + }, + { + "epoch": 5.100506756756757, + "grad_norm": 0.541260838508606, + "learning_rate": 1.8108e-05, + "loss": 0.0406, + "step": 6039 + }, + { + "epoch": 5.101351351351352, + "grad_norm": 0.6450951099395752, + "learning_rate": 1.8111e-05, + "loss": 0.0708, + "step": 6040 + }, + { + "epoch": 5.102195945945946, + "grad_norm": 0.4878927767276764, + "learning_rate": 1.8114e-05, + "loss": 0.0453, + "step": 6041 + }, + { + "epoch": 5.10304054054054, + "grad_norm": 0.47741425037384033, + "learning_rate": 1.8117e-05, + "loss": 0.0269, + "step": 6042 + }, + { + "epoch": 5.103885135135135, + "grad_norm": 0.45516833662986755, + "learning_rate": 1.812e-05, + "loss": 0.0492, + "step": 6043 + }, + { + "epoch": 5.10472972972973, + "grad_norm": 0.4968496263027191, + "learning_rate": 1.8123e-05, + "loss": 0.0366, + "step": 6044 + }, + { + "epoch": 5.105574324324325, + "grad_norm": 0.6257831454277039, + "learning_rate": 1.8126e-05, + "loss": 0.0322, + "step": 6045 + }, + { + "epoch": 5.106418918918919, + "grad_norm": 0.5421961545944214, + "learning_rate": 1.8129e-05, + "loss": 0.0289, + "step": 6046 + }, + { + "epoch": 5.107263513513513, + "grad_norm": 0.5860650539398193, + "learning_rate": 1.8132000000000002e-05, + "loss": 0.0514, + "step": 6047 + }, + { + "epoch": 5.108108108108108, + "grad_norm": 0.4933140277862549, + "learning_rate": 1.8135000000000002e-05, + "loss": 0.0439, + "step": 6048 + }, + { + "epoch": 5.108952702702703, + "grad_norm": 0.5068004727363586, + "learning_rate": 1.8138e-05, + "loss": 0.0399, + "step": 6049 + }, + { + "epoch": 5.109797297297297, + "grad_norm": 0.4740648567676544, + "learning_rate": 1.8141e-05, + "loss": 0.0449, + "step": 6050 + }, + { + "epoch": 5.110641891891892, + "grad_norm": 0.38258713483810425, + "learning_rate": 1.8144e-05, + "loss": 0.0342, + "step": 6051 + }, + { + "epoch": 5.111486486486487, + "grad_norm": 0.9249502420425415, + "learning_rate": 1.8147e-05, + "loss": 0.0618, + "step": 6052 + }, + { + "epoch": 5.112331081081081, + "grad_norm": 0.4820403754711151, + "learning_rate": 1.815e-05, + "loss": 0.0419, + "step": 6053 + }, + { + "epoch": 5.113175675675675, + "grad_norm": 1.2158780097961426, + "learning_rate": 1.8153e-05, + "loss": 0.0563, + "step": 6054 + }, + { + "epoch": 5.11402027027027, + "grad_norm": 0.6574094891548157, + "learning_rate": 1.8156e-05, + "loss": 0.0701, + "step": 6055 + }, + { + "epoch": 5.114864864864865, + "grad_norm": 0.49592116475105286, + "learning_rate": 1.8159e-05, + "loss": 0.0515, + "step": 6056 + }, + { + "epoch": 5.11570945945946, + "grad_norm": 0.512376070022583, + "learning_rate": 1.8162000000000003e-05, + "loss": 0.055, + "step": 6057 + }, + { + "epoch": 5.116554054054054, + "grad_norm": 0.49842289090156555, + "learning_rate": 1.8165000000000003e-05, + "loss": 0.0501, + "step": 6058 + }, + { + "epoch": 5.117398648648648, + "grad_norm": 0.7991566061973572, + "learning_rate": 1.8168000000000003e-05, + "loss": 0.062, + "step": 6059 + }, + { + "epoch": 5.118243243243243, + "grad_norm": 0.7241093516349792, + "learning_rate": 1.8171e-05, + "loss": 0.067, + "step": 6060 + }, + { + "epoch": 5.119087837837838, + "grad_norm": 0.7108708024024963, + "learning_rate": 1.8174e-05, + "loss": 0.0643, + "step": 6061 + }, + { + "epoch": 5.1199324324324325, + "grad_norm": 1.069469928741455, + "learning_rate": 1.8177e-05, + "loss": 0.0518, + "step": 6062 + }, + { + "epoch": 5.120777027027027, + "grad_norm": 0.38520681858062744, + "learning_rate": 1.818e-05, + "loss": 0.0313, + "step": 6063 + }, + { + "epoch": 5.121621621621622, + "grad_norm": 0.8172963261604309, + "learning_rate": 1.8183e-05, + "loss": 0.0594, + "step": 6064 + }, + { + "epoch": 5.122466216216216, + "grad_norm": 0.6507301926612854, + "learning_rate": 1.8186e-05, + "loss": 0.0488, + "step": 6065 + }, + { + "epoch": 5.1233108108108105, + "grad_norm": 0.7310038208961487, + "learning_rate": 1.8188999999999998e-05, + "loss": 0.0583, + "step": 6066 + }, + { + "epoch": 5.124155405405405, + "grad_norm": 0.6058720946311951, + "learning_rate": 1.8192e-05, + "loss": 0.0368, + "step": 6067 + }, + { + "epoch": 5.125, + "grad_norm": 0.9325514435768127, + "learning_rate": 1.8195e-05, + "loss": 0.0631, + "step": 6068 + }, + { + "epoch": 5.125844594594595, + "grad_norm": 0.8385271430015564, + "learning_rate": 1.8198e-05, + "loss": 0.1071, + "step": 6069 + }, + { + "epoch": 5.1266891891891895, + "grad_norm": 2.7453315258026123, + "learning_rate": 1.8201e-05, + "loss": 0.0904, + "step": 6070 + }, + { + "epoch": 5.127533783783784, + "grad_norm": 1.2409218549728394, + "learning_rate": 1.8204e-05, + "loss": 0.3654, + "step": 6071 + }, + { + "epoch": 5.128378378378378, + "grad_norm": 1.2056021690368652, + "learning_rate": 1.8207e-05, + "loss": 0.2291, + "step": 6072 + }, + { + "epoch": 5.129222972972973, + "grad_norm": 0.5319679379463196, + "learning_rate": 1.821e-05, + "loss": 0.0789, + "step": 6073 + }, + { + "epoch": 5.1300675675675675, + "grad_norm": 0.6808795928955078, + "learning_rate": 1.8213e-05, + "loss": 0.1557, + "step": 6074 + }, + { + "epoch": 5.130912162162162, + "grad_norm": 0.5662849545478821, + "learning_rate": 1.8216e-05, + "loss": 0.0645, + "step": 6075 + }, + { + "epoch": 5.131756756756757, + "grad_norm": 0.3921866714954376, + "learning_rate": 1.8219e-05, + "loss": 0.0487, + "step": 6076 + }, + { + "epoch": 5.132601351351352, + "grad_norm": 0.4725070893764496, + "learning_rate": 1.8222000000000003e-05, + "loss": 0.0484, + "step": 6077 + }, + { + "epoch": 5.133445945945946, + "grad_norm": 0.3875507712364197, + "learning_rate": 1.8225000000000003e-05, + "loss": 0.0403, + "step": 6078 + }, + { + "epoch": 5.13429054054054, + "grad_norm": 0.5310465097427368, + "learning_rate": 1.8228000000000002e-05, + "loss": 0.0548, + "step": 6079 + }, + { + "epoch": 5.135135135135135, + "grad_norm": 0.40497270226478577, + "learning_rate": 1.8231000000000002e-05, + "loss": 0.0518, + "step": 6080 + }, + { + "epoch": 5.13597972972973, + "grad_norm": 0.4483572840690613, + "learning_rate": 1.8234000000000002e-05, + "loss": 0.0412, + "step": 6081 + }, + { + "epoch": 5.136824324324325, + "grad_norm": 0.4542190730571747, + "learning_rate": 1.8237000000000002e-05, + "loss": 0.0499, + "step": 6082 + }, + { + "epoch": 5.137668918918919, + "grad_norm": 0.4022200107574463, + "learning_rate": 1.824e-05, + "loss": 0.0387, + "step": 6083 + }, + { + "epoch": 5.138513513513513, + "grad_norm": 0.48171767592430115, + "learning_rate": 1.8243e-05, + "loss": 0.0416, + "step": 6084 + }, + { + "epoch": 5.139358108108108, + "grad_norm": 0.3484119176864624, + "learning_rate": 1.8245999999999998e-05, + "loss": 0.0319, + "step": 6085 + }, + { + "epoch": 5.140202702702703, + "grad_norm": 0.533812403678894, + "learning_rate": 1.8248999999999998e-05, + "loss": 0.0415, + "step": 6086 + }, + { + "epoch": 5.141047297297297, + "grad_norm": 0.6610385775566101, + "learning_rate": 1.8252e-05, + "loss": 0.0441, + "step": 6087 + }, + { + "epoch": 5.141891891891892, + "grad_norm": 0.48334163427352905, + "learning_rate": 1.8255e-05, + "loss": 0.063, + "step": 6088 + }, + { + "epoch": 5.142736486486487, + "grad_norm": 0.6962344646453857, + "learning_rate": 1.8258e-05, + "loss": 0.0412, + "step": 6089 + }, + { + "epoch": 5.143581081081081, + "grad_norm": 0.5371875762939453, + "learning_rate": 1.8261e-05, + "loss": 0.0488, + "step": 6090 + }, + { + "epoch": 5.144425675675675, + "grad_norm": 0.4279254972934723, + "learning_rate": 1.8264e-05, + "loss": 0.0445, + "step": 6091 + }, + { + "epoch": 5.14527027027027, + "grad_norm": 0.37190669775009155, + "learning_rate": 1.8267e-05, + "loss": 0.0344, + "step": 6092 + }, + { + "epoch": 5.146114864864865, + "grad_norm": 0.4053208529949188, + "learning_rate": 1.827e-05, + "loss": 0.038, + "step": 6093 + }, + { + "epoch": 5.14695945945946, + "grad_norm": 0.43686890602111816, + "learning_rate": 1.8273e-05, + "loss": 0.0455, + "step": 6094 + }, + { + "epoch": 5.147804054054054, + "grad_norm": 0.401054322719574, + "learning_rate": 1.8276e-05, + "loss": 0.0412, + "step": 6095 + }, + { + "epoch": 5.148648648648648, + "grad_norm": 0.6263048648834229, + "learning_rate": 1.8279e-05, + "loss": 0.0634, + "step": 6096 + }, + { + "epoch": 5.149493243243243, + "grad_norm": 0.5003920197486877, + "learning_rate": 1.8282000000000002e-05, + "loss": 0.0472, + "step": 6097 + }, + { + "epoch": 5.150337837837838, + "grad_norm": 0.4939815104007721, + "learning_rate": 1.8285000000000002e-05, + "loss": 0.0355, + "step": 6098 + }, + { + "epoch": 5.1511824324324325, + "grad_norm": 0.75388103723526, + "learning_rate": 1.8288000000000002e-05, + "loss": 0.047, + "step": 6099 + }, + { + "epoch": 5.152027027027027, + "grad_norm": 0.48498255014419556, + "learning_rate": 1.8291e-05, + "loss": 0.0448, + "step": 6100 + }, + { + "epoch": 5.152871621621622, + "grad_norm": 0.5131135582923889, + "learning_rate": 1.8294e-05, + "loss": 0.0455, + "step": 6101 + }, + { + "epoch": 5.153716216216216, + "grad_norm": 0.5513062477111816, + "learning_rate": 1.8297e-05, + "loss": 0.0559, + "step": 6102 + }, + { + "epoch": 5.1545608108108105, + "grad_norm": 0.40330561995506287, + "learning_rate": 1.83e-05, + "loss": 0.0311, + "step": 6103 + }, + { + "epoch": 5.155405405405405, + "grad_norm": 0.7348480820655823, + "learning_rate": 1.8303e-05, + "loss": 0.0649, + "step": 6104 + }, + { + "epoch": 5.15625, + "grad_norm": 0.5092321038246155, + "learning_rate": 1.8306e-05, + "loss": 0.0498, + "step": 6105 + }, + { + "epoch": 5.157094594594595, + "grad_norm": 0.4822603166103363, + "learning_rate": 1.8309e-05, + "loss": 0.0401, + "step": 6106 + }, + { + "epoch": 5.1579391891891895, + "grad_norm": 0.5959089994430542, + "learning_rate": 1.8312000000000004e-05, + "loss": 0.0525, + "step": 6107 + }, + { + "epoch": 5.158783783783784, + "grad_norm": 0.6997390389442444, + "learning_rate": 1.8315000000000003e-05, + "loss": 0.0488, + "step": 6108 + }, + { + "epoch": 5.159628378378378, + "grad_norm": 0.5711942911148071, + "learning_rate": 1.8318e-05, + "loss": 0.0432, + "step": 6109 + }, + { + "epoch": 5.160472972972973, + "grad_norm": 0.704304575920105, + "learning_rate": 1.8321e-05, + "loss": 0.0555, + "step": 6110 + }, + { + "epoch": 5.1613175675675675, + "grad_norm": 0.545961320400238, + "learning_rate": 1.8324e-05, + "loss": 0.042, + "step": 6111 + }, + { + "epoch": 5.162162162162162, + "grad_norm": 0.6400869488716125, + "learning_rate": 1.8327e-05, + "loss": 0.0539, + "step": 6112 + }, + { + "epoch": 5.163006756756757, + "grad_norm": 0.5220774412155151, + "learning_rate": 1.833e-05, + "loss": 0.0393, + "step": 6113 + }, + { + "epoch": 5.163851351351352, + "grad_norm": 0.8200057148933411, + "learning_rate": 1.8333e-05, + "loss": 0.0814, + "step": 6114 + }, + { + "epoch": 5.164695945945946, + "grad_norm": 0.6463815569877625, + "learning_rate": 1.8336e-05, + "loss": 0.0538, + "step": 6115 + }, + { + "epoch": 5.16554054054054, + "grad_norm": 0.7345590591430664, + "learning_rate": 1.8339e-05, + "loss": 0.0581, + "step": 6116 + }, + { + "epoch": 5.166385135135135, + "grad_norm": 0.9819759726524353, + "learning_rate": 1.8342e-05, + "loss": 0.0736, + "step": 6117 + }, + { + "epoch": 5.16722972972973, + "grad_norm": 0.6886054277420044, + "learning_rate": 1.8345e-05, + "loss": 0.0515, + "step": 6118 + }, + { + "epoch": 5.168074324324325, + "grad_norm": 1.4646044969558716, + "learning_rate": 1.8348e-05, + "loss": 0.0853, + "step": 6119 + }, + { + "epoch": 5.168918918918919, + "grad_norm": 1.2200727462768555, + "learning_rate": 1.8351e-05, + "loss": 0.0933, + "step": 6120 + }, + { + "epoch": 5.169763513513513, + "grad_norm": 1.1518499851226807, + "learning_rate": 1.8354e-05, + "loss": 0.3495, + "step": 6121 + }, + { + "epoch": 5.170608108108108, + "grad_norm": 0.6654052734375, + "learning_rate": 1.8357e-05, + "loss": 0.2154, + "step": 6122 + }, + { + "epoch": 5.171452702702703, + "grad_norm": 0.5980299711227417, + "learning_rate": 1.836e-05, + "loss": 0.1185, + "step": 6123 + }, + { + "epoch": 5.172297297297297, + "grad_norm": 0.4324348270893097, + "learning_rate": 1.8363e-05, + "loss": 0.0549, + "step": 6124 + }, + { + "epoch": 5.173141891891892, + "grad_norm": 0.43259677290916443, + "learning_rate": 1.8366e-05, + "loss": 0.0503, + "step": 6125 + }, + { + "epoch": 5.173986486486487, + "grad_norm": 0.5841068029403687, + "learning_rate": 1.8369e-05, + "loss": 0.072, + "step": 6126 + }, + { + "epoch": 5.174831081081081, + "grad_norm": 0.6168647408485413, + "learning_rate": 1.8372000000000003e-05, + "loss": 0.0569, + "step": 6127 + }, + { + "epoch": 5.175675675675675, + "grad_norm": 0.46819403767585754, + "learning_rate": 1.8375000000000003e-05, + "loss": 0.0585, + "step": 6128 + }, + { + "epoch": 5.17652027027027, + "grad_norm": 0.45297425985336304, + "learning_rate": 1.8378000000000003e-05, + "loss": 0.0415, + "step": 6129 + }, + { + "epoch": 5.177364864864865, + "grad_norm": 0.8696893453598022, + "learning_rate": 1.8381000000000002e-05, + "loss": 0.076, + "step": 6130 + }, + { + "epoch": 5.17820945945946, + "grad_norm": 0.6493806838989258, + "learning_rate": 1.8384000000000002e-05, + "loss": 0.0447, + "step": 6131 + }, + { + "epoch": 5.179054054054054, + "grad_norm": 0.4149948060512543, + "learning_rate": 1.8387000000000002e-05, + "loss": 0.0432, + "step": 6132 + }, + { + "epoch": 5.179898648648648, + "grad_norm": 0.4724709093570709, + "learning_rate": 1.8390000000000002e-05, + "loss": 0.0498, + "step": 6133 + }, + { + "epoch": 5.180743243243243, + "grad_norm": 0.4539729952812195, + "learning_rate": 1.8392999999999998e-05, + "loss": 0.0406, + "step": 6134 + }, + { + "epoch": 5.181587837837838, + "grad_norm": 0.44544023275375366, + "learning_rate": 1.8395999999999998e-05, + "loss": 0.0555, + "step": 6135 + }, + { + "epoch": 5.1824324324324325, + "grad_norm": 0.5713903903961182, + "learning_rate": 1.8398999999999998e-05, + "loss": 0.0414, + "step": 6136 + }, + { + "epoch": 5.183277027027027, + "grad_norm": 0.346924364566803, + "learning_rate": 1.8401999999999998e-05, + "loss": 0.033, + "step": 6137 + }, + { + "epoch": 5.184121621621622, + "grad_norm": 0.4806607961654663, + "learning_rate": 1.8405e-05, + "loss": 0.04, + "step": 6138 + }, + { + "epoch": 5.184966216216216, + "grad_norm": 0.4385968744754791, + "learning_rate": 1.8408e-05, + "loss": 0.0451, + "step": 6139 + }, + { + "epoch": 5.1858108108108105, + "grad_norm": 0.4874754846096039, + "learning_rate": 1.8411e-05, + "loss": 0.0446, + "step": 6140 + }, + { + "epoch": 5.186655405405405, + "grad_norm": 0.40838101506233215, + "learning_rate": 1.8414e-05, + "loss": 0.0389, + "step": 6141 + }, + { + "epoch": 5.1875, + "grad_norm": 0.6811216473579407, + "learning_rate": 1.8417e-05, + "loss": 0.0573, + "step": 6142 + }, + { + "epoch": 5.188344594594595, + "grad_norm": 0.5706747770309448, + "learning_rate": 1.842e-05, + "loss": 0.0629, + "step": 6143 + }, + { + "epoch": 5.1891891891891895, + "grad_norm": 0.721744179725647, + "learning_rate": 1.8423e-05, + "loss": 0.0457, + "step": 6144 + }, + { + "epoch": 5.190033783783784, + "grad_norm": 0.4751221239566803, + "learning_rate": 1.8426e-05, + "loss": 0.0519, + "step": 6145 + }, + { + "epoch": 5.190878378378378, + "grad_norm": 0.5742515325546265, + "learning_rate": 1.8429e-05, + "loss": 0.0638, + "step": 6146 + }, + { + "epoch": 5.191722972972973, + "grad_norm": 0.2696020007133484, + "learning_rate": 1.8432e-05, + "loss": 0.0194, + "step": 6147 + }, + { + "epoch": 5.1925675675675675, + "grad_norm": 0.4260650873184204, + "learning_rate": 1.8435000000000002e-05, + "loss": 0.0354, + "step": 6148 + }, + { + "epoch": 5.193412162162162, + "grad_norm": 0.4724421203136444, + "learning_rate": 1.8438000000000002e-05, + "loss": 0.051, + "step": 6149 + }, + { + "epoch": 5.194256756756757, + "grad_norm": 0.670819878578186, + "learning_rate": 1.8441000000000002e-05, + "loss": 0.0394, + "step": 6150 + }, + { + "epoch": 5.195101351351352, + "grad_norm": 0.5483358502388, + "learning_rate": 1.8444e-05, + "loss": 0.0427, + "step": 6151 + }, + { + "epoch": 5.195945945945946, + "grad_norm": 0.45821425318717957, + "learning_rate": 1.8447e-05, + "loss": 0.0348, + "step": 6152 + }, + { + "epoch": 5.19679054054054, + "grad_norm": 0.2849996089935303, + "learning_rate": 1.845e-05, + "loss": 0.0167, + "step": 6153 + }, + { + "epoch": 5.197635135135135, + "grad_norm": 0.6832044124603271, + "learning_rate": 1.8453e-05, + "loss": 0.1163, + "step": 6154 + }, + { + "epoch": 5.19847972972973, + "grad_norm": 0.3706860840320587, + "learning_rate": 1.8456e-05, + "loss": 0.0196, + "step": 6155 + }, + { + "epoch": 5.199324324324325, + "grad_norm": 0.5517539978027344, + "learning_rate": 1.8459e-05, + "loss": 0.0329, + "step": 6156 + }, + { + "epoch": 5.200168918918919, + "grad_norm": 0.6359196305274963, + "learning_rate": 1.8462e-05, + "loss": 0.0469, + "step": 6157 + }, + { + "epoch": 5.201013513513513, + "grad_norm": 0.6240790486335754, + "learning_rate": 1.8465e-05, + "loss": 0.0352, + "step": 6158 + }, + { + "epoch": 5.201858108108108, + "grad_norm": 1.1401273012161255, + "learning_rate": 1.8468e-05, + "loss": 0.0486, + "step": 6159 + }, + { + "epoch": 5.202702702702703, + "grad_norm": 0.8742199540138245, + "learning_rate": 1.8471e-05, + "loss": 0.0478, + "step": 6160 + }, + { + "epoch": 5.203547297297297, + "grad_norm": 0.4970768094062805, + "learning_rate": 1.8474e-05, + "loss": 0.0373, + "step": 6161 + }, + { + "epoch": 5.204391891891892, + "grad_norm": 0.6910507678985596, + "learning_rate": 1.8477e-05, + "loss": 0.0573, + "step": 6162 + }, + { + "epoch": 5.205236486486487, + "grad_norm": 0.6615931987762451, + "learning_rate": 1.848e-05, + "loss": 0.038, + "step": 6163 + }, + { + "epoch": 5.206081081081081, + "grad_norm": 0.5653809309005737, + "learning_rate": 1.8483e-05, + "loss": 0.0386, + "step": 6164 + }, + { + "epoch": 5.206925675675675, + "grad_norm": 0.7859644889831543, + "learning_rate": 1.8486e-05, + "loss": 0.0435, + "step": 6165 + }, + { + "epoch": 5.20777027027027, + "grad_norm": 0.5777198076248169, + "learning_rate": 1.8489e-05, + "loss": 0.046, + "step": 6166 + }, + { + "epoch": 5.208614864864865, + "grad_norm": 0.8238911032676697, + "learning_rate": 1.8492e-05, + "loss": 0.061, + "step": 6167 + }, + { + "epoch": 5.20945945945946, + "grad_norm": 1.0648860931396484, + "learning_rate": 1.8495e-05, + "loss": 0.0795, + "step": 6168 + }, + { + "epoch": 5.210304054054054, + "grad_norm": 1.4663978815078735, + "learning_rate": 1.8498e-05, + "loss": 0.0719, + "step": 6169 + }, + { + "epoch": 5.211148648648648, + "grad_norm": 2.549006700515747, + "learning_rate": 1.8501e-05, + "loss": 0.1489, + "step": 6170 + }, + { + "epoch": 5.211993243243243, + "grad_norm": 2.569354772567749, + "learning_rate": 1.8504e-05, + "loss": 0.4244, + "step": 6171 + }, + { + "epoch": 5.212837837837838, + "grad_norm": 1.2114704847335815, + "learning_rate": 1.8507e-05, + "loss": 0.2446, + "step": 6172 + }, + { + "epoch": 5.2136824324324325, + "grad_norm": 0.9022679328918457, + "learning_rate": 1.851e-05, + "loss": 0.1427, + "step": 6173 + }, + { + "epoch": 5.214527027027027, + "grad_norm": 0.7167341709136963, + "learning_rate": 1.8513e-05, + "loss": 0.0659, + "step": 6174 + }, + { + "epoch": 5.215371621621622, + "grad_norm": 0.5090349912643433, + "learning_rate": 1.8516e-05, + "loss": 0.066, + "step": 6175 + }, + { + "epoch": 5.216216216216216, + "grad_norm": 0.4305361807346344, + "learning_rate": 1.8519e-05, + "loss": 0.0562, + "step": 6176 + }, + { + "epoch": 5.2170608108108105, + "grad_norm": 0.44354039430618286, + "learning_rate": 1.8522e-05, + "loss": 0.0481, + "step": 6177 + }, + { + "epoch": 5.217905405405405, + "grad_norm": 0.6196386814117432, + "learning_rate": 1.8525000000000003e-05, + "loss": 0.0538, + "step": 6178 + }, + { + "epoch": 5.21875, + "grad_norm": 0.561539888381958, + "learning_rate": 1.8528000000000003e-05, + "loss": 0.0491, + "step": 6179 + }, + { + "epoch": 5.219594594594595, + "grad_norm": 0.5059329867362976, + "learning_rate": 1.8531000000000003e-05, + "loss": 0.0569, + "step": 6180 + }, + { + "epoch": 5.2204391891891895, + "grad_norm": 0.5387657880783081, + "learning_rate": 1.8534000000000002e-05, + "loss": 0.0479, + "step": 6181 + }, + { + "epoch": 5.221283783783784, + "grad_norm": 0.5091668367385864, + "learning_rate": 1.8537000000000002e-05, + "loss": 0.0565, + "step": 6182 + }, + { + "epoch": 5.222128378378378, + "grad_norm": 0.6554510593414307, + "learning_rate": 1.854e-05, + "loss": 0.0659, + "step": 6183 + }, + { + "epoch": 5.222972972972973, + "grad_norm": 0.5668081641197205, + "learning_rate": 1.8543e-05, + "loss": 0.0445, + "step": 6184 + }, + { + "epoch": 5.2238175675675675, + "grad_norm": 0.4400665760040283, + "learning_rate": 1.8545999999999998e-05, + "loss": 0.0486, + "step": 6185 + }, + { + "epoch": 5.224662162162162, + "grad_norm": 0.4936438202857971, + "learning_rate": 1.8548999999999998e-05, + "loss": 0.0503, + "step": 6186 + }, + { + "epoch": 5.225506756756757, + "grad_norm": 0.6630100011825562, + "learning_rate": 1.8551999999999998e-05, + "loss": 0.0445, + "step": 6187 + }, + { + "epoch": 5.226351351351352, + "grad_norm": 0.5422500967979431, + "learning_rate": 1.8555e-05, + "loss": 0.0443, + "step": 6188 + }, + { + "epoch": 5.227195945945946, + "grad_norm": 0.42054978013038635, + "learning_rate": 1.8558e-05, + "loss": 0.0367, + "step": 6189 + }, + { + "epoch": 5.22804054054054, + "grad_norm": 0.3974725306034088, + "learning_rate": 1.8561e-05, + "loss": 0.036, + "step": 6190 + }, + { + "epoch": 5.228885135135135, + "grad_norm": 0.8563280701637268, + "learning_rate": 1.8564e-05, + "loss": 0.0469, + "step": 6191 + }, + { + "epoch": 5.22972972972973, + "grad_norm": 0.3922276794910431, + "learning_rate": 1.8567e-05, + "loss": 0.0309, + "step": 6192 + }, + { + "epoch": 5.230574324324325, + "grad_norm": 0.5082103610038757, + "learning_rate": 1.857e-05, + "loss": 0.0396, + "step": 6193 + }, + { + "epoch": 5.231418918918919, + "grad_norm": 0.5483917593955994, + "learning_rate": 1.8573e-05, + "loss": 0.0642, + "step": 6194 + }, + { + "epoch": 5.232263513513513, + "grad_norm": 0.5411818027496338, + "learning_rate": 1.8576e-05, + "loss": 0.0543, + "step": 6195 + }, + { + "epoch": 5.233108108108108, + "grad_norm": 0.37526458501815796, + "learning_rate": 1.8579e-05, + "loss": 0.0428, + "step": 6196 + }, + { + "epoch": 5.233952702702703, + "grad_norm": 0.4386172890663147, + "learning_rate": 1.8582e-05, + "loss": 0.0489, + "step": 6197 + }, + { + "epoch": 5.234797297297297, + "grad_norm": 0.4688875079154968, + "learning_rate": 1.8585000000000002e-05, + "loss": 0.0358, + "step": 6198 + }, + { + "epoch": 5.235641891891892, + "grad_norm": 0.6505007147789001, + "learning_rate": 1.8588000000000002e-05, + "loss": 0.0527, + "step": 6199 + }, + { + "epoch": 5.236486486486487, + "grad_norm": 0.49408262968063354, + "learning_rate": 1.8591000000000002e-05, + "loss": 0.0528, + "step": 6200 + }, + { + "epoch": 5.237331081081081, + "grad_norm": 0.5665692687034607, + "learning_rate": 1.8594000000000002e-05, + "loss": 0.0277, + "step": 6201 + }, + { + "epoch": 5.238175675675675, + "grad_norm": 0.46356794238090515, + "learning_rate": 1.8597e-05, + "loss": 0.0452, + "step": 6202 + }, + { + "epoch": 5.23902027027027, + "grad_norm": 1.0340064764022827, + "learning_rate": 1.86e-05, + "loss": 0.0546, + "step": 6203 + }, + { + "epoch": 5.239864864864865, + "grad_norm": 0.4792279601097107, + "learning_rate": 1.8603e-05, + "loss": 0.0455, + "step": 6204 + }, + { + "epoch": 5.24070945945946, + "grad_norm": 0.46365270018577576, + "learning_rate": 1.8606e-05, + "loss": 0.0307, + "step": 6205 + }, + { + "epoch": 5.241554054054054, + "grad_norm": 0.6124324202537537, + "learning_rate": 1.8609e-05, + "loss": 0.0573, + "step": 6206 + }, + { + "epoch": 5.242398648648648, + "grad_norm": 0.7252229452133179, + "learning_rate": 1.8612e-05, + "loss": 0.0509, + "step": 6207 + }, + { + "epoch": 5.243243243243243, + "grad_norm": 0.9782439470291138, + "learning_rate": 1.8615e-05, + "loss": 0.0679, + "step": 6208 + }, + { + "epoch": 5.244087837837838, + "grad_norm": 0.6570045351982117, + "learning_rate": 1.8618e-05, + "loss": 0.0375, + "step": 6209 + }, + { + "epoch": 5.2449324324324325, + "grad_norm": 0.6773396730422974, + "learning_rate": 1.8621e-05, + "loss": 0.0542, + "step": 6210 + }, + { + "epoch": 5.245777027027027, + "grad_norm": 0.5855635404586792, + "learning_rate": 1.8624e-05, + "loss": 0.05, + "step": 6211 + }, + { + "epoch": 5.246621621621622, + "grad_norm": 0.7347997426986694, + "learning_rate": 1.8627e-05, + "loss": 0.0518, + "step": 6212 + }, + { + "epoch": 5.247466216216216, + "grad_norm": 0.6820216774940491, + "learning_rate": 1.863e-05, + "loss": 0.0681, + "step": 6213 + }, + { + "epoch": 5.2483108108108105, + "grad_norm": 0.7282247543334961, + "learning_rate": 1.8633e-05, + "loss": 0.063, + "step": 6214 + }, + { + "epoch": 5.249155405405405, + "grad_norm": 0.4047255218029022, + "learning_rate": 1.8636e-05, + "loss": 0.0303, + "step": 6215 + }, + { + "epoch": 5.25, + "grad_norm": 1.828578233718872, + "learning_rate": 1.8639e-05, + "loss": 0.0465, + "step": 6216 + }, + { + "epoch": 5.250844594594595, + "grad_norm": 0.6400279402732849, + "learning_rate": 1.8642e-05, + "loss": 0.051, + "step": 6217 + }, + { + "epoch": 5.2516891891891895, + "grad_norm": 0.8496790528297424, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0661, + "step": 6218 + }, + { + "epoch": 5.252533783783784, + "grad_norm": 0.725494384765625, + "learning_rate": 1.8648000000000002e-05, + "loss": 0.047, + "step": 6219 + }, + { + "epoch": 5.253378378378378, + "grad_norm": 1.2988886833190918, + "learning_rate": 1.8651e-05, + "loss": 0.0776, + "step": 6220 + }, + { + "epoch": 5.254222972972973, + "grad_norm": 1.286712884902954, + "learning_rate": 1.8654e-05, + "loss": 0.398, + "step": 6221 + }, + { + "epoch": 5.2550675675675675, + "grad_norm": 0.9246966242790222, + "learning_rate": 1.8657e-05, + "loss": 0.2411, + "step": 6222 + }, + { + "epoch": 5.255912162162162, + "grad_norm": 0.6540345549583435, + "learning_rate": 1.866e-05, + "loss": 0.1034, + "step": 6223 + }, + { + "epoch": 5.256756756756757, + "grad_norm": 0.654694139957428, + "learning_rate": 1.8663e-05, + "loss": 0.1297, + "step": 6224 + }, + { + "epoch": 5.257601351351352, + "grad_norm": 0.5072597861289978, + "learning_rate": 1.8666e-05, + "loss": 0.0381, + "step": 6225 + }, + { + "epoch": 5.258445945945946, + "grad_norm": 0.5373795032501221, + "learning_rate": 1.8669e-05, + "loss": 0.055, + "step": 6226 + }, + { + "epoch": 5.25929054054054, + "grad_norm": 0.4374450445175171, + "learning_rate": 1.8672e-05, + "loss": 0.0421, + "step": 6227 + }, + { + "epoch": 5.260135135135135, + "grad_norm": 0.5289456844329834, + "learning_rate": 1.8675000000000003e-05, + "loss": 0.0543, + "step": 6228 + }, + { + "epoch": 5.26097972972973, + "grad_norm": 0.49113133549690247, + "learning_rate": 1.8678000000000003e-05, + "loss": 0.061, + "step": 6229 + }, + { + "epoch": 5.261824324324325, + "grad_norm": 0.6383668780326843, + "learning_rate": 1.8681000000000003e-05, + "loss": 0.0588, + "step": 6230 + }, + { + "epoch": 5.262668918918919, + "grad_norm": 0.5301132202148438, + "learning_rate": 1.8684000000000003e-05, + "loss": 0.049, + "step": 6231 + }, + { + "epoch": 5.263513513513513, + "grad_norm": 0.4912107288837433, + "learning_rate": 1.8687e-05, + "loss": 0.0438, + "step": 6232 + }, + { + "epoch": 5.264358108108108, + "grad_norm": 0.40606963634490967, + "learning_rate": 1.869e-05, + "loss": 0.0407, + "step": 6233 + }, + { + "epoch": 5.265202702702703, + "grad_norm": 0.45827019214630127, + "learning_rate": 1.8693e-05, + "loss": 0.0539, + "step": 6234 + }, + { + "epoch": 5.266047297297297, + "grad_norm": 0.5499840974807739, + "learning_rate": 1.8696e-05, + "loss": 0.0451, + "step": 6235 + }, + { + "epoch": 5.266891891891892, + "grad_norm": 0.47248366475105286, + "learning_rate": 1.8699e-05, + "loss": 0.0359, + "step": 6236 + }, + { + "epoch": 5.267736486486487, + "grad_norm": 1.698944330215454, + "learning_rate": 1.8701999999999998e-05, + "loss": 0.0544, + "step": 6237 + }, + { + "epoch": 5.268581081081081, + "grad_norm": 0.5897634625434875, + "learning_rate": 1.8705e-05, + "loss": 0.0497, + "step": 6238 + }, + { + "epoch": 5.269425675675675, + "grad_norm": 0.3836686909198761, + "learning_rate": 1.8708e-05, + "loss": 0.0302, + "step": 6239 + }, + { + "epoch": 5.27027027027027, + "grad_norm": 0.7378163933753967, + "learning_rate": 1.8711e-05, + "loss": 0.0577, + "step": 6240 + }, + { + "epoch": 5.271114864864865, + "grad_norm": 0.6061714887619019, + "learning_rate": 1.8714e-05, + "loss": 0.0655, + "step": 6241 + }, + { + "epoch": 5.27195945945946, + "grad_norm": 0.42821377515792847, + "learning_rate": 1.8717e-05, + "loss": 0.0458, + "step": 6242 + }, + { + "epoch": 5.272804054054054, + "grad_norm": 0.5508794784545898, + "learning_rate": 1.872e-05, + "loss": 0.0419, + "step": 6243 + }, + { + "epoch": 5.273648648648648, + "grad_norm": 0.5286490321159363, + "learning_rate": 1.8723e-05, + "loss": 0.0395, + "step": 6244 + }, + { + "epoch": 5.274493243243243, + "grad_norm": 0.4394782781600952, + "learning_rate": 1.8726e-05, + "loss": 0.0517, + "step": 6245 + }, + { + "epoch": 5.275337837837838, + "grad_norm": 0.5440753102302551, + "learning_rate": 1.8729e-05, + "loss": 0.0426, + "step": 6246 + }, + { + "epoch": 5.2761824324324325, + "grad_norm": 0.46428024768829346, + "learning_rate": 1.8732e-05, + "loss": 0.038, + "step": 6247 + }, + { + "epoch": 5.277027027027027, + "grad_norm": 0.8558477759361267, + "learning_rate": 1.8735000000000003e-05, + "loss": 0.0503, + "step": 6248 + }, + { + "epoch": 5.277871621621622, + "grad_norm": 0.40809640288352966, + "learning_rate": 1.8738000000000003e-05, + "loss": 0.0434, + "step": 6249 + }, + { + "epoch": 5.278716216216216, + "grad_norm": 0.4923362135887146, + "learning_rate": 1.8741000000000002e-05, + "loss": 0.0306, + "step": 6250 + }, + { + "epoch": 5.2795608108108105, + "grad_norm": 0.7472586631774902, + "learning_rate": 1.8744000000000002e-05, + "loss": 0.0652, + "step": 6251 + }, + { + "epoch": 5.280405405405405, + "grad_norm": 0.4477623701095581, + "learning_rate": 1.8747000000000002e-05, + "loss": 0.035, + "step": 6252 + }, + { + "epoch": 5.28125, + "grad_norm": 0.5465777516365051, + "learning_rate": 1.8750000000000002e-05, + "loss": 0.0408, + "step": 6253 + }, + { + "epoch": 5.282094594594595, + "grad_norm": 0.5588352680206299, + "learning_rate": 1.8753e-05, + "loss": 0.0404, + "step": 6254 + }, + { + "epoch": 5.2829391891891895, + "grad_norm": 0.520092248916626, + "learning_rate": 1.8756e-05, + "loss": 0.033, + "step": 6255 + }, + { + "epoch": 5.283783783783784, + "grad_norm": 0.6113447546958923, + "learning_rate": 1.8759e-05, + "loss": 0.0507, + "step": 6256 + }, + { + "epoch": 5.284628378378378, + "grad_norm": 0.5078818798065186, + "learning_rate": 1.8761999999999998e-05, + "loss": 0.0391, + "step": 6257 + }, + { + "epoch": 5.285472972972973, + "grad_norm": 0.8112066388130188, + "learning_rate": 1.8764999999999997e-05, + "loss": 0.0498, + "step": 6258 + }, + { + "epoch": 5.2863175675675675, + "grad_norm": 0.4634603261947632, + "learning_rate": 1.8768e-05, + "loss": 0.0329, + "step": 6259 + }, + { + "epoch": 5.287162162162162, + "grad_norm": 0.5959618091583252, + "learning_rate": 1.8771e-05, + "loss": 0.0454, + "step": 6260 + }, + { + "epoch": 5.288006756756757, + "grad_norm": 0.6256728172302246, + "learning_rate": 1.8774e-05, + "loss": 0.0541, + "step": 6261 + }, + { + "epoch": 5.288851351351352, + "grad_norm": 0.5233123898506165, + "learning_rate": 1.8777e-05, + "loss": 0.0372, + "step": 6262 + }, + { + "epoch": 5.289695945945946, + "grad_norm": 0.5167284607887268, + "learning_rate": 1.878e-05, + "loss": 0.0411, + "step": 6263 + }, + { + "epoch": 5.29054054054054, + "grad_norm": 0.7372802495956421, + "learning_rate": 1.8783e-05, + "loss": 0.0796, + "step": 6264 + }, + { + "epoch": 5.291385135135135, + "grad_norm": 0.5230129361152649, + "learning_rate": 1.8786e-05, + "loss": 0.0445, + "step": 6265 + }, + { + "epoch": 5.29222972972973, + "grad_norm": 0.783164381980896, + "learning_rate": 1.8789e-05, + "loss": 0.0954, + "step": 6266 + }, + { + "epoch": 5.293074324324325, + "grad_norm": 1.8039840459823608, + "learning_rate": 1.8792e-05, + "loss": 0.0688, + "step": 6267 + }, + { + "epoch": 5.293918918918919, + "grad_norm": 0.6050888299942017, + "learning_rate": 1.8795e-05, + "loss": 0.0461, + "step": 6268 + }, + { + "epoch": 5.294763513513513, + "grad_norm": 0.9252327084541321, + "learning_rate": 1.8798000000000002e-05, + "loss": 0.0635, + "step": 6269 + }, + { + "epoch": 5.295608108108108, + "grad_norm": 1.1939932107925415, + "learning_rate": 1.8801000000000002e-05, + "loss": 0.1558, + "step": 6270 + }, + { + "epoch": 5.296452702702703, + "grad_norm": 4.385910987854004, + "learning_rate": 1.8804e-05, + "loss": 0.5049, + "step": 6271 + }, + { + "epoch": 5.297297297297297, + "grad_norm": 0.9362618923187256, + "learning_rate": 1.8807e-05, + "loss": 0.1469, + "step": 6272 + }, + { + "epoch": 5.298141891891892, + "grad_norm": 0.6208640933036804, + "learning_rate": 1.881e-05, + "loss": 0.1131, + "step": 6273 + }, + { + "epoch": 5.298986486486487, + "grad_norm": 0.7202158570289612, + "learning_rate": 1.8813e-05, + "loss": 0.0989, + "step": 6274 + }, + { + "epoch": 5.299831081081081, + "grad_norm": 0.9260695576667786, + "learning_rate": 1.8816e-05, + "loss": 0.0674, + "step": 6275 + }, + { + "epoch": 5.300675675675675, + "grad_norm": 0.5021882057189941, + "learning_rate": 1.8819e-05, + "loss": 0.0614, + "step": 6276 + }, + { + "epoch": 5.30152027027027, + "grad_norm": 0.5877271890640259, + "learning_rate": 1.8822e-05, + "loss": 0.0449, + "step": 6277 + }, + { + "epoch": 5.302364864864865, + "grad_norm": 0.6930246353149414, + "learning_rate": 1.8825e-05, + "loss": 0.0669, + "step": 6278 + }, + { + "epoch": 5.30320945945946, + "grad_norm": 0.6835706830024719, + "learning_rate": 1.8828000000000003e-05, + "loss": 0.065, + "step": 6279 + }, + { + "epoch": 5.304054054054054, + "grad_norm": 0.43299412727355957, + "learning_rate": 1.8831000000000003e-05, + "loss": 0.0391, + "step": 6280 + }, + { + "epoch": 5.304898648648648, + "grad_norm": 0.3568536341190338, + "learning_rate": 1.8834e-05, + "loss": 0.0464, + "step": 6281 + }, + { + "epoch": 5.305743243243243, + "grad_norm": 0.3509628474712372, + "learning_rate": 1.8837e-05, + "loss": 0.0399, + "step": 6282 + }, + { + "epoch": 5.306587837837838, + "grad_norm": 0.5343244075775146, + "learning_rate": 1.884e-05, + "loss": 0.0658, + "step": 6283 + }, + { + "epoch": 5.3074324324324325, + "grad_norm": 0.5054064393043518, + "learning_rate": 1.8843e-05, + "loss": 0.0543, + "step": 6284 + }, + { + "epoch": 5.308277027027027, + "grad_norm": 0.48153799772262573, + "learning_rate": 1.8846e-05, + "loss": 0.0459, + "step": 6285 + }, + { + "epoch": 5.309121621621622, + "grad_norm": 0.5785962343215942, + "learning_rate": 1.8849e-05, + "loss": 0.0613, + "step": 6286 + }, + { + "epoch": 5.309966216216216, + "grad_norm": 0.5532504916191101, + "learning_rate": 1.8852e-05, + "loss": 0.0415, + "step": 6287 + }, + { + "epoch": 5.3108108108108105, + "grad_norm": 0.7265825867652893, + "learning_rate": 1.8854999999999998e-05, + "loss": 0.0701, + "step": 6288 + }, + { + "epoch": 5.311655405405405, + "grad_norm": 0.48315054178237915, + "learning_rate": 1.8858e-05, + "loss": 0.0451, + "step": 6289 + }, + { + "epoch": 5.3125, + "grad_norm": 0.5123523473739624, + "learning_rate": 1.8861e-05, + "loss": 0.0632, + "step": 6290 + }, + { + "epoch": 5.313344594594595, + "grad_norm": 0.5411537289619446, + "learning_rate": 1.8864e-05, + "loss": 0.0441, + "step": 6291 + }, + { + "epoch": 5.3141891891891895, + "grad_norm": 0.4863673448562622, + "learning_rate": 1.8867e-05, + "loss": 0.041, + "step": 6292 + }, + { + "epoch": 5.315033783783784, + "grad_norm": 0.5968215465545654, + "learning_rate": 1.887e-05, + "loss": 0.0475, + "step": 6293 + }, + { + "epoch": 5.315878378378378, + "grad_norm": 0.5549246668815613, + "learning_rate": 1.8873e-05, + "loss": 0.0292, + "step": 6294 + }, + { + "epoch": 5.316722972972973, + "grad_norm": 0.4530723989009857, + "learning_rate": 1.8876e-05, + "loss": 0.0362, + "step": 6295 + }, + { + "epoch": 5.3175675675675675, + "grad_norm": 0.42148613929748535, + "learning_rate": 1.8879e-05, + "loss": 0.0386, + "step": 6296 + }, + { + "epoch": 5.318412162162162, + "grad_norm": 0.4347010850906372, + "learning_rate": 1.8882e-05, + "loss": 0.0309, + "step": 6297 + }, + { + "epoch": 5.319256756756757, + "grad_norm": 0.6901813745498657, + "learning_rate": 1.8885e-05, + "loss": 0.0452, + "step": 6298 + }, + { + "epoch": 5.320101351351352, + "grad_norm": 0.5406534075737, + "learning_rate": 1.8888000000000003e-05, + "loss": 0.0354, + "step": 6299 + }, + { + "epoch": 5.320945945945946, + "grad_norm": 0.5998587012290955, + "learning_rate": 1.8891000000000003e-05, + "loss": 0.0443, + "step": 6300 + }, + { + "epoch": 5.32179054054054, + "grad_norm": 0.375460684299469, + "learning_rate": 1.8894000000000002e-05, + "loss": 0.028, + "step": 6301 + }, + { + "epoch": 5.322635135135135, + "grad_norm": 0.6904634833335876, + "learning_rate": 1.8897000000000002e-05, + "loss": 0.0441, + "step": 6302 + }, + { + "epoch": 5.32347972972973, + "grad_norm": 0.4171353876590729, + "learning_rate": 1.8900000000000002e-05, + "loss": 0.032, + "step": 6303 + }, + { + "epoch": 5.324324324324325, + "grad_norm": 0.6074157953262329, + "learning_rate": 1.8903000000000002e-05, + "loss": 0.0441, + "step": 6304 + }, + { + "epoch": 5.325168918918919, + "grad_norm": 0.7440653443336487, + "learning_rate": 1.8906e-05, + "loss": 0.0623, + "step": 6305 + }, + { + "epoch": 5.326013513513513, + "grad_norm": 0.6808591485023499, + "learning_rate": 1.8908999999999998e-05, + "loss": 0.0457, + "step": 6306 + }, + { + "epoch": 5.326858108108108, + "grad_norm": 0.5666131377220154, + "learning_rate": 1.8911999999999998e-05, + "loss": 0.0591, + "step": 6307 + }, + { + "epoch": 5.327702702702703, + "grad_norm": 0.6009809970855713, + "learning_rate": 1.8914999999999998e-05, + "loss": 0.0381, + "step": 6308 + }, + { + "epoch": 5.328547297297297, + "grad_norm": 0.6068600416183472, + "learning_rate": 1.8918e-05, + "loss": 0.0427, + "step": 6309 + }, + { + "epoch": 5.329391891891892, + "grad_norm": 0.6499283909797668, + "learning_rate": 1.8921e-05, + "loss": 0.0468, + "step": 6310 + }, + { + "epoch": 5.330236486486487, + "grad_norm": 0.5901891589164734, + "learning_rate": 1.8924e-05, + "loss": 0.0569, + "step": 6311 + }, + { + "epoch": 5.331081081081081, + "grad_norm": 0.7608896493911743, + "learning_rate": 1.8927e-05, + "loss": 0.0418, + "step": 6312 + }, + { + "epoch": 5.331925675675675, + "grad_norm": 0.7206678986549377, + "learning_rate": 1.893e-05, + "loss": 0.0459, + "step": 6313 + }, + { + "epoch": 5.33277027027027, + "grad_norm": 0.5585493445396423, + "learning_rate": 1.8933e-05, + "loss": 0.0675, + "step": 6314 + }, + { + "epoch": 5.333614864864865, + "grad_norm": 0.5709409713745117, + "learning_rate": 1.8936e-05, + "loss": 0.0447, + "step": 6315 + }, + { + "epoch": 5.33445945945946, + "grad_norm": 0.5460528135299683, + "learning_rate": 1.8939e-05, + "loss": 0.0439, + "step": 6316 + }, + { + "epoch": 5.335304054054054, + "grad_norm": 0.9388118982315063, + "learning_rate": 1.8942e-05, + "loss": 0.0391, + "step": 6317 + }, + { + "epoch": 5.336148648648648, + "grad_norm": 0.6672376990318298, + "learning_rate": 1.8945e-05, + "loss": 0.0639, + "step": 6318 + }, + { + "epoch": 5.336993243243243, + "grad_norm": 1.2592092752456665, + "learning_rate": 1.8948000000000002e-05, + "loss": 0.0937, + "step": 6319 + }, + { + "epoch": 5.337837837837838, + "grad_norm": 1.266367793083191, + "learning_rate": 1.8951000000000002e-05, + "loss": 0.0959, + "step": 6320 + }, + { + "epoch": 5.3386824324324325, + "grad_norm": 3.5469095706939697, + "learning_rate": 1.8954000000000002e-05, + "loss": 0.5372, + "step": 6321 + }, + { + "epoch": 5.339527027027027, + "grad_norm": 1.6443347930908203, + "learning_rate": 1.8957e-05, + "loss": 0.3124, + "step": 6322 + }, + { + "epoch": 5.340371621621622, + "grad_norm": 1.2239731550216675, + "learning_rate": 1.896e-05, + "loss": 0.1498, + "step": 6323 + }, + { + "epoch": 5.341216216216216, + "grad_norm": 1.2300798892974854, + "learning_rate": 1.8963e-05, + "loss": 0.1376, + "step": 6324 + }, + { + "epoch": 5.3420608108108105, + "grad_norm": 0.7903221845626831, + "learning_rate": 1.8966e-05, + "loss": 0.0621, + "step": 6325 + }, + { + "epoch": 5.342905405405405, + "grad_norm": 0.4821651875972748, + "learning_rate": 1.8969e-05, + "loss": 0.0512, + "step": 6326 + }, + { + "epoch": 5.34375, + "grad_norm": 0.491866797208786, + "learning_rate": 1.8972e-05, + "loss": 0.0558, + "step": 6327 + }, + { + "epoch": 5.344594594594595, + "grad_norm": 0.5533397793769836, + "learning_rate": 1.8975e-05, + "loss": 0.0451, + "step": 6328 + }, + { + "epoch": 5.3454391891891895, + "grad_norm": 0.6627824902534485, + "learning_rate": 1.8978000000000004e-05, + "loss": 0.0543, + "step": 6329 + }, + { + "epoch": 5.346283783783784, + "grad_norm": 0.7546939849853516, + "learning_rate": 1.8981e-05, + "loss": 0.0735, + "step": 6330 + }, + { + "epoch": 5.347128378378378, + "grad_norm": 0.43148988485336304, + "learning_rate": 1.8984e-05, + "loss": 0.0452, + "step": 6331 + }, + { + "epoch": 5.347972972972973, + "grad_norm": 0.5703436136245728, + "learning_rate": 1.8987e-05, + "loss": 0.0508, + "step": 6332 + }, + { + "epoch": 5.3488175675675675, + "grad_norm": 0.5780089497566223, + "learning_rate": 1.899e-05, + "loss": 0.0793, + "step": 6333 + }, + { + "epoch": 5.349662162162162, + "grad_norm": 0.5100356340408325, + "learning_rate": 1.8993e-05, + "loss": 0.0531, + "step": 6334 + }, + { + "epoch": 5.350506756756757, + "grad_norm": 0.4095100462436676, + "learning_rate": 1.8996e-05, + "loss": 0.0524, + "step": 6335 + }, + { + "epoch": 5.351351351351352, + "grad_norm": 0.4429928660392761, + "learning_rate": 1.8999e-05, + "loss": 0.041, + "step": 6336 + }, + { + "epoch": 5.352195945945946, + "grad_norm": 0.46551766991615295, + "learning_rate": 1.9002e-05, + "loss": 0.0475, + "step": 6337 + }, + { + "epoch": 5.35304054054054, + "grad_norm": 0.5455310940742493, + "learning_rate": 1.9005e-05, + "loss": 0.068, + "step": 6338 + }, + { + "epoch": 5.353885135135135, + "grad_norm": 0.48961392045021057, + "learning_rate": 1.9008e-05, + "loss": 0.0431, + "step": 6339 + }, + { + "epoch": 5.35472972972973, + "grad_norm": 0.5682375431060791, + "learning_rate": 1.9011e-05, + "loss": 0.0589, + "step": 6340 + }, + { + "epoch": 5.355574324324325, + "grad_norm": 0.5267724990844727, + "learning_rate": 1.9014e-05, + "loss": 0.0492, + "step": 6341 + }, + { + "epoch": 5.356418918918919, + "grad_norm": 0.6374741196632385, + "learning_rate": 1.9017e-05, + "loss": 0.0654, + "step": 6342 + }, + { + "epoch": 5.357263513513513, + "grad_norm": 0.5512130856513977, + "learning_rate": 1.902e-05, + "loss": 0.0537, + "step": 6343 + }, + { + "epoch": 5.358108108108108, + "grad_norm": 0.47098761796951294, + "learning_rate": 1.9023e-05, + "loss": 0.0556, + "step": 6344 + }, + { + "epoch": 5.358952702702703, + "grad_norm": 0.6985143423080444, + "learning_rate": 1.9026e-05, + "loss": 0.0909, + "step": 6345 + }, + { + "epoch": 5.359797297297297, + "grad_norm": 0.40413427352905273, + "learning_rate": 1.9029e-05, + "loss": 0.0353, + "step": 6346 + }, + { + "epoch": 5.360641891891892, + "grad_norm": 0.49162814021110535, + "learning_rate": 1.9032e-05, + "loss": 0.0358, + "step": 6347 + }, + { + "epoch": 5.361486486486487, + "grad_norm": 0.5514740347862244, + "learning_rate": 1.9035e-05, + "loss": 0.063, + "step": 6348 + }, + { + "epoch": 5.362331081081081, + "grad_norm": 0.6857358813285828, + "learning_rate": 1.9038000000000003e-05, + "loss": 0.0608, + "step": 6349 + }, + { + "epoch": 5.363175675675675, + "grad_norm": 0.6793399453163147, + "learning_rate": 1.9041000000000003e-05, + "loss": 0.0598, + "step": 6350 + }, + { + "epoch": 5.36402027027027, + "grad_norm": 0.37525349855422974, + "learning_rate": 1.9044000000000003e-05, + "loss": 0.0343, + "step": 6351 + }, + { + "epoch": 5.364864864864865, + "grad_norm": 0.5920498967170715, + "learning_rate": 1.9047000000000002e-05, + "loss": 0.0625, + "step": 6352 + }, + { + "epoch": 5.36570945945946, + "grad_norm": 0.3472271263599396, + "learning_rate": 1.9050000000000002e-05, + "loss": 0.0355, + "step": 6353 + }, + { + "epoch": 5.366554054054054, + "grad_norm": 0.5145653486251831, + "learning_rate": 1.9053000000000002e-05, + "loss": 0.0447, + "step": 6354 + }, + { + "epoch": 5.367398648648648, + "grad_norm": 0.5298900604248047, + "learning_rate": 1.9056e-05, + "loss": 0.0481, + "step": 6355 + }, + { + "epoch": 5.368243243243243, + "grad_norm": 0.9166790843009949, + "learning_rate": 1.9058999999999998e-05, + "loss": 0.0609, + "step": 6356 + }, + { + "epoch": 5.369087837837838, + "grad_norm": 0.4895477890968323, + "learning_rate": 1.9061999999999998e-05, + "loss": 0.0391, + "step": 6357 + }, + { + "epoch": 5.3699324324324325, + "grad_norm": 1.0181907415390015, + "learning_rate": 1.9064999999999998e-05, + "loss": 0.1178, + "step": 6358 + }, + { + "epoch": 5.370777027027027, + "grad_norm": 0.5979455709457397, + "learning_rate": 1.9068e-05, + "loss": 0.0281, + "step": 6359 + }, + { + "epoch": 5.371621621621622, + "grad_norm": 0.8881919384002686, + "learning_rate": 1.9071e-05, + "loss": 0.0596, + "step": 6360 + }, + { + "epoch": 5.372466216216216, + "grad_norm": 0.7509370446205139, + "learning_rate": 1.9074e-05, + "loss": 0.0562, + "step": 6361 + }, + { + "epoch": 5.3733108108108105, + "grad_norm": 0.570557713508606, + "learning_rate": 1.9077e-05, + "loss": 0.0344, + "step": 6362 + }, + { + "epoch": 5.374155405405405, + "grad_norm": 0.5400524735450745, + "learning_rate": 1.908e-05, + "loss": 0.0541, + "step": 6363 + }, + { + "epoch": 5.375, + "grad_norm": 0.5760430097579956, + "learning_rate": 1.9083e-05, + "loss": 0.0282, + "step": 6364 + }, + { + "epoch": 5.375844594594595, + "grad_norm": 0.9126818180084229, + "learning_rate": 1.9086e-05, + "loss": 0.0957, + "step": 6365 + }, + { + "epoch": 5.3766891891891895, + "grad_norm": 0.9730751514434814, + "learning_rate": 1.9089e-05, + "loss": 0.0623, + "step": 6366 + }, + { + "epoch": 5.377533783783784, + "grad_norm": 0.6224112510681152, + "learning_rate": 1.9092e-05, + "loss": 0.0675, + "step": 6367 + }, + { + "epoch": 5.378378378378378, + "grad_norm": 0.8667709231376648, + "learning_rate": 1.9095e-05, + "loss": 0.0904, + "step": 6368 + }, + { + "epoch": 5.379222972972973, + "grad_norm": 0.9729495048522949, + "learning_rate": 1.9098000000000002e-05, + "loss": 0.0973, + "step": 6369 + }, + { + "epoch": 5.3800675675675675, + "grad_norm": 2.995615243911743, + "learning_rate": 1.9101000000000002e-05, + "loss": 0.1032, + "step": 6370 + }, + { + "epoch": 5.380912162162162, + "grad_norm": 1.6188149452209473, + "learning_rate": 1.9104000000000002e-05, + "loss": 0.414, + "step": 6371 + }, + { + "epoch": 5.381756756756757, + "grad_norm": 0.6784438490867615, + "learning_rate": 1.9107000000000002e-05, + "loss": 0.1849, + "step": 6372 + }, + { + "epoch": 5.382601351351352, + "grad_norm": 0.4860365688800812, + "learning_rate": 1.911e-05, + "loss": 0.0796, + "step": 6373 + }, + { + "epoch": 5.383445945945946, + "grad_norm": 0.5507664680480957, + "learning_rate": 1.9113e-05, + "loss": 0.0636, + "step": 6374 + }, + { + "epoch": 5.38429054054054, + "grad_norm": 0.4155215620994568, + "learning_rate": 1.9116e-05, + "loss": 0.0636, + "step": 6375 + }, + { + "epoch": 5.385135135135135, + "grad_norm": 0.609125554561615, + "learning_rate": 1.9119e-05, + "loss": 0.0568, + "step": 6376 + }, + { + "epoch": 5.38597972972973, + "grad_norm": 0.5227804780006409, + "learning_rate": 1.9122e-05, + "loss": 0.0481, + "step": 6377 + }, + { + "epoch": 5.386824324324325, + "grad_norm": 0.48235055804252625, + "learning_rate": 1.9125e-05, + "loss": 0.0578, + "step": 6378 + }, + { + "epoch": 5.387668918918919, + "grad_norm": 0.5165147185325623, + "learning_rate": 1.9128e-05, + "loss": 0.0541, + "step": 6379 + }, + { + "epoch": 5.388513513513513, + "grad_norm": 0.4976240396499634, + "learning_rate": 1.9131e-05, + "loss": 0.037, + "step": 6380 + }, + { + "epoch": 5.389358108108108, + "grad_norm": 0.8102014064788818, + "learning_rate": 1.9134e-05, + "loss": 0.0809, + "step": 6381 + }, + { + "epoch": 5.390202702702703, + "grad_norm": 0.3839742839336395, + "learning_rate": 1.9137e-05, + "loss": 0.0331, + "step": 6382 + }, + { + "epoch": 5.391047297297297, + "grad_norm": 0.5020960569381714, + "learning_rate": 1.914e-05, + "loss": 0.0544, + "step": 6383 + }, + { + "epoch": 5.391891891891892, + "grad_norm": 0.42554858326911926, + "learning_rate": 1.9143e-05, + "loss": 0.0864, + "step": 6384 + }, + { + "epoch": 5.392736486486487, + "grad_norm": 0.44742605090141296, + "learning_rate": 1.9146e-05, + "loss": 0.0395, + "step": 6385 + }, + { + "epoch": 5.393581081081081, + "grad_norm": 0.5175040364265442, + "learning_rate": 1.9149e-05, + "loss": 0.0439, + "step": 6386 + }, + { + "epoch": 5.394425675675675, + "grad_norm": 0.5173166990280151, + "learning_rate": 1.9152e-05, + "loss": 0.0526, + "step": 6387 + }, + { + "epoch": 5.39527027027027, + "grad_norm": 0.3458120822906494, + "learning_rate": 1.9155e-05, + "loss": 0.0347, + "step": 6388 + }, + { + "epoch": 5.396114864864865, + "grad_norm": 0.46553274989128113, + "learning_rate": 1.9158e-05, + "loss": 0.0422, + "step": 6389 + }, + { + "epoch": 5.39695945945946, + "grad_norm": 0.5905307531356812, + "learning_rate": 1.9161000000000002e-05, + "loss": 0.0587, + "step": 6390 + }, + { + "epoch": 5.397804054054054, + "grad_norm": 0.40400922298431396, + "learning_rate": 1.9164e-05, + "loss": 0.0424, + "step": 6391 + }, + { + "epoch": 5.398648648648648, + "grad_norm": 0.813678503036499, + "learning_rate": 1.9167e-05, + "loss": 0.1054, + "step": 6392 + }, + { + "epoch": 5.399493243243243, + "grad_norm": 0.3542092442512512, + "learning_rate": 1.917e-05, + "loss": 0.0439, + "step": 6393 + }, + { + "epoch": 5.400337837837838, + "grad_norm": 0.4669809639453888, + "learning_rate": 1.9173e-05, + "loss": 0.048, + "step": 6394 + }, + { + "epoch": 5.4011824324324325, + "grad_norm": 0.4087640047073364, + "learning_rate": 1.9176e-05, + "loss": 0.0457, + "step": 6395 + }, + { + "epoch": 5.402027027027027, + "grad_norm": 0.3790331482887268, + "learning_rate": 1.9179e-05, + "loss": 0.0495, + "step": 6396 + }, + { + "epoch": 5.402871621621622, + "grad_norm": 0.5376933217048645, + "learning_rate": 1.9182e-05, + "loss": 0.048, + "step": 6397 + }, + { + "epoch": 5.403716216216216, + "grad_norm": 0.7200658321380615, + "learning_rate": 1.9185e-05, + "loss": 0.0391, + "step": 6398 + }, + { + "epoch": 5.4045608108108105, + "grad_norm": 0.36489155888557434, + "learning_rate": 1.9188e-05, + "loss": 0.0293, + "step": 6399 + }, + { + "epoch": 5.405405405405405, + "grad_norm": 0.4536289572715759, + "learning_rate": 1.9191000000000003e-05, + "loss": 0.0443, + "step": 6400 + }, + { + "epoch": 5.40625, + "grad_norm": 0.3297941982746124, + "learning_rate": 1.9194000000000003e-05, + "loss": 0.0368, + "step": 6401 + }, + { + "epoch": 5.407094594594595, + "grad_norm": 0.40674299001693726, + "learning_rate": 1.9197000000000003e-05, + "loss": 0.0413, + "step": 6402 + }, + { + "epoch": 5.4079391891891895, + "grad_norm": 0.5099718570709229, + "learning_rate": 1.9200000000000003e-05, + "loss": 0.0393, + "step": 6403 + }, + { + "epoch": 5.408783783783784, + "grad_norm": 0.7122861742973328, + "learning_rate": 1.9203e-05, + "loss": 0.0626, + "step": 6404 + }, + { + "epoch": 5.409628378378378, + "grad_norm": 0.5002649426460266, + "learning_rate": 1.9206e-05, + "loss": 0.0381, + "step": 6405 + }, + { + "epoch": 5.410472972972973, + "grad_norm": 0.7165715098381042, + "learning_rate": 1.9209e-05, + "loss": 0.0992, + "step": 6406 + }, + { + "epoch": 5.4113175675675675, + "grad_norm": 0.438909649848938, + "learning_rate": 1.9212e-05, + "loss": 0.0554, + "step": 6407 + }, + { + "epoch": 5.412162162162162, + "grad_norm": 0.47860971093177795, + "learning_rate": 1.9214999999999998e-05, + "loss": 0.0339, + "step": 6408 + }, + { + "epoch": 5.413006756756757, + "grad_norm": 0.541526734828949, + "learning_rate": 1.9217999999999998e-05, + "loss": 0.0428, + "step": 6409 + }, + { + "epoch": 5.413851351351352, + "grad_norm": 0.5281547904014587, + "learning_rate": 1.9221e-05, + "loss": 0.0396, + "step": 6410 + }, + { + "epoch": 5.414695945945946, + "grad_norm": 0.565081000328064, + "learning_rate": 1.9224e-05, + "loss": 0.0447, + "step": 6411 + }, + { + "epoch": 5.41554054054054, + "grad_norm": 0.7355669140815735, + "learning_rate": 1.9227e-05, + "loss": 0.0368, + "step": 6412 + }, + { + "epoch": 5.416385135135135, + "grad_norm": 0.7573081851005554, + "learning_rate": 1.923e-05, + "loss": 0.0668, + "step": 6413 + }, + { + "epoch": 5.41722972972973, + "grad_norm": 0.6661837697029114, + "learning_rate": 1.9233e-05, + "loss": 0.0513, + "step": 6414 + }, + { + "epoch": 5.418074324324325, + "grad_norm": 0.5313683152198792, + "learning_rate": 1.9236e-05, + "loss": 0.0458, + "step": 6415 + }, + { + "epoch": 5.418918918918919, + "grad_norm": 0.7362865209579468, + "learning_rate": 1.9239e-05, + "loss": 0.0556, + "step": 6416 + }, + { + "epoch": 5.419763513513513, + "grad_norm": 0.6298061609268188, + "learning_rate": 1.9242e-05, + "loss": 0.061, + "step": 6417 + }, + { + "epoch": 5.420608108108108, + "grad_norm": 0.6616401076316833, + "learning_rate": 1.9245e-05, + "loss": 0.0836, + "step": 6418 + }, + { + "epoch": 5.421452702702703, + "grad_norm": 0.9502959251403809, + "learning_rate": 1.9248e-05, + "loss": 0.0752, + "step": 6419 + }, + { + "epoch": 5.422297297297297, + "grad_norm": 1.0511201620101929, + "learning_rate": 1.9251000000000003e-05, + "loss": 0.0812, + "step": 6420 + }, + { + "epoch": 5.423141891891892, + "grad_norm": 2.855971097946167, + "learning_rate": 1.9254000000000002e-05, + "loss": 0.4845, + "step": 6421 + }, + { + "epoch": 5.423986486486487, + "grad_norm": 0.7512732148170471, + "learning_rate": 1.9257000000000002e-05, + "loss": 0.2654, + "step": 6422 + }, + { + "epoch": 5.424831081081081, + "grad_norm": 0.7223790287971497, + "learning_rate": 1.9260000000000002e-05, + "loss": 0.1347, + "step": 6423 + }, + { + "epoch": 5.425675675675675, + "grad_norm": 0.7884863615036011, + "learning_rate": 1.9263000000000002e-05, + "loss": 0.085, + "step": 6424 + }, + { + "epoch": 5.42652027027027, + "grad_norm": 0.6060654520988464, + "learning_rate": 1.9266e-05, + "loss": 0.0546, + "step": 6425 + }, + { + "epoch": 5.427364864864865, + "grad_norm": 0.47712618112564087, + "learning_rate": 1.9269e-05, + "loss": 0.0639, + "step": 6426 + }, + { + "epoch": 5.42820945945946, + "grad_norm": 0.4455801248550415, + "learning_rate": 1.9272e-05, + "loss": 0.0502, + "step": 6427 + }, + { + "epoch": 5.429054054054054, + "grad_norm": 0.45190510153770447, + "learning_rate": 1.9275e-05, + "loss": 0.0432, + "step": 6428 + }, + { + "epoch": 5.429898648648648, + "grad_norm": 0.6276513338088989, + "learning_rate": 1.9277999999999997e-05, + "loss": 0.0618, + "step": 6429 + }, + { + "epoch": 5.430743243243243, + "grad_norm": 0.42013460397720337, + "learning_rate": 1.9281e-05, + "loss": 0.0403, + "step": 6430 + }, + { + "epoch": 5.431587837837838, + "grad_norm": 0.29418209195137024, + "learning_rate": 1.9284e-05, + "loss": 0.0298, + "step": 6431 + }, + { + "epoch": 5.4324324324324325, + "grad_norm": 0.7005083560943604, + "learning_rate": 1.9287e-05, + "loss": 0.0427, + "step": 6432 + }, + { + "epoch": 5.433277027027027, + "grad_norm": 0.48569154739379883, + "learning_rate": 1.929e-05, + "loss": 0.0362, + "step": 6433 + }, + { + "epoch": 5.434121621621622, + "grad_norm": 0.6812341809272766, + "learning_rate": 1.9293e-05, + "loss": 0.0458, + "step": 6434 + }, + { + "epoch": 5.434966216216216, + "grad_norm": 0.462686687707901, + "learning_rate": 1.9296e-05, + "loss": 0.0447, + "step": 6435 + }, + { + "epoch": 5.4358108108108105, + "grad_norm": 0.39527273178100586, + "learning_rate": 1.9299e-05, + "loss": 0.0406, + "step": 6436 + }, + { + "epoch": 5.436655405405405, + "grad_norm": 0.49797171354293823, + "learning_rate": 1.9302e-05, + "loss": 0.0379, + "step": 6437 + }, + { + "epoch": 5.4375, + "grad_norm": 0.4722552001476288, + "learning_rate": 1.9305e-05, + "loss": 0.032, + "step": 6438 + }, + { + "epoch": 5.438344594594595, + "grad_norm": 0.5562309622764587, + "learning_rate": 1.9308e-05, + "loss": 0.0357, + "step": 6439 + }, + { + "epoch": 5.4391891891891895, + "grad_norm": 0.5008873343467712, + "learning_rate": 1.9311000000000002e-05, + "loss": 0.0329, + "step": 6440 + }, + { + "epoch": 5.440033783783784, + "grad_norm": 0.6035574078559875, + "learning_rate": 1.9314000000000002e-05, + "loss": 0.0639, + "step": 6441 + }, + { + "epoch": 5.440878378378378, + "grad_norm": 0.6117819547653198, + "learning_rate": 1.9317e-05, + "loss": 0.0526, + "step": 6442 + }, + { + "epoch": 5.441722972972973, + "grad_norm": 0.6670690178871155, + "learning_rate": 1.932e-05, + "loss": 0.0423, + "step": 6443 + }, + { + "epoch": 5.4425675675675675, + "grad_norm": 0.424923837184906, + "learning_rate": 1.9323e-05, + "loss": 0.0365, + "step": 6444 + }, + { + "epoch": 5.443412162162162, + "grad_norm": 0.6517522931098938, + "learning_rate": 1.9326e-05, + "loss": 0.0607, + "step": 6445 + }, + { + "epoch": 5.444256756756757, + "grad_norm": 0.6908397674560547, + "learning_rate": 1.9329e-05, + "loss": 0.0306, + "step": 6446 + }, + { + "epoch": 5.445101351351352, + "grad_norm": 0.7210440039634705, + "learning_rate": 1.9332e-05, + "loss": 0.0511, + "step": 6447 + }, + { + "epoch": 5.445945945945946, + "grad_norm": 0.42844557762145996, + "learning_rate": 1.9335e-05, + "loss": 0.05, + "step": 6448 + }, + { + "epoch": 5.44679054054054, + "grad_norm": 0.5789732933044434, + "learning_rate": 1.9338e-05, + "loss": 0.0426, + "step": 6449 + }, + { + "epoch": 5.447635135135135, + "grad_norm": 0.3433507978916168, + "learning_rate": 1.9341000000000003e-05, + "loss": 0.0321, + "step": 6450 + }, + { + "epoch": 5.44847972972973, + "grad_norm": 0.5539942979812622, + "learning_rate": 1.9344000000000003e-05, + "loss": 0.0529, + "step": 6451 + }, + { + "epoch": 5.449324324324325, + "grad_norm": 0.4774697721004486, + "learning_rate": 1.9347000000000003e-05, + "loss": 0.0557, + "step": 6452 + }, + { + "epoch": 5.450168918918919, + "grad_norm": 0.6081690192222595, + "learning_rate": 1.935e-05, + "loss": 0.0558, + "step": 6453 + }, + { + "epoch": 5.451013513513513, + "grad_norm": 0.4926130473613739, + "learning_rate": 1.9353e-05, + "loss": 0.0404, + "step": 6454 + }, + { + "epoch": 5.451858108108108, + "grad_norm": 0.5122441649436951, + "learning_rate": 1.9356e-05, + "loss": 0.0379, + "step": 6455 + }, + { + "epoch": 5.452702702702703, + "grad_norm": 0.7733946442604065, + "learning_rate": 1.9359e-05, + "loss": 0.0477, + "step": 6456 + }, + { + "epoch": 5.453547297297297, + "grad_norm": 0.5134831070899963, + "learning_rate": 1.9362e-05, + "loss": 0.0457, + "step": 6457 + }, + { + "epoch": 5.454391891891892, + "grad_norm": 0.5537050366401672, + "learning_rate": 1.9365e-05, + "loss": 0.0479, + "step": 6458 + }, + { + "epoch": 5.455236486486487, + "grad_norm": 0.7427040934562683, + "learning_rate": 1.9367999999999998e-05, + "loss": 0.0699, + "step": 6459 + }, + { + "epoch": 5.456081081081081, + "grad_norm": 1.3338677883148193, + "learning_rate": 1.9371e-05, + "loss": 0.0639, + "step": 6460 + }, + { + "epoch": 5.456925675675675, + "grad_norm": 0.5196845531463623, + "learning_rate": 1.9374e-05, + "loss": 0.044, + "step": 6461 + }, + { + "epoch": 5.45777027027027, + "grad_norm": 0.5742113590240479, + "learning_rate": 1.9377e-05, + "loss": 0.0481, + "step": 6462 + }, + { + "epoch": 5.458614864864865, + "grad_norm": 0.5246044993400574, + "learning_rate": 1.938e-05, + "loss": 0.0284, + "step": 6463 + }, + { + "epoch": 5.45945945945946, + "grad_norm": 0.930637538433075, + "learning_rate": 1.9383e-05, + "loss": 0.0515, + "step": 6464 + }, + { + "epoch": 5.460304054054054, + "grad_norm": 0.8001793622970581, + "learning_rate": 1.9386e-05, + "loss": 0.0663, + "step": 6465 + }, + { + "epoch": 5.461148648648648, + "grad_norm": 1.300156593322754, + "learning_rate": 1.9389e-05, + "loss": 0.0755, + "step": 6466 + }, + { + "epoch": 5.461993243243243, + "grad_norm": 0.6965791583061218, + "learning_rate": 1.9392e-05, + "loss": 0.0722, + "step": 6467 + }, + { + "epoch": 5.462837837837838, + "grad_norm": 0.7320055365562439, + "learning_rate": 1.9395e-05, + "loss": 0.0567, + "step": 6468 + }, + { + "epoch": 5.4636824324324325, + "grad_norm": 1.2786450386047363, + "learning_rate": 1.9398e-05, + "loss": 0.0838, + "step": 6469 + }, + { + "epoch": 5.464527027027027, + "grad_norm": 1.2958370447158813, + "learning_rate": 1.9401000000000003e-05, + "loss": 0.1109, + "step": 6470 + }, + { + "epoch": 5.465371621621622, + "grad_norm": 1.6266676187515259, + "learning_rate": 1.9404000000000003e-05, + "loss": 0.3612, + "step": 6471 + }, + { + "epoch": 5.466216216216216, + "grad_norm": 0.65400630235672, + "learning_rate": 1.9407000000000002e-05, + "loss": 0.2065, + "step": 6472 + }, + { + "epoch": 5.4670608108108105, + "grad_norm": 0.5098658800125122, + "learning_rate": 1.9410000000000002e-05, + "loss": 0.1023, + "step": 6473 + }, + { + "epoch": 5.467905405405405, + "grad_norm": 0.42521655559539795, + "learning_rate": 1.9413000000000002e-05, + "loss": 0.0525, + "step": 6474 + }, + { + "epoch": 5.46875, + "grad_norm": 0.5635232329368591, + "learning_rate": 1.9416000000000002e-05, + "loss": 0.0773, + "step": 6475 + }, + { + "epoch": 5.469594594594595, + "grad_norm": 0.8345545530319214, + "learning_rate": 1.9419e-05, + "loss": 0.0597, + "step": 6476 + }, + { + "epoch": 5.4704391891891895, + "grad_norm": 0.34253862500190735, + "learning_rate": 1.9422e-05, + "loss": 0.0357, + "step": 6477 + }, + { + "epoch": 5.471283783783784, + "grad_norm": 0.6801544427871704, + "learning_rate": 1.9424999999999998e-05, + "loss": 0.0567, + "step": 6478 + }, + { + "epoch": 5.472128378378378, + "grad_norm": 1.0520272254943848, + "learning_rate": 1.9427999999999998e-05, + "loss": 0.0722, + "step": 6479 + }, + { + "epoch": 5.472972972972973, + "grad_norm": 0.523476243019104, + "learning_rate": 1.9431e-05, + "loss": 0.0629, + "step": 6480 + }, + { + "epoch": 5.4738175675675675, + "grad_norm": 0.36708930134773254, + "learning_rate": 1.9434e-05, + "loss": 0.0388, + "step": 6481 + }, + { + "epoch": 5.474662162162162, + "grad_norm": 0.5589361786842346, + "learning_rate": 1.9437e-05, + "loss": 0.0566, + "step": 6482 + }, + { + "epoch": 5.475506756756757, + "grad_norm": 0.4646223187446594, + "learning_rate": 1.944e-05, + "loss": 0.0407, + "step": 6483 + }, + { + "epoch": 5.476351351351352, + "grad_norm": 0.5828756093978882, + "learning_rate": 1.9443e-05, + "loss": 0.0597, + "step": 6484 + }, + { + "epoch": 5.477195945945946, + "grad_norm": 0.3924199938774109, + "learning_rate": 1.9446e-05, + "loss": 0.039, + "step": 6485 + }, + { + "epoch": 5.47804054054054, + "grad_norm": 0.4880965054035187, + "learning_rate": 1.9449e-05, + "loss": 0.0429, + "step": 6486 + }, + { + "epoch": 5.478885135135135, + "grad_norm": 0.6951649785041809, + "learning_rate": 1.9452e-05, + "loss": 0.0622, + "step": 6487 + }, + { + "epoch": 5.47972972972973, + "grad_norm": 0.3420795500278473, + "learning_rate": 1.9455e-05, + "loss": 0.0267, + "step": 6488 + }, + { + "epoch": 5.480574324324325, + "grad_norm": 0.5788539052009583, + "learning_rate": 1.9458e-05, + "loss": 0.0755, + "step": 6489 + }, + { + "epoch": 5.481418918918919, + "grad_norm": 0.43914613127708435, + "learning_rate": 1.9461000000000002e-05, + "loss": 0.029, + "step": 6490 + }, + { + "epoch": 5.482263513513513, + "grad_norm": 0.5962207317352295, + "learning_rate": 1.9464000000000002e-05, + "loss": 0.0459, + "step": 6491 + }, + { + "epoch": 5.483108108108108, + "grad_norm": 0.5170795321464539, + "learning_rate": 1.9467000000000002e-05, + "loss": 0.0531, + "step": 6492 + }, + { + "epoch": 5.483952702702703, + "grad_norm": 0.48642706871032715, + "learning_rate": 1.947e-05, + "loss": 0.055, + "step": 6493 + }, + { + "epoch": 5.484797297297297, + "grad_norm": 0.5158630013465881, + "learning_rate": 1.9473e-05, + "loss": 0.0534, + "step": 6494 + }, + { + "epoch": 5.485641891891892, + "grad_norm": 0.45713549852371216, + "learning_rate": 1.9476e-05, + "loss": 0.0451, + "step": 6495 + }, + { + "epoch": 5.486486486486487, + "grad_norm": 0.7150502800941467, + "learning_rate": 1.9479e-05, + "loss": 0.0673, + "step": 6496 + }, + { + "epoch": 5.487331081081081, + "grad_norm": 0.6619002819061279, + "learning_rate": 1.9482e-05, + "loss": 0.0623, + "step": 6497 + }, + { + "epoch": 5.488175675675675, + "grad_norm": 0.532810389995575, + "learning_rate": 1.9485e-05, + "loss": 0.0334, + "step": 6498 + }, + { + "epoch": 5.48902027027027, + "grad_norm": 0.49802082777023315, + "learning_rate": 1.9488e-05, + "loss": 0.0571, + "step": 6499 + }, + { + "epoch": 5.489864864864865, + "grad_norm": 1.2990118265151978, + "learning_rate": 1.9491000000000004e-05, + "loss": 0.0611, + "step": 6500 + }, + { + "epoch": 5.49070945945946, + "grad_norm": 0.6681447625160217, + "learning_rate": 1.9494000000000003e-05, + "loss": 0.0521, + "step": 6501 + }, + { + "epoch": 5.491554054054054, + "grad_norm": 0.5758073329925537, + "learning_rate": 1.9497e-05, + "loss": 0.0674, + "step": 6502 + }, + { + "epoch": 5.492398648648648, + "grad_norm": 0.6697978973388672, + "learning_rate": 1.95e-05, + "loss": 0.0334, + "step": 6503 + }, + { + "epoch": 5.493243243243243, + "grad_norm": 0.48419857025146484, + "learning_rate": 1.9503e-05, + "loss": 0.0519, + "step": 6504 + }, + { + "epoch": 5.494087837837838, + "grad_norm": 0.37002524733543396, + "learning_rate": 1.9506e-05, + "loss": 0.0404, + "step": 6505 + }, + { + "epoch": 5.4949324324324325, + "grad_norm": 0.5350241661071777, + "learning_rate": 1.9509e-05, + "loss": 0.0453, + "step": 6506 + }, + { + "epoch": 5.495777027027027, + "grad_norm": 0.5617478489875793, + "learning_rate": 1.9512e-05, + "loss": 0.0451, + "step": 6507 + }, + { + "epoch": 5.496621621621622, + "grad_norm": 0.31552958488464355, + "learning_rate": 1.9515e-05, + "loss": 0.0291, + "step": 6508 + }, + { + "epoch": 5.497466216216216, + "grad_norm": 0.6423860788345337, + "learning_rate": 1.9518e-05, + "loss": 0.0347, + "step": 6509 + }, + { + "epoch": 5.4983108108108105, + "grad_norm": 0.5513625741004944, + "learning_rate": 1.9520999999999998e-05, + "loss": 0.0534, + "step": 6510 + }, + { + "epoch": 5.499155405405405, + "grad_norm": 0.5276400446891785, + "learning_rate": 1.9524e-05, + "loss": 0.0471, + "step": 6511 + }, + { + "epoch": 5.5, + "grad_norm": 0.6975421905517578, + "learning_rate": 1.9527e-05, + "loss": 0.0491, + "step": 6512 + }, + { + "epoch": 5.500844594594595, + "grad_norm": 1.3677353858947754, + "learning_rate": 1.953e-05, + "loss": 0.0653, + "step": 6513 + }, + { + "epoch": 5.5016891891891895, + "grad_norm": 0.7232567667961121, + "learning_rate": 1.9533e-05, + "loss": 0.081, + "step": 6514 + }, + { + "epoch": 5.502533783783784, + "grad_norm": 0.7727094888687134, + "learning_rate": 1.9536e-05, + "loss": 0.0577, + "step": 6515 + }, + { + "epoch": 5.503378378378378, + "grad_norm": 1.4879467487335205, + "learning_rate": 1.9539e-05, + "loss": 0.0465, + "step": 6516 + }, + { + "epoch": 5.504222972972973, + "grad_norm": 0.5472809076309204, + "learning_rate": 1.9542e-05, + "loss": 0.0392, + "step": 6517 + }, + { + "epoch": 5.5050675675675675, + "grad_norm": 0.6140777468681335, + "learning_rate": 1.9545e-05, + "loss": 0.0536, + "step": 6518 + }, + { + "epoch": 5.505912162162162, + "grad_norm": 0.6519005298614502, + "learning_rate": 1.9548e-05, + "loss": 0.046, + "step": 6519 + }, + { + "epoch": 5.506756756756757, + "grad_norm": 1.315064549446106, + "learning_rate": 1.9551e-05, + "loss": 0.1002, + "step": 6520 + }, + { + "epoch": 5.507601351351351, + "grad_norm": 2.166264533996582, + "learning_rate": 1.9554000000000003e-05, + "loss": 0.3714, + "step": 6521 + }, + { + "epoch": 5.508445945945946, + "grad_norm": 1.090726613998413, + "learning_rate": 1.9557000000000003e-05, + "loss": 0.2133, + "step": 6522 + }, + { + "epoch": 5.50929054054054, + "grad_norm": 0.6395775079727173, + "learning_rate": 1.9560000000000002e-05, + "loss": 0.1056, + "step": 6523 + }, + { + "epoch": 5.510135135135135, + "grad_norm": 0.6844499111175537, + "learning_rate": 1.9563000000000002e-05, + "loss": 0.1266, + "step": 6524 + }, + { + "epoch": 5.51097972972973, + "grad_norm": 0.8021019101142883, + "learning_rate": 1.9566000000000002e-05, + "loss": 0.0838, + "step": 6525 + }, + { + "epoch": 5.511824324324325, + "grad_norm": 0.5553011894226074, + "learning_rate": 1.9569000000000002e-05, + "loss": 0.0561, + "step": 6526 + }, + { + "epoch": 5.512668918918919, + "grad_norm": 0.47164952754974365, + "learning_rate": 1.9571999999999998e-05, + "loss": 0.0482, + "step": 6527 + }, + { + "epoch": 5.513513513513513, + "grad_norm": 0.40721356868743896, + "learning_rate": 1.9574999999999998e-05, + "loss": 0.0375, + "step": 6528 + }, + { + "epoch": 5.514358108108108, + "grad_norm": 0.4098554253578186, + "learning_rate": 1.9577999999999998e-05, + "loss": 0.0486, + "step": 6529 + }, + { + "epoch": 5.515202702702703, + "grad_norm": 0.5440728664398193, + "learning_rate": 1.9580999999999998e-05, + "loss": 0.037, + "step": 6530 + }, + { + "epoch": 5.516047297297297, + "grad_norm": 0.4769497811794281, + "learning_rate": 1.9584e-05, + "loss": 0.0444, + "step": 6531 + }, + { + "epoch": 5.516891891891892, + "grad_norm": 0.601378321647644, + "learning_rate": 1.9587e-05, + "loss": 0.0797, + "step": 6532 + }, + { + "epoch": 5.517736486486487, + "grad_norm": 0.6946000456809998, + "learning_rate": 1.959e-05, + "loss": 0.0379, + "step": 6533 + }, + { + "epoch": 5.518581081081081, + "grad_norm": 0.6193103790283203, + "learning_rate": 1.9593e-05, + "loss": 0.0631, + "step": 6534 + }, + { + "epoch": 5.519425675675675, + "grad_norm": 1.117037296295166, + "learning_rate": 1.9596e-05, + "loss": 0.083, + "step": 6535 + }, + { + "epoch": 5.52027027027027, + "grad_norm": 0.6932393908500671, + "learning_rate": 1.9599e-05, + "loss": 0.0549, + "step": 6536 + }, + { + "epoch": 5.521114864864865, + "grad_norm": 0.46349969506263733, + "learning_rate": 1.9602e-05, + "loss": 0.0586, + "step": 6537 + }, + { + "epoch": 5.52195945945946, + "grad_norm": 0.46539273858070374, + "learning_rate": 1.9605e-05, + "loss": 0.07, + "step": 6538 + }, + { + "epoch": 5.522804054054054, + "grad_norm": 0.4211996793746948, + "learning_rate": 1.9608e-05, + "loss": 0.0319, + "step": 6539 + }, + { + "epoch": 5.523648648648649, + "grad_norm": 0.7100182771682739, + "learning_rate": 1.9611e-05, + "loss": 0.0617, + "step": 6540 + }, + { + "epoch": 5.524493243243243, + "grad_norm": 0.3878450393676758, + "learning_rate": 1.9614000000000002e-05, + "loss": 0.0349, + "step": 6541 + }, + { + "epoch": 5.525337837837838, + "grad_norm": 0.5776789784431458, + "learning_rate": 1.9617000000000002e-05, + "loss": 0.0591, + "step": 6542 + }, + { + "epoch": 5.5261824324324325, + "grad_norm": 0.6897437572479248, + "learning_rate": 1.9620000000000002e-05, + "loss": 0.0405, + "step": 6543 + }, + { + "epoch": 5.527027027027027, + "grad_norm": 0.4868685305118561, + "learning_rate": 1.9623e-05, + "loss": 0.041, + "step": 6544 + }, + { + "epoch": 5.527871621621622, + "grad_norm": 0.6176590323448181, + "learning_rate": 1.9626e-05, + "loss": 0.098, + "step": 6545 + }, + { + "epoch": 5.528716216216216, + "grad_norm": 0.3734625577926636, + "learning_rate": 1.9629e-05, + "loss": 0.0365, + "step": 6546 + }, + { + "epoch": 5.5295608108108105, + "grad_norm": 0.589849054813385, + "learning_rate": 1.9632e-05, + "loss": 0.0494, + "step": 6547 + }, + { + "epoch": 5.530405405405405, + "grad_norm": 0.4525817334651947, + "learning_rate": 1.9635e-05, + "loss": 0.0462, + "step": 6548 + }, + { + "epoch": 5.53125, + "grad_norm": 0.3289918899536133, + "learning_rate": 1.9638e-05, + "loss": 0.0275, + "step": 6549 + }, + { + "epoch": 5.532094594594595, + "grad_norm": 0.5968455672264099, + "learning_rate": 1.9641e-05, + "loss": 0.0434, + "step": 6550 + }, + { + "epoch": 5.5329391891891895, + "grad_norm": 0.49565020203590393, + "learning_rate": 1.9644e-05, + "loss": 0.0481, + "step": 6551 + }, + { + "epoch": 5.533783783783784, + "grad_norm": 0.6474403738975525, + "learning_rate": 1.9647e-05, + "loss": 0.0612, + "step": 6552 + }, + { + "epoch": 5.534628378378378, + "grad_norm": 0.9851705431938171, + "learning_rate": 1.965e-05, + "loss": 0.0434, + "step": 6553 + }, + { + "epoch": 5.535472972972973, + "grad_norm": 0.49845126271247864, + "learning_rate": 1.9653e-05, + "loss": 0.0406, + "step": 6554 + }, + { + "epoch": 5.5363175675675675, + "grad_norm": 0.42701297998428345, + "learning_rate": 1.9656e-05, + "loss": 0.03, + "step": 6555 + }, + { + "epoch": 5.537162162162162, + "grad_norm": 0.5490127205848694, + "learning_rate": 1.9659e-05, + "loss": 0.039, + "step": 6556 + }, + { + "epoch": 5.538006756756757, + "grad_norm": 0.8233230113983154, + "learning_rate": 1.9662e-05, + "loss": 0.0658, + "step": 6557 + }, + { + "epoch": 5.538851351351351, + "grad_norm": 1.4103522300720215, + "learning_rate": 1.9665e-05, + "loss": 0.1002, + "step": 6558 + }, + { + "epoch": 5.539695945945946, + "grad_norm": 0.5925779938697815, + "learning_rate": 1.9668e-05, + "loss": 0.0429, + "step": 6559 + }, + { + "epoch": 5.54054054054054, + "grad_norm": 0.4085564911365509, + "learning_rate": 1.9671e-05, + "loss": 0.0477, + "step": 6560 + }, + { + "epoch": 5.541385135135135, + "grad_norm": 0.9780254364013672, + "learning_rate": 1.9674000000000002e-05, + "loss": 0.0663, + "step": 6561 + }, + { + "epoch": 5.54222972972973, + "grad_norm": 0.4515748620033264, + "learning_rate": 1.9677e-05, + "loss": 0.0477, + "step": 6562 + }, + { + "epoch": 5.543074324324325, + "grad_norm": 0.6012203097343445, + "learning_rate": 1.968e-05, + "loss": 0.0668, + "step": 6563 + }, + { + "epoch": 5.543918918918919, + "grad_norm": 0.8620942831039429, + "learning_rate": 1.9683e-05, + "loss": 0.0514, + "step": 6564 + }, + { + "epoch": 5.544763513513513, + "grad_norm": 0.5541222095489502, + "learning_rate": 1.9686e-05, + "loss": 0.0414, + "step": 6565 + }, + { + "epoch": 5.545608108108108, + "grad_norm": 1.1510182619094849, + "learning_rate": 1.9689e-05, + "loss": 0.0438, + "step": 6566 + }, + { + "epoch": 5.546452702702703, + "grad_norm": 0.7411449551582336, + "learning_rate": 1.9692e-05, + "loss": 0.064, + "step": 6567 + }, + { + "epoch": 5.547297297297297, + "grad_norm": 0.8466769456863403, + "learning_rate": 1.9695e-05, + "loss": 0.0514, + "step": 6568 + }, + { + "epoch": 5.548141891891892, + "grad_norm": 0.7180286049842834, + "learning_rate": 1.9698e-05, + "loss": 0.071, + "step": 6569 + }, + { + "epoch": 5.548986486486487, + "grad_norm": 1.4425907135009766, + "learning_rate": 1.9701e-05, + "loss": 0.1245, + "step": 6570 + }, + { + "epoch": 5.549831081081081, + "grad_norm": 1.1505212783813477, + "learning_rate": 1.9704000000000003e-05, + "loss": 0.2932, + "step": 6571 + }, + { + "epoch": 5.550675675675675, + "grad_norm": 1.017476201057434, + "learning_rate": 1.9707000000000003e-05, + "loss": 0.2235, + "step": 6572 + }, + { + "epoch": 5.55152027027027, + "grad_norm": 0.7462249398231506, + "learning_rate": 1.9710000000000003e-05, + "loss": 0.106, + "step": 6573 + }, + { + "epoch": 5.552364864864865, + "grad_norm": 0.6072161197662354, + "learning_rate": 1.9713000000000003e-05, + "loss": 0.0705, + "step": 6574 + }, + { + "epoch": 5.55320945945946, + "grad_norm": 0.5019283890724182, + "learning_rate": 1.9716000000000002e-05, + "loss": 0.0533, + "step": 6575 + }, + { + "epoch": 5.554054054054054, + "grad_norm": 0.4155936539173126, + "learning_rate": 1.9719e-05, + "loss": 0.0495, + "step": 6576 + }, + { + "epoch": 5.554898648648649, + "grad_norm": 0.7058822512626648, + "learning_rate": 1.9722e-05, + "loss": 0.0711, + "step": 6577 + }, + { + "epoch": 5.555743243243243, + "grad_norm": 0.41225549578666687, + "learning_rate": 1.9725e-05, + "loss": 0.0516, + "step": 6578 + }, + { + "epoch": 5.556587837837838, + "grad_norm": 0.3901579976081848, + "learning_rate": 1.9727999999999998e-05, + "loss": 0.0479, + "step": 6579 + }, + { + "epoch": 5.5574324324324325, + "grad_norm": 0.3729487657546997, + "learning_rate": 1.9730999999999998e-05, + "loss": 0.0433, + "step": 6580 + }, + { + "epoch": 5.558277027027027, + "grad_norm": 0.5702905654907227, + "learning_rate": 1.9734e-05, + "loss": 0.0528, + "step": 6581 + }, + { + "epoch": 5.559121621621622, + "grad_norm": 0.4571053385734558, + "learning_rate": 1.9737e-05, + "loss": 0.0413, + "step": 6582 + }, + { + "epoch": 5.559966216216216, + "grad_norm": 0.3767678141593933, + "learning_rate": 1.974e-05, + "loss": 0.0433, + "step": 6583 + }, + { + "epoch": 5.5608108108108105, + "grad_norm": 0.29034170508384705, + "learning_rate": 1.9743e-05, + "loss": 0.0409, + "step": 6584 + }, + { + "epoch": 5.561655405405405, + "grad_norm": 0.3870098292827606, + "learning_rate": 1.9746e-05, + "loss": 0.0513, + "step": 6585 + }, + { + "epoch": 5.5625, + "grad_norm": 0.5087055563926697, + "learning_rate": 1.9749e-05, + "loss": 0.0486, + "step": 6586 + }, + { + "epoch": 5.563344594594595, + "grad_norm": 0.8743130564689636, + "learning_rate": 1.9752e-05, + "loss": 0.0391, + "step": 6587 + }, + { + "epoch": 5.5641891891891895, + "grad_norm": 0.6184678077697754, + "learning_rate": 1.9755e-05, + "loss": 0.043, + "step": 6588 + }, + { + "epoch": 5.565033783783784, + "grad_norm": 0.8969951272010803, + "learning_rate": 1.9758e-05, + "loss": 0.0576, + "step": 6589 + }, + { + "epoch": 5.565878378378378, + "grad_norm": 0.6006982326507568, + "learning_rate": 1.9761e-05, + "loss": 0.1291, + "step": 6590 + }, + { + "epoch": 5.566722972972973, + "grad_norm": 0.8213615417480469, + "learning_rate": 1.9764000000000003e-05, + "loss": 0.0614, + "step": 6591 + }, + { + "epoch": 5.5675675675675675, + "grad_norm": 0.5374020338058472, + "learning_rate": 1.9767000000000002e-05, + "loss": 0.0433, + "step": 6592 + }, + { + "epoch": 5.568412162162162, + "grad_norm": 0.5379117727279663, + "learning_rate": 1.9770000000000002e-05, + "loss": 0.052, + "step": 6593 + }, + { + "epoch": 5.569256756756757, + "grad_norm": 0.44571730494499207, + "learning_rate": 1.9773000000000002e-05, + "loss": 0.0537, + "step": 6594 + }, + { + "epoch": 5.570101351351351, + "grad_norm": 0.513942539691925, + "learning_rate": 1.9776000000000002e-05, + "loss": 0.0388, + "step": 6595 + }, + { + "epoch": 5.570945945945946, + "grad_norm": 0.5815109610557556, + "learning_rate": 1.9779e-05, + "loss": 0.0538, + "step": 6596 + }, + { + "epoch": 5.57179054054054, + "grad_norm": 0.6653335690498352, + "learning_rate": 1.9782e-05, + "loss": 0.0712, + "step": 6597 + }, + { + "epoch": 5.572635135135135, + "grad_norm": 0.514365017414093, + "learning_rate": 1.9785e-05, + "loss": 0.0515, + "step": 6598 + }, + { + "epoch": 5.57347972972973, + "grad_norm": 0.430789053440094, + "learning_rate": 1.9788e-05, + "loss": 0.0332, + "step": 6599 + }, + { + "epoch": 5.574324324324325, + "grad_norm": 0.690294086933136, + "learning_rate": 1.9791e-05, + "loss": 0.0611, + "step": 6600 + }, + { + "epoch": 5.575168918918919, + "grad_norm": 0.33545029163360596, + "learning_rate": 1.9794e-05, + "loss": 0.0343, + "step": 6601 + }, + { + "epoch": 5.576013513513513, + "grad_norm": 0.5779942274093628, + "learning_rate": 1.9797e-05, + "loss": 0.0748, + "step": 6602 + }, + { + "epoch": 5.576858108108108, + "grad_norm": 0.5058507323265076, + "learning_rate": 1.98e-05, + "loss": 0.0408, + "step": 6603 + }, + { + "epoch": 5.577702702702703, + "grad_norm": 0.5784268975257874, + "learning_rate": 1.9803e-05, + "loss": 0.0697, + "step": 6604 + }, + { + "epoch": 5.578547297297297, + "grad_norm": 0.5732730627059937, + "learning_rate": 1.9806e-05, + "loss": 0.0456, + "step": 6605 + }, + { + "epoch": 5.579391891891892, + "grad_norm": 0.7148005962371826, + "learning_rate": 1.9809e-05, + "loss": 0.0578, + "step": 6606 + }, + { + "epoch": 5.580236486486487, + "grad_norm": 0.8262354135513306, + "learning_rate": 1.9812e-05, + "loss": 0.0429, + "step": 6607 + }, + { + "epoch": 5.581081081081081, + "grad_norm": 0.6078066825866699, + "learning_rate": 1.9815e-05, + "loss": 0.0494, + "step": 6608 + }, + { + "epoch": 5.581925675675675, + "grad_norm": 0.7824728488922119, + "learning_rate": 1.9818e-05, + "loss": 0.066, + "step": 6609 + }, + { + "epoch": 5.58277027027027, + "grad_norm": 0.6447984576225281, + "learning_rate": 1.9821e-05, + "loss": 0.0579, + "step": 6610 + }, + { + "epoch": 5.583614864864865, + "grad_norm": 0.30119726061820984, + "learning_rate": 1.9824000000000002e-05, + "loss": 0.0321, + "step": 6611 + }, + { + "epoch": 5.58445945945946, + "grad_norm": 0.9028931856155396, + "learning_rate": 1.9827000000000002e-05, + "loss": 0.0507, + "step": 6612 + }, + { + "epoch": 5.585304054054054, + "grad_norm": 0.42978954315185547, + "learning_rate": 1.983e-05, + "loss": 0.0327, + "step": 6613 + }, + { + "epoch": 5.586148648648649, + "grad_norm": 0.7695998549461365, + "learning_rate": 1.9833e-05, + "loss": 0.0597, + "step": 6614 + }, + { + "epoch": 5.586993243243243, + "grad_norm": 0.7771931886672974, + "learning_rate": 1.9836e-05, + "loss": 0.0613, + "step": 6615 + }, + { + "epoch": 5.587837837837838, + "grad_norm": 0.868556797504425, + "learning_rate": 1.9839e-05, + "loss": 0.0776, + "step": 6616 + }, + { + "epoch": 5.5886824324324325, + "grad_norm": 0.9233261942863464, + "learning_rate": 1.9842e-05, + "loss": 0.0713, + "step": 6617 + }, + { + "epoch": 5.589527027027027, + "grad_norm": 0.825441300868988, + "learning_rate": 1.9845e-05, + "loss": 0.0797, + "step": 6618 + }, + { + "epoch": 5.590371621621622, + "grad_norm": 0.6964433789253235, + "learning_rate": 1.9848e-05, + "loss": 0.0647, + "step": 6619 + }, + { + "epoch": 5.591216216216216, + "grad_norm": 1.0121238231658936, + "learning_rate": 1.9851e-05, + "loss": 0.0931, + "step": 6620 + }, + { + "epoch": 5.5920608108108105, + "grad_norm": 1.7439346313476562, + "learning_rate": 1.9854000000000003e-05, + "loss": 0.3305, + "step": 6621 + }, + { + "epoch": 5.592905405405405, + "grad_norm": 0.8460265398025513, + "learning_rate": 1.9857000000000003e-05, + "loss": 0.2529, + "step": 6622 + }, + { + "epoch": 5.59375, + "grad_norm": 0.5218455791473389, + "learning_rate": 1.9860000000000003e-05, + "loss": 0.1413, + "step": 6623 + }, + { + "epoch": 5.594594594594595, + "grad_norm": 0.6578313708305359, + "learning_rate": 1.9863000000000003e-05, + "loss": 0.0722, + "step": 6624 + }, + { + "epoch": 5.5954391891891895, + "grad_norm": 0.6083439588546753, + "learning_rate": 1.9866e-05, + "loss": 0.0834, + "step": 6625 + }, + { + "epoch": 5.596283783783784, + "grad_norm": 0.7306382656097412, + "learning_rate": 1.9869e-05, + "loss": 0.0901, + "step": 6626 + }, + { + "epoch": 5.597128378378378, + "grad_norm": 0.6043151617050171, + "learning_rate": 1.9872e-05, + "loss": 0.0676, + "step": 6627 + }, + { + "epoch": 5.597972972972973, + "grad_norm": 0.3979010283946991, + "learning_rate": 1.9875e-05, + "loss": 0.0374, + "step": 6628 + }, + { + "epoch": 5.5988175675675675, + "grad_norm": 0.6715452075004578, + "learning_rate": 1.9878e-05, + "loss": 0.0611, + "step": 6629 + }, + { + "epoch": 5.599662162162162, + "grad_norm": 0.506389856338501, + "learning_rate": 1.9880999999999998e-05, + "loss": 0.0544, + "step": 6630 + }, + { + "epoch": 5.600506756756757, + "grad_norm": 0.3263860046863556, + "learning_rate": 1.9883999999999998e-05, + "loss": 0.0405, + "step": 6631 + }, + { + "epoch": 5.601351351351351, + "grad_norm": 0.5984847545623779, + "learning_rate": 1.9887e-05, + "loss": 0.054, + "step": 6632 + }, + { + "epoch": 5.602195945945946, + "grad_norm": 0.6844760775566101, + "learning_rate": 1.989e-05, + "loss": 0.0403, + "step": 6633 + }, + { + "epoch": 5.60304054054054, + "grad_norm": 0.5948529243469238, + "learning_rate": 1.9893e-05, + "loss": 0.0592, + "step": 6634 + }, + { + "epoch": 5.603885135135135, + "grad_norm": 0.5138547420501709, + "learning_rate": 1.9896e-05, + "loss": 0.0481, + "step": 6635 + }, + { + "epoch": 5.60472972972973, + "grad_norm": 0.7029980421066284, + "learning_rate": 1.9899e-05, + "loss": 0.0627, + "step": 6636 + }, + { + "epoch": 5.605574324324325, + "grad_norm": 0.49128782749176025, + "learning_rate": 1.9902e-05, + "loss": 0.0542, + "step": 6637 + }, + { + "epoch": 5.606418918918919, + "grad_norm": 0.48678824305534363, + "learning_rate": 1.9905e-05, + "loss": 0.0654, + "step": 6638 + }, + { + "epoch": 5.607263513513513, + "grad_norm": 0.3381386399269104, + "learning_rate": 1.9908e-05, + "loss": 0.0256, + "step": 6639 + }, + { + "epoch": 5.608108108108108, + "grad_norm": 0.41976168751716614, + "learning_rate": 1.9911e-05, + "loss": 0.0418, + "step": 6640 + }, + { + "epoch": 5.608952702702703, + "grad_norm": 0.42623665928840637, + "learning_rate": 1.9914e-05, + "loss": 0.0368, + "step": 6641 + }, + { + "epoch": 5.609797297297297, + "grad_norm": 0.5241727828979492, + "learning_rate": 1.9917000000000003e-05, + "loss": 0.046, + "step": 6642 + }, + { + "epoch": 5.610641891891892, + "grad_norm": 0.5352593660354614, + "learning_rate": 1.9920000000000002e-05, + "loss": 0.0675, + "step": 6643 + }, + { + "epoch": 5.611486486486487, + "grad_norm": 0.7860026359558105, + "learning_rate": 1.9923000000000002e-05, + "loss": 0.0492, + "step": 6644 + }, + { + "epoch": 5.612331081081081, + "grad_norm": 0.752876877784729, + "learning_rate": 1.9926000000000002e-05, + "loss": 0.0661, + "step": 6645 + }, + { + "epoch": 5.613175675675675, + "grad_norm": 0.5825811624526978, + "learning_rate": 1.9929000000000002e-05, + "loss": 0.0559, + "step": 6646 + }, + { + "epoch": 5.61402027027027, + "grad_norm": 0.4264107346534729, + "learning_rate": 1.9932e-05, + "loss": 0.0459, + "step": 6647 + }, + { + "epoch": 5.614864864864865, + "grad_norm": 0.6561483144760132, + "learning_rate": 1.9935e-05, + "loss": 0.0401, + "step": 6648 + }, + { + "epoch": 5.61570945945946, + "grad_norm": 0.38008081912994385, + "learning_rate": 1.9938e-05, + "loss": 0.0335, + "step": 6649 + }, + { + "epoch": 5.616554054054054, + "grad_norm": 0.6528774499893188, + "learning_rate": 1.9940999999999998e-05, + "loss": 0.0648, + "step": 6650 + }, + { + "epoch": 5.617398648648649, + "grad_norm": 0.6048599481582642, + "learning_rate": 1.9943999999999997e-05, + "loss": 0.0458, + "step": 6651 + }, + { + "epoch": 5.618243243243243, + "grad_norm": 0.49134084582328796, + "learning_rate": 1.9947e-05, + "loss": 0.0522, + "step": 6652 + }, + { + "epoch": 5.619087837837838, + "grad_norm": 0.5589291453361511, + "learning_rate": 1.995e-05, + "loss": 0.0538, + "step": 6653 + }, + { + "epoch": 5.6199324324324325, + "grad_norm": 0.5274555683135986, + "learning_rate": 1.9953e-05, + "loss": 0.0583, + "step": 6654 + }, + { + "epoch": 5.620777027027027, + "grad_norm": 0.6085047721862793, + "learning_rate": 1.9956e-05, + "loss": 0.0413, + "step": 6655 + }, + { + "epoch": 5.621621621621622, + "grad_norm": 0.7553943395614624, + "learning_rate": 1.9959e-05, + "loss": 0.0497, + "step": 6656 + }, + { + "epoch": 5.622466216216216, + "grad_norm": 0.5338515639305115, + "learning_rate": 1.9962e-05, + "loss": 0.0404, + "step": 6657 + }, + { + "epoch": 5.6233108108108105, + "grad_norm": 0.47742682695388794, + "learning_rate": 1.9965e-05, + "loss": 0.0537, + "step": 6658 + }, + { + "epoch": 5.624155405405405, + "grad_norm": 0.642815113067627, + "learning_rate": 1.9968e-05, + "loss": 0.0705, + "step": 6659 + }, + { + "epoch": 5.625, + "grad_norm": 0.5885049104690552, + "learning_rate": 1.9971e-05, + "loss": 0.0655, + "step": 6660 + }, + { + "epoch": 5.625844594594595, + "grad_norm": 0.6304976940155029, + "learning_rate": 1.9974e-05, + "loss": 0.0587, + "step": 6661 + }, + { + "epoch": 5.6266891891891895, + "grad_norm": 0.46030256152153015, + "learning_rate": 1.9977000000000002e-05, + "loss": 0.0402, + "step": 6662 + }, + { + "epoch": 5.627533783783784, + "grad_norm": 0.6797904968261719, + "learning_rate": 1.9980000000000002e-05, + "loss": 0.0787, + "step": 6663 + }, + { + "epoch": 5.628378378378378, + "grad_norm": 0.597058117389679, + "learning_rate": 1.9983e-05, + "loss": 0.0451, + "step": 6664 + }, + { + "epoch": 5.629222972972973, + "grad_norm": 0.37991049885749817, + "learning_rate": 1.9986e-05, + "loss": 0.0375, + "step": 6665 + }, + { + "epoch": 5.6300675675675675, + "grad_norm": 0.7003069519996643, + "learning_rate": 1.9989e-05, + "loss": 0.0586, + "step": 6666 + }, + { + "epoch": 5.630912162162162, + "grad_norm": 0.6934742331504822, + "learning_rate": 1.9992e-05, + "loss": 0.0683, + "step": 6667 + }, + { + "epoch": 5.631756756756757, + "grad_norm": 9.167176246643066, + "learning_rate": 1.9995e-05, + "loss": 0.1978, + "step": 6668 + }, + { + "epoch": 5.632601351351351, + "grad_norm": 0.9193220138549805, + "learning_rate": 1.9998e-05, + "loss": 0.0877, + "step": 6669 + }, + { + "epoch": 5.633445945945946, + "grad_norm": 1.1086786985397339, + "learning_rate": 2.0001e-05, + "loss": 0.0661, + "step": 6670 + }, + { + "epoch": 5.63429054054054, + "grad_norm": 1.4942606687545776, + "learning_rate": 2.0004e-05, + "loss": 0.3679, + "step": 6671 + }, + { + "epoch": 5.635135135135135, + "grad_norm": 0.8641507625579834, + "learning_rate": 2.0007000000000003e-05, + "loss": 0.2277, + "step": 6672 + }, + { + "epoch": 5.63597972972973, + "grad_norm": 0.6516402959823608, + "learning_rate": 2.0010000000000003e-05, + "loss": 0.1028, + "step": 6673 + }, + { + "epoch": 5.636824324324325, + "grad_norm": 0.48130345344543457, + "learning_rate": 2.0013e-05, + "loss": 0.0674, + "step": 6674 + }, + { + "epoch": 5.637668918918919, + "grad_norm": 0.4221130907535553, + "learning_rate": 2.0016e-05, + "loss": 0.0607, + "step": 6675 + }, + { + "epoch": 5.638513513513513, + "grad_norm": 0.5435937643051147, + "learning_rate": 2.0019e-05, + "loss": 0.0561, + "step": 6676 + }, + { + "epoch": 5.639358108108108, + "grad_norm": 0.46711352467536926, + "learning_rate": 2.0022e-05, + "loss": 0.0394, + "step": 6677 + }, + { + "epoch": 5.640202702702703, + "grad_norm": 0.6058545708656311, + "learning_rate": 2.0025e-05, + "loss": 0.0515, + "step": 6678 + }, + { + "epoch": 5.641047297297297, + "grad_norm": 0.37414395809173584, + "learning_rate": 2.0028e-05, + "loss": 0.039, + "step": 6679 + }, + { + "epoch": 5.641891891891892, + "grad_norm": 0.5037145614624023, + "learning_rate": 2.0031e-05, + "loss": 0.0624, + "step": 6680 + }, + { + "epoch": 5.642736486486487, + "grad_norm": 0.4930229187011719, + "learning_rate": 2.0033999999999998e-05, + "loss": 0.0574, + "step": 6681 + }, + { + "epoch": 5.643581081081081, + "grad_norm": 0.7075209021568298, + "learning_rate": 2.0037e-05, + "loss": 0.0427, + "step": 6682 + }, + { + "epoch": 5.644425675675675, + "grad_norm": 1.6448034048080444, + "learning_rate": 2.004e-05, + "loss": 0.05, + "step": 6683 + }, + { + "epoch": 5.64527027027027, + "grad_norm": 0.5274102091789246, + "learning_rate": 2.0043e-05, + "loss": 0.0468, + "step": 6684 + }, + { + "epoch": 5.646114864864865, + "grad_norm": 0.5253546833992004, + "learning_rate": 2.0046e-05, + "loss": 0.0497, + "step": 6685 + }, + { + "epoch": 5.64695945945946, + "grad_norm": 0.31066974997520447, + "learning_rate": 2.0049e-05, + "loss": 0.0386, + "step": 6686 + }, + { + "epoch": 5.647804054054054, + "grad_norm": 1.1935912370681763, + "learning_rate": 2.0052e-05, + "loss": 0.0489, + "step": 6687 + }, + { + "epoch": 5.648648648648649, + "grad_norm": 0.6452167630195618, + "learning_rate": 2.0055e-05, + "loss": 0.0455, + "step": 6688 + }, + { + "epoch": 5.649493243243243, + "grad_norm": 0.5117774605751038, + "learning_rate": 2.0058e-05, + "loss": 0.0422, + "step": 6689 + }, + { + "epoch": 5.650337837837838, + "grad_norm": 0.776564359664917, + "learning_rate": 2.0061e-05, + "loss": 0.0982, + "step": 6690 + }, + { + "epoch": 5.6511824324324325, + "grad_norm": 0.46034547686576843, + "learning_rate": 2.0064e-05, + "loss": 0.0623, + "step": 6691 + }, + { + "epoch": 5.652027027027027, + "grad_norm": 0.4675822854042053, + "learning_rate": 2.0067000000000003e-05, + "loss": 0.0428, + "step": 6692 + }, + { + "epoch": 5.652871621621622, + "grad_norm": 0.5049993991851807, + "learning_rate": 2.0070000000000003e-05, + "loss": 0.0533, + "step": 6693 + }, + { + "epoch": 5.653716216216216, + "grad_norm": 0.4505496919155121, + "learning_rate": 2.0073000000000002e-05, + "loss": 0.0644, + "step": 6694 + }, + { + "epoch": 5.6545608108108105, + "grad_norm": 0.5535851716995239, + "learning_rate": 2.0076000000000002e-05, + "loss": 0.033, + "step": 6695 + }, + { + "epoch": 5.655405405405405, + "grad_norm": 1.2415578365325928, + "learning_rate": 2.0079000000000002e-05, + "loss": 0.056, + "step": 6696 + }, + { + "epoch": 5.65625, + "grad_norm": 0.5506049394607544, + "learning_rate": 2.0082000000000002e-05, + "loss": 0.0484, + "step": 6697 + }, + { + "epoch": 5.657094594594595, + "grad_norm": 0.4708053171634674, + "learning_rate": 2.0085e-05, + "loss": 0.0384, + "step": 6698 + }, + { + "epoch": 5.6579391891891895, + "grad_norm": 0.6394397616386414, + "learning_rate": 2.0087999999999998e-05, + "loss": 0.0587, + "step": 6699 + }, + { + "epoch": 5.658783783783784, + "grad_norm": 0.38965070247650146, + "learning_rate": 2.0090999999999998e-05, + "loss": 0.0447, + "step": 6700 + }, + { + "epoch": 5.659628378378378, + "grad_norm": 0.842036247253418, + "learning_rate": 2.0093999999999998e-05, + "loss": 0.0598, + "step": 6701 + }, + { + "epoch": 5.660472972972973, + "grad_norm": 0.6459710597991943, + "learning_rate": 2.0097e-05, + "loss": 0.0606, + "step": 6702 + }, + { + "epoch": 5.6613175675675675, + "grad_norm": 0.8153022527694702, + "learning_rate": 2.01e-05, + "loss": 0.0697, + "step": 6703 + }, + { + "epoch": 5.662162162162162, + "grad_norm": 0.5042495131492615, + "learning_rate": 2.0103e-05, + "loss": 0.043, + "step": 6704 + }, + { + "epoch": 5.663006756756757, + "grad_norm": 0.5109137296676636, + "learning_rate": 2.0106e-05, + "loss": 0.0435, + "step": 6705 + }, + { + "epoch": 5.663851351351351, + "grad_norm": 0.461593896150589, + "learning_rate": 2.0109e-05, + "loss": 0.0381, + "step": 6706 + }, + { + "epoch": 5.664695945945946, + "grad_norm": 0.4493881165981293, + "learning_rate": 2.0112e-05, + "loss": 0.0375, + "step": 6707 + }, + { + "epoch": 5.66554054054054, + "grad_norm": 0.49371010065078735, + "learning_rate": 2.0115e-05, + "loss": 0.0485, + "step": 6708 + }, + { + "epoch": 5.666385135135135, + "grad_norm": 0.6915813684463501, + "learning_rate": 2.0118e-05, + "loss": 0.0535, + "step": 6709 + }, + { + "epoch": 5.66722972972973, + "grad_norm": 0.4641967713832855, + "learning_rate": 2.0121e-05, + "loss": 0.0382, + "step": 6710 + }, + { + "epoch": 5.668074324324325, + "grad_norm": 0.548123836517334, + "learning_rate": 2.0124e-05, + "loss": 0.0309, + "step": 6711 + }, + { + "epoch": 5.668918918918919, + "grad_norm": 0.9668378233909607, + "learning_rate": 2.0127000000000002e-05, + "loss": 0.0606, + "step": 6712 + }, + { + "epoch": 5.669763513513513, + "grad_norm": 0.8192136883735657, + "learning_rate": 2.0130000000000002e-05, + "loss": 0.0551, + "step": 6713 + }, + { + "epoch": 5.670608108108108, + "grad_norm": 0.6719036102294922, + "learning_rate": 2.0133000000000002e-05, + "loss": 0.0745, + "step": 6714 + }, + { + "epoch": 5.671452702702703, + "grad_norm": 0.5304752588272095, + "learning_rate": 2.0136e-05, + "loss": 0.042, + "step": 6715 + }, + { + "epoch": 5.672297297297297, + "grad_norm": 0.6863628625869751, + "learning_rate": 2.0139e-05, + "loss": 0.0553, + "step": 6716 + }, + { + "epoch": 5.673141891891892, + "grad_norm": 0.6904192566871643, + "learning_rate": 2.0142e-05, + "loss": 0.0479, + "step": 6717 + }, + { + "epoch": 5.673986486486487, + "grad_norm": 1.1618640422821045, + "learning_rate": 2.0145e-05, + "loss": 0.0681, + "step": 6718 + }, + { + "epoch": 5.674831081081081, + "grad_norm": 0.9503903985023499, + "learning_rate": 2.0148e-05, + "loss": 0.0695, + "step": 6719 + }, + { + "epoch": 5.675675675675675, + "grad_norm": 1.3484811782836914, + "learning_rate": 2.0151e-05, + "loss": 0.0818, + "step": 6720 + }, + { + "epoch": 5.67652027027027, + "grad_norm": 1.0137057304382324, + "learning_rate": 2.0154e-05, + "loss": 0.3435, + "step": 6721 + }, + { + "epoch": 5.677364864864865, + "grad_norm": 0.8746610879898071, + "learning_rate": 2.0157000000000004e-05, + "loss": 0.2538, + "step": 6722 + }, + { + "epoch": 5.67820945945946, + "grad_norm": 0.7386916875839233, + "learning_rate": 2.016e-05, + "loss": 0.1236, + "step": 6723 + }, + { + "epoch": 5.679054054054054, + "grad_norm": 0.7325348258018494, + "learning_rate": 2.0163e-05, + "loss": 0.0782, + "step": 6724 + }, + { + "epoch": 5.679898648648649, + "grad_norm": 0.40273529291152954, + "learning_rate": 2.0166e-05, + "loss": 0.0705, + "step": 6725 + }, + { + "epoch": 5.680743243243243, + "grad_norm": 0.4372738301753998, + "learning_rate": 2.0169e-05, + "loss": 0.0642, + "step": 6726 + }, + { + "epoch": 5.681587837837838, + "grad_norm": 0.42882537841796875, + "learning_rate": 2.0172e-05, + "loss": 0.0495, + "step": 6727 + }, + { + "epoch": 5.6824324324324325, + "grad_norm": 0.49898383021354675, + "learning_rate": 2.0175e-05, + "loss": 0.0458, + "step": 6728 + }, + { + "epoch": 5.683277027027027, + "grad_norm": 0.4305238723754883, + "learning_rate": 2.0178e-05, + "loss": 0.0526, + "step": 6729 + }, + { + "epoch": 5.684121621621622, + "grad_norm": 0.39258238673210144, + "learning_rate": 2.0181e-05, + "loss": 0.0636, + "step": 6730 + }, + { + "epoch": 5.684966216216216, + "grad_norm": 0.4693121612071991, + "learning_rate": 2.0184e-05, + "loss": 0.0522, + "step": 6731 + }, + { + "epoch": 5.6858108108108105, + "grad_norm": 0.45171797275543213, + "learning_rate": 2.0187000000000002e-05, + "loss": 0.035, + "step": 6732 + }, + { + "epoch": 5.686655405405405, + "grad_norm": 0.32924652099609375, + "learning_rate": 2.019e-05, + "loss": 0.0349, + "step": 6733 + }, + { + "epoch": 5.6875, + "grad_norm": 0.8292099237442017, + "learning_rate": 2.0193e-05, + "loss": 0.0477, + "step": 6734 + }, + { + "epoch": 5.688344594594595, + "grad_norm": 0.48309218883514404, + "learning_rate": 2.0196e-05, + "loss": 0.0528, + "step": 6735 + }, + { + "epoch": 5.6891891891891895, + "grad_norm": 0.456332266330719, + "learning_rate": 2.0199e-05, + "loss": 0.0413, + "step": 6736 + }, + { + "epoch": 5.690033783783784, + "grad_norm": 0.45119068026542664, + "learning_rate": 2.0202e-05, + "loss": 0.0348, + "step": 6737 + }, + { + "epoch": 5.690878378378378, + "grad_norm": 0.3501463532447815, + "learning_rate": 2.0205e-05, + "loss": 0.0315, + "step": 6738 + }, + { + "epoch": 5.691722972972973, + "grad_norm": 0.4484836757183075, + "learning_rate": 2.0208e-05, + "loss": 0.0336, + "step": 6739 + }, + { + "epoch": 5.6925675675675675, + "grad_norm": 0.921169102191925, + "learning_rate": 2.0211e-05, + "loss": 0.0679, + "step": 6740 + }, + { + "epoch": 5.693412162162162, + "grad_norm": 0.5402729511260986, + "learning_rate": 2.0214e-05, + "loss": 0.043, + "step": 6741 + }, + { + "epoch": 5.694256756756757, + "grad_norm": 0.38100454211235046, + "learning_rate": 2.0217000000000003e-05, + "loss": 0.0333, + "step": 6742 + }, + { + "epoch": 5.695101351351351, + "grad_norm": 0.5485439896583557, + "learning_rate": 2.0220000000000003e-05, + "loss": 0.0448, + "step": 6743 + }, + { + "epoch": 5.695945945945946, + "grad_norm": 0.6059674620628357, + "learning_rate": 2.0223000000000003e-05, + "loss": 0.0526, + "step": 6744 + }, + { + "epoch": 5.69679054054054, + "grad_norm": 0.5099702477455139, + "learning_rate": 2.0226000000000003e-05, + "loss": 0.058, + "step": 6745 + }, + { + "epoch": 5.697635135135135, + "grad_norm": 0.41103625297546387, + "learning_rate": 2.0229000000000002e-05, + "loss": 0.0386, + "step": 6746 + }, + { + "epoch": 5.69847972972973, + "grad_norm": 0.6908190250396729, + "learning_rate": 2.0232000000000002e-05, + "loss": 0.0553, + "step": 6747 + }, + { + "epoch": 5.699324324324325, + "grad_norm": 0.47479450702667236, + "learning_rate": 2.0235e-05, + "loss": 0.0366, + "step": 6748 + }, + { + "epoch": 5.700168918918919, + "grad_norm": 0.48764634132385254, + "learning_rate": 2.0238e-05, + "loss": 0.0497, + "step": 6749 + }, + { + "epoch": 5.701013513513513, + "grad_norm": 0.48860567808151245, + "learning_rate": 2.0240999999999998e-05, + "loss": 0.043, + "step": 6750 + }, + { + "epoch": 5.701858108108108, + "grad_norm": 0.8157185912132263, + "learning_rate": 2.0243999999999998e-05, + "loss": 0.0529, + "step": 6751 + }, + { + "epoch": 5.702702702702703, + "grad_norm": 0.42971229553222656, + "learning_rate": 2.0247e-05, + "loss": 0.0447, + "step": 6752 + }, + { + "epoch": 5.703547297297297, + "grad_norm": 0.6050480604171753, + "learning_rate": 2.025e-05, + "loss": 0.0512, + "step": 6753 + }, + { + "epoch": 5.704391891891892, + "grad_norm": 0.5166911482810974, + "learning_rate": 2.0253e-05, + "loss": 0.0494, + "step": 6754 + }, + { + "epoch": 5.705236486486487, + "grad_norm": 0.789109468460083, + "learning_rate": 2.0256e-05, + "loss": 0.0682, + "step": 6755 + }, + { + "epoch": 5.706081081081081, + "grad_norm": 0.45943483710289, + "learning_rate": 2.0259e-05, + "loss": 0.0543, + "step": 6756 + }, + { + "epoch": 5.706925675675675, + "grad_norm": 0.5211117267608643, + "learning_rate": 2.0262e-05, + "loss": 0.0447, + "step": 6757 + }, + { + "epoch": 5.70777027027027, + "grad_norm": 0.5557420253753662, + "learning_rate": 2.0265e-05, + "loss": 0.0399, + "step": 6758 + }, + { + "epoch": 5.708614864864865, + "grad_norm": 0.5027178525924683, + "learning_rate": 2.0268e-05, + "loss": 0.0404, + "step": 6759 + }, + { + "epoch": 5.70945945945946, + "grad_norm": 0.7884660959243774, + "learning_rate": 2.0271e-05, + "loss": 0.0685, + "step": 6760 + }, + { + "epoch": 5.710304054054054, + "grad_norm": 0.3554600775241852, + "learning_rate": 2.0274e-05, + "loss": 0.0241, + "step": 6761 + }, + { + "epoch": 5.711148648648649, + "grad_norm": 0.5216452479362488, + "learning_rate": 2.0277e-05, + "loss": 0.036, + "step": 6762 + }, + { + "epoch": 5.711993243243243, + "grad_norm": 0.41590696573257446, + "learning_rate": 2.0280000000000002e-05, + "loss": 0.0395, + "step": 6763 + }, + { + "epoch": 5.712837837837838, + "grad_norm": 0.5986453294754028, + "learning_rate": 2.0283000000000002e-05, + "loss": 0.0407, + "step": 6764 + }, + { + "epoch": 5.7136824324324325, + "grad_norm": 0.8308255672454834, + "learning_rate": 2.0286000000000002e-05, + "loss": 0.0451, + "step": 6765 + }, + { + "epoch": 5.714527027027027, + "grad_norm": 0.85358726978302, + "learning_rate": 2.0289000000000002e-05, + "loss": 0.0546, + "step": 6766 + }, + { + "epoch": 5.715371621621622, + "grad_norm": 1.0777003765106201, + "learning_rate": 2.0292e-05, + "loss": 0.0443, + "step": 6767 + }, + { + "epoch": 5.716216216216216, + "grad_norm": 0.59566730260849, + "learning_rate": 2.0295e-05, + "loss": 0.0557, + "step": 6768 + }, + { + "epoch": 5.7170608108108105, + "grad_norm": 0.9447014331817627, + "learning_rate": 2.0298e-05, + "loss": 0.0933, + "step": 6769 + }, + { + "epoch": 5.717905405405405, + "grad_norm": 2.2091503143310547, + "learning_rate": 2.0301e-05, + "loss": 0.2328, + "step": 6770 + }, + { + "epoch": 5.71875, + "grad_norm": 1.2723544836044312, + "learning_rate": 2.0304e-05, + "loss": 0.4448, + "step": 6771 + }, + { + "epoch": 5.719594594594595, + "grad_norm": 0.7558942437171936, + "learning_rate": 2.0307e-05, + "loss": 0.1533, + "step": 6772 + }, + { + "epoch": 5.7204391891891895, + "grad_norm": 0.4302918016910553, + "learning_rate": 2.031e-05, + "loss": 0.0803, + "step": 6773 + }, + { + "epoch": 5.721283783783784, + "grad_norm": 0.45220470428466797, + "learning_rate": 2.0313e-05, + "loss": 0.0486, + "step": 6774 + }, + { + "epoch": 5.722128378378378, + "grad_norm": 0.49551287293434143, + "learning_rate": 2.0316e-05, + "loss": 0.0715, + "step": 6775 + }, + { + "epoch": 5.722972972972973, + "grad_norm": 0.5593451857566833, + "learning_rate": 2.0319e-05, + "loss": 0.0569, + "step": 6776 + }, + { + "epoch": 5.7238175675675675, + "grad_norm": 0.450895756483078, + "learning_rate": 2.0322e-05, + "loss": 0.0463, + "step": 6777 + }, + { + "epoch": 5.724662162162162, + "grad_norm": 0.44562146067619324, + "learning_rate": 2.0325e-05, + "loss": 0.0448, + "step": 6778 + }, + { + "epoch": 5.725506756756757, + "grad_norm": 0.4237382709980011, + "learning_rate": 2.0328e-05, + "loss": 0.0555, + "step": 6779 + }, + { + "epoch": 5.726351351351351, + "grad_norm": 0.3779258131980896, + "learning_rate": 2.0331e-05, + "loss": 0.0471, + "step": 6780 + }, + { + "epoch": 5.727195945945946, + "grad_norm": 0.5800846219062805, + "learning_rate": 2.0334e-05, + "loss": 0.0538, + "step": 6781 + }, + { + "epoch": 5.72804054054054, + "grad_norm": 0.4306369423866272, + "learning_rate": 2.0337e-05, + "loss": 0.0612, + "step": 6782 + }, + { + "epoch": 5.728885135135135, + "grad_norm": 0.5955209732055664, + "learning_rate": 2.0340000000000002e-05, + "loss": 0.0524, + "step": 6783 + }, + { + "epoch": 5.72972972972973, + "grad_norm": 0.45639216899871826, + "learning_rate": 2.0343e-05, + "loss": 0.0528, + "step": 6784 + }, + { + "epoch": 5.730574324324325, + "grad_norm": 0.3957684338092804, + "learning_rate": 2.0346e-05, + "loss": 0.0359, + "step": 6785 + }, + { + "epoch": 5.731418918918919, + "grad_norm": 0.5030094385147095, + "learning_rate": 2.0349e-05, + "loss": 0.0461, + "step": 6786 + }, + { + "epoch": 5.732263513513513, + "grad_norm": 0.5363333821296692, + "learning_rate": 2.0352e-05, + "loss": 0.0645, + "step": 6787 + }, + { + "epoch": 5.733108108108108, + "grad_norm": 0.4241180419921875, + "learning_rate": 2.0355e-05, + "loss": 0.0395, + "step": 6788 + }, + { + "epoch": 5.733952702702703, + "grad_norm": 0.48685744404792786, + "learning_rate": 2.0358e-05, + "loss": 0.0385, + "step": 6789 + }, + { + "epoch": 5.734797297297297, + "grad_norm": 0.32195234298706055, + "learning_rate": 2.0361e-05, + "loss": 0.0299, + "step": 6790 + }, + { + "epoch": 5.735641891891892, + "grad_norm": 0.48270153999328613, + "learning_rate": 2.0364e-05, + "loss": 0.0497, + "step": 6791 + }, + { + "epoch": 5.736486486486487, + "grad_norm": 0.5781535506248474, + "learning_rate": 2.0367e-05, + "loss": 0.0598, + "step": 6792 + }, + { + "epoch": 5.737331081081081, + "grad_norm": 0.4080573320388794, + "learning_rate": 2.0370000000000003e-05, + "loss": 0.0485, + "step": 6793 + }, + { + "epoch": 5.738175675675675, + "grad_norm": 0.471408486366272, + "learning_rate": 2.0373000000000003e-05, + "loss": 0.0393, + "step": 6794 + }, + { + "epoch": 5.73902027027027, + "grad_norm": 0.508826494216919, + "learning_rate": 2.0376000000000003e-05, + "loss": 0.0469, + "step": 6795 + }, + { + "epoch": 5.739864864864865, + "grad_norm": 0.546181857585907, + "learning_rate": 2.0379000000000003e-05, + "loss": 0.0504, + "step": 6796 + }, + { + "epoch": 5.74070945945946, + "grad_norm": 0.43524274230003357, + "learning_rate": 2.0382e-05, + "loss": 0.0418, + "step": 6797 + }, + { + "epoch": 5.741554054054054, + "grad_norm": 0.449017196893692, + "learning_rate": 2.0385e-05, + "loss": 0.0479, + "step": 6798 + }, + { + "epoch": 5.742398648648649, + "grad_norm": 0.6107879281044006, + "learning_rate": 2.0388e-05, + "loss": 0.0607, + "step": 6799 + }, + { + "epoch": 5.743243243243243, + "grad_norm": 0.4140191376209259, + "learning_rate": 2.0391e-05, + "loss": 0.0369, + "step": 6800 + }, + { + "epoch": 5.744087837837838, + "grad_norm": 0.7760442495346069, + "learning_rate": 2.0393999999999998e-05, + "loss": 0.0628, + "step": 6801 + }, + { + "epoch": 5.7449324324324325, + "grad_norm": 0.4265798032283783, + "learning_rate": 2.0396999999999998e-05, + "loss": 0.0471, + "step": 6802 + }, + { + "epoch": 5.745777027027027, + "grad_norm": 0.40531763434410095, + "learning_rate": 2.04e-05, + "loss": 0.0389, + "step": 6803 + }, + { + "epoch": 5.746621621621622, + "grad_norm": 0.4033690392971039, + "learning_rate": 2.0403e-05, + "loss": 0.0367, + "step": 6804 + }, + { + "epoch": 5.747466216216216, + "grad_norm": 0.5742476582527161, + "learning_rate": 2.0406e-05, + "loss": 0.0554, + "step": 6805 + }, + { + "epoch": 5.7483108108108105, + "grad_norm": 0.3511142432689667, + "learning_rate": 2.0409e-05, + "loss": 0.0347, + "step": 6806 + }, + { + "epoch": 5.749155405405405, + "grad_norm": 0.6328772306442261, + "learning_rate": 2.0412e-05, + "loss": 0.0423, + "step": 6807 + }, + { + "epoch": 5.75, + "grad_norm": 0.6597820520401001, + "learning_rate": 2.0415e-05, + "loss": 0.0514, + "step": 6808 + }, + { + "epoch": 5.750844594594595, + "grad_norm": 0.4379272162914276, + "learning_rate": 2.0418e-05, + "loss": 0.0304, + "step": 6809 + }, + { + "epoch": 5.7516891891891895, + "grad_norm": 0.9123837947845459, + "learning_rate": 2.0421e-05, + "loss": 0.0637, + "step": 6810 + }, + { + "epoch": 5.752533783783784, + "grad_norm": 0.8735905289649963, + "learning_rate": 2.0424e-05, + "loss": 0.0422, + "step": 6811 + }, + { + "epoch": 5.753378378378378, + "grad_norm": 0.5995604395866394, + "learning_rate": 2.0427e-05, + "loss": 0.0456, + "step": 6812 + }, + { + "epoch": 5.754222972972973, + "grad_norm": 0.44039782881736755, + "learning_rate": 2.0430000000000003e-05, + "loss": 0.0447, + "step": 6813 + }, + { + "epoch": 5.7550675675675675, + "grad_norm": 1.2391546964645386, + "learning_rate": 2.0433000000000002e-05, + "loss": 0.0597, + "step": 6814 + }, + { + "epoch": 5.755912162162162, + "grad_norm": 0.7831755876541138, + "learning_rate": 2.0436000000000002e-05, + "loss": 0.0619, + "step": 6815 + }, + { + "epoch": 5.756756756756757, + "grad_norm": 0.6239975690841675, + "learning_rate": 2.0439000000000002e-05, + "loss": 0.0591, + "step": 6816 + }, + { + "epoch": 5.757601351351351, + "grad_norm": 0.5598180294036865, + "learning_rate": 2.0442000000000002e-05, + "loss": 0.0448, + "step": 6817 + }, + { + "epoch": 5.758445945945946, + "grad_norm": 0.8174459338188171, + "learning_rate": 2.0445e-05, + "loss": 0.094, + "step": 6818 + }, + { + "epoch": 5.75929054054054, + "grad_norm": 0.7759846448898315, + "learning_rate": 2.0448e-05, + "loss": 0.0827, + "step": 6819 + }, + { + "epoch": 5.760135135135135, + "grad_norm": 1.0767951011657715, + "learning_rate": 2.0451e-05, + "loss": 0.1007, + "step": 6820 + }, + { + "epoch": 5.76097972972973, + "grad_norm": 1.2259929180145264, + "learning_rate": 2.0454e-05, + "loss": 0.3828, + "step": 6821 + }, + { + "epoch": 5.761824324324325, + "grad_norm": 0.831653892993927, + "learning_rate": 2.0456999999999997e-05, + "loss": 0.1709, + "step": 6822 + }, + { + "epoch": 5.762668918918919, + "grad_norm": 0.4902372658252716, + "learning_rate": 2.046e-05, + "loss": 0.0971, + "step": 6823 + }, + { + "epoch": 5.763513513513513, + "grad_norm": 0.6038952469825745, + "learning_rate": 2.0463e-05, + "loss": 0.0966, + "step": 6824 + }, + { + "epoch": 5.764358108108108, + "grad_norm": 0.6636776924133301, + "learning_rate": 2.0466e-05, + "loss": 0.078, + "step": 6825 + }, + { + "epoch": 5.765202702702703, + "grad_norm": 0.3051656186580658, + "learning_rate": 2.0469e-05, + "loss": 0.0413, + "step": 6826 + }, + { + "epoch": 5.766047297297297, + "grad_norm": 0.43865567445755005, + "learning_rate": 2.0472e-05, + "loss": 0.0412, + "step": 6827 + }, + { + "epoch": 5.766891891891892, + "grad_norm": 0.44363752007484436, + "learning_rate": 2.0475e-05, + "loss": 0.0388, + "step": 6828 + }, + { + "epoch": 5.767736486486487, + "grad_norm": 0.4226745665073395, + "learning_rate": 2.0478e-05, + "loss": 0.0419, + "step": 6829 + }, + { + "epoch": 5.768581081081081, + "grad_norm": 0.5790736675262451, + "learning_rate": 2.0481e-05, + "loss": 0.0456, + "step": 6830 + }, + { + "epoch": 5.769425675675675, + "grad_norm": 0.5321681499481201, + "learning_rate": 2.0484e-05, + "loss": 0.0572, + "step": 6831 + }, + { + "epoch": 5.77027027027027, + "grad_norm": 0.3357318341732025, + "learning_rate": 2.0487e-05, + "loss": 0.0289, + "step": 6832 + }, + { + "epoch": 5.771114864864865, + "grad_norm": 0.4817851185798645, + "learning_rate": 2.0490000000000002e-05, + "loss": 0.0327, + "step": 6833 + }, + { + "epoch": 5.77195945945946, + "grad_norm": 0.5546959638595581, + "learning_rate": 2.0493000000000002e-05, + "loss": 0.066, + "step": 6834 + }, + { + "epoch": 5.772804054054054, + "grad_norm": 0.3620668947696686, + "learning_rate": 2.0496e-05, + "loss": 0.0431, + "step": 6835 + }, + { + "epoch": 5.773648648648649, + "grad_norm": 0.37393519282341003, + "learning_rate": 2.0499e-05, + "loss": 0.0423, + "step": 6836 + }, + { + "epoch": 5.774493243243243, + "grad_norm": 0.4386485517024994, + "learning_rate": 2.0502e-05, + "loss": 0.0423, + "step": 6837 + }, + { + "epoch": 5.775337837837838, + "grad_norm": 0.3736984133720398, + "learning_rate": 2.0505e-05, + "loss": 0.0398, + "step": 6838 + }, + { + "epoch": 5.7761824324324325, + "grad_norm": 0.680182158946991, + "learning_rate": 2.0508e-05, + "loss": 0.0614, + "step": 6839 + }, + { + "epoch": 5.777027027027027, + "grad_norm": 0.42977333068847656, + "learning_rate": 2.0511e-05, + "loss": 0.0471, + "step": 6840 + }, + { + "epoch": 5.777871621621622, + "grad_norm": 0.412321001291275, + "learning_rate": 2.0514e-05, + "loss": 0.0443, + "step": 6841 + }, + { + "epoch": 5.778716216216216, + "grad_norm": 0.6640656590461731, + "learning_rate": 2.0517e-05, + "loss": 0.0469, + "step": 6842 + }, + { + "epoch": 5.7795608108108105, + "grad_norm": 0.3238629102706909, + "learning_rate": 2.0520000000000003e-05, + "loss": 0.0327, + "step": 6843 + }, + { + "epoch": 5.780405405405405, + "grad_norm": 0.363751620054245, + "learning_rate": 2.0523000000000003e-05, + "loss": 0.0377, + "step": 6844 + }, + { + "epoch": 5.78125, + "grad_norm": 0.4406898021697998, + "learning_rate": 2.0526000000000003e-05, + "loss": 0.0464, + "step": 6845 + }, + { + "epoch": 5.782094594594595, + "grad_norm": 0.5216416120529175, + "learning_rate": 2.0529e-05, + "loss": 0.0635, + "step": 6846 + }, + { + "epoch": 5.7829391891891895, + "grad_norm": 0.6113055944442749, + "learning_rate": 2.0532e-05, + "loss": 0.0561, + "step": 6847 + }, + { + "epoch": 5.783783783783784, + "grad_norm": 0.38943207263946533, + "learning_rate": 2.0535e-05, + "loss": 0.0358, + "step": 6848 + }, + { + "epoch": 5.784628378378378, + "grad_norm": 0.4765883982181549, + "learning_rate": 2.0538e-05, + "loss": 0.0548, + "step": 6849 + }, + { + "epoch": 5.785472972972973, + "grad_norm": 0.3783988654613495, + "learning_rate": 2.0541e-05, + "loss": 0.0452, + "step": 6850 + }, + { + "epoch": 5.7863175675675675, + "grad_norm": 0.4073716700077057, + "learning_rate": 2.0544e-05, + "loss": 0.0388, + "step": 6851 + }, + { + "epoch": 5.787162162162162, + "grad_norm": 0.4291004538536072, + "learning_rate": 2.0546999999999998e-05, + "loss": 0.0418, + "step": 6852 + }, + { + "epoch": 5.788006756756757, + "grad_norm": 0.6028547883033752, + "learning_rate": 2.055e-05, + "loss": 0.0482, + "step": 6853 + }, + { + "epoch": 5.788851351351351, + "grad_norm": 0.5950875282287598, + "learning_rate": 2.0553e-05, + "loss": 0.0542, + "step": 6854 + }, + { + "epoch": 5.789695945945946, + "grad_norm": 0.48752927780151367, + "learning_rate": 2.0556e-05, + "loss": 0.0327, + "step": 6855 + }, + { + "epoch": 5.79054054054054, + "grad_norm": 0.4852105975151062, + "learning_rate": 2.0559e-05, + "loss": 0.0511, + "step": 6856 + }, + { + "epoch": 5.791385135135135, + "grad_norm": 0.4256399869918823, + "learning_rate": 2.0562e-05, + "loss": 0.0465, + "step": 6857 + }, + { + "epoch": 5.79222972972973, + "grad_norm": 0.6160696148872375, + "learning_rate": 2.0565e-05, + "loss": 0.0549, + "step": 6858 + }, + { + "epoch": 5.793074324324325, + "grad_norm": 0.746242880821228, + "learning_rate": 2.0568e-05, + "loss": 0.0497, + "step": 6859 + }, + { + "epoch": 5.793918918918919, + "grad_norm": 0.5082012414932251, + "learning_rate": 2.0571e-05, + "loss": 0.0624, + "step": 6860 + }, + { + "epoch": 5.794763513513513, + "grad_norm": 0.4672819674015045, + "learning_rate": 2.0574e-05, + "loss": 0.051, + "step": 6861 + }, + { + "epoch": 5.795608108108108, + "grad_norm": 0.7485505938529968, + "learning_rate": 2.0577e-05, + "loss": 0.0568, + "step": 6862 + }, + { + "epoch": 5.796452702702703, + "grad_norm": 0.5749267339706421, + "learning_rate": 2.0580000000000003e-05, + "loss": 0.0512, + "step": 6863 + }, + { + "epoch": 5.797297297297297, + "grad_norm": 0.84360271692276, + "learning_rate": 2.0583000000000003e-05, + "loss": 0.1282, + "step": 6864 + }, + { + "epoch": 5.798141891891892, + "grad_norm": 0.7527876496315002, + "learning_rate": 2.0586000000000002e-05, + "loss": 0.0538, + "step": 6865 + }, + { + "epoch": 5.798986486486487, + "grad_norm": 1.0037578344345093, + "learning_rate": 2.0589000000000002e-05, + "loss": 0.0972, + "step": 6866 + }, + { + "epoch": 5.799831081081081, + "grad_norm": 0.609735906124115, + "learning_rate": 2.0592000000000002e-05, + "loss": 0.0629, + "step": 6867 + }, + { + "epoch": 5.800675675675675, + "grad_norm": 0.43227899074554443, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0487, + "step": 6868 + }, + { + "epoch": 5.80152027027027, + "grad_norm": 0.8076423406600952, + "learning_rate": 2.0598e-05, + "loss": 0.0901, + "step": 6869 + }, + { + "epoch": 5.802364864864865, + "grad_norm": 1.1955314874649048, + "learning_rate": 2.0601e-05, + "loss": 0.0958, + "step": 6870 + }, + { + "epoch": 5.80320945945946, + "grad_norm": 1.0822582244873047, + "learning_rate": 2.0603999999999998e-05, + "loss": 0.3765, + "step": 6871 + }, + { + "epoch": 5.804054054054054, + "grad_norm": 0.6380276679992676, + "learning_rate": 2.0606999999999998e-05, + "loss": 0.2086, + "step": 6872 + }, + { + "epoch": 5.804898648648649, + "grad_norm": 0.46940696239471436, + "learning_rate": 2.061e-05, + "loss": 0.0759, + "step": 6873 + }, + { + "epoch": 5.805743243243243, + "grad_norm": 0.5695683360099792, + "learning_rate": 2.0613e-05, + "loss": 0.0643, + "step": 6874 + }, + { + "epoch": 5.806587837837838, + "grad_norm": 0.7956430315971375, + "learning_rate": 2.0616e-05, + "loss": 0.0717, + "step": 6875 + }, + { + "epoch": 5.8074324324324325, + "grad_norm": 0.595691978931427, + "learning_rate": 2.0619e-05, + "loss": 0.0719, + "step": 6876 + }, + { + "epoch": 5.808277027027027, + "grad_norm": 0.5520039796829224, + "learning_rate": 2.0622e-05, + "loss": 0.0853, + "step": 6877 + }, + { + "epoch": 5.809121621621622, + "grad_norm": 0.4157843291759491, + "learning_rate": 2.0625e-05, + "loss": 0.0499, + "step": 6878 + }, + { + "epoch": 5.809966216216216, + "grad_norm": 0.6219871640205383, + "learning_rate": 2.0628e-05, + "loss": 0.045, + "step": 6879 + }, + { + "epoch": 5.8108108108108105, + "grad_norm": 0.4190238416194916, + "learning_rate": 2.0631e-05, + "loss": 0.0529, + "step": 6880 + }, + { + "epoch": 5.811655405405405, + "grad_norm": 0.5159740447998047, + "learning_rate": 2.0634e-05, + "loss": 0.0358, + "step": 6881 + }, + { + "epoch": 5.8125, + "grad_norm": 0.43261781334877014, + "learning_rate": 2.0637e-05, + "loss": 0.0545, + "step": 6882 + }, + { + "epoch": 5.813344594594595, + "grad_norm": 0.4392542541027069, + "learning_rate": 2.064e-05, + "loss": 0.0457, + "step": 6883 + }, + { + "epoch": 5.8141891891891895, + "grad_norm": 0.3316205143928528, + "learning_rate": 2.0643000000000002e-05, + "loss": 0.0321, + "step": 6884 + }, + { + "epoch": 5.815033783783784, + "grad_norm": 0.8024634122848511, + "learning_rate": 2.0646000000000002e-05, + "loss": 0.0456, + "step": 6885 + }, + { + "epoch": 5.815878378378378, + "grad_norm": 0.8329036831855774, + "learning_rate": 2.0649e-05, + "loss": 0.0485, + "step": 6886 + }, + { + "epoch": 5.816722972972973, + "grad_norm": 0.4963737726211548, + "learning_rate": 2.0652e-05, + "loss": 0.0501, + "step": 6887 + }, + { + "epoch": 5.8175675675675675, + "grad_norm": 0.4653836190700531, + "learning_rate": 2.0655e-05, + "loss": 0.0467, + "step": 6888 + }, + { + "epoch": 5.818412162162162, + "grad_norm": 0.4359171986579895, + "learning_rate": 2.0658e-05, + "loss": 0.044, + "step": 6889 + }, + { + "epoch": 5.819256756756757, + "grad_norm": 0.4166301488876343, + "learning_rate": 2.0661e-05, + "loss": 0.0354, + "step": 6890 + }, + { + "epoch": 5.820101351351351, + "grad_norm": 0.5780697464942932, + "learning_rate": 2.0664e-05, + "loss": 0.0521, + "step": 6891 + }, + { + "epoch": 5.820945945945946, + "grad_norm": 0.5544420480728149, + "learning_rate": 2.0667e-05, + "loss": 0.0454, + "step": 6892 + }, + { + "epoch": 5.82179054054054, + "grad_norm": 0.5617170929908752, + "learning_rate": 2.067e-05, + "loss": 0.0332, + "step": 6893 + }, + { + "epoch": 5.822635135135135, + "grad_norm": 0.4044084846973419, + "learning_rate": 2.0673000000000003e-05, + "loss": 0.0364, + "step": 6894 + }, + { + "epoch": 5.82347972972973, + "grad_norm": 0.45917806029319763, + "learning_rate": 2.0676e-05, + "loss": 0.0505, + "step": 6895 + }, + { + "epoch": 5.824324324324325, + "grad_norm": 0.5592328310012817, + "learning_rate": 2.0679e-05, + "loss": 0.0464, + "step": 6896 + }, + { + "epoch": 5.825168918918919, + "grad_norm": 0.540211021900177, + "learning_rate": 2.0682e-05, + "loss": 0.0574, + "step": 6897 + }, + { + "epoch": 5.826013513513513, + "grad_norm": 0.4369381070137024, + "learning_rate": 2.0685e-05, + "loss": 0.0509, + "step": 6898 + }, + { + "epoch": 5.826858108108108, + "grad_norm": 0.45377689599990845, + "learning_rate": 2.0688e-05, + "loss": 0.0436, + "step": 6899 + }, + { + "epoch": 5.827702702702703, + "grad_norm": 0.5303455591201782, + "learning_rate": 2.0691e-05, + "loss": 0.0539, + "step": 6900 + }, + { + "epoch": 5.828547297297297, + "grad_norm": 0.5170179605484009, + "learning_rate": 2.0694e-05, + "loss": 0.0395, + "step": 6901 + }, + { + "epoch": 5.829391891891892, + "grad_norm": 0.5890648365020752, + "learning_rate": 2.0697e-05, + "loss": 0.059, + "step": 6902 + }, + { + "epoch": 5.830236486486487, + "grad_norm": 0.4257182776927948, + "learning_rate": 2.07e-05, + "loss": 0.025, + "step": 6903 + }, + { + "epoch": 5.831081081081081, + "grad_norm": 0.560196042060852, + "learning_rate": 2.0703e-05, + "loss": 0.0406, + "step": 6904 + }, + { + "epoch": 5.831925675675675, + "grad_norm": 0.39471709728240967, + "learning_rate": 2.0706e-05, + "loss": 0.0429, + "step": 6905 + }, + { + "epoch": 5.83277027027027, + "grad_norm": 0.5693705081939697, + "learning_rate": 2.0709e-05, + "loss": 0.0557, + "step": 6906 + }, + { + "epoch": 5.833614864864865, + "grad_norm": 0.7745842337608337, + "learning_rate": 2.0712e-05, + "loss": 0.0358, + "step": 6907 + }, + { + "epoch": 5.83445945945946, + "grad_norm": 0.5925846695899963, + "learning_rate": 2.0715e-05, + "loss": 0.0534, + "step": 6908 + }, + { + "epoch": 5.835304054054054, + "grad_norm": 0.4744938910007477, + "learning_rate": 2.0718e-05, + "loss": 0.0412, + "step": 6909 + }, + { + "epoch": 5.836148648648649, + "grad_norm": 0.4460128843784332, + "learning_rate": 2.0721e-05, + "loss": 0.0339, + "step": 6910 + }, + { + "epoch": 5.836993243243243, + "grad_norm": 0.5588361620903015, + "learning_rate": 2.0724e-05, + "loss": 0.0433, + "step": 6911 + }, + { + "epoch": 5.837837837837838, + "grad_norm": 0.4617765247821808, + "learning_rate": 2.0727e-05, + "loss": 0.0469, + "step": 6912 + }, + { + "epoch": 5.8386824324324325, + "grad_norm": 0.47304990887641907, + "learning_rate": 2.073e-05, + "loss": 0.0461, + "step": 6913 + }, + { + "epoch": 5.839527027027027, + "grad_norm": 1.1786717176437378, + "learning_rate": 2.0733000000000003e-05, + "loss": 0.0704, + "step": 6914 + }, + { + "epoch": 5.840371621621622, + "grad_norm": 1.0105838775634766, + "learning_rate": 2.0736000000000003e-05, + "loss": 0.0474, + "step": 6915 + }, + { + "epoch": 5.841216216216216, + "grad_norm": 0.5675122141838074, + "learning_rate": 2.0739000000000003e-05, + "loss": 0.0445, + "step": 6916 + }, + { + "epoch": 5.8420608108108105, + "grad_norm": 1.59304940700531, + "learning_rate": 2.0742000000000002e-05, + "loss": 0.0621, + "step": 6917 + }, + { + "epoch": 5.842905405405405, + "grad_norm": 0.5923209190368652, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0522, + "step": 6918 + }, + { + "epoch": 5.84375, + "grad_norm": 0.6656051874160767, + "learning_rate": 2.0748000000000002e-05, + "loss": 0.0624, + "step": 6919 + }, + { + "epoch": 5.844594594594595, + "grad_norm": 1.1822679042816162, + "learning_rate": 2.0751e-05, + "loss": 0.094, + "step": 6920 + }, + { + "epoch": 5.8454391891891895, + "grad_norm": 1.6211254596710205, + "learning_rate": 2.0753999999999998e-05, + "loss": 0.3602, + "step": 6921 + }, + { + "epoch": 5.846283783783784, + "grad_norm": 0.996417760848999, + "learning_rate": 2.0756999999999998e-05, + "loss": 0.2636, + "step": 6922 + }, + { + "epoch": 5.847128378378378, + "grad_norm": 1.134889006614685, + "learning_rate": 2.0759999999999998e-05, + "loss": 0.1082, + "step": 6923 + }, + { + "epoch": 5.847972972972973, + "grad_norm": 0.7200987935066223, + "learning_rate": 2.0763e-05, + "loss": 0.0708, + "step": 6924 + }, + { + "epoch": 5.8488175675675675, + "grad_norm": 0.6157593131065369, + "learning_rate": 2.0766e-05, + "loss": 0.0708, + "step": 6925 + }, + { + "epoch": 5.849662162162162, + "grad_norm": 0.5320636630058289, + "learning_rate": 2.0769e-05, + "loss": 0.0663, + "step": 6926 + }, + { + "epoch": 5.850506756756757, + "grad_norm": 0.878194272518158, + "learning_rate": 2.0772e-05, + "loss": 0.0689, + "step": 6927 + }, + { + "epoch": 5.851351351351351, + "grad_norm": 0.4391617178916931, + "learning_rate": 2.0775e-05, + "loss": 0.0559, + "step": 6928 + }, + { + "epoch": 5.852195945945946, + "grad_norm": 0.4946390688419342, + "learning_rate": 2.0778e-05, + "loss": 0.0484, + "step": 6929 + }, + { + "epoch": 5.85304054054054, + "grad_norm": 0.4612930417060852, + "learning_rate": 2.0781e-05, + "loss": 0.0624, + "step": 6930 + }, + { + "epoch": 5.853885135135135, + "grad_norm": 0.637730062007904, + "learning_rate": 2.0784e-05, + "loss": 0.0692, + "step": 6931 + }, + { + "epoch": 5.85472972972973, + "grad_norm": 0.45965397357940674, + "learning_rate": 2.0787e-05, + "loss": 0.0433, + "step": 6932 + }, + { + "epoch": 5.855574324324325, + "grad_norm": 0.5560342669487, + "learning_rate": 2.079e-05, + "loss": 0.0392, + "step": 6933 + }, + { + "epoch": 5.856418918918919, + "grad_norm": 0.679189145565033, + "learning_rate": 2.0793000000000002e-05, + "loss": 0.0565, + "step": 6934 + }, + { + "epoch": 5.857263513513513, + "grad_norm": 0.4231336712837219, + "learning_rate": 2.0796000000000002e-05, + "loss": 0.0445, + "step": 6935 + }, + { + "epoch": 5.858108108108108, + "grad_norm": 0.4289316236972809, + "learning_rate": 2.0799000000000002e-05, + "loss": 0.04, + "step": 6936 + }, + { + "epoch": 5.858952702702703, + "grad_norm": 0.46417251229286194, + "learning_rate": 2.0802000000000002e-05, + "loss": 0.0391, + "step": 6937 + }, + { + "epoch": 5.859797297297297, + "grad_norm": 0.43214523792266846, + "learning_rate": 2.0805e-05, + "loss": 0.0379, + "step": 6938 + }, + { + "epoch": 5.860641891891892, + "grad_norm": 0.4649614989757538, + "learning_rate": 2.0808e-05, + "loss": 0.048, + "step": 6939 + }, + { + "epoch": 5.861486486486487, + "grad_norm": 0.682086169719696, + "learning_rate": 2.0811e-05, + "loss": 0.0382, + "step": 6940 + }, + { + "epoch": 5.862331081081081, + "grad_norm": 0.46759921312332153, + "learning_rate": 2.0814e-05, + "loss": 0.0577, + "step": 6941 + }, + { + "epoch": 5.863175675675675, + "grad_norm": 0.5790491104125977, + "learning_rate": 2.0817e-05, + "loss": 0.0601, + "step": 6942 + }, + { + "epoch": 5.86402027027027, + "grad_norm": 0.7860535979270935, + "learning_rate": 2.082e-05, + "loss": 0.0627, + "step": 6943 + }, + { + "epoch": 5.864864864864865, + "grad_norm": 0.429055780172348, + "learning_rate": 2.0823e-05, + "loss": 0.0445, + "step": 6944 + }, + { + "epoch": 5.86570945945946, + "grad_norm": 0.3800155818462372, + "learning_rate": 2.0826e-05, + "loss": 0.0323, + "step": 6945 + }, + { + "epoch": 5.866554054054054, + "grad_norm": 0.5367060303688049, + "learning_rate": 2.0829e-05, + "loss": 0.0532, + "step": 6946 + }, + { + "epoch": 5.867398648648649, + "grad_norm": 0.5389602184295654, + "learning_rate": 2.0832e-05, + "loss": 0.0775, + "step": 6947 + }, + { + "epoch": 5.868243243243243, + "grad_norm": 0.4197222888469696, + "learning_rate": 2.0835e-05, + "loss": 0.0582, + "step": 6948 + }, + { + "epoch": 5.869087837837838, + "grad_norm": 0.48012956976890564, + "learning_rate": 2.0838e-05, + "loss": 0.0525, + "step": 6949 + }, + { + "epoch": 5.8699324324324325, + "grad_norm": 0.4155469536781311, + "learning_rate": 2.0841e-05, + "loss": 0.0407, + "step": 6950 + }, + { + "epoch": 5.870777027027027, + "grad_norm": 0.3267977237701416, + "learning_rate": 2.0844e-05, + "loss": 0.0414, + "step": 6951 + }, + { + "epoch": 5.871621621621622, + "grad_norm": 0.3613460958003998, + "learning_rate": 2.0847e-05, + "loss": 0.0418, + "step": 6952 + }, + { + "epoch": 5.872466216216216, + "grad_norm": 0.4452787935733795, + "learning_rate": 2.085e-05, + "loss": 0.0514, + "step": 6953 + }, + { + "epoch": 5.8733108108108105, + "grad_norm": 0.5786432027816772, + "learning_rate": 2.0853000000000002e-05, + "loss": 0.0595, + "step": 6954 + }, + { + "epoch": 5.874155405405405, + "grad_norm": 0.5528143644332886, + "learning_rate": 2.0856e-05, + "loss": 0.0606, + "step": 6955 + }, + { + "epoch": 5.875, + "grad_norm": 0.5385833978652954, + "learning_rate": 2.0859e-05, + "loss": 0.0392, + "step": 6956 + }, + { + "epoch": 5.875844594594595, + "grad_norm": 0.6575754284858704, + "learning_rate": 2.0862e-05, + "loss": 0.0646, + "step": 6957 + }, + { + "epoch": 5.8766891891891895, + "grad_norm": 0.6162706017494202, + "learning_rate": 2.0865e-05, + "loss": 0.0373, + "step": 6958 + }, + { + "epoch": 5.877533783783784, + "grad_norm": 0.5437148213386536, + "learning_rate": 2.0868e-05, + "loss": 0.0591, + "step": 6959 + }, + { + "epoch": 5.878378378378378, + "grad_norm": 0.5530276298522949, + "learning_rate": 2.0871e-05, + "loss": 0.0676, + "step": 6960 + }, + { + "epoch": 5.879222972972973, + "grad_norm": 0.5368350148200989, + "learning_rate": 2.0874e-05, + "loss": 0.0527, + "step": 6961 + }, + { + "epoch": 5.8800675675675675, + "grad_norm": 0.8895555734634399, + "learning_rate": 2.0877e-05, + "loss": 0.0433, + "step": 6962 + }, + { + "epoch": 5.880912162162162, + "grad_norm": 1.0263879299163818, + "learning_rate": 2.088e-05, + "loss": 0.0596, + "step": 6963 + }, + { + "epoch": 5.881756756756757, + "grad_norm": 0.6584741473197937, + "learning_rate": 2.0883000000000003e-05, + "loss": 0.0624, + "step": 6964 + }, + { + "epoch": 5.882601351351351, + "grad_norm": 0.6083936095237732, + "learning_rate": 2.0886000000000003e-05, + "loss": 0.0508, + "step": 6965 + }, + { + "epoch": 5.883445945945946, + "grad_norm": 0.7627350687980652, + "learning_rate": 2.0889000000000003e-05, + "loss": 0.081, + "step": 6966 + }, + { + "epoch": 5.88429054054054, + "grad_norm": 0.8313725590705872, + "learning_rate": 2.0892000000000003e-05, + "loss": 0.0716, + "step": 6967 + }, + { + "epoch": 5.885135135135135, + "grad_norm": 0.6311262249946594, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0552, + "step": 6968 + }, + { + "epoch": 5.88597972972973, + "grad_norm": 0.740914523601532, + "learning_rate": 2.0898e-05, + "loss": 0.0799, + "step": 6969 + }, + { + "epoch": 5.886824324324325, + "grad_norm": 1.0052469968795776, + "learning_rate": 2.0901e-05, + "loss": 0.0725, + "step": 6970 + }, + { + "epoch": 5.887668918918919, + "grad_norm": 1.1651674509048462, + "learning_rate": 2.0904e-05, + "loss": 0.3162, + "step": 6971 + }, + { + "epoch": 5.888513513513513, + "grad_norm": 0.8683027029037476, + "learning_rate": 2.0906999999999998e-05, + "loss": 0.2144, + "step": 6972 + }, + { + "epoch": 5.889358108108108, + "grad_norm": 0.6723748445510864, + "learning_rate": 2.0909999999999998e-05, + "loss": 0.1277, + "step": 6973 + }, + { + "epoch": 5.890202702702703, + "grad_norm": 0.6278271675109863, + "learning_rate": 2.0913e-05, + "loss": 0.0859, + "step": 6974 + }, + { + "epoch": 5.891047297297297, + "grad_norm": 0.5073860287666321, + "learning_rate": 2.0916e-05, + "loss": 0.0979, + "step": 6975 + }, + { + "epoch": 5.891891891891892, + "grad_norm": 0.424366295337677, + "learning_rate": 2.0919e-05, + "loss": 0.0428, + "step": 6976 + }, + { + "epoch": 5.892736486486487, + "grad_norm": 0.5691332817077637, + "learning_rate": 2.0922e-05, + "loss": 0.0521, + "step": 6977 + }, + { + "epoch": 5.893581081081081, + "grad_norm": 0.49481630325317383, + "learning_rate": 2.0925e-05, + "loss": 0.0687, + "step": 6978 + }, + { + "epoch": 5.894425675675675, + "grad_norm": 0.40479981899261475, + "learning_rate": 2.0928e-05, + "loss": 0.0322, + "step": 6979 + }, + { + "epoch": 5.89527027027027, + "grad_norm": 0.5070905685424805, + "learning_rate": 2.0931e-05, + "loss": 0.052, + "step": 6980 + }, + { + "epoch": 5.896114864864865, + "grad_norm": 0.393593430519104, + "learning_rate": 2.0934e-05, + "loss": 0.0331, + "step": 6981 + }, + { + "epoch": 5.89695945945946, + "grad_norm": 0.3499518930912018, + "learning_rate": 2.0937e-05, + "loss": 0.0431, + "step": 6982 + }, + { + "epoch": 5.897804054054054, + "grad_norm": 0.5950884819030762, + "learning_rate": 2.094e-05, + "loss": 0.0532, + "step": 6983 + }, + { + "epoch": 5.898648648648649, + "grad_norm": 0.3273756206035614, + "learning_rate": 2.0943000000000003e-05, + "loss": 0.0374, + "step": 6984 + }, + { + "epoch": 5.899493243243243, + "grad_norm": 0.43758490681648254, + "learning_rate": 2.0946000000000002e-05, + "loss": 0.0444, + "step": 6985 + }, + { + "epoch": 5.900337837837838, + "grad_norm": 0.6232397556304932, + "learning_rate": 2.0949000000000002e-05, + "loss": 0.0916, + "step": 6986 + }, + { + "epoch": 5.9011824324324325, + "grad_norm": 0.4785841405391693, + "learning_rate": 2.0952000000000002e-05, + "loss": 0.0412, + "step": 6987 + }, + { + "epoch": 5.902027027027027, + "grad_norm": 0.5496357083320618, + "learning_rate": 2.0955000000000002e-05, + "loss": 0.0584, + "step": 6988 + }, + { + "epoch": 5.902871621621622, + "grad_norm": 0.7477105855941772, + "learning_rate": 2.0958e-05, + "loss": 0.0533, + "step": 6989 + }, + { + "epoch": 5.903716216216216, + "grad_norm": 0.546600341796875, + "learning_rate": 2.0961e-05, + "loss": 0.0457, + "step": 6990 + }, + { + "epoch": 5.9045608108108105, + "grad_norm": 0.4244646430015564, + "learning_rate": 2.0964e-05, + "loss": 0.0398, + "step": 6991 + }, + { + "epoch": 5.905405405405405, + "grad_norm": 0.4733007550239563, + "learning_rate": 2.0967e-05, + "loss": 0.053, + "step": 6992 + }, + { + "epoch": 5.90625, + "grad_norm": 0.4389418959617615, + "learning_rate": 2.097e-05, + "loss": 0.0408, + "step": 6993 + }, + { + "epoch": 5.907094594594595, + "grad_norm": 0.43889695405960083, + "learning_rate": 2.0973e-05, + "loss": 0.0445, + "step": 6994 + }, + { + "epoch": 5.9079391891891895, + "grad_norm": 0.5528516173362732, + "learning_rate": 2.0976e-05, + "loss": 0.0452, + "step": 6995 + }, + { + "epoch": 5.908783783783784, + "grad_norm": 0.48333263397216797, + "learning_rate": 2.0979e-05, + "loss": 0.0591, + "step": 6996 + }, + { + "epoch": 5.909628378378378, + "grad_norm": 0.38480299711227417, + "learning_rate": 2.0982e-05, + "loss": 0.0482, + "step": 6997 + }, + { + "epoch": 5.910472972972973, + "grad_norm": 0.44184452295303345, + "learning_rate": 2.0985e-05, + "loss": 0.0489, + "step": 6998 + }, + { + "epoch": 5.9113175675675675, + "grad_norm": 0.6922483444213867, + "learning_rate": 2.0988e-05, + "loss": 0.0508, + "step": 6999 + }, + { + "epoch": 5.912162162162162, + "grad_norm": 0.6981926560401917, + "learning_rate": 2.0991e-05, + "loss": 0.0444, + "step": 7000 + }, + { + "epoch": 5.912162162162162, + "eval_cer": 0.09806441249698819, + "eval_loss": 0.2983246147632599, + "eval_runtime": 8.6059, + "eval_samples_per_second": 11.736, + "eval_steps_per_second": 0.465, + "eval_wer": 0.3082010582010582, + "step": 7000 + }, + { + "epoch": 5.913006756756757, + "grad_norm": 0.548762857913971, + "learning_rate": 2.0994e-05, + "loss": 0.0404, + "step": 7001 + }, + { + "epoch": 5.913851351351351, + "grad_norm": 0.5087748765945435, + "learning_rate": 2.0997e-05, + "loss": 0.0519, + "step": 7002 + }, + { + "epoch": 5.914695945945946, + "grad_norm": 0.6638221144676208, + "learning_rate": 2.1e-05, + "loss": 0.0533, + "step": 7003 + }, + { + "epoch": 5.91554054054054, + "grad_norm": 0.4979211091995239, + "learning_rate": 2.1003e-05, + "loss": 0.0398, + "step": 7004 + }, + { + "epoch": 5.916385135135135, + "grad_norm": 0.45254942774772644, + "learning_rate": 2.1006000000000002e-05, + "loss": 0.037, + "step": 7005 + }, + { + "epoch": 5.91722972972973, + "grad_norm": 0.7962271571159363, + "learning_rate": 2.1009e-05, + "loss": 0.0489, + "step": 7006 + }, + { + "epoch": 5.918074324324325, + "grad_norm": 0.8281707763671875, + "learning_rate": 2.1012e-05, + "loss": 0.0422, + "step": 7007 + }, + { + "epoch": 5.918918918918919, + "grad_norm": 0.6889925003051758, + "learning_rate": 2.1015e-05, + "loss": 0.0475, + "step": 7008 + }, + { + "epoch": 5.919763513513513, + "grad_norm": 0.5159591436386108, + "learning_rate": 2.1018e-05, + "loss": 0.041, + "step": 7009 + }, + { + "epoch": 5.920608108108108, + "grad_norm": 0.3867497742176056, + "learning_rate": 2.1021e-05, + "loss": 0.0405, + "step": 7010 + }, + { + "epoch": 5.921452702702703, + "grad_norm": 0.8296692967414856, + "learning_rate": 2.1024e-05, + "loss": 0.0412, + "step": 7011 + }, + { + "epoch": 5.922297297297297, + "grad_norm": 1.1806339025497437, + "learning_rate": 2.1027e-05, + "loss": 0.0677, + "step": 7012 + }, + { + "epoch": 5.923141891891892, + "grad_norm": 0.7004194259643555, + "learning_rate": 2.103e-05, + "loss": 0.058, + "step": 7013 + }, + { + "epoch": 5.923986486486487, + "grad_norm": 0.633762538433075, + "learning_rate": 2.1033e-05, + "loss": 0.0511, + "step": 7014 + }, + { + "epoch": 5.924831081081081, + "grad_norm": 0.9700497984886169, + "learning_rate": 2.1036000000000003e-05, + "loss": 0.0686, + "step": 7015 + }, + { + "epoch": 5.925675675675675, + "grad_norm": 0.748296856880188, + "learning_rate": 2.1039000000000003e-05, + "loss": 0.0472, + "step": 7016 + }, + { + "epoch": 5.92652027027027, + "grad_norm": 0.7901691198348999, + "learning_rate": 2.1042000000000003e-05, + "loss": 0.0675, + "step": 7017 + }, + { + "epoch": 5.927364864864865, + "grad_norm": 0.7575404644012451, + "learning_rate": 2.1045e-05, + "loss": 0.0526, + "step": 7018 + }, + { + "epoch": 5.92820945945946, + "grad_norm": 1.0611921548843384, + "learning_rate": 2.1048e-05, + "loss": 0.1116, + "step": 7019 + }, + { + "epoch": 5.929054054054054, + "grad_norm": 1.2073249816894531, + "learning_rate": 2.1051e-05, + "loss": 0.0855, + "step": 7020 + }, + { + "epoch": 5.929898648648649, + "grad_norm": 1.385974645614624, + "learning_rate": 2.1054e-05, + "loss": 0.3813, + "step": 7021 + }, + { + "epoch": 5.930743243243243, + "grad_norm": 1.1161866188049316, + "learning_rate": 2.1057e-05, + "loss": 0.3264, + "step": 7022 + }, + { + "epoch": 5.931587837837838, + "grad_norm": 0.7131554484367371, + "learning_rate": 2.1059999999999998e-05, + "loss": 0.1284, + "step": 7023 + }, + { + "epoch": 5.9324324324324325, + "grad_norm": 0.8642660975456238, + "learning_rate": 2.1062999999999998e-05, + "loss": 0.103, + "step": 7024 + }, + { + "epoch": 5.933277027027027, + "grad_norm": 0.5398420691490173, + "learning_rate": 2.1066e-05, + "loss": 0.0665, + "step": 7025 + }, + { + "epoch": 5.934121621621622, + "grad_norm": 0.7162913680076599, + "learning_rate": 2.1069e-05, + "loss": 0.0629, + "step": 7026 + }, + { + "epoch": 5.934966216216216, + "grad_norm": 0.37237709760665894, + "learning_rate": 2.1072e-05, + "loss": 0.0434, + "step": 7027 + }, + { + "epoch": 5.9358108108108105, + "grad_norm": 0.4860078692436218, + "learning_rate": 2.1075e-05, + "loss": 0.0698, + "step": 7028 + }, + { + "epoch": 5.936655405405405, + "grad_norm": 0.48478835821151733, + "learning_rate": 2.1078e-05, + "loss": 0.0604, + "step": 7029 + }, + { + "epoch": 5.9375, + "grad_norm": 0.39613914489746094, + "learning_rate": 2.1081e-05, + "loss": 0.0403, + "step": 7030 + }, + { + "epoch": 5.938344594594595, + "grad_norm": 0.5843745470046997, + "learning_rate": 2.1084e-05, + "loss": 0.0584, + "step": 7031 + }, + { + "epoch": 5.9391891891891895, + "grad_norm": 0.4131008982658386, + "learning_rate": 2.1087e-05, + "loss": 0.0491, + "step": 7032 + }, + { + "epoch": 5.940033783783784, + "grad_norm": 0.5108950734138489, + "learning_rate": 2.109e-05, + "loss": 0.0562, + "step": 7033 + }, + { + "epoch": 5.940878378378378, + "grad_norm": 0.5765527486801147, + "learning_rate": 2.1093e-05, + "loss": 0.0424, + "step": 7034 + }, + { + "epoch": 5.941722972972973, + "grad_norm": 0.400917649269104, + "learning_rate": 2.1096000000000003e-05, + "loss": 0.0499, + "step": 7035 + }, + { + "epoch": 5.9425675675675675, + "grad_norm": 0.5462203621864319, + "learning_rate": 2.1099000000000002e-05, + "loss": 0.0482, + "step": 7036 + }, + { + "epoch": 5.943412162162162, + "grad_norm": 0.4534173011779785, + "learning_rate": 2.1102000000000002e-05, + "loss": 0.0511, + "step": 7037 + }, + { + "epoch": 5.944256756756757, + "grad_norm": 0.4887387454509735, + "learning_rate": 2.1105000000000002e-05, + "loss": 0.0509, + "step": 7038 + }, + { + "epoch": 5.945101351351351, + "grad_norm": 0.5164000988006592, + "learning_rate": 2.1108000000000002e-05, + "loss": 0.0407, + "step": 7039 + }, + { + "epoch": 5.945945945945946, + "grad_norm": 0.6495285034179688, + "learning_rate": 2.1111e-05, + "loss": 0.0481, + "step": 7040 + }, + { + "epoch": 5.94679054054054, + "grad_norm": 0.8300374746322632, + "learning_rate": 2.1114e-05, + "loss": 0.0515, + "step": 7041 + }, + { + "epoch": 5.947635135135135, + "grad_norm": 0.29357782006263733, + "learning_rate": 2.1117e-05, + "loss": 0.0296, + "step": 7042 + }, + { + "epoch": 5.94847972972973, + "grad_norm": 0.5295534133911133, + "learning_rate": 2.1119999999999998e-05, + "loss": 0.0486, + "step": 7043 + }, + { + "epoch": 5.949324324324325, + "grad_norm": 0.8441759943962097, + "learning_rate": 2.1122999999999997e-05, + "loss": 0.1518, + "step": 7044 + }, + { + "epoch": 5.950168918918919, + "grad_norm": 0.3546115458011627, + "learning_rate": 2.1126e-05, + "loss": 0.028, + "step": 7045 + }, + { + "epoch": 5.951013513513513, + "grad_norm": 0.5066583156585693, + "learning_rate": 2.1129e-05, + "loss": 0.0601, + "step": 7046 + }, + { + "epoch": 5.951858108108108, + "grad_norm": 0.6414474844932556, + "learning_rate": 2.1132e-05, + "loss": 0.0752, + "step": 7047 + }, + { + "epoch": 5.952702702702703, + "grad_norm": 0.2498338520526886, + "learning_rate": 2.1135e-05, + "loss": 0.0235, + "step": 7048 + }, + { + "epoch": 5.953547297297297, + "grad_norm": 0.4928322732448578, + "learning_rate": 2.1138e-05, + "loss": 0.0445, + "step": 7049 + }, + { + "epoch": 5.954391891891892, + "grad_norm": 0.53165602684021, + "learning_rate": 2.1141e-05, + "loss": 0.0837, + "step": 7050 + }, + { + "epoch": 5.955236486486487, + "grad_norm": 0.6754796504974365, + "learning_rate": 2.1144e-05, + "loss": 0.0444, + "step": 7051 + }, + { + "epoch": 5.956081081081081, + "grad_norm": 0.5738010406494141, + "learning_rate": 2.1147e-05, + "loss": 0.0387, + "step": 7052 + }, + { + "epoch": 5.956925675675675, + "grad_norm": 0.4308251440525055, + "learning_rate": 2.115e-05, + "loss": 0.0481, + "step": 7053 + }, + { + "epoch": 5.95777027027027, + "grad_norm": 0.41573870182037354, + "learning_rate": 2.1153e-05, + "loss": 0.0447, + "step": 7054 + }, + { + "epoch": 5.958614864864865, + "grad_norm": 0.5673714876174927, + "learning_rate": 2.1156000000000002e-05, + "loss": 0.0532, + "step": 7055 + }, + { + "epoch": 5.95945945945946, + "grad_norm": 0.5897387862205505, + "learning_rate": 2.1159000000000002e-05, + "loss": 0.0579, + "step": 7056 + }, + { + "epoch": 5.960304054054054, + "grad_norm": 0.4923725426197052, + "learning_rate": 2.1162e-05, + "loss": 0.033, + "step": 7057 + }, + { + "epoch": 5.961148648648649, + "grad_norm": 0.5670744180679321, + "learning_rate": 2.1165e-05, + "loss": 0.0547, + "step": 7058 + }, + { + "epoch": 5.961993243243243, + "grad_norm": 0.36321207880973816, + "learning_rate": 2.1168e-05, + "loss": 0.0357, + "step": 7059 + }, + { + "epoch": 5.962837837837838, + "grad_norm": 0.5485374927520752, + "learning_rate": 2.1171e-05, + "loss": 0.0398, + "step": 7060 + }, + { + "epoch": 5.9636824324324325, + "grad_norm": 0.5071348547935486, + "learning_rate": 2.1174e-05, + "loss": 0.0549, + "step": 7061 + }, + { + "epoch": 5.964527027027027, + "grad_norm": 0.45522740483283997, + "learning_rate": 2.1177e-05, + "loss": 0.047, + "step": 7062 + }, + { + "epoch": 5.965371621621622, + "grad_norm": 0.6206090450286865, + "learning_rate": 2.118e-05, + "loss": 0.0514, + "step": 7063 + }, + { + "epoch": 5.966216216216216, + "grad_norm": 0.8161625266075134, + "learning_rate": 2.1183e-05, + "loss": 0.0557, + "step": 7064 + }, + { + "epoch": 5.9670608108108105, + "grad_norm": 0.5723823308944702, + "learning_rate": 2.1186000000000003e-05, + "loss": 0.0632, + "step": 7065 + }, + { + "epoch": 5.967905405405405, + "grad_norm": 0.49489861726760864, + "learning_rate": 2.1189000000000003e-05, + "loss": 0.0386, + "step": 7066 + }, + { + "epoch": 5.96875, + "grad_norm": 0.7162661552429199, + "learning_rate": 2.1192e-05, + "loss": 0.0611, + "step": 7067 + }, + { + "epoch": 5.969594594594595, + "grad_norm": 0.741417407989502, + "learning_rate": 2.1195e-05, + "loss": 0.0943, + "step": 7068 + }, + { + "epoch": 5.9704391891891895, + "grad_norm": 0.7243646383285522, + "learning_rate": 2.1198e-05, + "loss": 0.05, + "step": 7069 + }, + { + "epoch": 5.971283783783784, + "grad_norm": 1.8279483318328857, + "learning_rate": 2.1201e-05, + "loss": 0.1584, + "step": 7070 + }, + { + "epoch": 5.972128378378378, + "grad_norm": 0.9395966529846191, + "learning_rate": 2.1204e-05, + "loss": 0.3138, + "step": 7071 + }, + { + "epoch": 5.972972972972973, + "grad_norm": 0.7484576106071472, + "learning_rate": 2.1207e-05, + "loss": 0.2021, + "step": 7072 + }, + { + "epoch": 5.9738175675675675, + "grad_norm": 0.5029864311218262, + "learning_rate": 2.121e-05, + "loss": 0.0621, + "step": 7073 + }, + { + "epoch": 5.974662162162162, + "grad_norm": 0.40686237812042236, + "learning_rate": 2.1213e-05, + "loss": 0.0482, + "step": 7074 + }, + { + "epoch": 5.975506756756757, + "grad_norm": 0.40653014183044434, + "learning_rate": 2.1216e-05, + "loss": 0.0506, + "step": 7075 + }, + { + "epoch": 5.976351351351351, + "grad_norm": 0.4217800199985504, + "learning_rate": 2.1219e-05, + "loss": 0.0434, + "step": 7076 + }, + { + "epoch": 5.977195945945946, + "grad_norm": 0.3780747056007385, + "learning_rate": 2.1222e-05, + "loss": 0.0385, + "step": 7077 + }, + { + "epoch": 5.97804054054054, + "grad_norm": 0.48314571380615234, + "learning_rate": 2.1225e-05, + "loss": 0.056, + "step": 7078 + }, + { + "epoch": 5.978885135135135, + "grad_norm": 0.5375436544418335, + "learning_rate": 2.1228e-05, + "loss": 0.0488, + "step": 7079 + }, + { + "epoch": 5.97972972972973, + "grad_norm": 0.3116242289543152, + "learning_rate": 2.1231e-05, + "loss": 0.048, + "step": 7080 + }, + { + "epoch": 5.980574324324325, + "grad_norm": 0.38254663348197937, + "learning_rate": 2.1234e-05, + "loss": 0.0387, + "step": 7081 + }, + { + "epoch": 5.981418918918919, + "grad_norm": 0.5657963156700134, + "learning_rate": 2.1237e-05, + "loss": 0.0417, + "step": 7082 + }, + { + "epoch": 5.982263513513513, + "grad_norm": 0.4260709583759308, + "learning_rate": 2.124e-05, + "loss": 0.0525, + "step": 7083 + }, + { + "epoch": 5.983108108108108, + "grad_norm": 0.33055225014686584, + "learning_rate": 2.1243e-05, + "loss": 0.0268, + "step": 7084 + }, + { + "epoch": 5.983952702702703, + "grad_norm": 0.4892235994338989, + "learning_rate": 2.1246000000000003e-05, + "loss": 0.0583, + "step": 7085 + }, + { + "epoch": 5.984797297297297, + "grad_norm": 0.5002988576889038, + "learning_rate": 2.1249000000000003e-05, + "loss": 0.0484, + "step": 7086 + }, + { + "epoch": 5.985641891891892, + "grad_norm": 0.29134562611579895, + "learning_rate": 2.1252000000000003e-05, + "loss": 0.032, + "step": 7087 + }, + { + "epoch": 5.986486486486487, + "grad_norm": 0.4529285132884979, + "learning_rate": 2.1255000000000002e-05, + "loss": 0.0291, + "step": 7088 + }, + { + "epoch": 5.987331081081081, + "grad_norm": 0.6367219090461731, + "learning_rate": 2.1258000000000002e-05, + "loss": 0.0485, + "step": 7089 + }, + { + "epoch": 5.988175675675675, + "grad_norm": 0.5486769676208496, + "learning_rate": 2.1261000000000002e-05, + "loss": 0.0589, + "step": 7090 + }, + { + "epoch": 5.98902027027027, + "grad_norm": 0.6533282399177551, + "learning_rate": 2.1264000000000002e-05, + "loss": 0.0335, + "step": 7091 + }, + { + "epoch": 5.989864864864865, + "grad_norm": 0.422770231962204, + "learning_rate": 2.1266999999999998e-05, + "loss": 0.0563, + "step": 7092 + }, + { + "epoch": 5.99070945945946, + "grad_norm": 0.5736091732978821, + "learning_rate": 2.1269999999999998e-05, + "loss": 0.0511, + "step": 7093 + }, + { + "epoch": 5.991554054054054, + "grad_norm": 0.40345728397369385, + "learning_rate": 2.1272999999999998e-05, + "loss": 0.0328, + "step": 7094 + }, + { + "epoch": 5.992398648648649, + "grad_norm": 0.8548729419708252, + "learning_rate": 2.1276e-05, + "loss": 0.0599, + "step": 7095 + }, + { + "epoch": 5.993243243243243, + "grad_norm": 0.6267647743225098, + "learning_rate": 2.1279e-05, + "loss": 0.0563, + "step": 7096 + }, + { + "epoch": 5.994087837837838, + "grad_norm": 0.7647335529327393, + "learning_rate": 2.1282e-05, + "loss": 0.102, + "step": 7097 + }, + { + "epoch": 5.9949324324324325, + "grad_norm": 0.9430329203605652, + "learning_rate": 2.1285e-05, + "loss": 0.0574, + "step": 7098 + }, + { + "epoch": 5.995777027027027, + "grad_norm": 0.6189017295837402, + "learning_rate": 2.1288e-05, + "loss": 0.0564, + "step": 7099 + }, + { + "epoch": 5.996621621621622, + "grad_norm": 0.62629234790802, + "learning_rate": 2.1291e-05, + "loss": 0.0534, + "step": 7100 + }, + { + "epoch": 5.997466216216216, + "grad_norm": 0.7762913703918457, + "learning_rate": 2.1294e-05, + "loss": 0.0678, + "step": 7101 + }, + { + "epoch": 5.9983108108108105, + "grad_norm": 0.673443078994751, + "learning_rate": 2.1297e-05, + "loss": 0.0531, + "step": 7102 + }, + { + "epoch": 5.999155405405405, + "grad_norm": 0.9016832113265991, + "learning_rate": 2.13e-05, + "loss": 0.0908, + "step": 7103 + }, + { + "epoch": 6.0, + "grad_norm": 0.6491882801055908, + "learning_rate": 2.1303e-05, + "loss": 0.0653, + "step": 7104 + }, + { + "epoch": 6.000844594594595, + "grad_norm": 1.2411333322525024, + "learning_rate": 2.1306000000000002e-05, + "loss": 0.278, + "step": 7105 + }, + { + "epoch": 6.0016891891891895, + "grad_norm": 0.7942840456962585, + "learning_rate": 2.1309000000000002e-05, + "loss": 0.1923, + "step": 7106 + }, + { + "epoch": 6.002533783783784, + "grad_norm": 0.5062112808227539, + "learning_rate": 2.1312000000000002e-05, + "loss": 0.1004, + "step": 7107 + }, + { + "epoch": 6.003378378378378, + "grad_norm": 0.4481554925441742, + "learning_rate": 2.1315000000000002e-05, + "loss": 0.0665, + "step": 7108 + }, + { + "epoch": 6.004222972972973, + "grad_norm": 0.5391278862953186, + "learning_rate": 2.1318e-05, + "loss": 0.0582, + "step": 7109 + }, + { + "epoch": 6.0050675675675675, + "grad_norm": 0.37148717045783997, + "learning_rate": 2.1321e-05, + "loss": 0.0569, + "step": 7110 + }, + { + "epoch": 6.005912162162162, + "grad_norm": 0.5277659893035889, + "learning_rate": 2.1324e-05, + "loss": 0.0393, + "step": 7111 + }, + { + "epoch": 6.006756756756757, + "grad_norm": 0.4578780233860016, + "learning_rate": 2.1327e-05, + "loss": 0.0442, + "step": 7112 + }, + { + "epoch": 6.007601351351352, + "grad_norm": 0.37673312425613403, + "learning_rate": 2.133e-05, + "loss": 0.06, + "step": 7113 + }, + { + "epoch": 6.008445945945946, + "grad_norm": 0.602414608001709, + "learning_rate": 2.1333e-05, + "loss": 0.0578, + "step": 7114 + }, + { + "epoch": 6.00929054054054, + "grad_norm": 0.45835211873054504, + "learning_rate": 2.1336000000000004e-05, + "loss": 0.0358, + "step": 7115 + }, + { + "epoch": 6.010135135135135, + "grad_norm": 0.4822278320789337, + "learning_rate": 2.1339e-05, + "loss": 0.0387, + "step": 7116 + }, + { + "epoch": 6.01097972972973, + "grad_norm": 0.38691461086273193, + "learning_rate": 2.1342e-05, + "loss": 0.0535, + "step": 7117 + }, + { + "epoch": 6.011824324324325, + "grad_norm": 0.6639519333839417, + "learning_rate": 2.1345e-05, + "loss": 0.051, + "step": 7118 + }, + { + "epoch": 6.012668918918919, + "grad_norm": 0.44485706090927124, + "learning_rate": 2.1348e-05, + "loss": 0.0354, + "step": 7119 + }, + { + "epoch": 6.013513513513513, + "grad_norm": 0.3784298598766327, + "learning_rate": 2.1351e-05, + "loss": 0.0442, + "step": 7120 + }, + { + "epoch": 6.014358108108108, + "grad_norm": 0.35095635056495667, + "learning_rate": 2.1354e-05, + "loss": 0.024, + "step": 7121 + }, + { + "epoch": 6.015202702702703, + "grad_norm": 0.40942665934562683, + "learning_rate": 2.1357e-05, + "loss": 0.0378, + "step": 7122 + }, + { + "epoch": 6.016047297297297, + "grad_norm": 0.33963581919670105, + "learning_rate": 2.136e-05, + "loss": 0.0301, + "step": 7123 + }, + { + "epoch": 6.016891891891892, + "grad_norm": 0.41756191849708557, + "learning_rate": 2.1363e-05, + "loss": 0.048, + "step": 7124 + }, + { + "epoch": 6.017736486486487, + "grad_norm": 0.4907315969467163, + "learning_rate": 2.1366000000000002e-05, + "loss": 0.0391, + "step": 7125 + }, + { + "epoch": 6.018581081081081, + "grad_norm": 0.4435591697692871, + "learning_rate": 2.1369e-05, + "loss": 0.0352, + "step": 7126 + }, + { + "epoch": 6.019425675675675, + "grad_norm": 0.3241303563117981, + "learning_rate": 2.1372e-05, + "loss": 0.0336, + "step": 7127 + }, + { + "epoch": 6.02027027027027, + "grad_norm": 0.4369211792945862, + "learning_rate": 2.1375e-05, + "loss": 0.0428, + "step": 7128 + }, + { + "epoch": 6.021114864864865, + "grad_norm": 0.28060662746429443, + "learning_rate": 2.1378e-05, + "loss": 0.0308, + "step": 7129 + }, + { + "epoch": 6.02195945945946, + "grad_norm": 0.3290829658508301, + "learning_rate": 2.1381e-05, + "loss": 0.0271, + "step": 7130 + }, + { + "epoch": 6.022804054054054, + "grad_norm": 0.3954499661922455, + "learning_rate": 2.1384e-05, + "loss": 0.0585, + "step": 7131 + }, + { + "epoch": 6.023648648648648, + "grad_norm": 0.3221830725669861, + "learning_rate": 2.1387e-05, + "loss": 0.0305, + "step": 7132 + }, + { + "epoch": 6.024493243243243, + "grad_norm": 0.34210827946662903, + "learning_rate": 2.139e-05, + "loss": 0.0395, + "step": 7133 + }, + { + "epoch": 6.025337837837838, + "grad_norm": 0.6572176218032837, + "learning_rate": 2.1393e-05, + "loss": 0.0435, + "step": 7134 + }, + { + "epoch": 6.0261824324324325, + "grad_norm": 0.4065794050693512, + "learning_rate": 2.1396e-05, + "loss": 0.0328, + "step": 7135 + }, + { + "epoch": 6.027027027027027, + "grad_norm": 0.4615703821182251, + "learning_rate": 2.1399000000000003e-05, + "loss": 0.0305, + "step": 7136 + }, + { + "epoch": 6.027871621621622, + "grad_norm": 0.4881494343280792, + "learning_rate": 2.1402000000000003e-05, + "loss": 0.0422, + "step": 7137 + }, + { + "epoch": 6.028716216216216, + "grad_norm": 0.5296261310577393, + "learning_rate": 2.1405000000000003e-05, + "loss": 0.0577, + "step": 7138 + }, + { + "epoch": 6.0295608108108105, + "grad_norm": 0.4751262366771698, + "learning_rate": 2.1408000000000002e-05, + "loss": 0.0893, + "step": 7139 + }, + { + "epoch": 6.030405405405405, + "grad_norm": 0.3926694095134735, + "learning_rate": 2.1411000000000002e-05, + "loss": 0.0369, + "step": 7140 + }, + { + "epoch": 6.03125, + "grad_norm": 0.8599852919578552, + "learning_rate": 2.1414e-05, + "loss": 0.0413, + "step": 7141 + }, + { + "epoch": 6.032094594594595, + "grad_norm": 0.45471400022506714, + "learning_rate": 2.1417e-05, + "loss": 0.0362, + "step": 7142 + }, + { + "epoch": 6.0329391891891895, + "grad_norm": 0.8819229006767273, + "learning_rate": 2.1419999999999998e-05, + "loss": 0.04, + "step": 7143 + }, + { + "epoch": 6.033783783783784, + "grad_norm": 0.5895608067512512, + "learning_rate": 2.1422999999999998e-05, + "loss": 0.0521, + "step": 7144 + }, + { + "epoch": 6.034628378378378, + "grad_norm": 0.44367122650146484, + "learning_rate": 2.1425999999999998e-05, + "loss": 0.0379, + "step": 7145 + }, + { + "epoch": 6.035472972972973, + "grad_norm": 0.8214004635810852, + "learning_rate": 2.1429e-05, + "loss": 0.046, + "step": 7146 + }, + { + "epoch": 6.0363175675675675, + "grad_norm": 0.36391013860702515, + "learning_rate": 2.1432e-05, + "loss": 0.0214, + "step": 7147 + }, + { + "epoch": 6.037162162162162, + "grad_norm": 0.4709986746311188, + "learning_rate": 2.1435e-05, + "loss": 0.0336, + "step": 7148 + }, + { + "epoch": 6.038006756756757, + "grad_norm": 0.5555054545402527, + "learning_rate": 2.1438e-05, + "loss": 0.0486, + "step": 7149 + }, + { + "epoch": 6.038851351351352, + "grad_norm": 0.5946035981178284, + "learning_rate": 2.1441e-05, + "loss": 0.0566, + "step": 7150 + }, + { + "epoch": 6.039695945945946, + "grad_norm": 0.3785681128501892, + "learning_rate": 2.1444e-05, + "loss": 0.0356, + "step": 7151 + }, + { + "epoch": 6.04054054054054, + "grad_norm": 1.0810390710830688, + "learning_rate": 2.1447e-05, + "loss": 0.069, + "step": 7152 + }, + { + "epoch": 6.041385135135135, + "grad_norm": 0.8132941722869873, + "learning_rate": 2.145e-05, + "loss": 0.0927, + "step": 7153 + }, + { + "epoch": 6.04222972972973, + "grad_norm": 0.6419681906700134, + "learning_rate": 2.1453e-05, + "loss": 0.0346, + "step": 7154 + }, + { + "epoch": 6.043074324324325, + "grad_norm": 1.8086451292037964, + "learning_rate": 2.1456e-05, + "loss": 0.3378, + "step": 7155 + }, + { + "epoch": 6.043918918918919, + "grad_norm": 0.6355175971984863, + "learning_rate": 2.1459000000000002e-05, + "loss": 0.1461, + "step": 7156 + }, + { + "epoch": 6.044763513513513, + "grad_norm": 0.5732604265213013, + "learning_rate": 2.1462000000000002e-05, + "loss": 0.1184, + "step": 7157 + }, + { + "epoch": 6.045608108108108, + "grad_norm": 0.4874204695224762, + "learning_rate": 2.1465000000000002e-05, + "loss": 0.0894, + "step": 7158 + }, + { + "epoch": 6.046452702702703, + "grad_norm": 0.550930917263031, + "learning_rate": 2.1468000000000002e-05, + "loss": 0.0615, + "step": 7159 + }, + { + "epoch": 6.047297297297297, + "grad_norm": 0.5168009996414185, + "learning_rate": 2.1471e-05, + "loss": 0.0632, + "step": 7160 + }, + { + "epoch": 6.048141891891892, + "grad_norm": 0.42398932576179504, + "learning_rate": 2.1474e-05, + "loss": 0.0546, + "step": 7161 + }, + { + "epoch": 6.048986486486487, + "grad_norm": 0.3432709574699402, + "learning_rate": 2.1477e-05, + "loss": 0.03, + "step": 7162 + }, + { + "epoch": 6.049831081081081, + "grad_norm": 0.5134495496749878, + "learning_rate": 2.148e-05, + "loss": 0.0398, + "step": 7163 + }, + { + "epoch": 6.050675675675675, + "grad_norm": 0.40231114625930786, + "learning_rate": 2.1483e-05, + "loss": 0.0397, + "step": 7164 + }, + { + "epoch": 6.05152027027027, + "grad_norm": 0.2651325464248657, + "learning_rate": 2.1486e-05, + "loss": 0.0218, + "step": 7165 + }, + { + "epoch": 6.052364864864865, + "grad_norm": 0.5569732785224915, + "learning_rate": 2.1489e-05, + "loss": 0.0354, + "step": 7166 + }, + { + "epoch": 6.05320945945946, + "grad_norm": 0.45496705174446106, + "learning_rate": 2.1492e-05, + "loss": 0.0353, + "step": 7167 + }, + { + "epoch": 6.054054054054054, + "grad_norm": 0.5216231346130371, + "learning_rate": 2.1495e-05, + "loss": 0.0489, + "step": 7168 + }, + { + "epoch": 6.054898648648648, + "grad_norm": 0.6172152757644653, + "learning_rate": 2.1498e-05, + "loss": 0.0618, + "step": 7169 + }, + { + "epoch": 6.055743243243243, + "grad_norm": 0.3933444619178772, + "learning_rate": 2.1501e-05, + "loss": 0.0334, + "step": 7170 + }, + { + "epoch": 6.056587837837838, + "grad_norm": 0.486097127199173, + "learning_rate": 2.1504e-05, + "loss": 0.0609, + "step": 7171 + }, + { + "epoch": 6.0574324324324325, + "grad_norm": 0.3834947347640991, + "learning_rate": 2.1507e-05, + "loss": 0.0245, + "step": 7172 + }, + { + "epoch": 6.058277027027027, + "grad_norm": 0.5205850005149841, + "learning_rate": 2.151e-05, + "loss": 0.0331, + "step": 7173 + }, + { + "epoch": 6.059121621621622, + "grad_norm": 0.366386741399765, + "learning_rate": 2.1513e-05, + "loss": 0.0342, + "step": 7174 + }, + { + "epoch": 6.059966216216216, + "grad_norm": 0.5103170275688171, + "learning_rate": 2.1516e-05, + "loss": 0.0425, + "step": 7175 + }, + { + "epoch": 6.0608108108108105, + "grad_norm": 0.4486725628376007, + "learning_rate": 2.1519000000000002e-05, + "loss": 0.0276, + "step": 7176 + }, + { + "epoch": 6.061655405405405, + "grad_norm": 0.4174152612686157, + "learning_rate": 2.1522e-05, + "loss": 0.0521, + "step": 7177 + }, + { + "epoch": 6.0625, + "grad_norm": 0.5949570536613464, + "learning_rate": 2.1525e-05, + "loss": 0.0417, + "step": 7178 + }, + { + "epoch": 6.063344594594595, + "grad_norm": 0.5720686316490173, + "learning_rate": 2.1528e-05, + "loss": 0.0431, + "step": 7179 + }, + { + "epoch": 6.0641891891891895, + "grad_norm": 0.3456312119960785, + "learning_rate": 2.1531e-05, + "loss": 0.0233, + "step": 7180 + }, + { + "epoch": 6.065033783783784, + "grad_norm": 0.39094778895378113, + "learning_rate": 2.1534e-05, + "loss": 0.0389, + "step": 7181 + }, + { + "epoch": 6.065878378378378, + "grad_norm": 0.4919477701187134, + "learning_rate": 2.1537e-05, + "loss": 0.0379, + "step": 7182 + }, + { + "epoch": 6.066722972972973, + "grad_norm": 0.4690771996974945, + "learning_rate": 2.154e-05, + "loss": 0.0343, + "step": 7183 + }, + { + "epoch": 6.0675675675675675, + "grad_norm": 0.6049042344093323, + "learning_rate": 2.1543e-05, + "loss": 0.0273, + "step": 7184 + }, + { + "epoch": 6.068412162162162, + "grad_norm": 0.6376462578773499, + "learning_rate": 2.1546e-05, + "loss": 0.0551, + "step": 7185 + }, + { + "epoch": 6.069256756756757, + "grad_norm": 0.34818583726882935, + "learning_rate": 2.1549000000000003e-05, + "loss": 0.0257, + "step": 7186 + }, + { + "epoch": 6.070101351351352, + "grad_norm": 0.4838002324104309, + "learning_rate": 2.1552000000000003e-05, + "loss": 0.0439, + "step": 7187 + }, + { + "epoch": 6.070945945945946, + "grad_norm": 0.3391258716583252, + "learning_rate": 2.1555000000000003e-05, + "loss": 0.0262, + "step": 7188 + }, + { + "epoch": 6.07179054054054, + "grad_norm": 0.47655320167541504, + "learning_rate": 2.1558000000000003e-05, + "loss": 0.0251, + "step": 7189 + }, + { + "epoch": 6.072635135135135, + "grad_norm": 0.37938863039016724, + "learning_rate": 2.1561e-05, + "loss": 0.0307, + "step": 7190 + }, + { + "epoch": 6.07347972972973, + "grad_norm": 0.4485480487346649, + "learning_rate": 2.1564e-05, + "loss": 0.0351, + "step": 7191 + }, + { + "epoch": 6.074324324324325, + "grad_norm": 1.454395055770874, + "learning_rate": 2.1567e-05, + "loss": 0.0316, + "step": 7192 + }, + { + "epoch": 6.075168918918919, + "grad_norm": 0.6319937705993652, + "learning_rate": 2.157e-05, + "loss": 0.037, + "step": 7193 + }, + { + "epoch": 6.076013513513513, + "grad_norm": 0.4077817499637604, + "learning_rate": 2.1572999999999998e-05, + "loss": 0.0231, + "step": 7194 + }, + { + "epoch": 6.076858108108108, + "grad_norm": 0.5476170778274536, + "learning_rate": 2.1575999999999998e-05, + "loss": 0.046, + "step": 7195 + }, + { + "epoch": 6.077702702702703, + "grad_norm": 0.5420228242874146, + "learning_rate": 2.1579e-05, + "loss": 0.0461, + "step": 7196 + }, + { + "epoch": 6.078547297297297, + "grad_norm": 0.4314330816268921, + "learning_rate": 2.1582e-05, + "loss": 0.0373, + "step": 7197 + }, + { + "epoch": 6.079391891891892, + "grad_norm": 0.604451060295105, + "learning_rate": 2.1585e-05, + "loss": 0.032, + "step": 7198 + }, + { + "epoch": 6.080236486486487, + "grad_norm": 0.5859724283218384, + "learning_rate": 2.1588e-05, + "loss": 0.0434, + "step": 7199 + }, + { + "epoch": 6.081081081081081, + "grad_norm": 1.5754148960113525, + "learning_rate": 2.1591e-05, + "loss": 0.0771, + "step": 7200 + }, + { + "epoch": 6.081925675675675, + "grad_norm": 0.626247227191925, + "learning_rate": 2.1594e-05, + "loss": 0.0575, + "step": 7201 + }, + { + "epoch": 6.08277027027027, + "grad_norm": 0.6536280512809753, + "learning_rate": 2.1597e-05, + "loss": 0.0553, + "step": 7202 + }, + { + "epoch": 6.083614864864865, + "grad_norm": 0.9966816306114197, + "learning_rate": 2.16e-05, + "loss": 0.0793, + "step": 7203 + }, + { + "epoch": 6.08445945945946, + "grad_norm": 0.9550701975822449, + "learning_rate": 2.1603e-05, + "loss": 0.1008, + "step": 7204 + }, + { + "epoch": 6.085304054054054, + "grad_norm": 3.5875442028045654, + "learning_rate": 2.1606e-05, + "loss": 0.4589, + "step": 7205 + }, + { + "epoch": 6.086148648648648, + "grad_norm": 0.8752773404121399, + "learning_rate": 2.1609000000000003e-05, + "loss": 0.2175, + "step": 7206 + }, + { + "epoch": 6.086993243243243, + "grad_norm": 0.5049457550048828, + "learning_rate": 2.1612000000000002e-05, + "loss": 0.088, + "step": 7207 + }, + { + "epoch": 6.087837837837838, + "grad_norm": 0.4812326431274414, + "learning_rate": 2.1615000000000002e-05, + "loss": 0.0676, + "step": 7208 + }, + { + "epoch": 6.0886824324324325, + "grad_norm": 0.4351433217525482, + "learning_rate": 2.1618000000000002e-05, + "loss": 0.0461, + "step": 7209 + }, + { + "epoch": 6.089527027027027, + "grad_norm": 0.4514690041542053, + "learning_rate": 2.1621000000000002e-05, + "loss": 0.0694, + "step": 7210 + }, + { + "epoch": 6.090371621621622, + "grad_norm": 0.369441956281662, + "learning_rate": 2.1624e-05, + "loss": 0.038, + "step": 7211 + }, + { + "epoch": 6.091216216216216, + "grad_norm": 0.5883796215057373, + "learning_rate": 2.1627e-05, + "loss": 0.0495, + "step": 7212 + }, + { + "epoch": 6.0920608108108105, + "grad_norm": 0.3267923593521118, + "learning_rate": 2.163e-05, + "loss": 0.0357, + "step": 7213 + }, + { + "epoch": 6.092905405405405, + "grad_norm": 0.27459803223609924, + "learning_rate": 2.1633e-05, + "loss": 0.0345, + "step": 7214 + }, + { + "epoch": 6.09375, + "grad_norm": 0.29078614711761475, + "learning_rate": 2.1635999999999997e-05, + "loss": 0.0252, + "step": 7215 + }, + { + "epoch": 6.094594594594595, + "grad_norm": 0.3002908229827881, + "learning_rate": 2.1639e-05, + "loss": 0.027, + "step": 7216 + }, + { + "epoch": 6.0954391891891895, + "grad_norm": 0.5489783883094788, + "learning_rate": 2.1642e-05, + "loss": 0.0427, + "step": 7217 + }, + { + "epoch": 6.096283783783784, + "grad_norm": 0.9915933012962341, + "learning_rate": 2.1645e-05, + "loss": 0.0509, + "step": 7218 + }, + { + "epoch": 6.097128378378378, + "grad_norm": 0.6837965846061707, + "learning_rate": 2.1648e-05, + "loss": 0.0531, + "step": 7219 + }, + { + "epoch": 6.097972972972973, + "grad_norm": 0.4582420587539673, + "learning_rate": 2.1651e-05, + "loss": 0.0362, + "step": 7220 + }, + { + "epoch": 6.0988175675675675, + "grad_norm": 0.3927347660064697, + "learning_rate": 2.1654e-05, + "loss": 0.0479, + "step": 7221 + }, + { + "epoch": 6.099662162162162, + "grad_norm": 0.5037749409675598, + "learning_rate": 2.1657e-05, + "loss": 0.0333, + "step": 7222 + }, + { + "epoch": 6.100506756756757, + "grad_norm": 0.5298280119895935, + "learning_rate": 2.166e-05, + "loss": 0.049, + "step": 7223 + }, + { + "epoch": 6.101351351351352, + "grad_norm": 0.5915836691856384, + "learning_rate": 2.1663e-05, + "loss": 0.0405, + "step": 7224 + }, + { + "epoch": 6.102195945945946, + "grad_norm": 0.44966110587120056, + "learning_rate": 2.1666e-05, + "loss": 0.0296, + "step": 7225 + }, + { + "epoch": 6.10304054054054, + "grad_norm": 0.43570420145988464, + "learning_rate": 2.1669000000000002e-05, + "loss": 0.0329, + "step": 7226 + }, + { + "epoch": 6.103885135135135, + "grad_norm": 0.4337502121925354, + "learning_rate": 2.1672000000000002e-05, + "loss": 0.0344, + "step": 7227 + }, + { + "epoch": 6.10472972972973, + "grad_norm": 0.46281108260154724, + "learning_rate": 2.1675e-05, + "loss": 0.0423, + "step": 7228 + }, + { + "epoch": 6.105574324324325, + "grad_norm": 0.512098491191864, + "learning_rate": 2.1678e-05, + "loss": 0.0323, + "step": 7229 + }, + { + "epoch": 6.106418918918919, + "grad_norm": 0.5264639854431152, + "learning_rate": 2.1681e-05, + "loss": 0.0423, + "step": 7230 + }, + { + "epoch": 6.107263513513513, + "grad_norm": 0.7309818267822266, + "learning_rate": 2.1684e-05, + "loss": 0.054, + "step": 7231 + }, + { + "epoch": 6.108108108108108, + "grad_norm": 0.2167467176914215, + "learning_rate": 2.1687e-05, + "loss": 0.0222, + "step": 7232 + }, + { + "epoch": 6.108952702702703, + "grad_norm": 0.5971783399581909, + "learning_rate": 2.169e-05, + "loss": 0.0365, + "step": 7233 + }, + { + "epoch": 6.109797297297297, + "grad_norm": 0.32220643758773804, + "learning_rate": 2.1693e-05, + "loss": 0.024, + "step": 7234 + }, + { + "epoch": 6.110641891891892, + "grad_norm": 0.32243338227272034, + "learning_rate": 2.1696e-05, + "loss": 0.0298, + "step": 7235 + }, + { + "epoch": 6.111486486486487, + "grad_norm": 0.40906795859336853, + "learning_rate": 2.1699000000000003e-05, + "loss": 0.0343, + "step": 7236 + }, + { + "epoch": 6.112331081081081, + "grad_norm": 0.46812301874160767, + "learning_rate": 2.1702000000000003e-05, + "loss": 0.0319, + "step": 7237 + }, + { + "epoch": 6.113175675675675, + "grad_norm": 0.42001721262931824, + "learning_rate": 2.1705000000000003e-05, + "loss": 0.0322, + "step": 7238 + }, + { + "epoch": 6.11402027027027, + "grad_norm": 0.5733482241630554, + "learning_rate": 2.1708e-05, + "loss": 0.0382, + "step": 7239 + }, + { + "epoch": 6.114864864864865, + "grad_norm": 0.5185551643371582, + "learning_rate": 2.1711e-05, + "loss": 0.0647, + "step": 7240 + }, + { + "epoch": 6.11570945945946, + "grad_norm": 0.49260300397872925, + "learning_rate": 2.1714e-05, + "loss": 0.0387, + "step": 7241 + }, + { + "epoch": 6.116554054054054, + "grad_norm": 0.7390943765640259, + "learning_rate": 2.1717e-05, + "loss": 0.0331, + "step": 7242 + }, + { + "epoch": 6.117398648648648, + "grad_norm": 0.5366110801696777, + "learning_rate": 2.172e-05, + "loss": 0.0401, + "step": 7243 + }, + { + "epoch": 6.118243243243243, + "grad_norm": 0.5103926062583923, + "learning_rate": 2.1723e-05, + "loss": 0.0438, + "step": 7244 + }, + { + "epoch": 6.119087837837838, + "grad_norm": 0.552852988243103, + "learning_rate": 2.1726e-05, + "loss": 0.0365, + "step": 7245 + }, + { + "epoch": 6.1199324324324325, + "grad_norm": 1.075995683670044, + "learning_rate": 2.1729e-05, + "loss": 0.0406, + "step": 7246 + }, + { + "epoch": 6.120777027027027, + "grad_norm": 0.7220231294631958, + "learning_rate": 2.1732e-05, + "loss": 0.0511, + "step": 7247 + }, + { + "epoch": 6.121621621621622, + "grad_norm": 0.8781309127807617, + "learning_rate": 2.1735e-05, + "loss": 0.0527, + "step": 7248 + }, + { + "epoch": 6.122466216216216, + "grad_norm": 0.6346035599708557, + "learning_rate": 2.1738e-05, + "loss": 0.0422, + "step": 7249 + }, + { + "epoch": 6.1233108108108105, + "grad_norm": 0.4869632124900818, + "learning_rate": 2.1741e-05, + "loss": 0.041, + "step": 7250 + }, + { + "epoch": 6.124155405405405, + "grad_norm": 0.5678491592407227, + "learning_rate": 2.1744e-05, + "loss": 0.0421, + "step": 7251 + }, + { + "epoch": 6.125, + "grad_norm": 0.501241147518158, + "learning_rate": 2.1747e-05, + "loss": 0.0481, + "step": 7252 + }, + { + "epoch": 6.125844594594595, + "grad_norm": 0.7166866660118103, + "learning_rate": 2.175e-05, + "loss": 0.0408, + "step": 7253 + }, + { + "epoch": 6.1266891891891895, + "grad_norm": 0.6855593919754028, + "learning_rate": 2.1753e-05, + "loss": 0.0574, + "step": 7254 + }, + { + "epoch": 6.127533783783784, + "grad_norm": 1.604520320892334, + "learning_rate": 2.1756e-05, + "loss": 0.3492, + "step": 7255 + }, + { + "epoch": 6.128378378378378, + "grad_norm": 1.0386346578598022, + "learning_rate": 2.1759e-05, + "loss": 0.2162, + "step": 7256 + }, + { + "epoch": 6.129222972972973, + "grad_norm": 0.5436184406280518, + "learning_rate": 2.1762000000000003e-05, + "loss": 0.0875, + "step": 7257 + }, + { + "epoch": 6.1300675675675675, + "grad_norm": 0.47641921043395996, + "learning_rate": 2.1765000000000003e-05, + "loss": 0.0676, + "step": 7258 + }, + { + "epoch": 6.130912162162162, + "grad_norm": 0.5764700174331665, + "learning_rate": 2.1768000000000002e-05, + "loss": 0.0512, + "step": 7259 + }, + { + "epoch": 6.131756756756757, + "grad_norm": 0.4528517425060272, + "learning_rate": 2.1771000000000002e-05, + "loss": 0.0487, + "step": 7260 + }, + { + "epoch": 6.132601351351352, + "grad_norm": 0.44919729232788086, + "learning_rate": 2.1774000000000002e-05, + "loss": 0.0385, + "step": 7261 + }, + { + "epoch": 6.133445945945946, + "grad_norm": 0.3790380358695984, + "learning_rate": 2.1777000000000002e-05, + "loss": 0.048, + "step": 7262 + }, + { + "epoch": 6.13429054054054, + "grad_norm": 0.5574291348457336, + "learning_rate": 2.178e-05, + "loss": 0.0423, + "step": 7263 + }, + { + "epoch": 6.135135135135135, + "grad_norm": 0.4991079568862915, + "learning_rate": 2.1782999999999998e-05, + "loss": 0.0397, + "step": 7264 + }, + { + "epoch": 6.13597972972973, + "grad_norm": 0.49059030413627625, + "learning_rate": 2.1785999999999998e-05, + "loss": 0.0367, + "step": 7265 + }, + { + "epoch": 6.136824324324325, + "grad_norm": 0.24888092279434204, + "learning_rate": 2.1788999999999998e-05, + "loss": 0.0251, + "step": 7266 + }, + { + "epoch": 6.137668918918919, + "grad_norm": 0.4245147705078125, + "learning_rate": 2.1792e-05, + "loss": 0.0415, + "step": 7267 + }, + { + "epoch": 6.138513513513513, + "grad_norm": 0.3680381774902344, + "learning_rate": 2.1795e-05, + "loss": 0.0279, + "step": 7268 + }, + { + "epoch": 6.139358108108108, + "grad_norm": 0.4505334496498108, + "learning_rate": 2.1798e-05, + "loss": 0.0528, + "step": 7269 + }, + { + "epoch": 6.140202702702703, + "grad_norm": 0.5433356761932373, + "learning_rate": 2.1801e-05, + "loss": 0.0441, + "step": 7270 + }, + { + "epoch": 6.141047297297297, + "grad_norm": 0.4145091474056244, + "learning_rate": 2.1804e-05, + "loss": 0.0239, + "step": 7271 + }, + { + "epoch": 6.141891891891892, + "grad_norm": 0.4579560458660126, + "learning_rate": 2.1807e-05, + "loss": 0.03, + "step": 7272 + }, + { + "epoch": 6.142736486486487, + "grad_norm": 0.8120315670967102, + "learning_rate": 2.181e-05, + "loss": 0.0406, + "step": 7273 + }, + { + "epoch": 6.143581081081081, + "grad_norm": 0.4392738342285156, + "learning_rate": 2.1813e-05, + "loss": 0.0301, + "step": 7274 + }, + { + "epoch": 6.144425675675675, + "grad_norm": 0.3585166931152344, + "learning_rate": 2.1816e-05, + "loss": 0.0252, + "step": 7275 + }, + { + "epoch": 6.14527027027027, + "grad_norm": 0.5309386253356934, + "learning_rate": 2.1819e-05, + "loss": 0.0543, + "step": 7276 + }, + { + "epoch": 6.146114864864865, + "grad_norm": 0.8673108220100403, + "learning_rate": 2.1822000000000002e-05, + "loss": 0.0564, + "step": 7277 + }, + { + "epoch": 6.14695945945946, + "grad_norm": 0.8913524150848389, + "learning_rate": 2.1825000000000002e-05, + "loss": 0.0518, + "step": 7278 + }, + { + "epoch": 6.147804054054054, + "grad_norm": 0.4703996479511261, + "learning_rate": 2.1828000000000002e-05, + "loss": 0.0448, + "step": 7279 + }, + { + "epoch": 6.148648648648648, + "grad_norm": 0.45349064469337463, + "learning_rate": 2.1831e-05, + "loss": 0.0338, + "step": 7280 + }, + { + "epoch": 6.149493243243243, + "grad_norm": 0.4851091206073761, + "learning_rate": 2.1834e-05, + "loss": 0.0373, + "step": 7281 + }, + { + "epoch": 6.150337837837838, + "grad_norm": 0.37252506613731384, + "learning_rate": 2.1837e-05, + "loss": 0.0407, + "step": 7282 + }, + { + "epoch": 6.1511824324324325, + "grad_norm": 0.5685481429100037, + "learning_rate": 2.184e-05, + "loss": 0.0509, + "step": 7283 + }, + { + "epoch": 6.152027027027027, + "grad_norm": 0.5770229697227478, + "learning_rate": 2.1843e-05, + "loss": 0.0426, + "step": 7284 + }, + { + "epoch": 6.152871621621622, + "grad_norm": 0.40392395853996277, + "learning_rate": 2.1846e-05, + "loss": 0.0394, + "step": 7285 + }, + { + "epoch": 6.153716216216216, + "grad_norm": 0.601516842842102, + "learning_rate": 2.1849e-05, + "loss": 0.0495, + "step": 7286 + }, + { + "epoch": 6.1545608108108105, + "grad_norm": 0.4554702639579773, + "learning_rate": 2.1852000000000004e-05, + "loss": 0.0448, + "step": 7287 + }, + { + "epoch": 6.155405405405405, + "grad_norm": 0.4206721782684326, + "learning_rate": 2.1855e-05, + "loss": 0.0327, + "step": 7288 + }, + { + "epoch": 6.15625, + "grad_norm": 0.6717713475227356, + "learning_rate": 2.1858e-05, + "loss": 0.0382, + "step": 7289 + }, + { + "epoch": 6.157094594594595, + "grad_norm": 0.33498382568359375, + "learning_rate": 2.1861e-05, + "loss": 0.029, + "step": 7290 + }, + { + "epoch": 6.1579391891891895, + "grad_norm": 0.47883716225624084, + "learning_rate": 2.1864e-05, + "loss": 0.0439, + "step": 7291 + }, + { + "epoch": 6.158783783783784, + "grad_norm": 0.5241338610649109, + "learning_rate": 2.1867e-05, + "loss": 0.0354, + "step": 7292 + }, + { + "epoch": 6.159628378378378, + "grad_norm": 0.6170089840888977, + "learning_rate": 2.187e-05, + "loss": 0.0433, + "step": 7293 + }, + { + "epoch": 6.160472972972973, + "grad_norm": 0.6060220003128052, + "learning_rate": 2.1873e-05, + "loss": 0.0386, + "step": 7294 + }, + { + "epoch": 6.1613175675675675, + "grad_norm": 0.7179556488990784, + "learning_rate": 2.1876e-05, + "loss": 0.0568, + "step": 7295 + }, + { + "epoch": 6.162162162162162, + "grad_norm": 0.7638072371482849, + "learning_rate": 2.1879e-05, + "loss": 0.0562, + "step": 7296 + }, + { + "epoch": 6.163006756756757, + "grad_norm": 1.0945935249328613, + "learning_rate": 2.1882e-05, + "loss": 0.0466, + "step": 7297 + }, + { + "epoch": 6.163851351351352, + "grad_norm": 0.7947364449501038, + "learning_rate": 2.1885e-05, + "loss": 0.0569, + "step": 7298 + }, + { + "epoch": 6.164695945945946, + "grad_norm": 0.56634122133255, + "learning_rate": 2.1888e-05, + "loss": 0.0579, + "step": 7299 + }, + { + "epoch": 6.16554054054054, + "grad_norm": 1.3370869159698486, + "learning_rate": 2.1891e-05, + "loss": 0.0495, + "step": 7300 + }, + { + "epoch": 6.166385135135135, + "grad_norm": 0.5537945628166199, + "learning_rate": 2.1894e-05, + "loss": 0.0298, + "step": 7301 + }, + { + "epoch": 6.16722972972973, + "grad_norm": 1.6982508897781372, + "learning_rate": 2.1897e-05, + "loss": 0.0931, + "step": 7302 + }, + { + "epoch": 6.168074324324325, + "grad_norm": 1.136252522468567, + "learning_rate": 2.19e-05, + "loss": 0.1028, + "step": 7303 + }, + { + "epoch": 6.168918918918919, + "grad_norm": 0.7818033695220947, + "learning_rate": 2.1903e-05, + "loss": 0.0664, + "step": 7304 + }, + { + "epoch": 6.169763513513513, + "grad_norm": 1.4213405847549438, + "learning_rate": 2.1906e-05, + "loss": 0.4025, + "step": 7305 + }, + { + "epoch": 6.170608108108108, + "grad_norm": 1.6767258644104004, + "learning_rate": 2.1909e-05, + "loss": 0.2157, + "step": 7306 + }, + { + "epoch": 6.171452702702703, + "grad_norm": 0.8621162176132202, + "learning_rate": 2.1912000000000003e-05, + "loss": 0.1281, + "step": 7307 + }, + { + "epoch": 6.172297297297297, + "grad_norm": 0.5283829569816589, + "learning_rate": 2.1915000000000003e-05, + "loss": 0.0579, + "step": 7308 + }, + { + "epoch": 6.173141891891892, + "grad_norm": 0.8568291068077087, + "learning_rate": 2.1918000000000003e-05, + "loss": 0.0735, + "step": 7309 + }, + { + "epoch": 6.173986486486487, + "grad_norm": 0.5657371878623962, + "learning_rate": 2.1921000000000002e-05, + "loss": 0.0491, + "step": 7310 + }, + { + "epoch": 6.174831081081081, + "grad_norm": 0.3998936116695404, + "learning_rate": 2.1924000000000002e-05, + "loss": 0.0494, + "step": 7311 + }, + { + "epoch": 6.175675675675675, + "grad_norm": 0.3854387402534485, + "learning_rate": 2.1927000000000002e-05, + "loss": 0.0589, + "step": 7312 + }, + { + "epoch": 6.17652027027027, + "grad_norm": 0.48322629928588867, + "learning_rate": 2.193e-05, + "loss": 0.041, + "step": 7313 + }, + { + "epoch": 6.177364864864865, + "grad_norm": 0.4420948028564453, + "learning_rate": 2.1932999999999998e-05, + "loss": 0.0565, + "step": 7314 + }, + { + "epoch": 6.17820945945946, + "grad_norm": 0.4282139241695404, + "learning_rate": 2.1935999999999998e-05, + "loss": 0.0551, + "step": 7315 + }, + { + "epoch": 6.179054054054054, + "grad_norm": 0.37148791551589966, + "learning_rate": 2.1938999999999998e-05, + "loss": 0.0371, + "step": 7316 + }, + { + "epoch": 6.179898648648648, + "grad_norm": 0.3097013831138611, + "learning_rate": 2.1942e-05, + "loss": 0.0354, + "step": 7317 + }, + { + "epoch": 6.180743243243243, + "grad_norm": 0.48559924960136414, + "learning_rate": 2.1945e-05, + "loss": 0.0491, + "step": 7318 + }, + { + "epoch": 6.181587837837838, + "grad_norm": 0.44005510210990906, + "learning_rate": 2.1948e-05, + "loss": 0.0444, + "step": 7319 + }, + { + "epoch": 6.1824324324324325, + "grad_norm": 0.38889265060424805, + "learning_rate": 2.1951e-05, + "loss": 0.0345, + "step": 7320 + }, + { + "epoch": 6.183277027027027, + "grad_norm": 0.40958529710769653, + "learning_rate": 2.1954e-05, + "loss": 0.0385, + "step": 7321 + }, + { + "epoch": 6.184121621621622, + "grad_norm": 0.6368007063865662, + "learning_rate": 2.1957e-05, + "loss": 0.061, + "step": 7322 + }, + { + "epoch": 6.184966216216216, + "grad_norm": 0.7291532158851624, + "learning_rate": 2.196e-05, + "loss": 0.0531, + "step": 7323 + }, + { + "epoch": 6.1858108108108105, + "grad_norm": 0.31827399134635925, + "learning_rate": 2.1963e-05, + "loss": 0.0272, + "step": 7324 + }, + { + "epoch": 6.186655405405405, + "grad_norm": 0.44589293003082275, + "learning_rate": 2.1966e-05, + "loss": 0.0297, + "step": 7325 + }, + { + "epoch": 6.1875, + "grad_norm": 0.48497679829597473, + "learning_rate": 2.1969e-05, + "loss": 0.0421, + "step": 7326 + }, + { + "epoch": 6.188344594594595, + "grad_norm": 0.5793243050575256, + "learning_rate": 2.1972000000000002e-05, + "loss": 0.0274, + "step": 7327 + }, + { + "epoch": 6.1891891891891895, + "grad_norm": 0.6027854681015015, + "learning_rate": 2.1975000000000002e-05, + "loss": 0.0533, + "step": 7328 + }, + { + "epoch": 6.190033783783784, + "grad_norm": 0.4442838132381439, + "learning_rate": 2.1978000000000002e-05, + "loss": 0.035, + "step": 7329 + }, + { + "epoch": 6.190878378378378, + "grad_norm": 0.4592352509498596, + "learning_rate": 2.1981000000000002e-05, + "loss": 0.0405, + "step": 7330 + }, + { + "epoch": 6.191722972972973, + "grad_norm": 0.34401747584342957, + "learning_rate": 2.1984e-05, + "loss": 0.024, + "step": 7331 + }, + { + "epoch": 6.1925675675675675, + "grad_norm": 0.38328662514686584, + "learning_rate": 2.1987e-05, + "loss": 0.0398, + "step": 7332 + }, + { + "epoch": 6.193412162162162, + "grad_norm": 0.535849928855896, + "learning_rate": 2.199e-05, + "loss": 0.0337, + "step": 7333 + }, + { + "epoch": 6.194256756756757, + "grad_norm": 0.40938401222229004, + "learning_rate": 2.1993e-05, + "loss": 0.0325, + "step": 7334 + }, + { + "epoch": 6.195101351351352, + "grad_norm": 0.5594891309738159, + "learning_rate": 2.1996e-05, + "loss": 0.0611, + "step": 7335 + }, + { + "epoch": 6.195945945945946, + "grad_norm": 0.4077969491481781, + "learning_rate": 2.1999e-05, + "loss": 0.0406, + "step": 7336 + }, + { + "epoch": 6.19679054054054, + "grad_norm": 0.5827769041061401, + "learning_rate": 2.2002e-05, + "loss": 0.0507, + "step": 7337 + }, + { + "epoch": 6.197635135135135, + "grad_norm": 0.5352149605751038, + "learning_rate": 2.2005e-05, + "loss": 0.0389, + "step": 7338 + }, + { + "epoch": 6.19847972972973, + "grad_norm": 0.4923750162124634, + "learning_rate": 2.2008e-05, + "loss": 0.025, + "step": 7339 + }, + { + "epoch": 6.199324324324325, + "grad_norm": 0.3406704068183899, + "learning_rate": 2.2011e-05, + "loss": 0.0387, + "step": 7340 + }, + { + "epoch": 6.200168918918919, + "grad_norm": 0.5637153387069702, + "learning_rate": 2.2014e-05, + "loss": 0.0383, + "step": 7341 + }, + { + "epoch": 6.201013513513513, + "grad_norm": 0.3562237322330475, + "learning_rate": 2.2017e-05, + "loss": 0.0347, + "step": 7342 + }, + { + "epoch": 6.201858108108108, + "grad_norm": 0.5766254663467407, + "learning_rate": 2.202e-05, + "loss": 0.0424, + "step": 7343 + }, + { + "epoch": 6.202702702702703, + "grad_norm": 0.5017872452735901, + "learning_rate": 2.2023e-05, + "loss": 0.0315, + "step": 7344 + }, + { + "epoch": 6.203547297297297, + "grad_norm": 0.7003552317619324, + "learning_rate": 2.2026e-05, + "loss": 0.044, + "step": 7345 + }, + { + "epoch": 6.204391891891892, + "grad_norm": 0.8420554399490356, + "learning_rate": 2.2029e-05, + "loss": 0.0454, + "step": 7346 + }, + { + "epoch": 6.205236486486487, + "grad_norm": 0.66948002576828, + "learning_rate": 2.2032000000000002e-05, + "loss": 0.039, + "step": 7347 + }, + { + "epoch": 6.206081081081081, + "grad_norm": 0.6431235074996948, + "learning_rate": 2.2035e-05, + "loss": 0.034, + "step": 7348 + }, + { + "epoch": 6.206925675675675, + "grad_norm": 0.7385426163673401, + "learning_rate": 2.2038e-05, + "loss": 0.0472, + "step": 7349 + }, + { + "epoch": 6.20777027027027, + "grad_norm": 0.6742839813232422, + "learning_rate": 2.2041e-05, + "loss": 0.0668, + "step": 7350 + }, + { + "epoch": 6.208614864864865, + "grad_norm": 0.5831263065338135, + "learning_rate": 2.2044e-05, + "loss": 0.0509, + "step": 7351 + }, + { + "epoch": 6.20945945945946, + "grad_norm": 0.794870138168335, + "learning_rate": 2.2047e-05, + "loss": 0.0675, + "step": 7352 + }, + { + "epoch": 6.210304054054054, + "grad_norm": 0.5966523289680481, + "learning_rate": 2.205e-05, + "loss": 0.0531, + "step": 7353 + }, + { + "epoch": 6.211148648648648, + "grad_norm": 1.8126721382141113, + "learning_rate": 2.2053e-05, + "loss": 0.1041, + "step": 7354 + }, + { + "epoch": 6.211993243243243, + "grad_norm": 2.2509753704071045, + "learning_rate": 2.2056e-05, + "loss": 0.4357, + "step": 7355 + }, + { + "epoch": 6.212837837837838, + "grad_norm": 0.7308059334754944, + "learning_rate": 2.2059e-05, + "loss": 0.1832, + "step": 7356 + }, + { + "epoch": 6.2136824324324325, + "grad_norm": 0.6994130611419678, + "learning_rate": 2.2062000000000003e-05, + "loss": 0.1216, + "step": 7357 + }, + { + "epoch": 6.214527027027027, + "grad_norm": 0.43585699796676636, + "learning_rate": 2.2065000000000003e-05, + "loss": 0.0682, + "step": 7358 + }, + { + "epoch": 6.215371621621622, + "grad_norm": 0.5425644516944885, + "learning_rate": 2.2068000000000003e-05, + "loss": 0.0539, + "step": 7359 + }, + { + "epoch": 6.216216216216216, + "grad_norm": 0.504546046257019, + "learning_rate": 2.2071000000000003e-05, + "loss": 0.0664, + "step": 7360 + }, + { + "epoch": 6.2170608108108105, + "grad_norm": 0.5274033546447754, + "learning_rate": 2.2074000000000002e-05, + "loss": 0.0625, + "step": 7361 + }, + { + "epoch": 6.217905405405405, + "grad_norm": 0.42088520526885986, + "learning_rate": 2.2077e-05, + "loss": 0.0498, + "step": 7362 + }, + { + "epoch": 6.21875, + "grad_norm": 0.33464789390563965, + "learning_rate": 2.208e-05, + "loss": 0.0455, + "step": 7363 + }, + { + "epoch": 6.219594594594595, + "grad_norm": 0.4426455497741699, + "learning_rate": 2.2083e-05, + "loss": 0.0413, + "step": 7364 + }, + { + "epoch": 6.2204391891891895, + "grad_norm": 0.3978762924671173, + "learning_rate": 2.2085999999999998e-05, + "loss": 0.0471, + "step": 7365 + }, + { + "epoch": 6.221283783783784, + "grad_norm": 0.526739776134491, + "learning_rate": 2.2088999999999998e-05, + "loss": 0.0502, + "step": 7366 + }, + { + "epoch": 6.222128378378378, + "grad_norm": 0.49023211002349854, + "learning_rate": 2.2092e-05, + "loss": 0.0474, + "step": 7367 + }, + { + "epoch": 6.222972972972973, + "grad_norm": 0.5317709445953369, + "learning_rate": 2.2095e-05, + "loss": 0.0452, + "step": 7368 + }, + { + "epoch": 6.2238175675675675, + "grad_norm": 2.3061952590942383, + "learning_rate": 2.2098e-05, + "loss": 0.0499, + "step": 7369 + }, + { + "epoch": 6.224662162162162, + "grad_norm": 0.6076472401618958, + "learning_rate": 2.2101e-05, + "loss": 0.04, + "step": 7370 + }, + { + "epoch": 6.225506756756757, + "grad_norm": 0.6010141372680664, + "learning_rate": 2.2104e-05, + "loss": 0.0739, + "step": 7371 + }, + { + "epoch": 6.226351351351352, + "grad_norm": 0.5735245943069458, + "learning_rate": 2.2107e-05, + "loss": 0.0505, + "step": 7372 + }, + { + "epoch": 6.227195945945946, + "grad_norm": 0.4079718291759491, + "learning_rate": 2.211e-05, + "loss": 0.0412, + "step": 7373 + }, + { + "epoch": 6.22804054054054, + "grad_norm": 0.473026841878891, + "learning_rate": 2.2113e-05, + "loss": 0.04, + "step": 7374 + }, + { + "epoch": 6.228885135135135, + "grad_norm": 0.3443799614906311, + "learning_rate": 2.2116e-05, + "loss": 0.0262, + "step": 7375 + }, + { + "epoch": 6.22972972972973, + "grad_norm": 0.36844658851623535, + "learning_rate": 2.2119e-05, + "loss": 0.0332, + "step": 7376 + }, + { + "epoch": 6.230574324324325, + "grad_norm": 0.3814883232116699, + "learning_rate": 2.2122000000000003e-05, + "loss": 0.042, + "step": 7377 + }, + { + "epoch": 6.231418918918919, + "grad_norm": 0.2909019887447357, + "learning_rate": 2.2125000000000002e-05, + "loss": 0.0231, + "step": 7378 + }, + { + "epoch": 6.232263513513513, + "grad_norm": 0.5028897523880005, + "learning_rate": 2.2128000000000002e-05, + "loss": 0.054, + "step": 7379 + }, + { + "epoch": 6.233108108108108, + "grad_norm": 0.42748647928237915, + "learning_rate": 2.2131000000000002e-05, + "loss": 0.0419, + "step": 7380 + }, + { + "epoch": 6.233952702702703, + "grad_norm": 0.511820375919342, + "learning_rate": 2.2134000000000002e-05, + "loss": 0.0435, + "step": 7381 + }, + { + "epoch": 6.234797297297297, + "grad_norm": 0.42213231325149536, + "learning_rate": 2.2137e-05, + "loss": 0.0343, + "step": 7382 + }, + { + "epoch": 6.235641891891892, + "grad_norm": 0.5419277548789978, + "learning_rate": 2.214e-05, + "loss": 0.0351, + "step": 7383 + }, + { + "epoch": 6.236486486486487, + "grad_norm": 0.7870531678199768, + "learning_rate": 2.2143e-05, + "loss": 0.0483, + "step": 7384 + }, + { + "epoch": 6.237331081081081, + "grad_norm": 0.44139230251312256, + "learning_rate": 2.2146e-05, + "loss": 0.0354, + "step": 7385 + }, + { + "epoch": 6.238175675675675, + "grad_norm": 0.8096404671669006, + "learning_rate": 2.2149e-05, + "loss": 0.0369, + "step": 7386 + }, + { + "epoch": 6.23902027027027, + "grad_norm": 0.5305107831954956, + "learning_rate": 2.2151999999999997e-05, + "loss": 0.0263, + "step": 7387 + }, + { + "epoch": 6.239864864864865, + "grad_norm": 0.41933032870292664, + "learning_rate": 2.2155e-05, + "loss": 0.0291, + "step": 7388 + }, + { + "epoch": 6.24070945945946, + "grad_norm": 0.4963532090187073, + "learning_rate": 2.2158e-05, + "loss": 0.0526, + "step": 7389 + }, + { + "epoch": 6.241554054054054, + "grad_norm": 0.422830194234848, + "learning_rate": 2.2161e-05, + "loss": 0.0377, + "step": 7390 + }, + { + "epoch": 6.242398648648648, + "grad_norm": 0.6565656065940857, + "learning_rate": 2.2164e-05, + "loss": 0.0461, + "step": 7391 + }, + { + "epoch": 6.243243243243243, + "grad_norm": 0.4333410859107971, + "learning_rate": 2.2167e-05, + "loss": 0.037, + "step": 7392 + }, + { + "epoch": 6.244087837837838, + "grad_norm": 0.7978221774101257, + "learning_rate": 2.217e-05, + "loss": 0.0425, + "step": 7393 + }, + { + "epoch": 6.2449324324324325, + "grad_norm": 0.4409162104129791, + "learning_rate": 2.2173e-05, + "loss": 0.0398, + "step": 7394 + }, + { + "epoch": 6.245777027027027, + "grad_norm": 0.5762070417404175, + "learning_rate": 2.2176e-05, + "loss": 0.0441, + "step": 7395 + }, + { + "epoch": 6.246621621621622, + "grad_norm": 0.5216937065124512, + "learning_rate": 2.2179e-05, + "loss": 0.0325, + "step": 7396 + }, + { + "epoch": 6.247466216216216, + "grad_norm": 1.3473607301712036, + "learning_rate": 2.2182e-05, + "loss": 0.0492, + "step": 7397 + }, + { + "epoch": 6.2483108108108105, + "grad_norm": 1.1191946268081665, + "learning_rate": 2.2185000000000002e-05, + "loss": 0.0414, + "step": 7398 + }, + { + "epoch": 6.249155405405405, + "grad_norm": 0.497972309589386, + "learning_rate": 2.2188e-05, + "loss": 0.0306, + "step": 7399 + }, + { + "epoch": 6.25, + "grad_norm": 0.645417332649231, + "learning_rate": 2.2191e-05, + "loss": 0.0528, + "step": 7400 + }, + { + "epoch": 6.250844594594595, + "grad_norm": 0.5966577529907227, + "learning_rate": 2.2194e-05, + "loss": 0.0476, + "step": 7401 + }, + { + "epoch": 6.2516891891891895, + "grad_norm": 0.5385844707489014, + "learning_rate": 2.2197e-05, + "loss": 0.0552, + "step": 7402 + }, + { + "epoch": 6.252533783783784, + "grad_norm": 0.7313559055328369, + "learning_rate": 2.22e-05, + "loss": 0.0687, + "step": 7403 + }, + { + "epoch": 6.253378378378378, + "grad_norm": 1.3081868886947632, + "learning_rate": 2.2203e-05, + "loss": 0.1022, + "step": 7404 + }, + { + "epoch": 6.254222972972973, + "grad_norm": 1.4736145734786987, + "learning_rate": 2.2206e-05, + "loss": 0.331, + "step": 7405 + }, + { + "epoch": 6.2550675675675675, + "grad_norm": 0.7967844605445862, + "learning_rate": 2.2209e-05, + "loss": 0.2074, + "step": 7406 + }, + { + "epoch": 6.255912162162162, + "grad_norm": 0.8154956102371216, + "learning_rate": 2.2212e-05, + "loss": 0.1222, + "step": 7407 + }, + { + "epoch": 6.256756756756757, + "grad_norm": 0.6720501184463501, + "learning_rate": 2.2215000000000003e-05, + "loss": 0.0665, + "step": 7408 + }, + { + "epoch": 6.257601351351352, + "grad_norm": 0.5939589142799377, + "learning_rate": 2.2218000000000003e-05, + "loss": 0.0658, + "step": 7409 + }, + { + "epoch": 6.258445945945946, + "grad_norm": 0.6367092728614807, + "learning_rate": 2.2221000000000003e-05, + "loss": 0.0517, + "step": 7410 + }, + { + "epoch": 6.25929054054054, + "grad_norm": 0.5055375695228577, + "learning_rate": 2.2224e-05, + "loss": 0.046, + "step": 7411 + }, + { + "epoch": 6.260135135135135, + "grad_norm": 0.500284731388092, + "learning_rate": 2.2227e-05, + "loss": 0.0486, + "step": 7412 + }, + { + "epoch": 6.26097972972973, + "grad_norm": 0.5335966944694519, + "learning_rate": 2.223e-05, + "loss": 0.0481, + "step": 7413 + }, + { + "epoch": 6.261824324324325, + "grad_norm": 0.45301225781440735, + "learning_rate": 2.2233e-05, + "loss": 0.0392, + "step": 7414 + }, + { + "epoch": 6.262668918918919, + "grad_norm": 0.4102776050567627, + "learning_rate": 2.2236e-05, + "loss": 0.0371, + "step": 7415 + }, + { + "epoch": 6.263513513513513, + "grad_norm": 0.4832525849342346, + "learning_rate": 2.2239e-05, + "loss": 0.0486, + "step": 7416 + }, + { + "epoch": 6.264358108108108, + "grad_norm": 0.5591051578521729, + "learning_rate": 2.2241999999999998e-05, + "loss": 0.0386, + "step": 7417 + }, + { + "epoch": 6.265202702702703, + "grad_norm": 0.4114435017108917, + "learning_rate": 2.2245e-05, + "loss": 0.0443, + "step": 7418 + }, + { + "epoch": 6.266047297297297, + "grad_norm": 0.34536996483802795, + "learning_rate": 2.2248e-05, + "loss": 0.0336, + "step": 7419 + }, + { + "epoch": 6.266891891891892, + "grad_norm": 0.3917091190814972, + "learning_rate": 2.2251e-05, + "loss": 0.0396, + "step": 7420 + }, + { + "epoch": 6.267736486486487, + "grad_norm": 0.7149558663368225, + "learning_rate": 2.2254e-05, + "loss": 0.0319, + "step": 7421 + }, + { + "epoch": 6.268581081081081, + "grad_norm": 0.4186626970767975, + "learning_rate": 2.2257e-05, + "loss": 0.0411, + "step": 7422 + }, + { + "epoch": 6.269425675675675, + "grad_norm": 0.48372578620910645, + "learning_rate": 2.226e-05, + "loss": 0.0288, + "step": 7423 + }, + { + "epoch": 6.27027027027027, + "grad_norm": 0.4189143776893616, + "learning_rate": 2.2263e-05, + "loss": 0.0433, + "step": 7424 + }, + { + "epoch": 6.271114864864865, + "grad_norm": 0.6110847592353821, + "learning_rate": 2.2266e-05, + "loss": 0.0752, + "step": 7425 + }, + { + "epoch": 6.27195945945946, + "grad_norm": 0.5584635138511658, + "learning_rate": 2.2269e-05, + "loss": 0.0332, + "step": 7426 + }, + { + "epoch": 6.272804054054054, + "grad_norm": 0.49427372217178345, + "learning_rate": 2.2272e-05, + "loss": 0.0456, + "step": 7427 + }, + { + "epoch": 6.273648648648648, + "grad_norm": 0.5771967172622681, + "learning_rate": 2.2275000000000003e-05, + "loss": 0.0543, + "step": 7428 + }, + { + "epoch": 6.274493243243243, + "grad_norm": 0.36759817600250244, + "learning_rate": 2.2278000000000003e-05, + "loss": 0.0263, + "step": 7429 + }, + { + "epoch": 6.275337837837838, + "grad_norm": 0.4171786904335022, + "learning_rate": 2.2281000000000002e-05, + "loss": 0.0429, + "step": 7430 + }, + { + "epoch": 6.2761824324324325, + "grad_norm": 0.47359830141067505, + "learning_rate": 2.2284000000000002e-05, + "loss": 0.0399, + "step": 7431 + }, + { + "epoch": 6.277027027027027, + "grad_norm": 0.45260554552078247, + "learning_rate": 2.2287000000000002e-05, + "loss": 0.0505, + "step": 7432 + }, + { + "epoch": 6.277871621621622, + "grad_norm": 0.4287358522415161, + "learning_rate": 2.2290000000000002e-05, + "loss": 0.0378, + "step": 7433 + }, + { + "epoch": 6.278716216216216, + "grad_norm": 0.6022997498512268, + "learning_rate": 2.2293e-05, + "loss": 0.0516, + "step": 7434 + }, + { + "epoch": 6.2795608108108105, + "grad_norm": 0.3141004741191864, + "learning_rate": 2.2296e-05, + "loss": 0.0344, + "step": 7435 + }, + { + "epoch": 6.280405405405405, + "grad_norm": 0.5192921757698059, + "learning_rate": 2.2298999999999998e-05, + "loss": 0.0386, + "step": 7436 + }, + { + "epoch": 6.28125, + "grad_norm": 0.5436587929725647, + "learning_rate": 2.2301999999999998e-05, + "loss": 0.0373, + "step": 7437 + }, + { + "epoch": 6.282094594594595, + "grad_norm": 0.6435862183570862, + "learning_rate": 2.2305e-05, + "loss": 0.0576, + "step": 7438 + }, + { + "epoch": 6.2829391891891895, + "grad_norm": 0.5621216297149658, + "learning_rate": 2.2308e-05, + "loss": 0.0529, + "step": 7439 + }, + { + "epoch": 6.283783783783784, + "grad_norm": 0.35978463292121887, + "learning_rate": 2.2311e-05, + "loss": 0.0249, + "step": 7440 + }, + { + "epoch": 6.284628378378378, + "grad_norm": 0.566673994064331, + "learning_rate": 2.2314e-05, + "loss": 0.0367, + "step": 7441 + }, + { + "epoch": 6.285472972972973, + "grad_norm": 0.33678656816482544, + "learning_rate": 2.2317e-05, + "loss": 0.0292, + "step": 7442 + }, + { + "epoch": 6.2863175675675675, + "grad_norm": 0.5137674808502197, + "learning_rate": 2.232e-05, + "loss": 0.0339, + "step": 7443 + }, + { + "epoch": 6.287162162162162, + "grad_norm": 0.6146572828292847, + "learning_rate": 2.2323e-05, + "loss": 0.0614, + "step": 7444 + }, + { + "epoch": 6.288006756756757, + "grad_norm": 0.4884694218635559, + "learning_rate": 2.2326e-05, + "loss": 0.0422, + "step": 7445 + }, + { + "epoch": 6.288851351351352, + "grad_norm": 0.5302731990814209, + "learning_rate": 2.2329e-05, + "loss": 0.0423, + "step": 7446 + }, + { + "epoch": 6.289695945945946, + "grad_norm": 0.7155188322067261, + "learning_rate": 2.2332e-05, + "loss": 0.048, + "step": 7447 + }, + { + "epoch": 6.29054054054054, + "grad_norm": 0.4184887707233429, + "learning_rate": 2.2335000000000002e-05, + "loss": 0.0314, + "step": 7448 + }, + { + "epoch": 6.291385135135135, + "grad_norm": 0.5305351614952087, + "learning_rate": 2.2338000000000002e-05, + "loss": 0.0422, + "step": 7449 + }, + { + "epoch": 6.29222972972973, + "grad_norm": 0.5536882281303406, + "learning_rate": 2.2341000000000002e-05, + "loss": 0.0375, + "step": 7450 + }, + { + "epoch": 6.293074324324325, + "grad_norm": 1.0027408599853516, + "learning_rate": 2.2344e-05, + "loss": 0.0835, + "step": 7451 + }, + { + "epoch": 6.293918918918919, + "grad_norm": 0.8658357858657837, + "learning_rate": 2.2347e-05, + "loss": 0.0303, + "step": 7452 + }, + { + "epoch": 6.294763513513513, + "grad_norm": 0.8572049736976624, + "learning_rate": 2.235e-05, + "loss": 0.0686, + "step": 7453 + }, + { + "epoch": 6.295608108108108, + "grad_norm": 0.7958404421806335, + "learning_rate": 2.2353e-05, + "loss": 0.0636, + "step": 7454 + }, + { + "epoch": 6.296452702702703, + "grad_norm": 0.8934358358383179, + "learning_rate": 2.2356e-05, + "loss": 0.2951, + "step": 7455 + }, + { + "epoch": 6.297297297297297, + "grad_norm": 0.8383619785308838, + "learning_rate": 2.2359e-05, + "loss": 0.2105, + "step": 7456 + }, + { + "epoch": 6.298141891891892, + "grad_norm": 0.4927704930305481, + "learning_rate": 2.2362e-05, + "loss": 0.1146, + "step": 7457 + }, + { + "epoch": 6.298986486486487, + "grad_norm": 0.4215516448020935, + "learning_rate": 2.2365000000000004e-05, + "loss": 0.0533, + "step": 7458 + }, + { + "epoch": 6.299831081081081, + "grad_norm": 0.42987969517707825, + "learning_rate": 2.2368000000000003e-05, + "loss": 0.0578, + "step": 7459 + }, + { + "epoch": 6.300675675675675, + "grad_norm": 0.418258398771286, + "learning_rate": 2.2371e-05, + "loss": 0.073, + "step": 7460 + }, + { + "epoch": 6.30152027027027, + "grad_norm": 0.58125239610672, + "learning_rate": 2.2374e-05, + "loss": 0.0629, + "step": 7461 + }, + { + "epoch": 6.302364864864865, + "grad_norm": 0.5522691607475281, + "learning_rate": 2.2377e-05, + "loss": 0.0573, + "step": 7462 + }, + { + "epoch": 6.30320945945946, + "grad_norm": 0.4825402796268463, + "learning_rate": 2.238e-05, + "loss": 0.0468, + "step": 7463 + }, + { + "epoch": 6.304054054054054, + "grad_norm": 0.2933714687824249, + "learning_rate": 2.2383e-05, + "loss": 0.0343, + "step": 7464 + }, + { + "epoch": 6.304898648648648, + "grad_norm": 0.2827019691467285, + "learning_rate": 2.2386e-05, + "loss": 0.0292, + "step": 7465 + }, + { + "epoch": 6.305743243243243, + "grad_norm": 0.5139277577400208, + "learning_rate": 2.2389e-05, + "loss": 0.0401, + "step": 7466 + }, + { + "epoch": 6.306587837837838, + "grad_norm": 0.3762552738189697, + "learning_rate": 2.2392e-05, + "loss": 0.0248, + "step": 7467 + }, + { + "epoch": 6.3074324324324325, + "grad_norm": 0.35279110074043274, + "learning_rate": 2.2395e-05, + "loss": 0.0306, + "step": 7468 + }, + { + "epoch": 6.308277027027027, + "grad_norm": 0.3147989511489868, + "learning_rate": 2.2398e-05, + "loss": 0.0416, + "step": 7469 + }, + { + "epoch": 6.309121621621622, + "grad_norm": 0.5266231894493103, + "learning_rate": 2.2401e-05, + "loss": 0.0444, + "step": 7470 + }, + { + "epoch": 6.309966216216216, + "grad_norm": 0.7219563722610474, + "learning_rate": 2.2404e-05, + "loss": 0.0479, + "step": 7471 + }, + { + "epoch": 6.3108108108108105, + "grad_norm": 0.4175613820552826, + "learning_rate": 2.2407e-05, + "loss": 0.0397, + "step": 7472 + }, + { + "epoch": 6.311655405405405, + "grad_norm": 0.5892776846885681, + "learning_rate": 2.241e-05, + "loss": 0.0508, + "step": 7473 + }, + { + "epoch": 6.3125, + "grad_norm": 0.7039805054664612, + "learning_rate": 2.2413e-05, + "loss": 0.0429, + "step": 7474 + }, + { + "epoch": 6.313344594594595, + "grad_norm": 0.4407186508178711, + "learning_rate": 2.2416e-05, + "loss": 0.047, + "step": 7475 + }, + { + "epoch": 6.3141891891891895, + "grad_norm": 0.2962958812713623, + "learning_rate": 2.2419e-05, + "loss": 0.0277, + "step": 7476 + }, + { + "epoch": 6.315033783783784, + "grad_norm": 0.4799213111400604, + "learning_rate": 2.2422e-05, + "loss": 0.0463, + "step": 7477 + }, + { + "epoch": 6.315878378378378, + "grad_norm": 0.4931034743785858, + "learning_rate": 2.2425000000000003e-05, + "loss": 0.0489, + "step": 7478 + }, + { + "epoch": 6.316722972972973, + "grad_norm": 0.49522697925567627, + "learning_rate": 2.2428000000000003e-05, + "loss": 0.0335, + "step": 7479 + }, + { + "epoch": 6.3175675675675675, + "grad_norm": 0.3432067036628723, + "learning_rate": 2.2431000000000003e-05, + "loss": 0.0342, + "step": 7480 + }, + { + "epoch": 6.318412162162162, + "grad_norm": 0.4145641028881073, + "learning_rate": 2.2434000000000002e-05, + "loss": 0.0274, + "step": 7481 + }, + { + "epoch": 6.319256756756757, + "grad_norm": 0.4783467948436737, + "learning_rate": 2.2437000000000002e-05, + "loss": 0.043, + "step": 7482 + }, + { + "epoch": 6.320101351351352, + "grad_norm": 0.43737050890922546, + "learning_rate": 2.2440000000000002e-05, + "loss": 0.0346, + "step": 7483 + }, + { + "epoch": 6.320945945945946, + "grad_norm": 0.6362435817718506, + "learning_rate": 2.2443000000000002e-05, + "loss": 0.0646, + "step": 7484 + }, + { + "epoch": 6.32179054054054, + "grad_norm": 0.5999560952186584, + "learning_rate": 2.2445999999999998e-05, + "loss": 0.0488, + "step": 7485 + }, + { + "epoch": 6.322635135135135, + "grad_norm": 0.577404797077179, + "learning_rate": 2.2448999999999998e-05, + "loss": 0.0412, + "step": 7486 + }, + { + "epoch": 6.32347972972973, + "grad_norm": 0.39050909876823425, + "learning_rate": 2.2451999999999998e-05, + "loss": 0.0386, + "step": 7487 + }, + { + "epoch": 6.324324324324325, + "grad_norm": 0.6703923940658569, + "learning_rate": 2.2455e-05, + "loss": 0.0505, + "step": 7488 + }, + { + "epoch": 6.325168918918919, + "grad_norm": 0.5211614966392517, + "learning_rate": 2.2458e-05, + "loss": 0.0361, + "step": 7489 + }, + { + "epoch": 6.326013513513513, + "grad_norm": 0.6340376734733582, + "learning_rate": 2.2461e-05, + "loss": 0.0334, + "step": 7490 + }, + { + "epoch": 6.326858108108108, + "grad_norm": 0.6459017395973206, + "learning_rate": 2.2464e-05, + "loss": 0.0491, + "step": 7491 + }, + { + "epoch": 6.327702702702703, + "grad_norm": 1.0236971378326416, + "learning_rate": 2.2467e-05, + "loss": 0.0531, + "step": 7492 + }, + { + "epoch": 6.328547297297297, + "grad_norm": 0.5249583721160889, + "learning_rate": 2.247e-05, + "loss": 0.0396, + "step": 7493 + }, + { + "epoch": 6.329391891891892, + "grad_norm": 0.5142756104469299, + "learning_rate": 2.2473e-05, + "loss": 0.05, + "step": 7494 + }, + { + "epoch": 6.330236486486487, + "grad_norm": 0.6819520592689514, + "learning_rate": 2.2476e-05, + "loss": 0.0508, + "step": 7495 + }, + { + "epoch": 6.331081081081081, + "grad_norm": 0.5383079051971436, + "learning_rate": 2.2479e-05, + "loss": 0.0403, + "step": 7496 + }, + { + "epoch": 6.331925675675675, + "grad_norm": 0.5647111535072327, + "learning_rate": 2.2482e-05, + "loss": 0.0638, + "step": 7497 + }, + { + "epoch": 6.33277027027027, + "grad_norm": 0.8020392060279846, + "learning_rate": 2.2485000000000002e-05, + "loss": 0.0566, + "step": 7498 + }, + { + "epoch": 6.333614864864865, + "grad_norm": 0.8523931503295898, + "learning_rate": 2.2488000000000002e-05, + "loss": 0.0479, + "step": 7499 + }, + { + "epoch": 6.33445945945946, + "grad_norm": 0.7352905869483948, + "learning_rate": 2.2491000000000002e-05, + "loss": 0.0566, + "step": 7500 + }, + { + "epoch": 6.335304054054054, + "grad_norm": 0.6873253583908081, + "learning_rate": 2.2494000000000002e-05, + "loss": 0.0568, + "step": 7501 + }, + { + "epoch": 6.336148648648648, + "grad_norm": 0.5765884518623352, + "learning_rate": 2.2497e-05, + "loss": 0.0548, + "step": 7502 + }, + { + "epoch": 6.336993243243243, + "grad_norm": 0.8660240769386292, + "learning_rate": 2.25e-05, + "loss": 0.0668, + "step": 7503 + }, + { + "epoch": 6.337837837837838, + "grad_norm": 1.7742125988006592, + "learning_rate": 2.2503e-05, + "loss": 0.1046, + "step": 7504 + }, + { + "epoch": 6.3386824324324325, + "grad_norm": 1.319542646408081, + "learning_rate": 2.2506e-05, + "loss": 0.3513, + "step": 7505 + }, + { + "epoch": 6.339527027027027, + "grad_norm": 1.194499135017395, + "learning_rate": 2.2509e-05, + "loss": 0.2859, + "step": 7506 + }, + { + "epoch": 6.340371621621622, + "grad_norm": 0.7083882093429565, + "learning_rate": 2.2512e-05, + "loss": 0.156, + "step": 7507 + }, + { + "epoch": 6.341216216216216, + "grad_norm": 0.5185095071792603, + "learning_rate": 2.2515e-05, + "loss": 0.0655, + "step": 7508 + }, + { + "epoch": 6.3420608108108105, + "grad_norm": 0.43137601017951965, + "learning_rate": 2.2518e-05, + "loss": 0.0696, + "step": 7509 + }, + { + "epoch": 6.342905405405405, + "grad_norm": 0.5256198644638062, + "learning_rate": 2.2521e-05, + "loss": 0.075, + "step": 7510 + }, + { + "epoch": 6.34375, + "grad_norm": 0.4373866617679596, + "learning_rate": 2.2524e-05, + "loss": 0.0497, + "step": 7511 + }, + { + "epoch": 6.344594594594595, + "grad_norm": 0.4795936346054077, + "learning_rate": 2.2527e-05, + "loss": 0.0477, + "step": 7512 + }, + { + "epoch": 6.3454391891891895, + "grad_norm": 0.4222911596298218, + "learning_rate": 2.253e-05, + "loss": 0.0517, + "step": 7513 + }, + { + "epoch": 6.346283783783784, + "grad_norm": 0.32863694429397583, + "learning_rate": 2.2533e-05, + "loss": 0.03, + "step": 7514 + }, + { + "epoch": 6.347128378378378, + "grad_norm": 0.3713842034339905, + "learning_rate": 2.2536e-05, + "loss": 0.0351, + "step": 7515 + }, + { + "epoch": 6.347972972972973, + "grad_norm": 0.5173021554946899, + "learning_rate": 2.2539e-05, + "loss": 0.0415, + "step": 7516 + }, + { + "epoch": 6.3488175675675675, + "grad_norm": 0.42150744795799255, + "learning_rate": 2.2542e-05, + "loss": 0.0409, + "step": 7517 + }, + { + "epoch": 6.349662162162162, + "grad_norm": 0.37855690717697144, + "learning_rate": 2.2545e-05, + "loss": 0.0339, + "step": 7518 + }, + { + "epoch": 6.350506756756757, + "grad_norm": 0.9102452993392944, + "learning_rate": 2.2548e-05, + "loss": 0.0814, + "step": 7519 + }, + { + "epoch": 6.351351351351352, + "grad_norm": 0.6812641024589539, + "learning_rate": 2.2551e-05, + "loss": 0.064, + "step": 7520 + }, + { + "epoch": 6.352195945945946, + "grad_norm": 0.5158823132514954, + "learning_rate": 2.2554e-05, + "loss": 0.0248, + "step": 7521 + }, + { + "epoch": 6.35304054054054, + "grad_norm": 0.6863649487495422, + "learning_rate": 2.2557e-05, + "loss": 0.0453, + "step": 7522 + }, + { + "epoch": 6.353885135135135, + "grad_norm": 0.48326534032821655, + "learning_rate": 2.256e-05, + "loss": 0.0309, + "step": 7523 + }, + { + "epoch": 6.35472972972973, + "grad_norm": 0.856892466545105, + "learning_rate": 2.2563e-05, + "loss": 0.0628, + "step": 7524 + }, + { + "epoch": 6.355574324324325, + "grad_norm": 0.5710940957069397, + "learning_rate": 2.2566e-05, + "loss": 0.0388, + "step": 7525 + }, + { + "epoch": 6.356418918918919, + "grad_norm": 0.48855260014533997, + "learning_rate": 2.2569e-05, + "loss": 0.0438, + "step": 7526 + }, + { + "epoch": 6.357263513513513, + "grad_norm": 0.38429591059684753, + "learning_rate": 2.2572e-05, + "loss": 0.0331, + "step": 7527 + }, + { + "epoch": 6.358108108108108, + "grad_norm": 0.5911156535148621, + "learning_rate": 2.2575e-05, + "loss": 0.0524, + "step": 7528 + }, + { + "epoch": 6.358952702702703, + "grad_norm": 0.37075260281562805, + "learning_rate": 2.2578000000000003e-05, + "loss": 0.0419, + "step": 7529 + }, + { + "epoch": 6.359797297297297, + "grad_norm": 0.33340784907341003, + "learning_rate": 2.2581000000000003e-05, + "loss": 0.0344, + "step": 7530 + }, + { + "epoch": 6.360641891891892, + "grad_norm": 0.3697618842124939, + "learning_rate": 2.2584000000000003e-05, + "loss": 0.0294, + "step": 7531 + }, + { + "epoch": 6.361486486486487, + "grad_norm": 0.42075639963150024, + "learning_rate": 2.2587000000000002e-05, + "loss": 0.0313, + "step": 7532 + }, + { + "epoch": 6.362331081081081, + "grad_norm": 0.2879771888256073, + "learning_rate": 2.2590000000000002e-05, + "loss": 0.0285, + "step": 7533 + }, + { + "epoch": 6.363175675675675, + "grad_norm": 0.5750707983970642, + "learning_rate": 2.2593e-05, + "loss": 0.0571, + "step": 7534 + }, + { + "epoch": 6.36402027027027, + "grad_norm": 0.5361441969871521, + "learning_rate": 2.2596e-05, + "loss": 0.0405, + "step": 7535 + }, + { + "epoch": 6.364864864864865, + "grad_norm": 0.35820892453193665, + "learning_rate": 2.2598999999999998e-05, + "loss": 0.0432, + "step": 7536 + }, + { + "epoch": 6.36570945945946, + "grad_norm": 0.44154977798461914, + "learning_rate": 2.2601999999999998e-05, + "loss": 0.0402, + "step": 7537 + }, + { + "epoch": 6.366554054054054, + "grad_norm": 0.6710176467895508, + "learning_rate": 2.2604999999999998e-05, + "loss": 0.0772, + "step": 7538 + }, + { + "epoch": 6.367398648648648, + "grad_norm": 0.4802284240722656, + "learning_rate": 2.2608e-05, + "loss": 0.0337, + "step": 7539 + }, + { + "epoch": 6.368243243243243, + "grad_norm": 0.38020455837249756, + "learning_rate": 2.2611e-05, + "loss": 0.0292, + "step": 7540 + }, + { + "epoch": 6.369087837837838, + "grad_norm": 0.4477778971195221, + "learning_rate": 2.2614e-05, + "loss": 0.0313, + "step": 7541 + }, + { + "epoch": 6.3699324324324325, + "grad_norm": 0.6034426689147949, + "learning_rate": 2.2617e-05, + "loss": 0.0641, + "step": 7542 + }, + { + "epoch": 6.370777027027027, + "grad_norm": 0.49812012910842896, + "learning_rate": 2.262e-05, + "loss": 0.0503, + "step": 7543 + }, + { + "epoch": 6.371621621621622, + "grad_norm": 0.6583446264266968, + "learning_rate": 2.2623e-05, + "loss": 0.0425, + "step": 7544 + }, + { + "epoch": 6.372466216216216, + "grad_norm": 0.8407129049301147, + "learning_rate": 2.2626e-05, + "loss": 0.0305, + "step": 7545 + }, + { + "epoch": 6.3733108108108105, + "grad_norm": 0.569472074508667, + "learning_rate": 2.2629e-05, + "loss": 0.0463, + "step": 7546 + }, + { + "epoch": 6.374155405405405, + "grad_norm": 0.5414562225341797, + "learning_rate": 2.2632e-05, + "loss": 0.0531, + "step": 7547 + }, + { + "epoch": 6.375, + "grad_norm": 0.6547749638557434, + "learning_rate": 2.2635e-05, + "loss": 0.046, + "step": 7548 + }, + { + "epoch": 6.375844594594595, + "grad_norm": 0.649823784828186, + "learning_rate": 2.2638000000000002e-05, + "loss": 0.0397, + "step": 7549 + }, + { + "epoch": 6.3766891891891895, + "grad_norm": 1.028628945350647, + "learning_rate": 2.2641000000000002e-05, + "loss": 0.085, + "step": 7550 + }, + { + "epoch": 6.377533783783784, + "grad_norm": 0.7732828855514526, + "learning_rate": 2.2644000000000002e-05, + "loss": 0.0365, + "step": 7551 + }, + { + "epoch": 6.378378378378378, + "grad_norm": 1.039281964302063, + "learning_rate": 2.2647000000000002e-05, + "loss": 0.062, + "step": 7552 + }, + { + "epoch": 6.379222972972973, + "grad_norm": 0.8084739446640015, + "learning_rate": 2.265e-05, + "loss": 0.0729, + "step": 7553 + }, + { + "epoch": 6.3800675675675675, + "grad_norm": 1.3197283744812012, + "learning_rate": 2.2653e-05, + "loss": 0.0994, + "step": 7554 + }, + { + "epoch": 6.380912162162162, + "grad_norm": 1.4743988513946533, + "learning_rate": 2.2656e-05, + "loss": 0.3263, + "step": 7555 + }, + { + "epoch": 6.381756756756757, + "grad_norm": 0.619223952293396, + "learning_rate": 2.2659e-05, + "loss": 0.1495, + "step": 7556 + }, + { + "epoch": 6.382601351351352, + "grad_norm": 0.5773822665214539, + "learning_rate": 2.2662e-05, + "loss": 0.114, + "step": 7557 + }, + { + "epoch": 6.383445945945946, + "grad_norm": 0.5804970264434814, + "learning_rate": 2.2665e-05, + "loss": 0.073, + "step": 7558 + }, + { + "epoch": 6.38429054054054, + "grad_norm": 0.45818060636520386, + "learning_rate": 2.2668e-05, + "loss": 0.0541, + "step": 7559 + }, + { + "epoch": 6.385135135135135, + "grad_norm": 0.6693601608276367, + "learning_rate": 2.2671e-05, + "loss": 0.0455, + "step": 7560 + }, + { + "epoch": 6.38597972972973, + "grad_norm": 0.41461020708084106, + "learning_rate": 2.2674e-05, + "loss": 0.0509, + "step": 7561 + }, + { + "epoch": 6.386824324324325, + "grad_norm": 0.5318171381950378, + "learning_rate": 2.2677e-05, + "loss": 0.0651, + "step": 7562 + }, + { + "epoch": 6.387668918918919, + "grad_norm": 0.5305613875389099, + "learning_rate": 2.268e-05, + "loss": 0.0317, + "step": 7563 + }, + { + "epoch": 6.388513513513513, + "grad_norm": 0.35518237948417664, + "learning_rate": 2.2683e-05, + "loss": 0.0386, + "step": 7564 + }, + { + "epoch": 6.389358108108108, + "grad_norm": 0.46549341082572937, + "learning_rate": 2.2686e-05, + "loss": 0.0669, + "step": 7565 + }, + { + "epoch": 6.390202702702703, + "grad_norm": 0.37480008602142334, + "learning_rate": 2.2689e-05, + "loss": 0.0333, + "step": 7566 + }, + { + "epoch": 6.391047297297297, + "grad_norm": 0.43655481934547424, + "learning_rate": 2.2692e-05, + "loss": 0.0471, + "step": 7567 + }, + { + "epoch": 6.391891891891892, + "grad_norm": 0.5156072974205017, + "learning_rate": 2.2695e-05, + "loss": 0.0496, + "step": 7568 + }, + { + "epoch": 6.392736486486487, + "grad_norm": 0.4442504048347473, + "learning_rate": 2.2698000000000002e-05, + "loss": 0.0415, + "step": 7569 + }, + { + "epoch": 6.393581081081081, + "grad_norm": 0.3603479564189911, + "learning_rate": 2.2701000000000002e-05, + "loss": 0.0462, + "step": 7570 + }, + { + "epoch": 6.394425675675675, + "grad_norm": 0.3777558207511902, + "learning_rate": 2.2704e-05, + "loss": 0.0309, + "step": 7571 + }, + { + "epoch": 6.39527027027027, + "grad_norm": 0.3175908923149109, + "learning_rate": 2.2707e-05, + "loss": 0.0302, + "step": 7572 + }, + { + "epoch": 6.396114864864865, + "grad_norm": 0.38674014806747437, + "learning_rate": 2.271e-05, + "loss": 0.0376, + "step": 7573 + }, + { + "epoch": 6.39695945945946, + "grad_norm": 1.0826166868209839, + "learning_rate": 2.2713e-05, + "loss": 0.182, + "step": 7574 + }, + { + "epoch": 6.397804054054054, + "grad_norm": 0.4685635268688202, + "learning_rate": 2.2716e-05, + "loss": 0.0341, + "step": 7575 + }, + { + "epoch": 6.398648648648648, + "grad_norm": 0.48511916399002075, + "learning_rate": 2.2719e-05, + "loss": 0.0324, + "step": 7576 + }, + { + "epoch": 6.399493243243243, + "grad_norm": 0.4488259553909302, + "learning_rate": 2.2722e-05, + "loss": 0.0448, + "step": 7577 + }, + { + "epoch": 6.400337837837838, + "grad_norm": 0.5120110511779785, + "learning_rate": 2.2725e-05, + "loss": 0.0327, + "step": 7578 + }, + { + "epoch": 6.4011824324324325, + "grad_norm": 0.35014817118644714, + "learning_rate": 2.2728000000000003e-05, + "loss": 0.0302, + "step": 7579 + }, + { + "epoch": 6.402027027027027, + "grad_norm": 0.641083836555481, + "learning_rate": 2.2731000000000003e-05, + "loss": 0.0541, + "step": 7580 + }, + { + "epoch": 6.402871621621622, + "grad_norm": 0.34008124470710754, + "learning_rate": 2.2734000000000003e-05, + "loss": 0.0389, + "step": 7581 + }, + { + "epoch": 6.403716216216216, + "grad_norm": 0.44344890117645264, + "learning_rate": 2.2737000000000003e-05, + "loss": 0.0512, + "step": 7582 + }, + { + "epoch": 6.4045608108108105, + "grad_norm": 0.40876340866088867, + "learning_rate": 2.274e-05, + "loss": 0.032, + "step": 7583 + }, + { + "epoch": 6.405405405405405, + "grad_norm": 0.49636775255203247, + "learning_rate": 2.2743e-05, + "loss": 0.0325, + "step": 7584 + }, + { + "epoch": 6.40625, + "grad_norm": 0.37288185954093933, + "learning_rate": 2.2746e-05, + "loss": 0.0305, + "step": 7585 + }, + { + "epoch": 6.407094594594595, + "grad_norm": 0.7611053586006165, + "learning_rate": 2.2749e-05, + "loss": 0.0565, + "step": 7586 + }, + { + "epoch": 6.4079391891891895, + "grad_norm": 0.48055192828178406, + "learning_rate": 2.2752e-05, + "loss": 0.0337, + "step": 7587 + }, + { + "epoch": 6.408783783783784, + "grad_norm": 0.3658534288406372, + "learning_rate": 2.2754999999999998e-05, + "loss": 0.0298, + "step": 7588 + }, + { + "epoch": 6.409628378378378, + "grad_norm": 0.39493733644485474, + "learning_rate": 2.2758e-05, + "loss": 0.0309, + "step": 7589 + }, + { + "epoch": 6.410472972972973, + "grad_norm": 0.7311388254165649, + "learning_rate": 2.2761e-05, + "loss": 0.0407, + "step": 7590 + }, + { + "epoch": 6.4113175675675675, + "grad_norm": 0.5944271087646484, + "learning_rate": 2.2764e-05, + "loss": 0.0609, + "step": 7591 + }, + { + "epoch": 6.412162162162162, + "grad_norm": 0.5875877737998962, + "learning_rate": 2.2767e-05, + "loss": 0.0938, + "step": 7592 + }, + { + "epoch": 6.413006756756757, + "grad_norm": 0.4421088397502899, + "learning_rate": 2.277e-05, + "loss": 0.0404, + "step": 7593 + }, + { + "epoch": 6.413851351351352, + "grad_norm": 0.6407526135444641, + "learning_rate": 2.2773e-05, + "loss": 0.0439, + "step": 7594 + }, + { + "epoch": 6.414695945945946, + "grad_norm": 0.32436951994895935, + "learning_rate": 2.2776e-05, + "loss": 0.0192, + "step": 7595 + }, + { + "epoch": 6.41554054054054, + "grad_norm": 0.3847612142562866, + "learning_rate": 2.2779e-05, + "loss": 0.0271, + "step": 7596 + }, + { + "epoch": 6.416385135135135, + "grad_norm": 0.47022178769111633, + "learning_rate": 2.2782e-05, + "loss": 0.0449, + "step": 7597 + }, + { + "epoch": 6.41722972972973, + "grad_norm": 0.7477688789367676, + "learning_rate": 2.2785e-05, + "loss": 0.0537, + "step": 7598 + }, + { + "epoch": 6.418074324324325, + "grad_norm": 0.7018906474113464, + "learning_rate": 2.2788000000000003e-05, + "loss": 0.0679, + "step": 7599 + }, + { + "epoch": 6.418918918918919, + "grad_norm": 0.9371423125267029, + "learning_rate": 2.2791000000000003e-05, + "loss": 0.0508, + "step": 7600 + }, + { + "epoch": 6.419763513513513, + "grad_norm": 0.5434965491294861, + "learning_rate": 2.2794000000000002e-05, + "loss": 0.0437, + "step": 7601 + }, + { + "epoch": 6.420608108108108, + "grad_norm": 0.7825470566749573, + "learning_rate": 2.2797000000000002e-05, + "loss": 0.0489, + "step": 7602 + }, + { + "epoch": 6.421452702702703, + "grad_norm": 0.652159571647644, + "learning_rate": 2.2800000000000002e-05, + "loss": 0.0597, + "step": 7603 + }, + { + "epoch": 6.422297297297297, + "grad_norm": 0.7564002275466919, + "learning_rate": 2.2803000000000002e-05, + "loss": 0.0689, + "step": 7604 + }, + { + "epoch": 6.423141891891892, + "grad_norm": 1.2166898250579834, + "learning_rate": 2.2806e-05, + "loss": 0.3869, + "step": 7605 + }, + { + "epoch": 6.423986486486487, + "grad_norm": 0.8775268197059631, + "learning_rate": 2.2809e-05, + "loss": 0.2261, + "step": 7606 + }, + { + "epoch": 6.424831081081081, + "grad_norm": 0.6831047534942627, + "learning_rate": 2.2812e-05, + "loss": 0.1658, + "step": 7607 + }, + { + "epoch": 6.425675675675675, + "grad_norm": 0.764782190322876, + "learning_rate": 2.2814999999999998e-05, + "loss": 0.1097, + "step": 7608 + }, + { + "epoch": 6.42652027027027, + "grad_norm": 0.36399945616722107, + "learning_rate": 2.2818e-05, + "loss": 0.0556, + "step": 7609 + }, + { + "epoch": 6.427364864864865, + "grad_norm": 0.489568293094635, + "learning_rate": 2.2821e-05, + "loss": 0.0692, + "step": 7610 + }, + { + "epoch": 6.42820945945946, + "grad_norm": 0.4142359495162964, + "learning_rate": 2.2824e-05, + "loss": 0.0561, + "step": 7611 + }, + { + "epoch": 6.429054054054054, + "grad_norm": 0.5733292102813721, + "learning_rate": 2.2827e-05, + "loss": 0.045, + "step": 7612 + }, + { + "epoch": 6.429898648648648, + "grad_norm": 0.4601961374282837, + "learning_rate": 2.283e-05, + "loss": 0.0493, + "step": 7613 + }, + { + "epoch": 6.430743243243243, + "grad_norm": 0.44900161027908325, + "learning_rate": 2.2833e-05, + "loss": 0.0453, + "step": 7614 + }, + { + "epoch": 6.431587837837838, + "grad_norm": 0.4303470849990845, + "learning_rate": 2.2836e-05, + "loss": 0.0426, + "step": 7615 + }, + { + "epoch": 6.4324324324324325, + "grad_norm": 0.4005585014820099, + "learning_rate": 2.2839e-05, + "loss": 0.0417, + "step": 7616 + }, + { + "epoch": 6.433277027027027, + "grad_norm": 0.4285769462585449, + "learning_rate": 2.2842e-05, + "loss": 0.033, + "step": 7617 + }, + { + "epoch": 6.434121621621622, + "grad_norm": 0.48861944675445557, + "learning_rate": 2.2845e-05, + "loss": 0.0465, + "step": 7618 + }, + { + "epoch": 6.434966216216216, + "grad_norm": 0.4745907485485077, + "learning_rate": 2.2848000000000002e-05, + "loss": 0.0519, + "step": 7619 + }, + { + "epoch": 6.4358108108108105, + "grad_norm": 0.3448898494243622, + "learning_rate": 2.2851000000000002e-05, + "loss": 0.0319, + "step": 7620 + }, + { + "epoch": 6.436655405405405, + "grad_norm": 0.42125383019447327, + "learning_rate": 2.2854000000000002e-05, + "loss": 0.0337, + "step": 7621 + }, + { + "epoch": 6.4375, + "grad_norm": 0.4445746839046478, + "learning_rate": 2.2857e-05, + "loss": 0.0292, + "step": 7622 + }, + { + "epoch": 6.438344594594595, + "grad_norm": 0.5552256107330322, + "learning_rate": 2.286e-05, + "loss": 0.0617, + "step": 7623 + }, + { + "epoch": 6.4391891891891895, + "grad_norm": 0.8001555800437927, + "learning_rate": 2.2863e-05, + "loss": 0.0535, + "step": 7624 + }, + { + "epoch": 6.440033783783784, + "grad_norm": 0.6843758821487427, + "learning_rate": 2.2866e-05, + "loss": 0.0472, + "step": 7625 + }, + { + "epoch": 6.440878378378378, + "grad_norm": 0.36434173583984375, + "learning_rate": 2.2869e-05, + "loss": 0.0337, + "step": 7626 + }, + { + "epoch": 6.441722972972973, + "grad_norm": 0.44060707092285156, + "learning_rate": 2.2872e-05, + "loss": 0.0451, + "step": 7627 + }, + { + "epoch": 6.4425675675675675, + "grad_norm": 0.47410082817077637, + "learning_rate": 2.2875e-05, + "loss": 0.0393, + "step": 7628 + }, + { + "epoch": 6.443412162162162, + "grad_norm": 0.4046815037727356, + "learning_rate": 2.2878e-05, + "loss": 0.0415, + "step": 7629 + }, + { + "epoch": 6.444256756756757, + "grad_norm": 0.3387708365917206, + "learning_rate": 2.2881000000000003e-05, + "loss": 0.0377, + "step": 7630 + }, + { + "epoch": 6.445101351351352, + "grad_norm": 0.7111454010009766, + "learning_rate": 2.2884000000000003e-05, + "loss": 0.0421, + "step": 7631 + }, + { + "epoch": 6.445945945945946, + "grad_norm": 0.2900792062282562, + "learning_rate": 2.2887e-05, + "loss": 0.0211, + "step": 7632 + }, + { + "epoch": 6.44679054054054, + "grad_norm": 0.6054859757423401, + "learning_rate": 2.289e-05, + "loss": 0.0411, + "step": 7633 + }, + { + "epoch": 6.447635135135135, + "grad_norm": 0.5366570949554443, + "learning_rate": 2.2893e-05, + "loss": 0.045, + "step": 7634 + }, + { + "epoch": 6.44847972972973, + "grad_norm": 0.4739098846912384, + "learning_rate": 2.2896e-05, + "loss": 0.0436, + "step": 7635 + }, + { + "epoch": 6.449324324324325, + "grad_norm": 0.6801203489303589, + "learning_rate": 2.2899e-05, + "loss": 0.065, + "step": 7636 + }, + { + "epoch": 6.450168918918919, + "grad_norm": 0.5413877964019775, + "learning_rate": 2.2902e-05, + "loss": 0.043, + "step": 7637 + }, + { + "epoch": 6.451013513513513, + "grad_norm": 0.5203103423118591, + "learning_rate": 2.2905e-05, + "loss": 0.0646, + "step": 7638 + }, + { + "epoch": 6.451858108108108, + "grad_norm": 1.0384361743927002, + "learning_rate": 2.2907999999999998e-05, + "loss": 0.0305, + "step": 7639 + }, + { + "epoch": 6.452702702702703, + "grad_norm": 0.5492703318595886, + "learning_rate": 2.2911e-05, + "loss": 0.0345, + "step": 7640 + }, + { + "epoch": 6.453547297297297, + "grad_norm": 0.6830381751060486, + "learning_rate": 2.2914e-05, + "loss": 0.0635, + "step": 7641 + }, + { + "epoch": 6.454391891891892, + "grad_norm": 0.34961870312690735, + "learning_rate": 2.2917e-05, + "loss": 0.0297, + "step": 7642 + }, + { + "epoch": 6.455236486486487, + "grad_norm": 0.609046995639801, + "learning_rate": 2.292e-05, + "loss": 0.0371, + "step": 7643 + }, + { + "epoch": 6.456081081081081, + "grad_norm": 0.9619761109352112, + "learning_rate": 2.2923e-05, + "loss": 0.0413, + "step": 7644 + }, + { + "epoch": 6.456925675675675, + "grad_norm": 0.5745372772216797, + "learning_rate": 2.2926e-05, + "loss": 0.0544, + "step": 7645 + }, + { + "epoch": 6.45777027027027, + "grad_norm": 0.49042192101478577, + "learning_rate": 2.2929e-05, + "loss": 0.0476, + "step": 7646 + }, + { + "epoch": 6.458614864864865, + "grad_norm": 0.6496189832687378, + "learning_rate": 2.2932e-05, + "loss": 0.0415, + "step": 7647 + }, + { + "epoch": 6.45945945945946, + "grad_norm": 0.503977358341217, + "learning_rate": 2.2935e-05, + "loss": 0.051, + "step": 7648 + }, + { + "epoch": 6.460304054054054, + "grad_norm": 0.6377308964729309, + "learning_rate": 2.2938e-05, + "loss": 0.0441, + "step": 7649 + }, + { + "epoch": 6.461148648648648, + "grad_norm": 0.8374037742614746, + "learning_rate": 2.2941000000000003e-05, + "loss": 0.0355, + "step": 7650 + }, + { + "epoch": 6.461993243243243, + "grad_norm": 0.7948093414306641, + "learning_rate": 2.2944000000000003e-05, + "loss": 0.0559, + "step": 7651 + }, + { + "epoch": 6.462837837837838, + "grad_norm": 0.8668928742408752, + "learning_rate": 2.2947000000000002e-05, + "loss": 0.0523, + "step": 7652 + }, + { + "epoch": 6.4636824324324325, + "grad_norm": 0.7512110471725464, + "learning_rate": 2.2950000000000002e-05, + "loss": 0.0499, + "step": 7653 + }, + { + "epoch": 6.464527027027027, + "grad_norm": 1.5717618465423584, + "learning_rate": 2.2953000000000002e-05, + "loss": 0.07, + "step": 7654 + }, + { + "epoch": 6.465371621621622, + "grad_norm": 1.2716015577316284, + "learning_rate": 2.2956000000000002e-05, + "loss": 0.3178, + "step": 7655 + }, + { + "epoch": 6.466216216216216, + "grad_norm": 0.7970500588417053, + "learning_rate": 2.2959e-05, + "loss": 0.1999, + "step": 7656 + }, + { + "epoch": 6.4670608108108105, + "grad_norm": 0.6981584429740906, + "learning_rate": 2.2961999999999998e-05, + "loss": 0.1121, + "step": 7657 + }, + { + "epoch": 6.467905405405405, + "grad_norm": 0.39398619532585144, + "learning_rate": 2.2964999999999998e-05, + "loss": 0.0626, + "step": 7658 + }, + { + "epoch": 6.46875, + "grad_norm": 1.2388051748275757, + "learning_rate": 2.2967999999999998e-05, + "loss": 0.0519, + "step": 7659 + }, + { + "epoch": 6.469594594594595, + "grad_norm": 0.3987072706222534, + "learning_rate": 2.2971e-05, + "loss": 0.0423, + "step": 7660 + }, + { + "epoch": 6.4704391891891895, + "grad_norm": 0.8979229927062988, + "learning_rate": 2.2974e-05, + "loss": 0.0648, + "step": 7661 + }, + { + "epoch": 6.471283783783784, + "grad_norm": 1.1928008794784546, + "learning_rate": 2.2977e-05, + "loss": 0.0528, + "step": 7662 + }, + { + "epoch": 6.472128378378378, + "grad_norm": 0.4292657971382141, + "learning_rate": 2.298e-05, + "loss": 0.0465, + "step": 7663 + }, + { + "epoch": 6.472972972972973, + "grad_norm": 0.6304280161857605, + "learning_rate": 2.2983e-05, + "loss": 0.054, + "step": 7664 + }, + { + "epoch": 6.4738175675675675, + "grad_norm": 0.33679822087287903, + "learning_rate": 2.2986e-05, + "loss": 0.0312, + "step": 7665 + }, + { + "epoch": 6.474662162162162, + "grad_norm": 0.4734349250793457, + "learning_rate": 2.2989e-05, + "loss": 0.0426, + "step": 7666 + }, + { + "epoch": 6.475506756756757, + "grad_norm": 0.4836953580379486, + "learning_rate": 2.2992e-05, + "loss": 0.0429, + "step": 7667 + }, + { + "epoch": 6.476351351351352, + "grad_norm": 0.3670620024204254, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 7668 + }, + { + "epoch": 6.477195945945946, + "grad_norm": 0.4024556279182434, + "learning_rate": 2.2998e-05, + "loss": 0.0324, + "step": 7669 + }, + { + "epoch": 6.47804054054054, + "grad_norm": 0.478047639131546, + "learning_rate": 2.3001000000000002e-05, + "loss": 0.0437, + "step": 7670 + }, + { + "epoch": 6.478885135135135, + "grad_norm": 0.5303732752799988, + "learning_rate": 2.3004000000000002e-05, + "loss": 0.0495, + "step": 7671 + }, + { + "epoch": 6.47972972972973, + "grad_norm": 0.3349417746067047, + "learning_rate": 2.3007000000000002e-05, + "loss": 0.0335, + "step": 7672 + }, + { + "epoch": 6.480574324324325, + "grad_norm": 0.3574216961860657, + "learning_rate": 2.301e-05, + "loss": 0.0487, + "step": 7673 + }, + { + "epoch": 6.481418918918919, + "grad_norm": 0.5543693900108337, + "learning_rate": 2.3013e-05, + "loss": 0.0451, + "step": 7674 + }, + { + "epoch": 6.482263513513513, + "grad_norm": 0.5480382442474365, + "learning_rate": 2.3016e-05, + "loss": 0.0386, + "step": 7675 + }, + { + "epoch": 6.483108108108108, + "grad_norm": 0.31171151995658875, + "learning_rate": 2.3019e-05, + "loss": 0.0261, + "step": 7676 + }, + { + "epoch": 6.483952702702703, + "grad_norm": 0.27205726504325867, + "learning_rate": 2.3022e-05, + "loss": 0.0236, + "step": 7677 + }, + { + "epoch": 6.484797297297297, + "grad_norm": 0.6018331050872803, + "learning_rate": 2.3025e-05, + "loss": 0.0944, + "step": 7678 + }, + { + "epoch": 6.485641891891892, + "grad_norm": 0.5081701278686523, + "learning_rate": 2.3028e-05, + "loss": 0.0324, + "step": 7679 + }, + { + "epoch": 6.486486486486487, + "grad_norm": 0.39142200350761414, + "learning_rate": 2.3031000000000004e-05, + "loss": 0.0372, + "step": 7680 + }, + { + "epoch": 6.487331081081081, + "grad_norm": 0.4752267301082611, + "learning_rate": 2.3034e-05, + "loss": 0.0466, + "step": 7681 + }, + { + "epoch": 6.488175675675675, + "grad_norm": 0.4192509949207306, + "learning_rate": 2.3037e-05, + "loss": 0.0375, + "step": 7682 + }, + { + "epoch": 6.48902027027027, + "grad_norm": 0.5945051312446594, + "learning_rate": 2.304e-05, + "loss": 0.0406, + "step": 7683 + }, + { + "epoch": 6.489864864864865, + "grad_norm": 0.47928181290626526, + "learning_rate": 2.3043e-05, + "loss": 0.0379, + "step": 7684 + }, + { + "epoch": 6.49070945945946, + "grad_norm": 0.4935055077075958, + "learning_rate": 2.3046e-05, + "loss": 0.0472, + "step": 7685 + }, + { + "epoch": 6.491554054054054, + "grad_norm": 0.3477061688899994, + "learning_rate": 2.3049e-05, + "loss": 0.0367, + "step": 7686 + }, + { + "epoch": 6.492398648648648, + "grad_norm": 0.5376030206680298, + "learning_rate": 2.3052e-05, + "loss": 0.0349, + "step": 7687 + }, + { + "epoch": 6.493243243243243, + "grad_norm": 0.4412943720817566, + "learning_rate": 2.3055e-05, + "loss": 0.037, + "step": 7688 + }, + { + "epoch": 6.494087837837838, + "grad_norm": 0.5057927370071411, + "learning_rate": 2.3058e-05, + "loss": 0.0466, + "step": 7689 + }, + { + "epoch": 6.4949324324324325, + "grad_norm": 0.46692657470703125, + "learning_rate": 2.3061e-05, + "loss": 0.0309, + "step": 7690 + }, + { + "epoch": 6.495777027027027, + "grad_norm": 0.47503575682640076, + "learning_rate": 2.3064e-05, + "loss": 0.0352, + "step": 7691 + }, + { + "epoch": 6.496621621621622, + "grad_norm": 0.5085157752037048, + "learning_rate": 2.3067e-05, + "loss": 0.0533, + "step": 7692 + }, + { + "epoch": 6.497466216216216, + "grad_norm": 0.5338962078094482, + "learning_rate": 2.307e-05, + "loss": 0.0577, + "step": 7693 + }, + { + "epoch": 6.4983108108108105, + "grad_norm": 0.3563752770423889, + "learning_rate": 2.3073e-05, + "loss": 0.0304, + "step": 7694 + }, + { + "epoch": 6.499155405405405, + "grad_norm": 0.6837064623832703, + "learning_rate": 2.3076e-05, + "loss": 0.0419, + "step": 7695 + }, + { + "epoch": 6.5, + "grad_norm": 0.5023593902587891, + "learning_rate": 2.3079e-05, + "loss": 0.0552, + "step": 7696 + }, + { + "epoch": 6.500844594594595, + "grad_norm": 0.5933271050453186, + "learning_rate": 2.3082e-05, + "loss": 0.0522, + "step": 7697 + }, + { + "epoch": 6.5016891891891895, + "grad_norm": 0.5087243914604187, + "learning_rate": 2.3085e-05, + "loss": 0.0733, + "step": 7698 + }, + { + "epoch": 6.502533783783784, + "grad_norm": 0.5394738912582397, + "learning_rate": 2.3088e-05, + "loss": 0.04, + "step": 7699 + }, + { + "epoch": 6.503378378378378, + "grad_norm": 0.7892783880233765, + "learning_rate": 2.3091000000000003e-05, + "loss": 0.0462, + "step": 7700 + }, + { + "epoch": 6.504222972972973, + "grad_norm": 0.41279157996177673, + "learning_rate": 2.3094000000000003e-05, + "loss": 0.0408, + "step": 7701 + }, + { + "epoch": 6.5050675675675675, + "grad_norm": 0.8126270771026611, + "learning_rate": 2.3097000000000003e-05, + "loss": 0.0583, + "step": 7702 + }, + { + "epoch": 6.505912162162162, + "grad_norm": 2.3734242916107178, + "learning_rate": 2.3100000000000002e-05, + "loss": 0.0823, + "step": 7703 + }, + { + "epoch": 6.506756756756757, + "grad_norm": 1.5086308717727661, + "learning_rate": 2.3103000000000002e-05, + "loss": 0.0838, + "step": 7704 + }, + { + "epoch": 6.507601351351351, + "grad_norm": 2.12609601020813, + "learning_rate": 2.3106000000000002e-05, + "loss": 0.3714, + "step": 7705 + }, + { + "epoch": 6.508445945945946, + "grad_norm": 0.7553391456604004, + "learning_rate": 2.3109e-05, + "loss": 0.1765, + "step": 7706 + }, + { + "epoch": 6.50929054054054, + "grad_norm": 0.368868887424469, + "learning_rate": 2.3111999999999998e-05, + "loss": 0.071, + "step": 7707 + }, + { + "epoch": 6.510135135135135, + "grad_norm": 0.463012158870697, + "learning_rate": 2.3114999999999998e-05, + "loss": 0.0704, + "step": 7708 + }, + { + "epoch": 6.51097972972973, + "grad_norm": 0.5872842669487, + "learning_rate": 2.3117999999999998e-05, + "loss": 0.0604, + "step": 7709 + }, + { + "epoch": 6.511824324324325, + "grad_norm": 0.41065582633018494, + "learning_rate": 2.3121e-05, + "loss": 0.0462, + "step": 7710 + }, + { + "epoch": 6.512668918918919, + "grad_norm": 0.5627772212028503, + "learning_rate": 2.3124e-05, + "loss": 0.0571, + "step": 7711 + }, + { + "epoch": 6.513513513513513, + "grad_norm": 0.3915734887123108, + "learning_rate": 2.3127e-05, + "loss": 0.033, + "step": 7712 + }, + { + "epoch": 6.514358108108108, + "grad_norm": 0.4129081666469574, + "learning_rate": 2.313e-05, + "loss": 0.0474, + "step": 7713 + }, + { + "epoch": 6.515202702702703, + "grad_norm": 0.6611066460609436, + "learning_rate": 2.3133e-05, + "loss": 0.0583, + "step": 7714 + }, + { + "epoch": 6.516047297297297, + "grad_norm": 0.6393797397613525, + "learning_rate": 2.3136e-05, + "loss": 0.0625, + "step": 7715 + }, + { + "epoch": 6.516891891891892, + "grad_norm": 0.7222577333450317, + "learning_rate": 2.3139e-05, + "loss": 0.0402, + "step": 7716 + }, + { + "epoch": 6.517736486486487, + "grad_norm": 0.35814595222473145, + "learning_rate": 2.3142e-05, + "loss": 0.0471, + "step": 7717 + }, + { + "epoch": 6.518581081081081, + "grad_norm": 0.37494802474975586, + "learning_rate": 2.3145e-05, + "loss": 0.0228, + "step": 7718 + }, + { + "epoch": 6.519425675675675, + "grad_norm": 0.4215737581253052, + "learning_rate": 2.3148e-05, + "loss": 0.0332, + "step": 7719 + }, + { + "epoch": 6.52027027027027, + "grad_norm": 0.45475149154663086, + "learning_rate": 2.3151000000000002e-05, + "loss": 0.0435, + "step": 7720 + }, + { + "epoch": 6.521114864864865, + "grad_norm": 0.46575531363487244, + "learning_rate": 2.3154000000000002e-05, + "loss": 0.039, + "step": 7721 + }, + { + "epoch": 6.52195945945946, + "grad_norm": 0.36244428157806396, + "learning_rate": 2.3157000000000002e-05, + "loss": 0.0309, + "step": 7722 + }, + { + "epoch": 6.522804054054054, + "grad_norm": 0.47109806537628174, + "learning_rate": 2.3160000000000002e-05, + "loss": 0.0235, + "step": 7723 + }, + { + "epoch": 6.523648648648649, + "grad_norm": 0.4470530152320862, + "learning_rate": 2.3163e-05, + "loss": 0.0312, + "step": 7724 + }, + { + "epoch": 6.524493243243243, + "grad_norm": 0.4583154320716858, + "learning_rate": 2.3166e-05, + "loss": 0.0345, + "step": 7725 + }, + { + "epoch": 6.525337837837838, + "grad_norm": 0.44425156712532043, + "learning_rate": 2.3169e-05, + "loss": 0.0301, + "step": 7726 + }, + { + "epoch": 6.5261824324324325, + "grad_norm": 0.5818871259689331, + "learning_rate": 2.3172e-05, + "loss": 0.0433, + "step": 7727 + }, + { + "epoch": 6.527027027027027, + "grad_norm": 0.5827089548110962, + "learning_rate": 2.3175e-05, + "loss": 0.0395, + "step": 7728 + }, + { + "epoch": 6.527871621621622, + "grad_norm": 0.45634356141090393, + "learning_rate": 2.3178e-05, + "loss": 0.0336, + "step": 7729 + }, + { + "epoch": 6.528716216216216, + "grad_norm": 0.3844955861568451, + "learning_rate": 2.3181000000000004e-05, + "loss": 0.0373, + "step": 7730 + }, + { + "epoch": 6.5295608108108105, + "grad_norm": 0.7535548210144043, + "learning_rate": 2.3184e-05, + "loss": 0.058, + "step": 7731 + }, + { + "epoch": 6.530405405405405, + "grad_norm": 0.431026428937912, + "learning_rate": 2.3187e-05, + "loss": 0.0329, + "step": 7732 + }, + { + "epoch": 6.53125, + "grad_norm": 0.4957793951034546, + "learning_rate": 2.319e-05, + "loss": 0.0385, + "step": 7733 + }, + { + "epoch": 6.532094594594595, + "grad_norm": 0.8035506010055542, + "learning_rate": 2.3193e-05, + "loss": 0.0339, + "step": 7734 + }, + { + "epoch": 6.5329391891891895, + "grad_norm": 0.44161519408226013, + "learning_rate": 2.3196e-05, + "loss": 0.0265, + "step": 7735 + }, + { + "epoch": 6.533783783783784, + "grad_norm": 0.4731196165084839, + "learning_rate": 2.3199e-05, + "loss": 0.0358, + "step": 7736 + }, + { + "epoch": 6.534628378378378, + "grad_norm": 0.47310671210289, + "learning_rate": 2.3202e-05, + "loss": 0.0374, + "step": 7737 + }, + { + "epoch": 6.535472972972973, + "grad_norm": 1.1683365106582642, + "learning_rate": 2.3205e-05, + "loss": 0.0479, + "step": 7738 + }, + { + "epoch": 6.5363175675675675, + "grad_norm": 0.5604109764099121, + "learning_rate": 2.3208e-05, + "loss": 0.0377, + "step": 7739 + }, + { + "epoch": 6.537162162162162, + "grad_norm": 0.4835769236087799, + "learning_rate": 2.3211000000000002e-05, + "loss": 0.0469, + "step": 7740 + }, + { + "epoch": 6.538006756756757, + "grad_norm": 0.6738689541816711, + "learning_rate": 2.3214000000000002e-05, + "loss": 0.0445, + "step": 7741 + }, + { + "epoch": 6.538851351351351, + "grad_norm": 0.6477972269058228, + "learning_rate": 2.3217e-05, + "loss": 0.0351, + "step": 7742 + }, + { + "epoch": 6.539695945945946, + "grad_norm": 0.927934467792511, + "learning_rate": 2.322e-05, + "loss": 0.0578, + "step": 7743 + }, + { + "epoch": 6.54054054054054, + "grad_norm": 0.6601645350456238, + "learning_rate": 2.3223e-05, + "loss": 0.0514, + "step": 7744 + }, + { + "epoch": 6.541385135135135, + "grad_norm": 0.4345646798610687, + "learning_rate": 2.3226e-05, + "loss": 0.0368, + "step": 7745 + }, + { + "epoch": 6.54222972972973, + "grad_norm": 0.42035967111587524, + "learning_rate": 2.3229e-05, + "loss": 0.052, + "step": 7746 + }, + { + "epoch": 6.543074324324325, + "grad_norm": 0.5590560436248779, + "learning_rate": 2.3232e-05, + "loss": 0.0467, + "step": 7747 + }, + { + "epoch": 6.543918918918919, + "grad_norm": 0.40205103158950806, + "learning_rate": 2.3235e-05, + "loss": 0.0276, + "step": 7748 + }, + { + "epoch": 6.544763513513513, + "grad_norm": 1.1303030252456665, + "learning_rate": 2.3238e-05, + "loss": 0.0416, + "step": 7749 + }, + { + "epoch": 6.545608108108108, + "grad_norm": 0.6862250566482544, + "learning_rate": 2.3241000000000003e-05, + "loss": 0.0411, + "step": 7750 + }, + { + "epoch": 6.546452702702703, + "grad_norm": 0.5149973630905151, + "learning_rate": 2.3244000000000003e-05, + "loss": 0.0517, + "step": 7751 + }, + { + "epoch": 6.547297297297297, + "grad_norm": 0.6244577169418335, + "learning_rate": 2.3247000000000003e-05, + "loss": 0.0738, + "step": 7752 + }, + { + "epoch": 6.548141891891892, + "grad_norm": 1.2772465944290161, + "learning_rate": 2.3250000000000003e-05, + "loss": 0.093, + "step": 7753 + }, + { + "epoch": 6.548986486486487, + "grad_norm": 1.7335561513900757, + "learning_rate": 2.3253000000000003e-05, + "loss": 0.1274, + "step": 7754 + }, + { + "epoch": 6.549831081081081, + "grad_norm": 0.9470184445381165, + "learning_rate": 2.3256e-05, + "loss": 0.4028, + "step": 7755 + }, + { + "epoch": 6.550675675675675, + "grad_norm": 0.7491062879562378, + "learning_rate": 2.3259e-05, + "loss": 0.1824, + "step": 7756 + }, + { + "epoch": 6.55152027027027, + "grad_norm": 0.47146302461624146, + "learning_rate": 2.3262e-05, + "loss": 0.1112, + "step": 7757 + }, + { + "epoch": 6.552364864864865, + "grad_norm": 0.41863253712654114, + "learning_rate": 2.3265e-05, + "loss": 0.0697, + "step": 7758 + }, + { + "epoch": 6.55320945945946, + "grad_norm": 0.7216401100158691, + "learning_rate": 2.3267999999999998e-05, + "loss": 0.0521, + "step": 7759 + }, + { + "epoch": 6.554054054054054, + "grad_norm": 0.3433159589767456, + "learning_rate": 2.3270999999999998e-05, + "loss": 0.0488, + "step": 7760 + }, + { + "epoch": 6.554898648648649, + "grad_norm": 0.47545844316482544, + "learning_rate": 2.3274e-05, + "loss": 0.054, + "step": 7761 + }, + { + "epoch": 6.555743243243243, + "grad_norm": 0.29535534977912903, + "learning_rate": 2.3277e-05, + "loss": 0.0375, + "step": 7762 + }, + { + "epoch": 6.556587837837838, + "grad_norm": 0.5192347764968872, + "learning_rate": 2.328e-05, + "loss": 0.0593, + "step": 7763 + }, + { + "epoch": 6.5574324324324325, + "grad_norm": 0.7835689187049866, + "learning_rate": 2.3283e-05, + "loss": 0.0723, + "step": 7764 + }, + { + "epoch": 6.558277027027027, + "grad_norm": 0.5142146944999695, + "learning_rate": 2.3286e-05, + "loss": 0.0492, + "step": 7765 + }, + { + "epoch": 6.559121621621622, + "grad_norm": 0.3335695266723633, + "learning_rate": 2.3289e-05, + "loss": 0.0329, + "step": 7766 + }, + { + "epoch": 6.559966216216216, + "grad_norm": 0.2475631833076477, + "learning_rate": 2.3292e-05, + "loss": 0.0332, + "step": 7767 + }, + { + "epoch": 6.5608108108108105, + "grad_norm": 0.5449269413948059, + "learning_rate": 2.3295e-05, + "loss": 0.0321, + "step": 7768 + }, + { + "epoch": 6.561655405405405, + "grad_norm": 0.3786778748035431, + "learning_rate": 2.3298e-05, + "loss": 0.0303, + "step": 7769 + }, + { + "epoch": 6.5625, + "grad_norm": 0.49861425161361694, + "learning_rate": 2.3301e-05, + "loss": 0.0559, + "step": 7770 + }, + { + "epoch": 6.563344594594595, + "grad_norm": 0.4275512993335724, + "learning_rate": 2.3304000000000003e-05, + "loss": 0.0574, + "step": 7771 + }, + { + "epoch": 6.5641891891891895, + "grad_norm": 0.49520254135131836, + "learning_rate": 2.3307000000000002e-05, + "loss": 0.0481, + "step": 7772 + }, + { + "epoch": 6.565033783783784, + "grad_norm": 0.44011548161506653, + "learning_rate": 2.3310000000000002e-05, + "loss": 0.0512, + "step": 7773 + }, + { + "epoch": 6.565878378378378, + "grad_norm": 0.4072219431400299, + "learning_rate": 2.3313000000000002e-05, + "loss": 0.0318, + "step": 7774 + }, + { + "epoch": 6.566722972972973, + "grad_norm": 0.3388694226741791, + "learning_rate": 2.3316000000000002e-05, + "loss": 0.0416, + "step": 7775 + }, + { + "epoch": 6.5675675675675675, + "grad_norm": 0.3843449652194977, + "learning_rate": 2.3319e-05, + "loss": 0.0432, + "step": 7776 + }, + { + "epoch": 6.568412162162162, + "grad_norm": 0.3985867202281952, + "learning_rate": 2.3322e-05, + "loss": 0.0459, + "step": 7777 + }, + { + "epoch": 6.569256756756757, + "grad_norm": 0.44054558873176575, + "learning_rate": 2.3325e-05, + "loss": 0.0285, + "step": 7778 + }, + { + "epoch": 6.570101351351351, + "grad_norm": 0.48163580894470215, + "learning_rate": 2.3328e-05, + "loss": 0.042, + "step": 7779 + }, + { + "epoch": 6.570945945945946, + "grad_norm": 0.5253520011901855, + "learning_rate": 2.3330999999999997e-05, + "loss": 0.0474, + "step": 7780 + }, + { + "epoch": 6.57179054054054, + "grad_norm": 0.6040852069854736, + "learning_rate": 2.3334e-05, + "loss": 0.0467, + "step": 7781 + }, + { + "epoch": 6.572635135135135, + "grad_norm": 0.46410033106803894, + "learning_rate": 2.3337e-05, + "loss": 0.0382, + "step": 7782 + }, + { + "epoch": 6.57347972972973, + "grad_norm": 0.3135310113430023, + "learning_rate": 2.334e-05, + "loss": 0.0367, + "step": 7783 + }, + { + "epoch": 6.574324324324325, + "grad_norm": 0.3952442705631256, + "learning_rate": 2.3343e-05, + "loss": 0.0339, + "step": 7784 + }, + { + "epoch": 6.575168918918919, + "grad_norm": 0.28939592838287354, + "learning_rate": 2.3346e-05, + "loss": 0.0331, + "step": 7785 + }, + { + "epoch": 6.576013513513513, + "grad_norm": 0.5027974843978882, + "learning_rate": 2.3349e-05, + "loss": 0.0452, + "step": 7786 + }, + { + "epoch": 6.576858108108108, + "grad_norm": 0.5557368397712708, + "learning_rate": 2.3352e-05, + "loss": 0.058, + "step": 7787 + }, + { + "epoch": 6.577702702702703, + "grad_norm": 1.2274978160858154, + "learning_rate": 2.3355e-05, + "loss": 0.0825, + "step": 7788 + }, + { + "epoch": 6.578547297297297, + "grad_norm": 0.5303643345832825, + "learning_rate": 2.3358e-05, + "loss": 0.0545, + "step": 7789 + }, + { + "epoch": 6.579391891891892, + "grad_norm": 0.43878623843193054, + "learning_rate": 2.3361e-05, + "loss": 0.029, + "step": 7790 + }, + { + "epoch": 6.580236486486487, + "grad_norm": 0.6282500624656677, + "learning_rate": 2.3364000000000002e-05, + "loss": 0.0389, + "step": 7791 + }, + { + "epoch": 6.581081081081081, + "grad_norm": 0.5857342481613159, + "learning_rate": 2.3367000000000002e-05, + "loss": 0.0482, + "step": 7792 + }, + { + "epoch": 6.581925675675675, + "grad_norm": 0.42155709862709045, + "learning_rate": 2.337e-05, + "loss": 0.0366, + "step": 7793 + }, + { + "epoch": 6.58277027027027, + "grad_norm": 0.8205639123916626, + "learning_rate": 2.3373e-05, + "loss": 0.0524, + "step": 7794 + }, + { + "epoch": 6.583614864864865, + "grad_norm": 0.5912173986434937, + "learning_rate": 2.3376e-05, + "loss": 0.0347, + "step": 7795 + }, + { + "epoch": 6.58445945945946, + "grad_norm": 0.7127105593681335, + "learning_rate": 2.3379e-05, + "loss": 0.0418, + "step": 7796 + }, + { + "epoch": 6.585304054054054, + "grad_norm": 0.6921581029891968, + "learning_rate": 2.3382e-05, + "loss": 0.0514, + "step": 7797 + }, + { + "epoch": 6.586148648648649, + "grad_norm": 1.2404357194900513, + "learning_rate": 2.3385e-05, + "loss": 0.0484, + "step": 7798 + }, + { + "epoch": 6.586993243243243, + "grad_norm": 0.610345184803009, + "learning_rate": 2.3388e-05, + "loss": 0.0596, + "step": 7799 + }, + { + "epoch": 6.587837837837838, + "grad_norm": 0.5046082139015198, + "learning_rate": 2.3391e-05, + "loss": 0.055, + "step": 7800 + }, + { + "epoch": 6.5886824324324325, + "grad_norm": 0.6417945027351379, + "learning_rate": 2.3394000000000003e-05, + "loss": 0.0524, + "step": 7801 + }, + { + "epoch": 6.589527027027027, + "grad_norm": 1.201310396194458, + "learning_rate": 2.3397000000000003e-05, + "loss": 0.0861, + "step": 7802 + }, + { + "epoch": 6.590371621621622, + "grad_norm": 0.7024200558662415, + "learning_rate": 2.3400000000000003e-05, + "loss": 0.0662, + "step": 7803 + }, + { + "epoch": 6.591216216216216, + "grad_norm": 1.234525203704834, + "learning_rate": 2.3403e-05, + "loss": 0.0942, + "step": 7804 + }, + { + "epoch": 6.5920608108108105, + "grad_norm": 1.7821011543273926, + "learning_rate": 2.3406e-05, + "loss": 0.3689, + "step": 7805 + }, + { + "epoch": 6.592905405405405, + "grad_norm": Infinity, + "learning_rate": 2.3406e-05, + "loss": 0.2346, + "step": 7806 + }, + { + "epoch": 6.59375, + "grad_norm": 0.6419737935066223, + "learning_rate": 2.3409e-05, + "loss": 0.1332, + "step": 7807 + }, + { + "epoch": 6.594594594594595, + "grad_norm": 0.37909796833992004, + "learning_rate": 2.3412e-05, + "loss": 0.0604, + "step": 7808 + }, + { + "epoch": 6.5954391891891895, + "grad_norm": 0.391910582780838, + "learning_rate": 2.3415e-05, + "loss": 0.0439, + "step": 7809 + }, + { + "epoch": 6.596283783783784, + "grad_norm": 0.31456196308135986, + "learning_rate": 2.3418e-05, + "loss": 0.0468, + "step": 7810 + }, + { + "epoch": 6.597128378378378, + "grad_norm": 0.40337076783180237, + "learning_rate": 2.3420999999999998e-05, + "loss": 0.0409, + "step": 7811 + }, + { + "epoch": 6.597972972972973, + "grad_norm": 0.44750458002090454, + "learning_rate": 2.3424e-05, + "loss": 0.0357, + "step": 7812 + }, + { + "epoch": 6.5988175675675675, + "grad_norm": 0.4206535220146179, + "learning_rate": 2.3427e-05, + "loss": 0.0437, + "step": 7813 + }, + { + "epoch": 6.599662162162162, + "grad_norm": 0.8155301809310913, + "learning_rate": 2.343e-05, + "loss": 0.0494, + "step": 7814 + }, + { + "epoch": 6.600506756756757, + "grad_norm": 0.40583428740501404, + "learning_rate": 2.3433e-05, + "loss": 0.0406, + "step": 7815 + }, + { + "epoch": 6.601351351351351, + "grad_norm": 0.4444418251514435, + "learning_rate": 2.3436e-05, + "loss": 0.0631, + "step": 7816 + }, + { + "epoch": 6.602195945945946, + "grad_norm": 0.49379998445510864, + "learning_rate": 2.3439e-05, + "loss": 0.0502, + "step": 7817 + }, + { + "epoch": 6.60304054054054, + "grad_norm": 0.35031431913375854, + "learning_rate": 2.3442e-05, + "loss": 0.0398, + "step": 7818 + }, + { + "epoch": 6.603885135135135, + "grad_norm": 0.6774099469184875, + "learning_rate": 2.3445e-05, + "loss": 0.0329, + "step": 7819 + }, + { + "epoch": 6.60472972972973, + "grad_norm": 0.5271437764167786, + "learning_rate": 2.3448e-05, + "loss": 0.036, + "step": 7820 + }, + { + "epoch": 6.605574324324325, + "grad_norm": 0.48896175622940063, + "learning_rate": 2.3451e-05, + "loss": 0.0455, + "step": 7821 + }, + { + "epoch": 6.606418918918919, + "grad_norm": 0.8494086265563965, + "learning_rate": 2.3454000000000003e-05, + "loss": 0.0627, + "step": 7822 + }, + { + "epoch": 6.607263513513513, + "grad_norm": 0.5810129642486572, + "learning_rate": 2.3457000000000003e-05, + "loss": 0.0523, + "step": 7823 + }, + { + "epoch": 6.608108108108108, + "grad_norm": 0.332876980304718, + "learning_rate": 2.3460000000000002e-05, + "loss": 0.03, + "step": 7824 + }, + { + "epoch": 6.608952702702703, + "grad_norm": 0.5935805439949036, + "learning_rate": 2.3463000000000002e-05, + "loss": 0.0544, + "step": 7825 + }, + { + "epoch": 6.609797297297297, + "grad_norm": 0.4539819359779358, + "learning_rate": 2.3466000000000002e-05, + "loss": 0.0382, + "step": 7826 + }, + { + "epoch": 6.610641891891892, + "grad_norm": 0.5557988286018372, + "learning_rate": 2.3469000000000002e-05, + "loss": 0.0567, + "step": 7827 + }, + { + "epoch": 6.611486486486487, + "grad_norm": 0.40940427780151367, + "learning_rate": 2.3472e-05, + "loss": 0.0373, + "step": 7828 + }, + { + "epoch": 6.612331081081081, + "grad_norm": 0.49393942952156067, + "learning_rate": 2.3475e-05, + "loss": 0.0522, + "step": 7829 + }, + { + "epoch": 6.613175675675675, + "grad_norm": 0.4104006588459015, + "learning_rate": 2.3477999999999998e-05, + "loss": 0.0446, + "step": 7830 + }, + { + "epoch": 6.61402027027027, + "grad_norm": 0.8704939484596252, + "learning_rate": 2.3480999999999998e-05, + "loss": 0.0398, + "step": 7831 + }, + { + "epoch": 6.614864864864865, + "grad_norm": 0.3942711055278778, + "learning_rate": 2.3484e-05, + "loss": 0.0343, + "step": 7832 + }, + { + "epoch": 6.61570945945946, + "grad_norm": 0.4180881381034851, + "learning_rate": 2.3487e-05, + "loss": 0.029, + "step": 7833 + }, + { + "epoch": 6.616554054054054, + "grad_norm": 0.8309807181358337, + "learning_rate": 2.349e-05, + "loss": 0.0805, + "step": 7834 + }, + { + "epoch": 6.617398648648649, + "grad_norm": 0.33471858501434326, + "learning_rate": 2.3493e-05, + "loss": 0.0346, + "step": 7835 + }, + { + "epoch": 6.618243243243243, + "grad_norm": 0.6058258414268494, + "learning_rate": 2.3496e-05, + "loss": 0.0581, + "step": 7836 + }, + { + "epoch": 6.619087837837838, + "grad_norm": 0.4449347257614136, + "learning_rate": 2.3499e-05, + "loss": 0.0524, + "step": 7837 + }, + { + "epoch": 6.6199324324324325, + "grad_norm": 0.47179824113845825, + "learning_rate": 2.3502e-05, + "loss": 0.0477, + "step": 7838 + }, + { + "epoch": 6.620777027027027, + "grad_norm": 0.4063235819339752, + "learning_rate": 2.3505e-05, + "loss": 0.0449, + "step": 7839 + }, + { + "epoch": 6.621621621621622, + "grad_norm": 0.7093759179115295, + "learning_rate": 2.3508e-05, + "loss": 0.0415, + "step": 7840 + }, + { + "epoch": 6.622466216216216, + "grad_norm": 0.4883376359939575, + "learning_rate": 2.3511e-05, + "loss": 0.0429, + "step": 7841 + }, + { + "epoch": 6.6233108108108105, + "grad_norm": 0.7121303081512451, + "learning_rate": 2.3514000000000002e-05, + "loss": 0.0732, + "step": 7842 + }, + { + "epoch": 6.624155405405405, + "grad_norm": 1.1018394231796265, + "learning_rate": 2.3517000000000002e-05, + "loss": 0.1139, + "step": 7843 + }, + { + "epoch": 6.625, + "grad_norm": 0.29527369141578674, + "learning_rate": 2.3520000000000002e-05, + "loss": 0.0343, + "step": 7844 + }, + { + "epoch": 6.625844594594595, + "grad_norm": 0.3402997851371765, + "learning_rate": 2.3523e-05, + "loss": 0.0441, + "step": 7845 + }, + { + "epoch": 6.6266891891891895, + "grad_norm": 0.6348522901535034, + "learning_rate": 2.3526e-05, + "loss": 0.0659, + "step": 7846 + }, + { + "epoch": 6.627533783783784, + "grad_norm": 0.5410457849502563, + "learning_rate": 2.3529e-05, + "loss": 0.0458, + "step": 7847 + }, + { + "epoch": 6.628378378378378, + "grad_norm": 0.6527268886566162, + "learning_rate": 2.3532e-05, + "loss": 0.0567, + "step": 7848 + }, + { + "epoch": 6.629222972972973, + "grad_norm": 0.3924355208873749, + "learning_rate": 2.3535e-05, + "loss": 0.0382, + "step": 7849 + }, + { + "epoch": 6.6300675675675675, + "grad_norm": 0.47390615940093994, + "learning_rate": 2.3538e-05, + "loss": 0.0429, + "step": 7850 + }, + { + "epoch": 6.630912162162162, + "grad_norm": 0.6265420913696289, + "learning_rate": 2.3541e-05, + "loss": 0.059, + "step": 7851 + }, + { + "epoch": 6.631756756756757, + "grad_norm": 1.0769356489181519, + "learning_rate": 2.3544000000000004e-05, + "loss": 0.0572, + "step": 7852 + }, + { + "epoch": 6.632601351351351, + "grad_norm": 0.8569247722625732, + "learning_rate": 2.3547000000000003e-05, + "loss": 0.0717, + "step": 7853 + }, + { + "epoch": 6.633445945945946, + "grad_norm": 0.887245237827301, + "learning_rate": 2.3550000000000003e-05, + "loss": 0.0801, + "step": 7854 + }, + { + "epoch": 6.63429054054054, + "grad_norm": 1.5861082077026367, + "learning_rate": 2.3553e-05, + "loss": 0.315, + "step": 7855 + }, + { + "epoch": 6.635135135135135, + "grad_norm": 0.7606345415115356, + "learning_rate": 2.3556e-05, + "loss": 0.1964, + "step": 7856 + }, + { + "epoch": 6.63597972972973, + "grad_norm": 0.5568476319313049, + "learning_rate": 2.3559e-05, + "loss": 0.0968, + "step": 7857 + }, + { + "epoch": 6.636824324324325, + "grad_norm": 0.4784906208515167, + "learning_rate": 2.3562e-05, + "loss": 0.1004, + "step": 7858 + }, + { + "epoch": 6.637668918918919, + "grad_norm": 0.4716562330722809, + "learning_rate": 2.3565e-05, + "loss": 0.0743, + "step": 7859 + }, + { + "epoch": 6.638513513513513, + "grad_norm": 0.41028228402137756, + "learning_rate": 2.3568e-05, + "loss": 0.0606, + "step": 7860 + }, + { + "epoch": 6.639358108108108, + "grad_norm": 0.4371851980686188, + "learning_rate": 2.3571e-05, + "loss": 0.0453, + "step": 7861 + }, + { + "epoch": 6.640202702702703, + "grad_norm": 0.42675310373306274, + "learning_rate": 2.3574e-05, + "loss": 0.0476, + "step": 7862 + }, + { + "epoch": 6.641047297297297, + "grad_norm": 0.43350929021835327, + "learning_rate": 2.3577e-05, + "loss": 0.0565, + "step": 7863 + }, + { + "epoch": 6.641891891891892, + "grad_norm": 0.4200706481933594, + "learning_rate": 2.358e-05, + "loss": 0.0594, + "step": 7864 + }, + { + "epoch": 6.642736486486487, + "grad_norm": 0.3895255923271179, + "learning_rate": 2.3583e-05, + "loss": 0.054, + "step": 7865 + }, + { + "epoch": 6.643581081081081, + "grad_norm": 0.33602097630500793, + "learning_rate": 2.3586e-05, + "loss": 0.0362, + "step": 7866 + }, + { + "epoch": 6.644425675675675, + "grad_norm": 0.48384472727775574, + "learning_rate": 2.3589e-05, + "loss": 0.0475, + "step": 7867 + }, + { + "epoch": 6.64527027027027, + "grad_norm": 0.6640112400054932, + "learning_rate": 2.3592e-05, + "loss": 0.0454, + "step": 7868 + }, + { + "epoch": 6.646114864864865, + "grad_norm": 0.6257356405258179, + "learning_rate": 2.3595e-05, + "loss": 0.0381, + "step": 7869 + }, + { + "epoch": 6.64695945945946, + "grad_norm": 0.7452064156532288, + "learning_rate": 2.3598e-05, + "loss": 0.0401, + "step": 7870 + }, + { + "epoch": 6.647804054054054, + "grad_norm": 0.5554962754249573, + "learning_rate": 2.3601e-05, + "loss": 0.057, + "step": 7871 + }, + { + "epoch": 6.648648648648649, + "grad_norm": 0.7091006636619568, + "learning_rate": 2.3604000000000003e-05, + "loss": 0.0755, + "step": 7872 + }, + { + "epoch": 6.649493243243243, + "grad_norm": 0.5705171227455139, + "learning_rate": 2.3607000000000003e-05, + "loss": 0.0424, + "step": 7873 + }, + { + "epoch": 6.650337837837838, + "grad_norm": 0.434192419052124, + "learning_rate": 2.3610000000000003e-05, + "loss": 0.0464, + "step": 7874 + }, + { + "epoch": 6.6511824324324325, + "grad_norm": 0.5423678159713745, + "learning_rate": 2.3613000000000002e-05, + "loss": 0.0417, + "step": 7875 + }, + { + "epoch": 6.652027027027027, + "grad_norm": 0.3839511573314667, + "learning_rate": 2.3616000000000002e-05, + "loss": 0.0281, + "step": 7876 + }, + { + "epoch": 6.652871621621622, + "grad_norm": 0.522944450378418, + "learning_rate": 2.3619000000000002e-05, + "loss": 0.0509, + "step": 7877 + }, + { + "epoch": 6.653716216216216, + "grad_norm": 0.39247289299964905, + "learning_rate": 2.3622000000000002e-05, + "loss": 0.0364, + "step": 7878 + }, + { + "epoch": 6.6545608108108105, + "grad_norm": 0.5344017744064331, + "learning_rate": 2.3624999999999998e-05, + "loss": 0.0495, + "step": 7879 + }, + { + "epoch": 6.655405405405405, + "grad_norm": 0.7351146340370178, + "learning_rate": 2.3627999999999998e-05, + "loss": 0.0576, + "step": 7880 + }, + { + "epoch": 6.65625, + "grad_norm": 0.7042700052261353, + "learning_rate": 2.3630999999999998e-05, + "loss": 0.0321, + "step": 7881 + }, + { + "epoch": 6.657094594594595, + "grad_norm": 0.5313669443130493, + "learning_rate": 2.3633999999999998e-05, + "loss": 0.0405, + "step": 7882 + }, + { + "epoch": 6.6579391891891895, + "grad_norm": 0.37574586272239685, + "learning_rate": 2.3637e-05, + "loss": 0.0241, + "step": 7883 + }, + { + "epoch": 6.658783783783784, + "grad_norm": 0.5710436105728149, + "learning_rate": 2.364e-05, + "loss": 0.06, + "step": 7884 + }, + { + "epoch": 6.659628378378378, + "grad_norm": 0.4621107578277588, + "learning_rate": 2.3643e-05, + "loss": 0.0471, + "step": 7885 + }, + { + "epoch": 6.660472972972973, + "grad_norm": 0.5319846868515015, + "learning_rate": 2.3646e-05, + "loss": 0.0448, + "step": 7886 + }, + { + "epoch": 6.6613175675675675, + "grad_norm": 0.4431034028530121, + "learning_rate": 2.3649e-05, + "loss": 0.0316, + "step": 7887 + }, + { + "epoch": 6.662162162162162, + "grad_norm": 0.9290576577186584, + "learning_rate": 2.3652e-05, + "loss": 0.1159, + "step": 7888 + }, + { + "epoch": 6.663006756756757, + "grad_norm": 0.342990905046463, + "learning_rate": 2.3655e-05, + "loss": 0.0276, + "step": 7889 + }, + { + "epoch": 6.663851351351351, + "grad_norm": 0.3842092752456665, + "learning_rate": 2.3658e-05, + "loss": 0.0351, + "step": 7890 + }, + { + "epoch": 6.664695945945946, + "grad_norm": 0.513401985168457, + "learning_rate": 2.3661e-05, + "loss": 0.0319, + "step": 7891 + }, + { + "epoch": 6.66554054054054, + "grad_norm": 0.6814237236976624, + "learning_rate": 2.3664e-05, + "loss": 0.0518, + "step": 7892 + }, + { + "epoch": 6.666385135135135, + "grad_norm": 1.468707799911499, + "learning_rate": 2.3667000000000002e-05, + "loss": 0.0511, + "step": 7893 + }, + { + "epoch": 6.66722972972973, + "grad_norm": 0.5083751678466797, + "learning_rate": 2.3670000000000002e-05, + "loss": 0.0417, + "step": 7894 + }, + { + "epoch": 6.668074324324325, + "grad_norm": 0.7901058793067932, + "learning_rate": 2.3673000000000002e-05, + "loss": 0.0522, + "step": 7895 + }, + { + "epoch": 6.668918918918919, + "grad_norm": 0.7829826474189758, + "learning_rate": 2.3676e-05, + "loss": 0.0412, + "step": 7896 + }, + { + "epoch": 6.669763513513513, + "grad_norm": 0.30395743250846863, + "learning_rate": 2.3679e-05, + "loss": 0.0289, + "step": 7897 + }, + { + "epoch": 6.670608108108108, + "grad_norm": 0.6999709010124207, + "learning_rate": 2.3682e-05, + "loss": 0.1021, + "step": 7898 + }, + { + "epoch": 6.671452702702703, + "grad_norm": 0.41097337007522583, + "learning_rate": 2.3685e-05, + "loss": 0.041, + "step": 7899 + }, + { + "epoch": 6.672297297297297, + "grad_norm": 0.41861385107040405, + "learning_rate": 2.3688e-05, + "loss": 0.036, + "step": 7900 + }, + { + "epoch": 6.673141891891892, + "grad_norm": 0.6194043159484863, + "learning_rate": 2.3691e-05, + "loss": 0.06, + "step": 7901 + }, + { + "epoch": 6.673986486486487, + "grad_norm": 1.0487157106399536, + "learning_rate": 2.3694e-05, + "loss": 0.0581, + "step": 7902 + }, + { + "epoch": 6.674831081081081, + "grad_norm": 0.8221165537834167, + "learning_rate": 2.3697000000000004e-05, + "loss": 0.0847, + "step": 7903 + }, + { + "epoch": 6.675675675675675, + "grad_norm": 1.04623281955719, + "learning_rate": 2.37e-05, + "loss": 0.0823, + "step": 7904 + }, + { + "epoch": 6.67652027027027, + "grad_norm": 3.185835361480713, + "learning_rate": 2.3703e-05, + "loss": 0.4376, + "step": 7905 + }, + { + "epoch": 6.677364864864865, + "grad_norm": 1.4107036590576172, + "learning_rate": 2.3706e-05, + "loss": 0.2554, + "step": 7906 + }, + { + "epoch": 6.67820945945946, + "grad_norm": 0.8462070822715759, + "learning_rate": 2.3709e-05, + "loss": 0.1296, + "step": 7907 + }, + { + "epoch": 6.679054054054054, + "grad_norm": 0.6940361261367798, + "learning_rate": 2.3712e-05, + "loss": 0.0686, + "step": 7908 + }, + { + "epoch": 6.679898648648649, + "grad_norm": 0.5947845578193665, + "learning_rate": 2.3715e-05, + "loss": 0.064, + "step": 7909 + }, + { + "epoch": 6.680743243243243, + "grad_norm": 0.4145672023296356, + "learning_rate": 2.3718e-05, + "loss": 0.0533, + "step": 7910 + }, + { + "epoch": 6.681587837837838, + "grad_norm": 0.33295488357543945, + "learning_rate": 2.3721e-05, + "loss": 0.0404, + "step": 7911 + }, + { + "epoch": 6.6824324324324325, + "grad_norm": 0.34794580936431885, + "learning_rate": 2.3724e-05, + "loss": 0.0392, + "step": 7912 + }, + { + "epoch": 6.683277027027027, + "grad_norm": 0.38753896951675415, + "learning_rate": 2.3727000000000002e-05, + "loss": 0.0466, + "step": 7913 + }, + { + "epoch": 6.684121621621622, + "grad_norm": 0.4955885410308838, + "learning_rate": 2.373e-05, + "loss": 0.0634, + "step": 7914 + }, + { + "epoch": 6.684966216216216, + "grad_norm": 0.5834840536117554, + "learning_rate": 2.3733e-05, + "loss": 0.0631, + "step": 7915 + }, + { + "epoch": 6.6858108108108105, + "grad_norm": 0.39592108130455017, + "learning_rate": 2.3736e-05, + "loss": 0.0539, + "step": 7916 + }, + { + "epoch": 6.686655405405405, + "grad_norm": 0.5847482681274414, + "learning_rate": 2.3739e-05, + "loss": 0.0473, + "step": 7917 + }, + { + "epoch": 6.6875, + "grad_norm": 0.4436328113079071, + "learning_rate": 2.3742e-05, + "loss": 0.0503, + "step": 7918 + }, + { + "epoch": 6.688344594594595, + "grad_norm": 0.32145369052886963, + "learning_rate": 2.3745e-05, + "loss": 0.0361, + "step": 7919 + }, + { + "epoch": 6.6891891891891895, + "grad_norm": 0.3701646625995636, + "learning_rate": 2.3748e-05, + "loss": 0.0314, + "step": 7920 + }, + { + "epoch": 6.690033783783784, + "grad_norm": 0.44720351696014404, + "learning_rate": 2.3751e-05, + "loss": 0.0469, + "step": 7921 + }, + { + "epoch": 6.690878378378378, + "grad_norm": 0.3157263398170471, + "learning_rate": 2.3754e-05, + "loss": 0.0315, + "step": 7922 + }, + { + "epoch": 6.691722972972973, + "grad_norm": 0.624514639377594, + "learning_rate": 2.3757000000000003e-05, + "loss": 0.0654, + "step": 7923 + }, + { + "epoch": 6.6925675675675675, + "grad_norm": 0.3958071768283844, + "learning_rate": 2.3760000000000003e-05, + "loss": 0.0409, + "step": 7924 + }, + { + "epoch": 6.693412162162162, + "grad_norm": 0.6174474358558655, + "learning_rate": 2.3763000000000003e-05, + "loss": 0.061, + "step": 7925 + }, + { + "epoch": 6.694256756756757, + "grad_norm": 0.4544570744037628, + "learning_rate": 2.3766000000000003e-05, + "loss": 0.037, + "step": 7926 + }, + { + "epoch": 6.695101351351351, + "grad_norm": 0.47464877367019653, + "learning_rate": 2.3769000000000002e-05, + "loss": 0.0568, + "step": 7927 + }, + { + "epoch": 6.695945945945946, + "grad_norm": 0.7219470143318176, + "learning_rate": 2.3772e-05, + "loss": 0.1325, + "step": 7928 + }, + { + "epoch": 6.69679054054054, + "grad_norm": 0.3955022692680359, + "learning_rate": 2.3775e-05, + "loss": 0.0692, + "step": 7929 + }, + { + "epoch": 6.697635135135135, + "grad_norm": 0.43559494614601135, + "learning_rate": 2.3778e-05, + "loss": 0.0359, + "step": 7930 + }, + { + "epoch": 6.69847972972973, + "grad_norm": 0.5472060441970825, + "learning_rate": 2.3780999999999998e-05, + "loss": 0.0653, + "step": 7931 + }, + { + "epoch": 6.699324324324325, + "grad_norm": 0.47819727659225464, + "learning_rate": 2.3783999999999998e-05, + "loss": 0.0404, + "step": 7932 + }, + { + "epoch": 6.700168918918919, + "grad_norm": 0.3225521147251129, + "learning_rate": 2.3787e-05, + "loss": 0.035, + "step": 7933 + }, + { + "epoch": 6.701013513513513, + "grad_norm": 0.3583749532699585, + "learning_rate": 2.379e-05, + "loss": 0.0414, + "step": 7934 + }, + { + "epoch": 6.701858108108108, + "grad_norm": 0.44690456986427307, + "learning_rate": 2.3793e-05, + "loss": 0.0513, + "step": 7935 + }, + { + "epoch": 6.702702702702703, + "grad_norm": 0.37176838517189026, + "learning_rate": 2.3796e-05, + "loss": 0.0427, + "step": 7936 + }, + { + "epoch": 6.703547297297297, + "grad_norm": 0.5371810793876648, + "learning_rate": 2.3799e-05, + "loss": 0.0445, + "step": 7937 + }, + { + "epoch": 6.704391891891892, + "grad_norm": 0.4688936471939087, + "learning_rate": 2.3802e-05, + "loss": 0.0287, + "step": 7938 + }, + { + "epoch": 6.705236486486487, + "grad_norm": 0.368728369474411, + "learning_rate": 2.3805e-05, + "loss": 0.0446, + "step": 7939 + }, + { + "epoch": 6.706081081081081, + "grad_norm": 0.3904072642326355, + "learning_rate": 2.3808e-05, + "loss": 0.043, + "step": 7940 + }, + { + "epoch": 6.706925675675675, + "grad_norm": 0.5052811503410339, + "learning_rate": 2.3811e-05, + "loss": 0.058, + "step": 7941 + }, + { + "epoch": 6.70777027027027, + "grad_norm": 0.4655936062335968, + "learning_rate": 2.3814e-05, + "loss": 0.0445, + "step": 7942 + }, + { + "epoch": 6.708614864864865, + "grad_norm": 0.42475420236587524, + "learning_rate": 2.3817000000000003e-05, + "loss": 0.0309, + "step": 7943 + }, + { + "epoch": 6.70945945945946, + "grad_norm": 0.6578283905982971, + "learning_rate": 2.3820000000000002e-05, + "loss": 0.0444, + "step": 7944 + }, + { + "epoch": 6.710304054054054, + "grad_norm": 0.6387541890144348, + "learning_rate": 2.3823000000000002e-05, + "loss": 0.0535, + "step": 7945 + }, + { + "epoch": 6.711148648648649, + "grad_norm": 0.6250705718994141, + "learning_rate": 2.3826000000000002e-05, + "loss": 0.04, + "step": 7946 + }, + { + "epoch": 6.711993243243243, + "grad_norm": 0.4692637622356415, + "learning_rate": 2.3829000000000002e-05, + "loss": 0.0278, + "step": 7947 + }, + { + "epoch": 6.712837837837838, + "grad_norm": 0.5051496028900146, + "learning_rate": 2.3832e-05, + "loss": 0.0456, + "step": 7948 + }, + { + "epoch": 6.7136824324324325, + "grad_norm": 0.9369716644287109, + "learning_rate": 2.3835e-05, + "loss": 0.0262, + "step": 7949 + }, + { + "epoch": 6.714527027027027, + "grad_norm": 0.7364141941070557, + "learning_rate": 2.3838e-05, + "loss": 0.0537, + "step": 7950 + }, + { + "epoch": 6.715371621621622, + "grad_norm": 0.4797641336917877, + "learning_rate": 2.3841e-05, + "loss": 0.0445, + "step": 7951 + }, + { + "epoch": 6.716216216216216, + "grad_norm": 0.6475617289543152, + "learning_rate": 2.3844e-05, + "loss": 0.043, + "step": 7952 + }, + { + "epoch": 6.7170608108108105, + "grad_norm": 0.5836294293403625, + "learning_rate": 2.3847e-05, + "loss": 0.061, + "step": 7953 + }, + { + "epoch": 6.717905405405405, + "grad_norm": 1.7124524116516113, + "learning_rate": 2.385e-05, + "loss": 0.076, + "step": 7954 + }, + { + "epoch": 6.71875, + "grad_norm": 1.011040449142456, + "learning_rate": 2.3853e-05, + "loss": 0.3527, + "step": 7955 + }, + { + "epoch": 6.719594594594595, + "grad_norm": 0.6260673999786377, + "learning_rate": 2.3856e-05, + "loss": 0.1509, + "step": 7956 + }, + { + "epoch": 6.7204391891891895, + "grad_norm": 0.7283392548561096, + "learning_rate": 2.3859e-05, + "loss": 0.0995, + "step": 7957 + }, + { + "epoch": 6.721283783783784, + "grad_norm": 0.560309886932373, + "learning_rate": 2.3862e-05, + "loss": 0.062, + "step": 7958 + }, + { + "epoch": 6.722128378378378, + "grad_norm": 0.47182348370552063, + "learning_rate": 2.3865e-05, + "loss": 0.0524, + "step": 7959 + }, + { + "epoch": 6.722972972972973, + "grad_norm": 0.44893771409988403, + "learning_rate": 2.3868e-05, + "loss": 0.0519, + "step": 7960 + }, + { + "epoch": 6.7238175675675675, + "grad_norm": 0.43095651268959045, + "learning_rate": 2.3871e-05, + "loss": 0.0532, + "step": 7961 + }, + { + "epoch": 6.724662162162162, + "grad_norm": 0.40875738859176636, + "learning_rate": 2.3874e-05, + "loss": 0.0489, + "step": 7962 + }, + { + "epoch": 6.725506756756757, + "grad_norm": 0.36171218752861023, + "learning_rate": 2.3877000000000002e-05, + "loss": 0.0388, + "step": 7963 + }, + { + "epoch": 6.726351351351351, + "grad_norm": 0.28395047783851624, + "learning_rate": 2.3880000000000002e-05, + "loss": 0.0251, + "step": 7964 + }, + { + "epoch": 6.727195945945946, + "grad_norm": 0.32228049635887146, + "learning_rate": 2.3883e-05, + "loss": 0.0468, + "step": 7965 + }, + { + "epoch": 6.72804054054054, + "grad_norm": 0.29778361320495605, + "learning_rate": 2.3886e-05, + "loss": 0.0321, + "step": 7966 + }, + { + "epoch": 6.728885135135135, + "grad_norm": 0.8862317800521851, + "learning_rate": 2.3889e-05, + "loss": 0.0624, + "step": 7967 + }, + { + "epoch": 6.72972972972973, + "grad_norm": 0.7577795386314392, + "learning_rate": 2.3892e-05, + "loss": 0.0651, + "step": 7968 + }, + { + "epoch": 6.730574324324325, + "grad_norm": 0.5131037831306458, + "learning_rate": 2.3895e-05, + "loss": 0.0449, + "step": 7969 + }, + { + "epoch": 6.731418918918919, + "grad_norm": 0.43260106444358826, + "learning_rate": 2.3898e-05, + "loss": 0.0322, + "step": 7970 + }, + { + "epoch": 6.732263513513513, + "grad_norm": 0.5967618227005005, + "learning_rate": 2.3901e-05, + "loss": 0.0663, + "step": 7971 + }, + { + "epoch": 6.733108108108108, + "grad_norm": 0.7026291489601135, + "learning_rate": 2.3904e-05, + "loss": 0.0419, + "step": 7972 + }, + { + "epoch": 6.733952702702703, + "grad_norm": 0.5362334847450256, + "learning_rate": 2.3907000000000003e-05, + "loss": 0.0483, + "step": 7973 + }, + { + "epoch": 6.734797297297297, + "grad_norm": 0.41238507628440857, + "learning_rate": 2.3910000000000003e-05, + "loss": 0.0397, + "step": 7974 + }, + { + "epoch": 6.735641891891892, + "grad_norm": 0.7877869009971619, + "learning_rate": 2.3913000000000003e-05, + "loss": 0.051, + "step": 7975 + }, + { + "epoch": 6.736486486486487, + "grad_norm": 0.3214462399482727, + "learning_rate": 2.3916000000000003e-05, + "loss": 0.0245, + "step": 7976 + }, + { + "epoch": 6.737331081081081, + "grad_norm": 0.4441215991973877, + "learning_rate": 2.3919e-05, + "loss": 0.0448, + "step": 7977 + }, + { + "epoch": 6.738175675675675, + "grad_norm": 0.5020055174827576, + "learning_rate": 2.3922e-05, + "loss": 0.0396, + "step": 7978 + }, + { + "epoch": 6.73902027027027, + "grad_norm": 0.330802857875824, + "learning_rate": 2.3925e-05, + "loss": 0.023, + "step": 7979 + }, + { + "epoch": 6.739864864864865, + "grad_norm": 0.5056531429290771, + "learning_rate": 2.3928e-05, + "loss": 0.0472, + "step": 7980 + }, + { + "epoch": 6.74070945945946, + "grad_norm": 0.4204583168029785, + "learning_rate": 2.3931e-05, + "loss": 0.0383, + "step": 7981 + }, + { + "epoch": 6.741554054054054, + "grad_norm": 0.36279550194740295, + "learning_rate": 2.3933999999999998e-05, + "loss": 0.032, + "step": 7982 + }, + { + "epoch": 6.742398648648649, + "grad_norm": 0.4076613187789917, + "learning_rate": 2.3937e-05, + "loss": 0.0333, + "step": 7983 + }, + { + "epoch": 6.743243243243243, + "grad_norm": 0.655422031879425, + "learning_rate": 2.394e-05, + "loss": 0.0479, + "step": 7984 + }, + { + "epoch": 6.744087837837838, + "grad_norm": 0.4481765925884247, + "learning_rate": 2.3943e-05, + "loss": 0.0418, + "step": 7985 + }, + { + "epoch": 6.7449324324324325, + "grad_norm": 1.4324251413345337, + "learning_rate": 2.3946e-05, + "loss": 0.0431, + "step": 7986 + }, + { + "epoch": 6.745777027027027, + "grad_norm": 0.5347203016281128, + "learning_rate": 2.3949e-05, + "loss": 0.0371, + "step": 7987 + }, + { + "epoch": 6.746621621621622, + "grad_norm": 0.6563568115234375, + "learning_rate": 2.3952e-05, + "loss": 0.0697, + "step": 7988 + }, + { + "epoch": 6.747466216216216, + "grad_norm": 0.8118888139724731, + "learning_rate": 2.3955e-05, + "loss": 0.0649, + "step": 7989 + }, + { + "epoch": 6.7483108108108105, + "grad_norm": 0.6382665634155273, + "learning_rate": 2.3958e-05, + "loss": 0.0646, + "step": 7990 + }, + { + "epoch": 6.749155405405405, + "grad_norm": 0.3853220045566559, + "learning_rate": 2.3961e-05, + "loss": 0.0439, + "step": 7991 + }, + { + "epoch": 6.75, + "grad_norm": 0.3668144941329956, + "learning_rate": 2.3964e-05, + "loss": 0.0339, + "step": 7992 + }, + { + "epoch": 6.750844594594595, + "grad_norm": 0.37278127670288086, + "learning_rate": 2.3967000000000003e-05, + "loss": 0.0333, + "step": 7993 + }, + { + "epoch": 6.7516891891891895, + "grad_norm": 0.6185402870178223, + "learning_rate": 2.3970000000000003e-05, + "loss": 0.0514, + "step": 7994 + }, + { + "epoch": 6.752533783783784, + "grad_norm": 0.6018336415290833, + "learning_rate": 2.3973000000000002e-05, + "loss": 0.0586, + "step": 7995 + }, + { + "epoch": 6.753378378378378, + "grad_norm": 0.4356682002544403, + "learning_rate": 2.3976000000000002e-05, + "loss": 0.0502, + "step": 7996 + }, + { + "epoch": 6.754222972972973, + "grad_norm": 0.5112601518630981, + "learning_rate": 2.3979000000000002e-05, + "loss": 0.0755, + "step": 7997 + }, + { + "epoch": 6.7550675675675675, + "grad_norm": 0.6434314250946045, + "learning_rate": 2.3982000000000002e-05, + "loss": 0.0522, + "step": 7998 + }, + { + "epoch": 6.755912162162162, + "grad_norm": 0.39840489625930786, + "learning_rate": 2.3985e-05, + "loss": 0.0515, + "step": 7999 + }, + { + "epoch": 6.756756756756757, + "grad_norm": 0.6303961277008057, + "learning_rate": 2.3988e-05, + "loss": 0.0704, + "step": 8000 + }, + { + "epoch": 6.756756756756757, + "eval_cer": 0.09677937515059032, + "eval_loss": 0.2802891433238983, + "eval_runtime": 8.4248, + "eval_samples_per_second": 11.988, + "eval_steps_per_second": 0.475, + "eval_wer": 0.29717813051146386, + "step": 8000 + }, + { + "epoch": 6.757601351351351, + "grad_norm": 0.44308581948280334, + "learning_rate": 2.3991e-05, + "loss": 0.0448, + "step": 8001 + }, + { + "epoch": 6.758445945945946, + "grad_norm": 0.4108806550502777, + "learning_rate": 2.3993999999999998e-05, + "loss": 0.0434, + "step": 8002 + }, + { + "epoch": 6.75929054054054, + "grad_norm": 1.3033620119094849, + "learning_rate": 2.3997e-05, + "loss": 0.0671, + "step": 8003 + }, + { + "epoch": 6.760135135135135, + "grad_norm": 1.004335880279541, + "learning_rate": 2.4e-05, + "loss": 0.0814, + "step": 8004 + }, + { + "epoch": 6.76097972972973, + "grad_norm": 1.416593313217163, + "learning_rate": 2.4003e-05, + "loss": 0.3451, + "step": 8005 + }, + { + "epoch": 6.761824324324325, + "grad_norm": 0.7006650567054749, + "learning_rate": 2.4006e-05, + "loss": 0.1987, + "step": 8006 + }, + { + "epoch": 6.762668918918919, + "grad_norm": 0.6587291955947876, + "learning_rate": 2.4009e-05, + "loss": 0.0966, + "step": 8007 + }, + { + "epoch": 6.763513513513513, + "grad_norm": 0.5363113284111023, + "learning_rate": 2.4012e-05, + "loss": 0.061, + "step": 8008 + }, + { + "epoch": 6.764358108108108, + "grad_norm": 0.32790517807006836, + "learning_rate": 2.4015e-05, + "loss": 0.0428, + "step": 8009 + }, + { + "epoch": 6.765202702702703, + "grad_norm": 0.48907914757728577, + "learning_rate": 2.4018e-05, + "loss": 0.0459, + "step": 8010 + }, + { + "epoch": 6.766047297297297, + "grad_norm": 0.5246337652206421, + "learning_rate": 2.4021e-05, + "loss": 0.0519, + "step": 8011 + }, + { + "epoch": 6.766891891891892, + "grad_norm": 0.296391099691391, + "learning_rate": 2.4024e-05, + "loss": 0.0359, + "step": 8012 + }, + { + "epoch": 6.767736486486487, + "grad_norm": 0.4817016124725342, + "learning_rate": 2.4027e-05, + "loss": 0.0433, + "step": 8013 + }, + { + "epoch": 6.768581081081081, + "grad_norm": 0.4367753565311432, + "learning_rate": 2.4030000000000002e-05, + "loss": 0.0397, + "step": 8014 + }, + { + "epoch": 6.769425675675675, + "grad_norm": 0.4429146349430084, + "learning_rate": 2.4033000000000002e-05, + "loss": 0.0337, + "step": 8015 + }, + { + "epoch": 6.77027027027027, + "grad_norm": 0.5270665287971497, + "learning_rate": 2.4036e-05, + "loss": 0.0533, + "step": 8016 + }, + { + "epoch": 6.771114864864865, + "grad_norm": 1.0480570793151855, + "learning_rate": 2.4039e-05, + "loss": 0.0481, + "step": 8017 + }, + { + "epoch": 6.77195945945946, + "grad_norm": 0.43193286657333374, + "learning_rate": 2.4042e-05, + "loss": 0.0414, + "step": 8018 + }, + { + "epoch": 6.772804054054054, + "grad_norm": 0.6094619035720825, + "learning_rate": 2.4045e-05, + "loss": 0.0323, + "step": 8019 + }, + { + "epoch": 6.773648648648649, + "grad_norm": 0.9599156975746155, + "learning_rate": 2.4048e-05, + "loss": 0.0384, + "step": 8020 + }, + { + "epoch": 6.774493243243243, + "grad_norm": 0.8494139313697815, + "learning_rate": 2.4051e-05, + "loss": 0.0569, + "step": 8021 + }, + { + "epoch": 6.775337837837838, + "grad_norm": 0.4733354449272156, + "learning_rate": 2.4054e-05, + "loss": 0.0386, + "step": 8022 + }, + { + "epoch": 6.7761824324324325, + "grad_norm": 0.5961107611656189, + "learning_rate": 2.4057e-05, + "loss": 0.0521, + "step": 8023 + }, + { + "epoch": 6.777027027027027, + "grad_norm": 0.5026003122329712, + "learning_rate": 2.4060000000000003e-05, + "loss": 0.0528, + "step": 8024 + }, + { + "epoch": 6.777871621621622, + "grad_norm": 0.41302597522735596, + "learning_rate": 2.4063000000000003e-05, + "loss": 0.0425, + "step": 8025 + }, + { + "epoch": 6.778716216216216, + "grad_norm": 0.39434587955474854, + "learning_rate": 2.4066000000000003e-05, + "loss": 0.0307, + "step": 8026 + }, + { + "epoch": 6.7795608108108105, + "grad_norm": 0.7852178812026978, + "learning_rate": 2.4069e-05, + "loss": 0.0524, + "step": 8027 + }, + { + "epoch": 6.780405405405405, + "grad_norm": 0.3433028757572174, + "learning_rate": 2.4072e-05, + "loss": 0.0367, + "step": 8028 + }, + { + "epoch": 6.78125, + "grad_norm": 0.45503973960876465, + "learning_rate": 2.4075e-05, + "loss": 0.0551, + "step": 8029 + }, + { + "epoch": 6.782094594594595, + "grad_norm": 0.49220892786979675, + "learning_rate": 2.4078e-05, + "loss": 0.0474, + "step": 8030 + }, + { + "epoch": 6.7829391891891895, + "grad_norm": 0.3981021046638489, + "learning_rate": 2.4081e-05, + "loss": 0.0467, + "step": 8031 + }, + { + "epoch": 6.783783783783784, + "grad_norm": 0.8574933409690857, + "learning_rate": 2.4084e-05, + "loss": 0.0525, + "step": 8032 + }, + { + "epoch": 6.784628378378378, + "grad_norm": 0.8363673686981201, + "learning_rate": 2.4086999999999998e-05, + "loss": 0.0346, + "step": 8033 + }, + { + "epoch": 6.785472972972973, + "grad_norm": 0.4767259359359741, + "learning_rate": 2.409e-05, + "loss": 0.0542, + "step": 8034 + }, + { + "epoch": 6.7863175675675675, + "grad_norm": 0.3833406865596771, + "learning_rate": 2.4093e-05, + "loss": 0.0414, + "step": 8035 + }, + { + "epoch": 6.787162162162162, + "grad_norm": 0.30581867694854736, + "learning_rate": 2.4096e-05, + "loss": 0.0351, + "step": 8036 + }, + { + "epoch": 6.788006756756757, + "grad_norm": 0.47514215111732483, + "learning_rate": 2.4099e-05, + "loss": 0.0402, + "step": 8037 + }, + { + "epoch": 6.788851351351351, + "grad_norm": 0.6622620820999146, + "learning_rate": 2.4102e-05, + "loss": 0.0572, + "step": 8038 + }, + { + "epoch": 6.789695945945946, + "grad_norm": 0.397880882024765, + "learning_rate": 2.4105e-05, + "loss": 0.0272, + "step": 8039 + }, + { + "epoch": 6.79054054054054, + "grad_norm": 0.7730638980865479, + "learning_rate": 2.4108e-05, + "loss": 0.0475, + "step": 8040 + }, + { + "epoch": 6.791385135135135, + "grad_norm": 0.38002026081085205, + "learning_rate": 2.4111e-05, + "loss": 0.029, + "step": 8041 + }, + { + "epoch": 6.79222972972973, + "grad_norm": 0.5147538781166077, + "learning_rate": 2.4114e-05, + "loss": 0.0607, + "step": 8042 + }, + { + "epoch": 6.793074324324325, + "grad_norm": 0.45160797238349915, + "learning_rate": 2.4117e-05, + "loss": 0.0293, + "step": 8043 + }, + { + "epoch": 6.793918918918919, + "grad_norm": 0.4729089140892029, + "learning_rate": 2.4120000000000003e-05, + "loss": 0.0425, + "step": 8044 + }, + { + "epoch": 6.794763513513513, + "grad_norm": 0.8643420934677124, + "learning_rate": 2.4123000000000003e-05, + "loss": 0.0619, + "step": 8045 + }, + { + "epoch": 6.795608108108108, + "grad_norm": 0.8888353705406189, + "learning_rate": 2.4126000000000002e-05, + "loss": 0.0516, + "step": 8046 + }, + { + "epoch": 6.796452702702703, + "grad_norm": 0.45463791489601135, + "learning_rate": 2.4129000000000002e-05, + "loss": 0.0324, + "step": 8047 + }, + { + "epoch": 6.797297297297297, + "grad_norm": 0.5901049971580505, + "learning_rate": 2.4132000000000002e-05, + "loss": 0.0395, + "step": 8048 + }, + { + "epoch": 6.798141891891892, + "grad_norm": 0.853671133518219, + "learning_rate": 2.4135000000000002e-05, + "loss": 0.0535, + "step": 8049 + }, + { + "epoch": 6.798986486486487, + "grad_norm": 0.6535190343856812, + "learning_rate": 2.4138e-05, + "loss": 0.0466, + "step": 8050 + }, + { + "epoch": 6.799831081081081, + "grad_norm": 0.8683130145072937, + "learning_rate": 2.4140999999999998e-05, + "loss": 0.0526, + "step": 8051 + }, + { + "epoch": 6.800675675675675, + "grad_norm": 0.8349970579147339, + "learning_rate": 2.4143999999999998e-05, + "loss": 0.0662, + "step": 8052 + }, + { + "epoch": 6.80152027027027, + "grad_norm": 0.9601642489433289, + "learning_rate": 2.4146999999999998e-05, + "loss": 0.0919, + "step": 8053 + }, + { + "epoch": 6.802364864864865, + "grad_norm": 1.4559308290481567, + "learning_rate": 2.415e-05, + "loss": 0.1433, + "step": 8054 + }, + { + "epoch": 6.80320945945946, + "grad_norm": 0.8781335949897766, + "learning_rate": 2.4153e-05, + "loss": 0.3229, + "step": 8055 + }, + { + "epoch": 6.804054054054054, + "grad_norm": 0.7387393712997437, + "learning_rate": 2.4156e-05, + "loss": 0.1855, + "step": 8056 + }, + { + "epoch": 6.804898648648649, + "grad_norm": 0.665205180644989, + "learning_rate": 2.4159e-05, + "loss": 0.1346, + "step": 8057 + }, + { + "epoch": 6.805743243243243, + "grad_norm": 0.4556819796562195, + "learning_rate": 2.4162e-05, + "loss": 0.051, + "step": 8058 + }, + { + "epoch": 6.806587837837838, + "grad_norm": 0.36555367708206177, + "learning_rate": 2.4165e-05, + "loss": 0.0396, + "step": 8059 + }, + { + "epoch": 6.8074324324324325, + "grad_norm": 0.4370061457157135, + "learning_rate": 2.4168e-05, + "loss": 0.0435, + "step": 8060 + }, + { + "epoch": 6.808277027027027, + "grad_norm": 0.4119870960712433, + "learning_rate": 2.4171e-05, + "loss": 0.0462, + "step": 8061 + }, + { + "epoch": 6.809121621621622, + "grad_norm": 0.39686378836631775, + "learning_rate": 2.4174e-05, + "loss": 0.038, + "step": 8062 + }, + { + "epoch": 6.809966216216216, + "grad_norm": 0.6326102018356323, + "learning_rate": 2.4177e-05, + "loss": 0.0606, + "step": 8063 + }, + { + "epoch": 6.8108108108108105, + "grad_norm": 0.37595877051353455, + "learning_rate": 2.4180000000000002e-05, + "loss": 0.0578, + "step": 8064 + }, + { + "epoch": 6.811655405405405, + "grad_norm": 0.33922579884529114, + "learning_rate": 2.4183000000000002e-05, + "loss": 0.0379, + "step": 8065 + }, + { + "epoch": 6.8125, + "grad_norm": 0.5752536654472351, + "learning_rate": 2.4186000000000002e-05, + "loss": 0.0664, + "step": 8066 + }, + { + "epoch": 6.813344594594595, + "grad_norm": 0.46604159474372864, + "learning_rate": 2.4189e-05, + "loss": 0.0543, + "step": 8067 + }, + { + "epoch": 6.8141891891891895, + "grad_norm": 0.5282436013221741, + "learning_rate": 2.4192e-05, + "loss": 0.0551, + "step": 8068 + }, + { + "epoch": 6.815033783783784, + "grad_norm": 0.3954164385795593, + "learning_rate": 2.4195e-05, + "loss": 0.0509, + "step": 8069 + }, + { + "epoch": 6.815878378378378, + "grad_norm": 0.4431390166282654, + "learning_rate": 2.4198e-05, + "loss": 0.0362, + "step": 8070 + }, + { + "epoch": 6.816722972972973, + "grad_norm": 0.32982349395751953, + "learning_rate": 2.4201e-05, + "loss": 0.0412, + "step": 8071 + }, + { + "epoch": 6.8175675675675675, + "grad_norm": 0.420118123292923, + "learning_rate": 2.4204e-05, + "loss": 0.057, + "step": 8072 + }, + { + "epoch": 6.818412162162162, + "grad_norm": 0.35787826776504517, + "learning_rate": 2.4207e-05, + "loss": 0.0324, + "step": 8073 + }, + { + "epoch": 6.819256756756757, + "grad_norm": 0.4289657771587372, + "learning_rate": 2.4210000000000004e-05, + "loss": 0.0489, + "step": 8074 + }, + { + "epoch": 6.820101351351351, + "grad_norm": 0.4023676812648773, + "learning_rate": 2.4213000000000003e-05, + "loss": 0.0347, + "step": 8075 + }, + { + "epoch": 6.820945945945946, + "grad_norm": 0.4888417720794678, + "learning_rate": 2.4216e-05, + "loss": 0.03, + "step": 8076 + }, + { + "epoch": 6.82179054054054, + "grad_norm": 0.366825670003891, + "learning_rate": 2.4219e-05, + "loss": 0.0321, + "step": 8077 + }, + { + "epoch": 6.822635135135135, + "grad_norm": 0.9606490135192871, + "learning_rate": 2.4222e-05, + "loss": 0.0448, + "step": 8078 + }, + { + "epoch": 6.82347972972973, + "grad_norm": 0.45618727803230286, + "learning_rate": 2.4225e-05, + "loss": 0.041, + "step": 8079 + }, + { + "epoch": 6.824324324324325, + "grad_norm": 0.6308690905570984, + "learning_rate": 2.4228e-05, + "loss": 0.0668, + "step": 8080 + }, + { + "epoch": 6.825168918918919, + "grad_norm": 0.5617489218711853, + "learning_rate": 2.4231e-05, + "loss": 0.0548, + "step": 8081 + }, + { + "epoch": 6.826013513513513, + "grad_norm": 0.4904876947402954, + "learning_rate": 2.4234e-05, + "loss": 0.0461, + "step": 8082 + }, + { + "epoch": 6.826858108108108, + "grad_norm": 0.43315187096595764, + "learning_rate": 2.4237e-05, + "loss": 0.0318, + "step": 8083 + }, + { + "epoch": 6.827702702702703, + "grad_norm": 0.46225738525390625, + "learning_rate": 2.4240000000000002e-05, + "loss": 0.0423, + "step": 8084 + }, + { + "epoch": 6.828547297297297, + "grad_norm": 0.42334893345832825, + "learning_rate": 2.4243e-05, + "loss": 0.0385, + "step": 8085 + }, + { + "epoch": 6.829391891891892, + "grad_norm": 0.49203959107398987, + "learning_rate": 2.4246e-05, + "loss": 0.0363, + "step": 8086 + }, + { + "epoch": 6.830236486486487, + "grad_norm": 1.2240482568740845, + "learning_rate": 2.4249e-05, + "loss": 0.0541, + "step": 8087 + }, + { + "epoch": 6.831081081081081, + "grad_norm": 0.3367334306240082, + "learning_rate": 2.4252e-05, + "loss": 0.0284, + "step": 8088 + }, + { + "epoch": 6.831925675675675, + "grad_norm": 0.5609108805656433, + "learning_rate": 2.4255e-05, + "loss": 0.0375, + "step": 8089 + }, + { + "epoch": 6.83277027027027, + "grad_norm": 0.6699854135513306, + "learning_rate": 2.4258e-05, + "loss": 0.0534, + "step": 8090 + }, + { + "epoch": 6.833614864864865, + "grad_norm": 0.7764503955841064, + "learning_rate": 2.4261e-05, + "loss": 0.0482, + "step": 8091 + }, + { + "epoch": 6.83445945945946, + "grad_norm": 0.41455212235450745, + "learning_rate": 2.4264e-05, + "loss": 0.0356, + "step": 8092 + }, + { + "epoch": 6.835304054054054, + "grad_norm": 0.6042985320091248, + "learning_rate": 2.4267e-05, + "loss": 0.0388, + "step": 8093 + }, + { + "epoch": 6.836148648648649, + "grad_norm": 0.7619005441665649, + "learning_rate": 2.4270000000000003e-05, + "loss": 0.0503, + "step": 8094 + }, + { + "epoch": 6.836993243243243, + "grad_norm": 0.6201984882354736, + "learning_rate": 2.4273000000000003e-05, + "loss": 0.0516, + "step": 8095 + }, + { + "epoch": 6.837837837837838, + "grad_norm": 0.5762540102005005, + "learning_rate": 2.4276000000000003e-05, + "loss": 0.0416, + "step": 8096 + }, + { + "epoch": 6.8386824324324325, + "grad_norm": 0.33187857270240784, + "learning_rate": 2.4279000000000003e-05, + "loss": 0.0235, + "step": 8097 + }, + { + "epoch": 6.839527027027027, + "grad_norm": 0.7174978852272034, + "learning_rate": 2.4282000000000002e-05, + "loss": 0.0417, + "step": 8098 + }, + { + "epoch": 6.840371621621622, + "grad_norm": 0.6136664152145386, + "learning_rate": 2.4285000000000002e-05, + "loss": 0.0355, + "step": 8099 + }, + { + "epoch": 6.841216216216216, + "grad_norm": 0.49558544158935547, + "learning_rate": 2.4288e-05, + "loss": 0.0423, + "step": 8100 + }, + { + "epoch": 6.8420608108108105, + "grad_norm": 0.6700972318649292, + "learning_rate": 2.4291e-05, + "loss": 0.0498, + "step": 8101 + }, + { + "epoch": 6.842905405405405, + "grad_norm": 0.867508590221405, + "learning_rate": 2.4293999999999998e-05, + "loss": 0.0519, + "step": 8102 + }, + { + "epoch": 6.84375, + "grad_norm": 0.8396888375282288, + "learning_rate": 2.4296999999999998e-05, + "loss": 0.0697, + "step": 8103 + }, + { + "epoch": 6.844594594594595, + "grad_norm": 1.268298625946045, + "learning_rate": 2.43e-05, + "loss": 0.1638, + "step": 8104 + }, + { + "epoch": 6.8454391891891895, + "grad_norm": 1.486007809638977, + "learning_rate": 2.4303e-05, + "loss": 0.3703, + "step": 8105 + }, + { + "epoch": 6.846283783783784, + "grad_norm": 0.8151599764823914, + "learning_rate": 2.4306e-05, + "loss": 0.2821, + "step": 8106 + }, + { + "epoch": 6.847128378378378, + "grad_norm": 0.5491477847099304, + "learning_rate": 2.4309e-05, + "loss": 0.0763, + "step": 8107 + }, + { + "epoch": 6.847972972972973, + "grad_norm": 0.46520736813545227, + "learning_rate": 2.4312e-05, + "loss": 0.0813, + "step": 8108 + }, + { + "epoch": 6.8488175675675675, + "grad_norm": 0.590123176574707, + "learning_rate": 2.4315e-05, + "loss": 0.0521, + "step": 8109 + }, + { + "epoch": 6.849662162162162, + "grad_norm": 0.3711545765399933, + "learning_rate": 2.4318e-05, + "loss": 0.046, + "step": 8110 + }, + { + "epoch": 6.850506756756757, + "grad_norm": 0.5010202527046204, + "learning_rate": 2.4321e-05, + "loss": 0.0478, + "step": 8111 + }, + { + "epoch": 6.851351351351351, + "grad_norm": 0.3912036418914795, + "learning_rate": 2.4324e-05, + "loss": 0.0563, + "step": 8112 + }, + { + "epoch": 6.852195945945946, + "grad_norm": 0.6257323622703552, + "learning_rate": 2.4327e-05, + "loss": 0.0585, + "step": 8113 + }, + { + "epoch": 6.85304054054054, + "grad_norm": 0.29826948046684265, + "learning_rate": 2.4330000000000003e-05, + "loss": 0.0286, + "step": 8114 + }, + { + "epoch": 6.853885135135135, + "grad_norm": 0.6219203472137451, + "learning_rate": 2.4333000000000002e-05, + "loss": 0.0599, + "step": 8115 + }, + { + "epoch": 6.85472972972973, + "grad_norm": 0.3901589810848236, + "learning_rate": 2.4336000000000002e-05, + "loss": 0.0439, + "step": 8116 + }, + { + "epoch": 6.855574324324325, + "grad_norm": 0.44871416687965393, + "learning_rate": 2.4339000000000002e-05, + "loss": 0.0531, + "step": 8117 + }, + { + "epoch": 6.856418918918919, + "grad_norm": 0.37633228302001953, + "learning_rate": 2.4342000000000002e-05, + "loss": 0.0378, + "step": 8118 + }, + { + "epoch": 6.857263513513513, + "grad_norm": 0.46283748745918274, + "learning_rate": 2.4345e-05, + "loss": 0.0561, + "step": 8119 + }, + { + "epoch": 6.858108108108108, + "grad_norm": 0.5332757830619812, + "learning_rate": 2.4348e-05, + "loss": 0.0376, + "step": 8120 + }, + { + "epoch": 6.858952702702703, + "grad_norm": 0.3683103024959564, + "learning_rate": 2.4351e-05, + "loss": 0.0331, + "step": 8121 + }, + { + "epoch": 6.859797297297297, + "grad_norm": 0.456961452960968, + "learning_rate": 2.4354e-05, + "loss": 0.0401, + "step": 8122 + }, + { + "epoch": 6.860641891891892, + "grad_norm": 0.5338718295097351, + "learning_rate": 2.4357e-05, + "loss": 0.0575, + "step": 8123 + }, + { + "epoch": 6.861486486486487, + "grad_norm": 0.7113693952560425, + "learning_rate": 2.4360000000000004e-05, + "loss": 0.0367, + "step": 8124 + }, + { + "epoch": 6.862331081081081, + "grad_norm": 0.4236084520816803, + "learning_rate": 2.4363e-05, + "loss": 0.0517, + "step": 8125 + }, + { + "epoch": 6.863175675675675, + "grad_norm": 0.37999290227890015, + "learning_rate": 2.4366e-05, + "loss": 0.0296, + "step": 8126 + }, + { + "epoch": 6.86402027027027, + "grad_norm": 0.3849978744983673, + "learning_rate": 2.4369e-05, + "loss": 0.0416, + "step": 8127 + }, + { + "epoch": 6.864864864864865, + "grad_norm": 0.6038616895675659, + "learning_rate": 2.4372e-05, + "loss": 0.0467, + "step": 8128 + }, + { + "epoch": 6.86570945945946, + "grad_norm": 0.39621153473854065, + "learning_rate": 2.4375e-05, + "loss": 0.0272, + "step": 8129 + }, + { + "epoch": 6.866554054054054, + "grad_norm": 0.5500006079673767, + "learning_rate": 2.4378e-05, + "loss": 0.0386, + "step": 8130 + }, + { + "epoch": 6.867398648648649, + "grad_norm": 0.4364812970161438, + "learning_rate": 2.4381e-05, + "loss": 0.0421, + "step": 8131 + }, + { + "epoch": 6.868243243243243, + "grad_norm": 0.3995268940925598, + "learning_rate": 2.4384e-05, + "loss": 0.0317, + "step": 8132 + }, + { + "epoch": 6.869087837837838, + "grad_norm": 0.5529139041900635, + "learning_rate": 2.4387e-05, + "loss": 0.0546, + "step": 8133 + }, + { + "epoch": 6.8699324324324325, + "grad_norm": 0.37985143065452576, + "learning_rate": 2.439e-05, + "loss": 0.0346, + "step": 8134 + }, + { + "epoch": 6.870777027027027, + "grad_norm": 0.6052584052085876, + "learning_rate": 2.4393000000000002e-05, + "loss": 0.0353, + "step": 8135 + }, + { + "epoch": 6.871621621621622, + "grad_norm": 0.5543026328086853, + "learning_rate": 2.4396e-05, + "loss": 0.0551, + "step": 8136 + }, + { + "epoch": 6.872466216216216, + "grad_norm": 0.4556068778038025, + "learning_rate": 2.4399e-05, + "loss": 0.0511, + "step": 8137 + }, + { + "epoch": 6.8733108108108105, + "grad_norm": 0.42236924171447754, + "learning_rate": 2.4402e-05, + "loss": 0.0425, + "step": 8138 + }, + { + "epoch": 6.874155405405405, + "grad_norm": 0.4618793725967407, + "learning_rate": 2.4405e-05, + "loss": 0.041, + "step": 8139 + }, + { + "epoch": 6.875, + "grad_norm": 0.5096422433853149, + "learning_rate": 2.4408e-05, + "loss": 0.0322, + "step": 8140 + }, + { + "epoch": 6.875844594594595, + "grad_norm": 0.8745798468589783, + "learning_rate": 2.4411e-05, + "loss": 0.0469, + "step": 8141 + }, + { + "epoch": 6.8766891891891895, + "grad_norm": 0.7393580675125122, + "learning_rate": 2.4414e-05, + "loss": 0.044, + "step": 8142 + }, + { + "epoch": 6.877533783783784, + "grad_norm": 0.3497968018054962, + "learning_rate": 2.4417e-05, + "loss": 0.0342, + "step": 8143 + }, + { + "epoch": 6.878378378378378, + "grad_norm": 0.536247193813324, + "learning_rate": 2.442e-05, + "loss": 0.0487, + "step": 8144 + }, + { + "epoch": 6.879222972972973, + "grad_norm": 0.6729299426078796, + "learning_rate": 2.4423000000000003e-05, + "loss": 0.069, + "step": 8145 + }, + { + "epoch": 6.8800675675675675, + "grad_norm": 0.564706563949585, + "learning_rate": 2.4426000000000003e-05, + "loss": 0.0494, + "step": 8146 + }, + { + "epoch": 6.880912162162162, + "grad_norm": 0.4719953238964081, + "learning_rate": 2.4429000000000003e-05, + "loss": 0.0506, + "step": 8147 + }, + { + "epoch": 6.881756756756757, + "grad_norm": 0.40152621269226074, + "learning_rate": 2.4432000000000003e-05, + "loss": 0.0385, + "step": 8148 + }, + { + "epoch": 6.882601351351351, + "grad_norm": 0.8213954567909241, + "learning_rate": 2.4435e-05, + "loss": 0.061, + "step": 8149 + }, + { + "epoch": 6.883445945945946, + "grad_norm": 0.902843177318573, + "learning_rate": 2.4438e-05, + "loss": 0.0617, + "step": 8150 + }, + { + "epoch": 6.88429054054054, + "grad_norm": 0.5885393023490906, + "learning_rate": 2.4441e-05, + "loss": 0.0487, + "step": 8151 + }, + { + "epoch": 6.885135135135135, + "grad_norm": 0.836760401725769, + "learning_rate": 2.4444e-05, + "loss": 0.053, + "step": 8152 + }, + { + "epoch": 6.88597972972973, + "grad_norm": 1.9723845720291138, + "learning_rate": 2.4446999999999998e-05, + "loss": 0.146, + "step": 8153 + }, + { + "epoch": 6.886824324324325, + "grad_norm": 1.1357905864715576, + "learning_rate": 2.4449999999999998e-05, + "loss": 0.1093, + "step": 8154 + }, + { + "epoch": 6.887668918918919, + "grad_norm": 1.2705168724060059, + "learning_rate": 2.4453e-05, + "loss": 0.4271, + "step": 8155 + }, + { + "epoch": 6.888513513513513, + "grad_norm": 0.7743123769760132, + "learning_rate": 2.4456e-05, + "loss": 0.2087, + "step": 8156 + }, + { + "epoch": 6.889358108108108, + "grad_norm": 0.5693255066871643, + "learning_rate": 2.4459e-05, + "loss": 0.0942, + "step": 8157 + }, + { + "epoch": 6.890202702702703, + "grad_norm": 0.34761664271354675, + "learning_rate": 2.4462e-05, + "loss": 0.0471, + "step": 8158 + }, + { + "epoch": 6.891047297297297, + "grad_norm": 0.4569321870803833, + "learning_rate": 2.4465e-05, + "loss": 0.0716, + "step": 8159 + }, + { + "epoch": 6.891891891891892, + "grad_norm": 0.42661648988723755, + "learning_rate": 2.4468e-05, + "loss": 0.0432, + "step": 8160 + }, + { + "epoch": 6.892736486486487, + "grad_norm": 0.4600014388561249, + "learning_rate": 2.4471e-05, + "loss": 0.0605, + "step": 8161 + }, + { + "epoch": 6.893581081081081, + "grad_norm": 0.3529330790042877, + "learning_rate": 2.4474e-05, + "loss": 0.0336, + "step": 8162 + }, + { + "epoch": 6.894425675675675, + "grad_norm": 0.4968368411064148, + "learning_rate": 2.4477e-05, + "loss": 0.0483, + "step": 8163 + }, + { + "epoch": 6.89527027027027, + "grad_norm": 0.3824132978916168, + "learning_rate": 2.448e-05, + "loss": 0.0486, + "step": 8164 + }, + { + "epoch": 6.896114864864865, + "grad_norm": 0.38609468936920166, + "learning_rate": 2.4483000000000003e-05, + "loss": 0.056, + "step": 8165 + }, + { + "epoch": 6.89695945945946, + "grad_norm": 0.4309416115283966, + "learning_rate": 2.4486000000000002e-05, + "loss": 0.0573, + "step": 8166 + }, + { + "epoch": 6.897804054054054, + "grad_norm": 0.6869108080863953, + "learning_rate": 2.4489000000000002e-05, + "loss": 0.034, + "step": 8167 + }, + { + "epoch": 6.898648648648649, + "grad_norm": 0.5496016144752502, + "learning_rate": 2.4492000000000002e-05, + "loss": 0.0444, + "step": 8168 + }, + { + "epoch": 6.899493243243243, + "grad_norm": 0.43887272477149963, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0573, + "step": 8169 + }, + { + "epoch": 6.900337837837838, + "grad_norm": 0.44583284854888916, + "learning_rate": 2.4498e-05, + "loss": 0.0504, + "step": 8170 + }, + { + "epoch": 6.9011824324324325, + "grad_norm": 0.41913485527038574, + "learning_rate": 2.4501e-05, + "loss": 0.0451, + "step": 8171 + }, + { + "epoch": 6.902027027027027, + "grad_norm": 0.4154692590236664, + "learning_rate": 2.4504e-05, + "loss": 0.0347, + "step": 8172 + }, + { + "epoch": 6.902871621621622, + "grad_norm": 0.5488682389259338, + "learning_rate": 2.4507e-05, + "loss": 0.0459, + "step": 8173 + }, + { + "epoch": 6.903716216216216, + "grad_norm": 0.48611709475517273, + "learning_rate": 2.4509999999999997e-05, + "loss": 0.0347, + "step": 8174 + }, + { + "epoch": 6.9045608108108105, + "grad_norm": 0.3440541625022888, + "learning_rate": 2.4513e-05, + "loss": 0.0387, + "step": 8175 + }, + { + "epoch": 6.905405405405405, + "grad_norm": 0.45880308747291565, + "learning_rate": 2.4516e-05, + "loss": 0.0439, + "step": 8176 + }, + { + "epoch": 6.90625, + "grad_norm": 0.5772331953048706, + "learning_rate": 2.4519e-05, + "loss": 0.0699, + "step": 8177 + }, + { + "epoch": 6.907094594594595, + "grad_norm": 0.45249369740486145, + "learning_rate": 2.4522e-05, + "loss": 0.0542, + "step": 8178 + }, + { + "epoch": 6.9079391891891895, + "grad_norm": 0.5484697818756104, + "learning_rate": 2.4525e-05, + "loss": 0.0432, + "step": 8179 + }, + { + "epoch": 6.908783783783784, + "grad_norm": 0.36019182205200195, + "learning_rate": 2.4528e-05, + "loss": 0.0362, + "step": 8180 + }, + { + "epoch": 6.909628378378378, + "grad_norm": 1.6283063888549805, + "learning_rate": 2.4531e-05, + "loss": 0.0864, + "step": 8181 + }, + { + "epoch": 6.910472972972973, + "grad_norm": 0.47667059302330017, + "learning_rate": 2.4534e-05, + "loss": 0.0344, + "step": 8182 + }, + { + "epoch": 6.9113175675675675, + "grad_norm": 0.2808787524700165, + "learning_rate": 2.4537e-05, + "loss": 0.0294, + "step": 8183 + }, + { + "epoch": 6.912162162162162, + "grad_norm": 0.4422146677970886, + "learning_rate": 2.454e-05, + "loss": 0.0353, + "step": 8184 + }, + { + "epoch": 6.913006756756757, + "grad_norm": 0.6303074359893799, + "learning_rate": 2.4543000000000002e-05, + "loss": 0.0439, + "step": 8185 + }, + { + "epoch": 6.913851351351351, + "grad_norm": 0.4465081989765167, + "learning_rate": 2.4546000000000002e-05, + "loss": 0.0498, + "step": 8186 + }, + { + "epoch": 6.914695945945946, + "grad_norm": 0.7821530103683472, + "learning_rate": 2.4549e-05, + "loss": 0.0345, + "step": 8187 + }, + { + "epoch": 6.91554054054054, + "grad_norm": 0.47680115699768066, + "learning_rate": 2.4552e-05, + "loss": 0.0468, + "step": 8188 + }, + { + "epoch": 6.916385135135135, + "grad_norm": 0.5085294842720032, + "learning_rate": 2.4555e-05, + "loss": 0.0534, + "step": 8189 + }, + { + "epoch": 6.91722972972973, + "grad_norm": 0.44386664032936096, + "learning_rate": 2.4558e-05, + "loss": 0.0354, + "step": 8190 + }, + { + "epoch": 6.918074324324325, + "grad_norm": 0.69710373878479, + "learning_rate": 2.4561e-05, + "loss": 0.0536, + "step": 8191 + }, + { + "epoch": 6.918918918918919, + "grad_norm": 0.549384355545044, + "learning_rate": 2.4564e-05, + "loss": 0.0568, + "step": 8192 + }, + { + "epoch": 6.919763513513513, + "grad_norm": 0.5571083426475525, + "learning_rate": 2.4567e-05, + "loss": 0.0459, + "step": 8193 + }, + { + "epoch": 6.920608108108108, + "grad_norm": 0.3982073962688446, + "learning_rate": 2.457e-05, + "loss": 0.0394, + "step": 8194 + }, + { + "epoch": 6.921452702702703, + "grad_norm": 0.5545638203620911, + "learning_rate": 2.4573000000000003e-05, + "loss": 0.0515, + "step": 8195 + }, + { + "epoch": 6.922297297297297, + "grad_norm": 0.511340856552124, + "learning_rate": 2.4576000000000003e-05, + "loss": 0.0423, + "step": 8196 + }, + { + "epoch": 6.923141891891892, + "grad_norm": 1.3885917663574219, + "learning_rate": 2.4579000000000003e-05, + "loss": 0.0559, + "step": 8197 + }, + { + "epoch": 6.923986486486487, + "grad_norm": 0.6082301735877991, + "learning_rate": 2.4582000000000003e-05, + "loss": 0.0532, + "step": 8198 + }, + { + "epoch": 6.924831081081081, + "grad_norm": 0.4722056984901428, + "learning_rate": 2.4585e-05, + "loss": 0.0439, + "step": 8199 + }, + { + "epoch": 6.925675675675675, + "grad_norm": 0.7967303395271301, + "learning_rate": 2.4588e-05, + "loss": 0.0523, + "step": 8200 + }, + { + "epoch": 6.92652027027027, + "grad_norm": 0.6749444603919983, + "learning_rate": 2.4591e-05, + "loss": 0.0689, + "step": 8201 + }, + { + "epoch": 6.927364864864865, + "grad_norm": 2.44087815284729, + "learning_rate": 2.4594e-05, + "loss": 0.1032, + "step": 8202 + }, + { + "epoch": 6.92820945945946, + "grad_norm": 0.6540454030036926, + "learning_rate": 2.4597e-05, + "loss": 0.0534, + "step": 8203 + }, + { + "epoch": 6.929054054054054, + "grad_norm": 1.140180230140686, + "learning_rate": 2.4599999999999998e-05, + "loss": 0.0623, + "step": 8204 + }, + { + "epoch": 6.929898648648649, + "grad_norm": 2.411782741546631, + "learning_rate": 2.4603e-05, + "loss": 0.3977, + "step": 8205 + }, + { + "epoch": 6.930743243243243, + "grad_norm": 0.7806223630905151, + "learning_rate": 2.4606e-05, + "loss": 0.1846, + "step": 8206 + }, + { + "epoch": 6.931587837837838, + "grad_norm": 0.6319096088409424, + "learning_rate": 2.4609e-05, + "loss": 0.1137, + "step": 8207 + }, + { + "epoch": 6.9324324324324325, + "grad_norm": 0.5567706823348999, + "learning_rate": 2.4612e-05, + "loss": 0.0757, + "step": 8208 + }, + { + "epoch": 6.933277027027027, + "grad_norm": 0.3808775246143341, + "learning_rate": 2.4615e-05, + "loss": 0.0545, + "step": 8209 + }, + { + "epoch": 6.934121621621622, + "grad_norm": 1.5910106897354126, + "learning_rate": 2.4618e-05, + "loss": 0.0437, + "step": 8210 + }, + { + "epoch": 6.934966216216216, + "grad_norm": 0.5412640571594238, + "learning_rate": 2.4621e-05, + "loss": 0.0431, + "step": 8211 + }, + { + "epoch": 6.9358108108108105, + "grad_norm": 0.4534092843532562, + "learning_rate": 2.4624e-05, + "loss": 0.0432, + "step": 8212 + }, + { + "epoch": 6.936655405405405, + "grad_norm": 0.500119149684906, + "learning_rate": 2.4627e-05, + "loss": 0.0644, + "step": 8213 + }, + { + "epoch": 6.9375, + "grad_norm": 0.3699922263622284, + "learning_rate": 2.463e-05, + "loss": 0.0504, + "step": 8214 + }, + { + "epoch": 6.938344594594595, + "grad_norm": 0.2711613178253174, + "learning_rate": 2.4633000000000003e-05, + "loss": 0.035, + "step": 8215 + }, + { + "epoch": 6.9391891891891895, + "grad_norm": 0.583046019077301, + "learning_rate": 2.4636000000000003e-05, + "loss": 0.0516, + "step": 8216 + }, + { + "epoch": 6.940033783783784, + "grad_norm": 0.37098127603530884, + "learning_rate": 2.4639000000000002e-05, + "loss": 0.0593, + "step": 8217 + }, + { + "epoch": 6.940878378378378, + "grad_norm": 0.4783220887184143, + "learning_rate": 2.4642000000000002e-05, + "loss": 0.0506, + "step": 8218 + }, + { + "epoch": 6.941722972972973, + "grad_norm": 0.7259538173675537, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0512, + "step": 8219 + }, + { + "epoch": 6.9425675675675675, + "grad_norm": 0.30167821049690247, + "learning_rate": 2.4648000000000002e-05, + "loss": 0.0386, + "step": 8220 + }, + { + "epoch": 6.943412162162162, + "grad_norm": 0.33457520604133606, + "learning_rate": 2.4651e-05, + "loss": 0.0377, + "step": 8221 + }, + { + "epoch": 6.944256756756757, + "grad_norm": 0.3200888931751251, + "learning_rate": 2.4654e-05, + "loss": 0.0247, + "step": 8222 + }, + { + "epoch": 6.945101351351351, + "grad_norm": 0.352251261472702, + "learning_rate": 2.4656999999999998e-05, + "loss": 0.0389, + "step": 8223 + }, + { + "epoch": 6.945945945945946, + "grad_norm": 0.31294530630111694, + "learning_rate": 2.4659999999999998e-05, + "loss": 0.0378, + "step": 8224 + }, + { + "epoch": 6.94679054054054, + "grad_norm": 0.41699305176734924, + "learning_rate": 2.4663e-05, + "loss": 0.0327, + "step": 8225 + }, + { + "epoch": 6.947635135135135, + "grad_norm": 0.47125980257987976, + "learning_rate": 2.4666e-05, + "loss": 0.0447, + "step": 8226 + }, + { + "epoch": 6.94847972972973, + "grad_norm": 0.6224949955940247, + "learning_rate": 2.4669e-05, + "loss": 0.0981, + "step": 8227 + }, + { + "epoch": 6.949324324324325, + "grad_norm": 0.3702024817466736, + "learning_rate": 2.4672e-05, + "loss": 0.035, + "step": 8228 + }, + { + "epoch": 6.950168918918919, + "grad_norm": 0.3734736740589142, + "learning_rate": 2.4675e-05, + "loss": 0.0366, + "step": 8229 + }, + { + "epoch": 6.951013513513513, + "grad_norm": 0.48466941714286804, + "learning_rate": 2.4678e-05, + "loss": 0.026, + "step": 8230 + }, + { + "epoch": 6.951858108108108, + "grad_norm": 0.5513573884963989, + "learning_rate": 2.4681e-05, + "loss": 0.0414, + "step": 8231 + }, + { + "epoch": 6.952702702702703, + "grad_norm": 0.40603873133659363, + "learning_rate": 2.4684e-05, + "loss": 0.0317, + "step": 8232 + }, + { + "epoch": 6.953547297297297, + "grad_norm": 0.5410335659980774, + "learning_rate": 2.4687e-05, + "loss": 0.052, + "step": 8233 + }, + { + "epoch": 6.954391891891892, + "grad_norm": 0.43915632367134094, + "learning_rate": 2.469e-05, + "loss": 0.0329, + "step": 8234 + }, + { + "epoch": 6.955236486486487, + "grad_norm": 0.3761819899082184, + "learning_rate": 2.4693000000000002e-05, + "loss": 0.0413, + "step": 8235 + }, + { + "epoch": 6.956081081081081, + "grad_norm": 0.5746178030967712, + "learning_rate": 2.4696000000000002e-05, + "loss": 0.0452, + "step": 8236 + }, + { + "epoch": 6.956925675675675, + "grad_norm": 0.5536448955535889, + "learning_rate": 2.4699000000000002e-05, + "loss": 0.0487, + "step": 8237 + }, + { + "epoch": 6.95777027027027, + "grad_norm": 0.46919557452201843, + "learning_rate": 2.4702e-05, + "loss": 0.04, + "step": 8238 + }, + { + "epoch": 6.958614864864865, + "grad_norm": 0.5060279965400696, + "learning_rate": 2.4705e-05, + "loss": 0.0577, + "step": 8239 + }, + { + "epoch": 6.95945945945946, + "grad_norm": 0.6719871759414673, + "learning_rate": 2.4708e-05, + "loss": 0.054, + "step": 8240 + }, + { + "epoch": 6.960304054054054, + "grad_norm": 0.9165152311325073, + "learning_rate": 2.4711e-05, + "loss": 0.1019, + "step": 8241 + }, + { + "epoch": 6.961148648648649, + "grad_norm": 0.5463565587997437, + "learning_rate": 2.4714e-05, + "loss": 0.0316, + "step": 8242 + }, + { + "epoch": 6.961993243243243, + "grad_norm": 0.5838072896003723, + "learning_rate": 2.4717e-05, + "loss": 0.0505, + "step": 8243 + }, + { + "epoch": 6.962837837837838, + "grad_norm": 0.7413327097892761, + "learning_rate": 2.472e-05, + "loss": 0.0487, + "step": 8244 + }, + { + "epoch": 6.9636824324324325, + "grad_norm": 0.38071945309638977, + "learning_rate": 2.4723000000000004e-05, + "loss": 0.036, + "step": 8245 + }, + { + "epoch": 6.964527027027027, + "grad_norm": 0.5405975580215454, + "learning_rate": 2.4726000000000003e-05, + "loss": 0.0472, + "step": 8246 + }, + { + "epoch": 6.965371621621622, + "grad_norm": 0.4303346574306488, + "learning_rate": 2.4729000000000003e-05, + "loss": 0.0489, + "step": 8247 + }, + { + "epoch": 6.966216216216216, + "grad_norm": 1.0110167264938354, + "learning_rate": 2.4732e-05, + "loss": 0.04, + "step": 8248 + }, + { + "epoch": 6.9670608108108105, + "grad_norm": 0.6750807762145996, + "learning_rate": 2.4735e-05, + "loss": 0.0932, + "step": 8249 + }, + { + "epoch": 6.967905405405405, + "grad_norm": 0.4715632200241089, + "learning_rate": 2.4738e-05, + "loss": 0.0472, + "step": 8250 + }, + { + "epoch": 6.96875, + "grad_norm": 0.8515667915344238, + "learning_rate": 2.4741e-05, + "loss": 0.0699, + "step": 8251 + }, + { + "epoch": 6.969594594594595, + "grad_norm": 0.5247920155525208, + "learning_rate": 2.4744e-05, + "loss": 0.0368, + "step": 8252 + }, + { + "epoch": 6.9704391891891895, + "grad_norm": 0.7389630675315857, + "learning_rate": 2.4747e-05, + "loss": 0.0498, + "step": 8253 + }, + { + "epoch": 6.971283783783784, + "grad_norm": 0.8444725275039673, + "learning_rate": 2.475e-05, + "loss": 0.0717, + "step": 8254 + }, + { + "epoch": 6.972128378378378, + "grad_norm": 1.8965981006622314, + "learning_rate": 2.4753e-05, + "loss": 0.378, + "step": 8255 + }, + { + "epoch": 6.972972972972973, + "grad_norm": 0.5655329823493958, + "learning_rate": 2.4756e-05, + "loss": 0.1191, + "step": 8256 + }, + { + "epoch": 6.9738175675675675, + "grad_norm": 0.6179541945457458, + "learning_rate": 2.4759e-05, + "loss": 0.0952, + "step": 8257 + }, + { + "epoch": 6.974662162162162, + "grad_norm": 0.7257128953933716, + "learning_rate": 2.4762e-05, + "loss": 0.0983, + "step": 8258 + }, + { + "epoch": 6.975506756756757, + "grad_norm": 1.6725519895553589, + "learning_rate": 2.4765e-05, + "loss": 0.0362, + "step": 8259 + }, + { + "epoch": 6.976351351351351, + "grad_norm": 0.5176898837089539, + "learning_rate": 2.4768e-05, + "loss": 0.0456, + "step": 8260 + }, + { + "epoch": 6.977195945945946, + "grad_norm": 0.45279377698898315, + "learning_rate": 2.4771e-05, + "loss": 0.0414, + "step": 8261 + }, + { + "epoch": 6.97804054054054, + "grad_norm": 0.3687394857406616, + "learning_rate": 2.4774e-05, + "loss": 0.029, + "step": 8262 + }, + { + "epoch": 6.978885135135135, + "grad_norm": 0.7954910397529602, + "learning_rate": 2.4777e-05, + "loss": 0.0444, + "step": 8263 + }, + { + "epoch": 6.97972972972973, + "grad_norm": 0.4671214520931244, + "learning_rate": 2.478e-05, + "loss": 0.0387, + "step": 8264 + }, + { + "epoch": 6.980574324324325, + "grad_norm": 0.3627639412879944, + "learning_rate": 2.4783e-05, + "loss": 0.0388, + "step": 8265 + }, + { + "epoch": 6.981418918918919, + "grad_norm": 0.5224405527114868, + "learning_rate": 2.4786000000000003e-05, + "loss": 0.037, + "step": 8266 + }, + { + "epoch": 6.982263513513513, + "grad_norm": 0.27985432744026184, + "learning_rate": 2.4789000000000003e-05, + "loss": 0.0249, + "step": 8267 + }, + { + "epoch": 6.983108108108108, + "grad_norm": 0.42628756165504456, + "learning_rate": 2.4792000000000003e-05, + "loss": 0.0432, + "step": 8268 + }, + { + "epoch": 6.983952702702703, + "grad_norm": 0.34234678745269775, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0411, + "step": 8269 + }, + { + "epoch": 6.984797297297297, + "grad_norm": 0.4670203626155853, + "learning_rate": 2.4798000000000002e-05, + "loss": 0.0469, + "step": 8270 + }, + { + "epoch": 6.985641891891892, + "grad_norm": 0.8155250549316406, + "learning_rate": 2.4801000000000002e-05, + "loss": 0.0493, + "step": 8271 + }, + { + "epoch": 6.986486486486487, + "grad_norm": 0.7815845608711243, + "learning_rate": 2.4804e-05, + "loss": 0.0625, + "step": 8272 + }, + { + "epoch": 6.987331081081081, + "grad_norm": 0.588569700717926, + "learning_rate": 2.4806999999999998e-05, + "loss": 0.0412, + "step": 8273 + }, + { + "epoch": 6.988175675675675, + "grad_norm": 0.6162354350090027, + "learning_rate": 2.4809999999999998e-05, + "loss": 0.0441, + "step": 8274 + }, + { + "epoch": 6.98902027027027, + "grad_norm": 0.8524506092071533, + "learning_rate": 2.4812999999999998e-05, + "loss": 0.0438, + "step": 8275 + }, + { + "epoch": 6.989864864864865, + "grad_norm": 0.9711463451385498, + "learning_rate": 2.4816e-05, + "loss": 0.0586, + "step": 8276 + }, + { + "epoch": 6.99070945945946, + "grad_norm": 0.5493976473808289, + "learning_rate": 2.4819e-05, + "loss": 0.0631, + "step": 8277 + }, + { + "epoch": 6.991554054054054, + "grad_norm": 0.3817479908466339, + "learning_rate": 2.4822e-05, + "loss": 0.0242, + "step": 8278 + }, + { + "epoch": 6.992398648648649, + "grad_norm": 0.6683887243270874, + "learning_rate": 2.4825e-05, + "loss": 0.049, + "step": 8279 + }, + { + "epoch": 6.993243243243243, + "grad_norm": 0.4465233087539673, + "learning_rate": 2.4828e-05, + "loss": 0.0465, + "step": 8280 + }, + { + "epoch": 6.994087837837838, + "grad_norm": 1.5614627599716187, + "learning_rate": 2.4831e-05, + "loss": 0.056, + "step": 8281 + }, + { + "epoch": 6.9949324324324325, + "grad_norm": 0.4614065885543823, + "learning_rate": 2.4834e-05, + "loss": 0.0631, + "step": 8282 + }, + { + "epoch": 6.995777027027027, + "grad_norm": 0.4148452579975128, + "learning_rate": 2.4837e-05, + "loss": 0.0356, + "step": 8283 + }, + { + "epoch": 6.996621621621622, + "grad_norm": 0.7042726874351501, + "learning_rate": 2.484e-05, + "loss": 0.0454, + "step": 8284 + }, + { + "epoch": 6.997466216216216, + "grad_norm": 0.5863615870475769, + "learning_rate": 2.4843e-05, + "loss": 0.0509, + "step": 8285 + }, + { + "epoch": 6.9983108108108105, + "grad_norm": 0.39169177412986755, + "learning_rate": 2.4846000000000002e-05, + "loss": 0.0358, + "step": 8286 + }, + { + "epoch": 6.999155405405405, + "grad_norm": 1.162367343902588, + "learning_rate": 2.4849000000000002e-05, + "loss": 0.1098, + "step": 8287 + }, + { + "epoch": 7.0, + "grad_norm": 1.0314301252365112, + "learning_rate": 2.4852000000000002e-05, + "loss": 0.077, + "step": 8288 + }, + { + "epoch": 7.000844594594595, + "grad_norm": 0.989345133304596, + "learning_rate": 2.4855000000000002e-05, + "loss": 0.3162, + "step": 8289 + }, + { + "epoch": 7.0016891891891895, + "grad_norm": 0.8297896385192871, + "learning_rate": 2.4858e-05, + "loss": 0.2114, + "step": 8290 + }, + { + "epoch": 7.002533783783784, + "grad_norm": 0.7512825727462769, + "learning_rate": 2.4861e-05, + "loss": 0.1338, + "step": 8291 + }, + { + "epoch": 7.003378378378378, + "grad_norm": 0.34687620401382446, + "learning_rate": 2.4864e-05, + "loss": 0.0548, + "step": 8292 + }, + { + "epoch": 7.004222972972973, + "grad_norm": 0.38198861479759216, + "learning_rate": 2.4867e-05, + "loss": 0.0475, + "step": 8293 + }, + { + "epoch": 7.0050675675675675, + "grad_norm": 0.31371837854385376, + "learning_rate": 2.487e-05, + "loss": 0.0409, + "step": 8294 + }, + { + "epoch": 7.005912162162162, + "grad_norm": 0.3124985694885254, + "learning_rate": 2.4873e-05, + "loss": 0.0336, + "step": 8295 + }, + { + "epoch": 7.006756756756757, + "grad_norm": 0.4345627725124359, + "learning_rate": 2.4876000000000004e-05, + "loss": 0.0377, + "step": 8296 + }, + { + "epoch": 7.007601351351352, + "grad_norm": 0.3678625524044037, + "learning_rate": 2.4879e-05, + "loss": 0.04, + "step": 8297 + }, + { + "epoch": 7.008445945945946, + "grad_norm": 0.47553062438964844, + "learning_rate": 2.4882e-05, + "loss": 0.0332, + "step": 8298 + }, + { + "epoch": 7.00929054054054, + "grad_norm": 0.4514593780040741, + "learning_rate": 2.4885e-05, + "loss": 0.0433, + "step": 8299 + }, + { + "epoch": 7.010135135135135, + "grad_norm": 0.4457988739013672, + "learning_rate": 2.4888e-05, + "loss": 0.0374, + "step": 8300 + }, + { + "epoch": 7.01097972972973, + "grad_norm": 0.4024510383605957, + "learning_rate": 2.4891e-05, + "loss": 0.0339, + "step": 8301 + }, + { + "epoch": 7.011824324324325, + "grad_norm": 0.41564443707466125, + "learning_rate": 2.4894e-05, + "loss": 0.0404, + "step": 8302 + }, + { + "epoch": 7.012668918918919, + "grad_norm": 0.27978476881980896, + "learning_rate": 2.4897e-05, + "loss": 0.0276, + "step": 8303 + }, + { + "epoch": 7.013513513513513, + "grad_norm": 0.4900560975074768, + "learning_rate": 2.49e-05, + "loss": 0.0501, + "step": 8304 + }, + { + "epoch": 7.014358108108108, + "grad_norm": 0.612104594707489, + "learning_rate": 2.4903e-05, + "loss": 0.0421, + "step": 8305 + }, + { + "epoch": 7.015202702702703, + "grad_norm": 0.40444937348365784, + "learning_rate": 2.4906000000000002e-05, + "loss": 0.0334, + "step": 8306 + }, + { + "epoch": 7.016047297297297, + "grad_norm": 0.34686195850372314, + "learning_rate": 2.4909e-05, + "loss": 0.0349, + "step": 8307 + }, + { + "epoch": 7.016891891891892, + "grad_norm": 0.3519292175769806, + "learning_rate": 2.4912e-05, + "loss": 0.0363, + "step": 8308 + }, + { + "epoch": 7.017736486486487, + "grad_norm": 1.0678298473358154, + "learning_rate": 2.4915e-05, + "loss": 0.0399, + "step": 8309 + }, + { + "epoch": 7.018581081081081, + "grad_norm": 0.4278223514556885, + "learning_rate": 2.4918e-05, + "loss": 0.0424, + "step": 8310 + }, + { + "epoch": 7.019425675675675, + "grad_norm": 0.4761946499347687, + "learning_rate": 2.4921e-05, + "loss": 0.0349, + "step": 8311 + }, + { + "epoch": 7.02027027027027, + "grad_norm": 0.3292560577392578, + "learning_rate": 2.4924e-05, + "loss": 0.03, + "step": 8312 + }, + { + "epoch": 7.021114864864865, + "grad_norm": 0.6588030457496643, + "learning_rate": 2.4927e-05, + "loss": 0.0465, + "step": 8313 + }, + { + "epoch": 7.02195945945946, + "grad_norm": 0.5582924485206604, + "learning_rate": 2.493e-05, + "loss": 0.0525, + "step": 8314 + }, + { + "epoch": 7.022804054054054, + "grad_norm": 0.2530175745487213, + "learning_rate": 2.4933e-05, + "loss": 0.0275, + "step": 8315 + }, + { + "epoch": 7.023648648648648, + "grad_norm": 0.37171125411987305, + "learning_rate": 2.4936000000000003e-05, + "loss": 0.0212, + "step": 8316 + }, + { + "epoch": 7.024493243243243, + "grad_norm": 0.6026996970176697, + "learning_rate": 2.4939000000000003e-05, + "loss": 0.0355, + "step": 8317 + }, + { + "epoch": 7.025337837837838, + "grad_norm": 0.3836705684661865, + "learning_rate": 2.4942000000000003e-05, + "loss": 0.0368, + "step": 8318 + }, + { + "epoch": 7.0261824324324325, + "grad_norm": 0.306535929441452, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0281, + "step": 8319 + }, + { + "epoch": 7.027027027027027, + "grad_norm": 0.5371415019035339, + "learning_rate": 2.4948000000000002e-05, + "loss": 0.0418, + "step": 8320 + }, + { + "epoch": 7.027871621621622, + "grad_norm": 0.3493037819862366, + "learning_rate": 2.4951e-05, + "loss": 0.032, + "step": 8321 + }, + { + "epoch": 7.028716216216216, + "grad_norm": 0.2839891016483307, + "learning_rate": 2.4954e-05, + "loss": 0.0303, + "step": 8322 + }, + { + "epoch": 7.0295608108108105, + "grad_norm": 0.4120151102542877, + "learning_rate": 2.4957e-05, + "loss": 0.0317, + "step": 8323 + }, + { + "epoch": 7.030405405405405, + "grad_norm": 0.43056073784828186, + "learning_rate": 2.4959999999999998e-05, + "loss": 0.0332, + "step": 8324 + }, + { + "epoch": 7.03125, + "grad_norm": 0.4376700222492218, + "learning_rate": 2.4962999999999998e-05, + "loss": 0.0277, + "step": 8325 + }, + { + "epoch": 7.032094594594595, + "grad_norm": 0.4546850025653839, + "learning_rate": 2.4966e-05, + "loss": 0.0307, + "step": 8326 + }, + { + "epoch": 7.0329391891891895, + "grad_norm": 0.4030517637729645, + "learning_rate": 2.4969e-05, + "loss": 0.0331, + "step": 8327 + }, + { + "epoch": 7.033783783783784, + "grad_norm": 1.0086994171142578, + "learning_rate": 2.4972e-05, + "loss": 0.0549, + "step": 8328 + }, + { + "epoch": 7.034628378378378, + "grad_norm": 0.5124897956848145, + "learning_rate": 2.4975e-05, + "loss": 0.0344, + "step": 8329 + }, + { + "epoch": 7.035472972972973, + "grad_norm": 0.5109788179397583, + "learning_rate": 2.4978e-05, + "loss": 0.0495, + "step": 8330 + }, + { + "epoch": 7.0363175675675675, + "grad_norm": 0.58246910572052, + "learning_rate": 2.4981e-05, + "loss": 0.0327, + "step": 8331 + }, + { + "epoch": 7.037162162162162, + "grad_norm": 0.5481300950050354, + "learning_rate": 2.4984e-05, + "loss": 0.0388, + "step": 8332 + }, + { + "epoch": 7.038006756756757, + "grad_norm": 0.7487584948539734, + "learning_rate": 2.4987e-05, + "loss": 0.0502, + "step": 8333 + }, + { + "epoch": 7.038851351351352, + "grad_norm": 0.8272464871406555, + "learning_rate": 2.499e-05, + "loss": 0.0642, + "step": 8334 + }, + { + "epoch": 7.039695945945946, + "grad_norm": 0.5213354229927063, + "learning_rate": 2.4993e-05, + "loss": 0.0279, + "step": 8335 + }, + { + "epoch": 7.04054054054054, + "grad_norm": 0.7635986804962158, + "learning_rate": 2.4996000000000003e-05, + "loss": 0.0847, + "step": 8336 + }, + { + "epoch": 7.041385135135135, + "grad_norm": 0.8802917003631592, + "learning_rate": 2.4999000000000002e-05, + "loss": 0.0535, + "step": 8337 + }, + { + "epoch": 7.04222972972973, + "grad_norm": 0.6915863752365112, + "learning_rate": 2.5002000000000002e-05, + "loss": 0.0646, + "step": 8338 + }, + { + "epoch": 7.043074324324325, + "grad_norm": 3.12076473236084, + "learning_rate": 2.5005000000000002e-05, + "loss": 0.3639, + "step": 8339 + }, + { + "epoch": 7.043918918918919, + "grad_norm": 0.9943249821662903, + "learning_rate": 2.5008000000000002e-05, + "loss": 0.1638, + "step": 8340 + }, + { + "epoch": 7.044763513513513, + "grad_norm": 0.605364978313446, + "learning_rate": 2.5011e-05, + "loss": 0.0722, + "step": 8341 + }, + { + "epoch": 7.045608108108108, + "grad_norm": 0.48359498381614685, + "learning_rate": 2.5014e-05, + "loss": 0.0771, + "step": 8342 + }, + { + "epoch": 7.046452702702703, + "grad_norm": 0.47687003016471863, + "learning_rate": 2.5017e-05, + "loss": 0.058, + "step": 8343 + }, + { + "epoch": 7.047297297297297, + "grad_norm": 0.46489760279655457, + "learning_rate": 2.502e-05, + "loss": 0.0503, + "step": 8344 + }, + { + "epoch": 7.048141891891892, + "grad_norm": 0.3561033308506012, + "learning_rate": 2.5023e-05, + "loss": 0.0493, + "step": 8345 + }, + { + "epoch": 7.048986486486487, + "grad_norm": 0.5929678678512573, + "learning_rate": 2.5026e-05, + "loss": 0.0518, + "step": 8346 + }, + { + "epoch": 7.049831081081081, + "grad_norm": 0.43521127104759216, + "learning_rate": 2.5029e-05, + "loss": 0.0344, + "step": 8347 + }, + { + "epoch": 7.050675675675675, + "grad_norm": 0.5676953196525574, + "learning_rate": 2.5032e-05, + "loss": 0.0368, + "step": 8348 + }, + { + "epoch": 7.05152027027027, + "grad_norm": 0.2851664125919342, + "learning_rate": 2.5035e-05, + "loss": 0.0329, + "step": 8349 + }, + { + "epoch": 7.052364864864865, + "grad_norm": 0.35954970121383667, + "learning_rate": 2.5038e-05, + "loss": 0.0317, + "step": 8350 + }, + { + "epoch": 7.05320945945946, + "grad_norm": 0.425824373960495, + "learning_rate": 2.5041e-05, + "loss": 0.0429, + "step": 8351 + }, + { + "epoch": 7.054054054054054, + "grad_norm": 0.49869057536125183, + "learning_rate": 2.5044e-05, + "loss": 0.0503, + "step": 8352 + }, + { + "epoch": 7.054898648648648, + "grad_norm": 0.4069646894931793, + "learning_rate": 2.5047e-05, + "loss": 0.0375, + "step": 8353 + }, + { + "epoch": 7.055743243243243, + "grad_norm": 0.515841007232666, + "learning_rate": 2.505e-05, + "loss": 0.0382, + "step": 8354 + }, + { + "epoch": 7.056587837837838, + "grad_norm": 0.3833809494972229, + "learning_rate": 2.5053e-05, + "loss": 0.0531, + "step": 8355 + }, + { + "epoch": 7.0574324324324325, + "grad_norm": 0.6143726706504822, + "learning_rate": 2.5056000000000002e-05, + "loss": 0.0474, + "step": 8356 + }, + { + "epoch": 7.058277027027027, + "grad_norm": 0.5147258043289185, + "learning_rate": 2.5059000000000002e-05, + "loss": 0.031, + "step": 8357 + }, + { + "epoch": 7.059121621621622, + "grad_norm": 0.3136848211288452, + "learning_rate": 2.5062e-05, + "loss": 0.0261, + "step": 8358 + }, + { + "epoch": 7.059966216216216, + "grad_norm": 0.38759562373161316, + "learning_rate": 2.5065e-05, + "loss": 0.0363, + "step": 8359 + }, + { + "epoch": 7.0608108108108105, + "grad_norm": 0.40039679408073425, + "learning_rate": 2.5068e-05, + "loss": 0.031, + "step": 8360 + }, + { + "epoch": 7.061655405405405, + "grad_norm": 0.3800030052661896, + "learning_rate": 2.5071e-05, + "loss": 0.039, + "step": 8361 + }, + { + "epoch": 7.0625, + "grad_norm": 0.5099263191223145, + "learning_rate": 2.5074e-05, + "loss": 0.0685, + "step": 8362 + }, + { + "epoch": 7.063344594594595, + "grad_norm": 0.4721314609050751, + "learning_rate": 2.5077e-05, + "loss": 0.0335, + "step": 8363 + }, + { + "epoch": 7.0641891891891895, + "grad_norm": 0.2827744483947754, + "learning_rate": 2.508e-05, + "loss": 0.0332, + "step": 8364 + }, + { + "epoch": 7.065033783783784, + "grad_norm": 0.2793208658695221, + "learning_rate": 2.5083e-05, + "loss": 0.0203, + "step": 8365 + }, + { + "epoch": 7.065878378378378, + "grad_norm": 0.4167604148387909, + "learning_rate": 2.5086000000000003e-05, + "loss": 0.044, + "step": 8366 + }, + { + "epoch": 7.066722972972973, + "grad_norm": 0.8628392815589905, + "learning_rate": 2.5089000000000003e-05, + "loss": 0.0468, + "step": 8367 + }, + { + "epoch": 7.0675675675675675, + "grad_norm": 0.3329401910305023, + "learning_rate": 2.5092000000000003e-05, + "loss": 0.0321, + "step": 8368 + }, + { + "epoch": 7.068412162162162, + "grad_norm": 0.49239784479141235, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0274, + "step": 8369 + }, + { + "epoch": 7.069256756756757, + "grad_norm": 0.4255485236644745, + "learning_rate": 2.5098000000000003e-05, + "loss": 0.0403, + "step": 8370 + }, + { + "epoch": 7.070101351351352, + "grad_norm": 0.40658697485923767, + "learning_rate": 2.5101e-05, + "loss": 0.0328, + "step": 8371 + }, + { + "epoch": 7.070945945945946, + "grad_norm": 0.3521552085876465, + "learning_rate": 2.5104e-05, + "loss": 0.032, + "step": 8372 + }, + { + "epoch": 7.07179054054054, + "grad_norm": 0.26033321022987366, + "learning_rate": 2.5107e-05, + "loss": 0.0276, + "step": 8373 + }, + { + "epoch": 7.072635135135135, + "grad_norm": 0.3925905227661133, + "learning_rate": 2.511e-05, + "loss": 0.0317, + "step": 8374 + }, + { + "epoch": 7.07347972972973, + "grad_norm": 0.2016783356666565, + "learning_rate": 2.5112999999999998e-05, + "loss": 0.0141, + "step": 8375 + }, + { + "epoch": 7.074324324324325, + "grad_norm": 0.3397044241428375, + "learning_rate": 2.5116e-05, + "loss": 0.0402, + "step": 8376 + }, + { + "epoch": 7.075168918918919, + "grad_norm": 0.35882654786109924, + "learning_rate": 2.5119e-05, + "loss": 0.027, + "step": 8377 + }, + { + "epoch": 7.076013513513513, + "grad_norm": 0.6584435701370239, + "learning_rate": 2.5122e-05, + "loss": 0.0663, + "step": 8378 + }, + { + "epoch": 7.076858108108108, + "grad_norm": 0.5617596507072449, + "learning_rate": 2.5125e-05, + "loss": 0.0519, + "step": 8379 + }, + { + "epoch": 7.077702702702703, + "grad_norm": 0.7044975757598877, + "learning_rate": 2.5128e-05, + "loss": 0.0382, + "step": 8380 + }, + { + "epoch": 7.078547297297297, + "grad_norm": 0.3850657641887665, + "learning_rate": 2.5131e-05, + "loss": 0.033, + "step": 8381 + }, + { + "epoch": 7.079391891891892, + "grad_norm": 1.0443687438964844, + "learning_rate": 2.5134e-05, + "loss": 0.0519, + "step": 8382 + }, + { + "epoch": 7.080236486486487, + "grad_norm": 0.4875355064868927, + "learning_rate": 2.5137e-05, + "loss": 0.0443, + "step": 8383 + }, + { + "epoch": 7.081081081081081, + "grad_norm": 0.4858391284942627, + "learning_rate": 2.514e-05, + "loss": 0.0452, + "step": 8384 + }, + { + "epoch": 7.081925675675675, + "grad_norm": 1.070759654045105, + "learning_rate": 2.5143e-05, + "loss": 0.0345, + "step": 8385 + }, + { + "epoch": 7.08277027027027, + "grad_norm": 0.5817584991455078, + "learning_rate": 2.5146e-05, + "loss": 0.0568, + "step": 8386 + }, + { + "epoch": 7.083614864864865, + "grad_norm": 1.252701759338379, + "learning_rate": 2.5149000000000003e-05, + "loss": 0.0759, + "step": 8387 + }, + { + "epoch": 7.08445945945946, + "grad_norm": 1.1501158475875854, + "learning_rate": 2.5152000000000002e-05, + "loss": 0.089, + "step": 8388 + }, + { + "epoch": 7.085304054054054, + "grad_norm": 0.7710330486297607, + "learning_rate": 2.5155000000000002e-05, + "loss": 0.354, + "step": 8389 + }, + { + "epoch": 7.086148648648648, + "grad_norm": 0.8470629453659058, + "learning_rate": 2.5158000000000002e-05, + "loss": 0.2404, + "step": 8390 + }, + { + "epoch": 7.086993243243243, + "grad_norm": 0.48840996623039246, + "learning_rate": 2.5161000000000002e-05, + "loss": 0.082, + "step": 8391 + }, + { + "epoch": 7.087837837837838, + "grad_norm": 0.38698911666870117, + "learning_rate": 2.5164e-05, + "loss": 0.0675, + "step": 8392 + }, + { + "epoch": 7.0886824324324325, + "grad_norm": 0.42777031660079956, + "learning_rate": 2.5167e-05, + "loss": 0.0537, + "step": 8393 + }, + { + "epoch": 7.089527027027027, + "grad_norm": 0.4056488275527954, + "learning_rate": 2.517e-05, + "loss": 0.0423, + "step": 8394 + }, + { + "epoch": 7.090371621621622, + "grad_norm": 0.5153475999832153, + "learning_rate": 2.5172999999999998e-05, + "loss": 0.0621, + "step": 8395 + }, + { + "epoch": 7.091216216216216, + "grad_norm": 0.4059523642063141, + "learning_rate": 2.5175999999999997e-05, + "loss": 0.0529, + "step": 8396 + }, + { + "epoch": 7.0920608108108105, + "grad_norm": 0.3104512691497803, + "learning_rate": 2.5179e-05, + "loss": 0.0309, + "step": 8397 + }, + { + "epoch": 7.092905405405405, + "grad_norm": 0.2993943691253662, + "learning_rate": 2.5182e-05, + "loss": 0.0497, + "step": 8398 + }, + { + "epoch": 7.09375, + "grad_norm": 0.39531171321868896, + "learning_rate": 2.5185e-05, + "loss": 0.0356, + "step": 8399 + }, + { + "epoch": 7.094594594594595, + "grad_norm": 0.2741668224334717, + "learning_rate": 2.5188e-05, + "loss": 0.0239, + "step": 8400 + }, + { + "epoch": 7.0954391891891895, + "grad_norm": 0.4295528531074524, + "learning_rate": 2.5191e-05, + "loss": 0.0337, + "step": 8401 + }, + { + "epoch": 7.096283783783784, + "grad_norm": 0.4718814492225647, + "learning_rate": 2.5194e-05, + "loss": 0.0444, + "step": 8402 + }, + { + "epoch": 7.097128378378378, + "grad_norm": 0.4777906537055969, + "learning_rate": 2.5197e-05, + "loss": 0.0389, + "step": 8403 + }, + { + "epoch": 7.097972972972973, + "grad_norm": 0.34867507219314575, + "learning_rate": 2.52e-05, + "loss": 0.0356, + "step": 8404 + }, + { + "epoch": 7.0988175675675675, + "grad_norm": 0.4281001389026642, + "learning_rate": 2.5203e-05, + "loss": 0.0327, + "step": 8405 + }, + { + "epoch": 7.099662162162162, + "grad_norm": 0.5878276824951172, + "learning_rate": 2.5206e-05, + "loss": 0.0734, + "step": 8406 + }, + { + "epoch": 7.100506756756757, + "grad_norm": 0.2957923114299774, + "learning_rate": 2.5209000000000002e-05, + "loss": 0.0298, + "step": 8407 + }, + { + "epoch": 7.101351351351352, + "grad_norm": 0.5757549405097961, + "learning_rate": 2.5212000000000002e-05, + "loss": 0.046, + "step": 8408 + }, + { + "epoch": 7.102195945945946, + "grad_norm": 0.34124985337257385, + "learning_rate": 2.5215e-05, + "loss": 0.0272, + "step": 8409 + }, + { + "epoch": 7.10304054054054, + "grad_norm": 0.4622043967247009, + "learning_rate": 2.5218e-05, + "loss": 0.0359, + "step": 8410 + }, + { + "epoch": 7.103885135135135, + "grad_norm": 0.5820953249931335, + "learning_rate": 2.5221e-05, + "loss": 0.0394, + "step": 8411 + }, + { + "epoch": 7.10472972972973, + "grad_norm": 0.3508901000022888, + "learning_rate": 2.5224e-05, + "loss": 0.0438, + "step": 8412 + }, + { + "epoch": 7.105574324324325, + "grad_norm": 0.28145918250083923, + "learning_rate": 2.5227e-05, + "loss": 0.0173, + "step": 8413 + }, + { + "epoch": 7.106418918918919, + "grad_norm": 0.27981340885162354, + "learning_rate": 2.523e-05, + "loss": 0.0217, + "step": 8414 + }, + { + "epoch": 7.107263513513513, + "grad_norm": 0.5430893301963806, + "learning_rate": 2.5233e-05, + "loss": 0.0436, + "step": 8415 + }, + { + "epoch": 7.108108108108108, + "grad_norm": 0.5729487538337708, + "learning_rate": 2.5236e-05, + "loss": 0.0401, + "step": 8416 + }, + { + "epoch": 7.108952702702703, + "grad_norm": 0.39996999502182007, + "learning_rate": 2.5239000000000003e-05, + "loss": 0.0345, + "step": 8417 + }, + { + "epoch": 7.109797297297297, + "grad_norm": 0.5286132097244263, + "learning_rate": 2.5242000000000003e-05, + "loss": 0.0376, + "step": 8418 + }, + { + "epoch": 7.110641891891892, + "grad_norm": 0.38844698667526245, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0303, + "step": 8419 + }, + { + "epoch": 7.111486486486487, + "grad_norm": 0.4210180342197418, + "learning_rate": 2.5248e-05, + "loss": 0.0256, + "step": 8420 + }, + { + "epoch": 7.112331081081081, + "grad_norm": 0.4125474989414215, + "learning_rate": 2.5251e-05, + "loss": 0.0353, + "step": 8421 + }, + { + "epoch": 7.113175675675675, + "grad_norm": 0.8300979137420654, + "learning_rate": 2.5254e-05, + "loss": 0.0352, + "step": 8422 + }, + { + "epoch": 7.11402027027027, + "grad_norm": 0.3740251958370209, + "learning_rate": 2.5257e-05, + "loss": 0.0212, + "step": 8423 + }, + { + "epoch": 7.114864864864865, + "grad_norm": 0.5554611086845398, + "learning_rate": 2.526e-05, + "loss": 0.0343, + "step": 8424 + }, + { + "epoch": 7.11570945945946, + "grad_norm": 0.3722030222415924, + "learning_rate": 2.5263e-05, + "loss": 0.0465, + "step": 8425 + }, + { + "epoch": 7.116554054054054, + "grad_norm": 0.2974865436553955, + "learning_rate": 2.5266e-05, + "loss": 0.0315, + "step": 8426 + }, + { + "epoch": 7.117398648648648, + "grad_norm": 0.6940622329711914, + "learning_rate": 2.5269e-05, + "loss": 0.0411, + "step": 8427 + }, + { + "epoch": 7.118243243243243, + "grad_norm": 0.6776379942893982, + "learning_rate": 2.5272e-05, + "loss": 0.0527, + "step": 8428 + }, + { + "epoch": 7.119087837837838, + "grad_norm": 0.8411216735839844, + "learning_rate": 2.5275e-05, + "loss": 0.0496, + "step": 8429 + }, + { + "epoch": 7.1199324324324325, + "grad_norm": 0.6961426138877869, + "learning_rate": 2.5278e-05, + "loss": 0.0321, + "step": 8430 + }, + { + "epoch": 7.120777027027027, + "grad_norm": 0.4817158877849579, + "learning_rate": 2.5281e-05, + "loss": 0.0269, + "step": 8431 + }, + { + "epoch": 7.121621621621622, + "grad_norm": 0.3415526747703552, + "learning_rate": 2.5284e-05, + "loss": 0.0305, + "step": 8432 + }, + { + "epoch": 7.122466216216216, + "grad_norm": 0.6307191848754883, + "learning_rate": 2.5287e-05, + "loss": 0.0371, + "step": 8433 + }, + { + "epoch": 7.1233108108108105, + "grad_norm": 0.37313607335090637, + "learning_rate": 2.529e-05, + "loss": 0.0284, + "step": 8434 + }, + { + "epoch": 7.124155405405405, + "grad_norm": 0.606883704662323, + "learning_rate": 2.5293e-05, + "loss": 0.0433, + "step": 8435 + }, + { + "epoch": 7.125, + "grad_norm": 1.142498254776001, + "learning_rate": 2.5296e-05, + "loss": 0.0637, + "step": 8436 + }, + { + "epoch": 7.125844594594595, + "grad_norm": 1.3456695079803467, + "learning_rate": 2.5299000000000003e-05, + "loss": 0.0865, + "step": 8437 + }, + { + "epoch": 7.1266891891891895, + "grad_norm": 1.0461338758468628, + "learning_rate": 2.5302000000000003e-05, + "loss": 0.0496, + "step": 8438 + }, + { + "epoch": 7.127533783783784, + "grad_norm": 0.8351903557777405, + "learning_rate": 2.5305000000000003e-05, + "loss": 0.3352, + "step": 8439 + }, + { + "epoch": 7.128378378378378, + "grad_norm": 0.9722549915313721, + "learning_rate": 2.5308000000000002e-05, + "loss": 0.24, + "step": 8440 + }, + { + "epoch": 7.129222972972973, + "grad_norm": 0.5254881381988525, + "learning_rate": 2.5311000000000002e-05, + "loss": 0.0619, + "step": 8441 + }, + { + "epoch": 7.1300675675675675, + "grad_norm": 0.46310415863990784, + "learning_rate": 2.5314000000000002e-05, + "loss": 0.0951, + "step": 8442 + }, + { + "epoch": 7.130912162162162, + "grad_norm": 0.41942742466926575, + "learning_rate": 2.5317000000000002e-05, + "loss": 0.0426, + "step": 8443 + }, + { + "epoch": 7.131756756756757, + "grad_norm": 0.5361645221710205, + "learning_rate": 2.5319999999999998e-05, + "loss": 0.0944, + "step": 8444 + }, + { + "epoch": 7.132601351351352, + "grad_norm": 0.4577176868915558, + "learning_rate": 2.5322999999999998e-05, + "loss": 0.0419, + "step": 8445 + }, + { + "epoch": 7.133445945945946, + "grad_norm": 0.4762841761112213, + "learning_rate": 2.5325999999999998e-05, + "loss": 0.0504, + "step": 8446 + }, + { + "epoch": 7.13429054054054, + "grad_norm": 0.41529497504234314, + "learning_rate": 2.5329e-05, + "loss": 0.0464, + "step": 8447 + }, + { + "epoch": 7.135135135135135, + "grad_norm": 0.5987734198570251, + "learning_rate": 2.5332e-05, + "loss": 0.0409, + "step": 8448 + }, + { + "epoch": 7.13597972972973, + "grad_norm": 0.2828116714954376, + "learning_rate": 2.5335e-05, + "loss": 0.0171, + "step": 8449 + }, + { + "epoch": 7.136824324324325, + "grad_norm": 0.3213757574558258, + "learning_rate": 2.5338e-05, + "loss": 0.0326, + "step": 8450 + }, + { + "epoch": 7.137668918918919, + "grad_norm": 0.85801762342453, + "learning_rate": 2.5341e-05, + "loss": 0.0353, + "step": 8451 + }, + { + "epoch": 7.138513513513513, + "grad_norm": 0.3510242998600006, + "learning_rate": 2.5344e-05, + "loss": 0.0361, + "step": 8452 + }, + { + "epoch": 7.139358108108108, + "grad_norm": 0.4263480007648468, + "learning_rate": 2.5347e-05, + "loss": 0.0434, + "step": 8453 + }, + { + "epoch": 7.140202702702703, + "grad_norm": 0.4097990095615387, + "learning_rate": 2.535e-05, + "loss": 0.0459, + "step": 8454 + }, + { + "epoch": 7.141047297297297, + "grad_norm": 0.316975861787796, + "learning_rate": 2.5353e-05, + "loss": 0.029, + "step": 8455 + }, + { + "epoch": 7.141891891891892, + "grad_norm": 0.5414454936981201, + "learning_rate": 2.5356e-05, + "loss": 0.0526, + "step": 8456 + }, + { + "epoch": 7.142736486486487, + "grad_norm": 0.445917546749115, + "learning_rate": 2.5359000000000002e-05, + "loss": 0.027, + "step": 8457 + }, + { + "epoch": 7.143581081081081, + "grad_norm": 0.43882325291633606, + "learning_rate": 2.5362000000000002e-05, + "loss": 0.0328, + "step": 8458 + }, + { + "epoch": 7.144425675675675, + "grad_norm": 0.47022607922554016, + "learning_rate": 2.5365000000000002e-05, + "loss": 0.0443, + "step": 8459 + }, + { + "epoch": 7.14527027027027, + "grad_norm": 0.2676221430301666, + "learning_rate": 2.5368000000000002e-05, + "loss": 0.0199, + "step": 8460 + }, + { + "epoch": 7.146114864864865, + "grad_norm": 0.5830633044242859, + "learning_rate": 2.5371e-05, + "loss": 0.0337, + "step": 8461 + }, + { + "epoch": 7.14695945945946, + "grad_norm": 0.3709107041358948, + "learning_rate": 2.5374e-05, + "loss": 0.0338, + "step": 8462 + }, + { + "epoch": 7.147804054054054, + "grad_norm": 0.48463061451911926, + "learning_rate": 2.5377e-05, + "loss": 0.0261, + "step": 8463 + }, + { + "epoch": 7.148648648648648, + "grad_norm": 0.5700327157974243, + "learning_rate": 2.538e-05, + "loss": 0.037, + "step": 8464 + }, + { + "epoch": 7.149493243243243, + "grad_norm": 0.29390648007392883, + "learning_rate": 2.5383e-05, + "loss": 0.0229, + "step": 8465 + }, + { + "epoch": 7.150337837837838, + "grad_norm": 0.419796884059906, + "learning_rate": 2.5386e-05, + "loss": 0.0331, + "step": 8466 + }, + { + "epoch": 7.1511824324324325, + "grad_norm": 0.5244752764701843, + "learning_rate": 2.5389000000000004e-05, + "loss": 0.0282, + "step": 8467 + }, + { + "epoch": 7.152027027027027, + "grad_norm": 0.5653504133224487, + "learning_rate": 2.5392000000000004e-05, + "loss": 0.0479, + "step": 8468 + }, + { + "epoch": 7.152871621621622, + "grad_norm": 0.4120278060436249, + "learning_rate": 2.5395e-05, + "loss": 0.0286, + "step": 8469 + }, + { + "epoch": 7.153716216216216, + "grad_norm": 0.6583344340324402, + "learning_rate": 2.5398e-05, + "loss": 0.0349, + "step": 8470 + }, + { + "epoch": 7.1545608108108105, + "grad_norm": 0.40198656916618347, + "learning_rate": 2.5401e-05, + "loss": 0.0283, + "step": 8471 + }, + { + "epoch": 7.155405405405405, + "grad_norm": 0.5465641617774963, + "learning_rate": 2.5404e-05, + "loss": 0.0449, + "step": 8472 + }, + { + "epoch": 7.15625, + "grad_norm": 0.5609258413314819, + "learning_rate": 2.5407e-05, + "loss": 0.031, + "step": 8473 + }, + { + "epoch": 7.157094594594595, + "grad_norm": 0.6179313063621521, + "learning_rate": 2.541e-05, + "loss": 0.04, + "step": 8474 + }, + { + "epoch": 7.1579391891891895, + "grad_norm": 0.4544624090194702, + "learning_rate": 2.5413e-05, + "loss": 0.0295, + "step": 8475 + }, + { + "epoch": 7.158783783783784, + "grad_norm": 0.48408886790275574, + "learning_rate": 2.5416e-05, + "loss": 0.0427, + "step": 8476 + }, + { + "epoch": 7.159628378378378, + "grad_norm": 0.324870228767395, + "learning_rate": 2.5419000000000002e-05, + "loss": 0.0265, + "step": 8477 + }, + { + "epoch": 7.160472972972973, + "grad_norm": 0.5513197779655457, + "learning_rate": 2.5422e-05, + "loss": 0.0578, + "step": 8478 + }, + { + "epoch": 7.1613175675675675, + "grad_norm": 0.7354120016098022, + "learning_rate": 2.5425e-05, + "loss": 0.0448, + "step": 8479 + }, + { + "epoch": 7.162162162162162, + "grad_norm": 0.5426446199417114, + "learning_rate": 2.5428e-05, + "loss": 0.033, + "step": 8480 + }, + { + "epoch": 7.163006756756757, + "grad_norm": 0.5987861752510071, + "learning_rate": 2.5431e-05, + "loss": 0.0476, + "step": 8481 + }, + { + "epoch": 7.163851351351352, + "grad_norm": 0.47132882475852966, + "learning_rate": 2.5434e-05, + "loss": 0.0409, + "step": 8482 + }, + { + "epoch": 7.164695945945946, + "grad_norm": 0.5232285261154175, + "learning_rate": 2.5437e-05, + "loss": 0.0474, + "step": 8483 + }, + { + "epoch": 7.16554054054054, + "grad_norm": 0.9577731490135193, + "learning_rate": 2.544e-05, + "loss": 0.0488, + "step": 8484 + }, + { + "epoch": 7.166385135135135, + "grad_norm": 0.788862943649292, + "learning_rate": 2.5443e-05, + "loss": 0.0315, + "step": 8485 + }, + { + "epoch": 7.16722972972973, + "grad_norm": 0.44605425000190735, + "learning_rate": 2.5446e-05, + "loss": 0.0458, + "step": 8486 + }, + { + "epoch": 7.168074324324325, + "grad_norm": 0.5922413468360901, + "learning_rate": 2.5449000000000003e-05, + "loss": 0.0697, + "step": 8487 + }, + { + "epoch": 7.168918918918919, + "grad_norm": 0.802810549736023, + "learning_rate": 2.5452000000000003e-05, + "loss": 0.0522, + "step": 8488 + }, + { + "epoch": 7.169763513513513, + "grad_norm": 0.9977692365646362, + "learning_rate": 2.5455000000000003e-05, + "loss": 0.3045, + "step": 8489 + }, + { + "epoch": 7.170608108108108, + "grad_norm": 0.6169528365135193, + "learning_rate": 2.5458000000000003e-05, + "loss": 0.1967, + "step": 8490 + }, + { + "epoch": 7.171452702702703, + "grad_norm": 0.42693543434143066, + "learning_rate": 2.5461000000000002e-05, + "loss": 0.0548, + "step": 8491 + }, + { + "epoch": 7.172297297297297, + "grad_norm": 0.5635342001914978, + "learning_rate": 2.5464000000000002e-05, + "loss": 0.0632, + "step": 8492 + }, + { + "epoch": 7.173141891891892, + "grad_norm": 0.3348971903324127, + "learning_rate": 2.5467e-05, + "loss": 0.0329, + "step": 8493 + }, + { + "epoch": 7.173986486486487, + "grad_norm": 0.38231661915779114, + "learning_rate": 2.547e-05, + "loss": 0.0475, + "step": 8494 + }, + { + "epoch": 7.174831081081081, + "grad_norm": 0.37203502655029297, + "learning_rate": 2.5472999999999998e-05, + "loss": 0.0487, + "step": 8495 + }, + { + "epoch": 7.175675675675675, + "grad_norm": 0.37148943543434143, + "learning_rate": 2.5475999999999998e-05, + "loss": 0.0369, + "step": 8496 + }, + { + "epoch": 7.17652027027027, + "grad_norm": 0.3042099177837372, + "learning_rate": 2.5479e-05, + "loss": 0.024, + "step": 8497 + }, + { + "epoch": 7.177364864864865, + "grad_norm": 0.37033811211586, + "learning_rate": 2.5482e-05, + "loss": 0.0442, + "step": 8498 + }, + { + "epoch": 7.17820945945946, + "grad_norm": 0.5357081890106201, + "learning_rate": 2.5485e-05, + "loss": 0.04, + "step": 8499 + }, + { + "epoch": 7.179054054054054, + "grad_norm": 0.5518682599067688, + "learning_rate": 2.5488e-05, + "loss": 0.0723, + "step": 8500 + }, + { + "epoch": 7.179898648648648, + "grad_norm": 1.3197474479675293, + "learning_rate": 2.5491e-05, + "loss": 0.0295, + "step": 8501 + }, + { + "epoch": 7.180743243243243, + "grad_norm": 0.3259567320346832, + "learning_rate": 2.5494e-05, + "loss": 0.0436, + "step": 8502 + }, + { + "epoch": 7.181587837837838, + "grad_norm": 0.3532215356826782, + "learning_rate": 2.5497e-05, + "loss": 0.032, + "step": 8503 + }, + { + "epoch": 7.1824324324324325, + "grad_norm": 0.5612914562225342, + "learning_rate": 2.55e-05, + "loss": 0.0401, + "step": 8504 + }, + { + "epoch": 7.183277027027027, + "grad_norm": 0.31603604555130005, + "learning_rate": 2.5503e-05, + "loss": 0.0268, + "step": 8505 + }, + { + "epoch": 7.184121621621622, + "grad_norm": 0.39376300573349, + "learning_rate": 2.5506e-05, + "loss": 0.0365, + "step": 8506 + }, + { + "epoch": 7.184966216216216, + "grad_norm": 0.25829118490219116, + "learning_rate": 2.5509e-05, + "loss": 0.0207, + "step": 8507 + }, + { + "epoch": 7.1858108108108105, + "grad_norm": 0.3685970902442932, + "learning_rate": 2.5512000000000002e-05, + "loss": 0.0277, + "step": 8508 + }, + { + "epoch": 7.186655405405405, + "grad_norm": 0.48777610063552856, + "learning_rate": 2.5515000000000002e-05, + "loss": 0.0293, + "step": 8509 + }, + { + "epoch": 7.1875, + "grad_norm": 0.6827481985092163, + "learning_rate": 2.5518000000000002e-05, + "loss": 0.0318, + "step": 8510 + }, + { + "epoch": 7.188344594594595, + "grad_norm": 0.504529595375061, + "learning_rate": 2.5521000000000002e-05, + "loss": 0.0341, + "step": 8511 + }, + { + "epoch": 7.1891891891891895, + "grad_norm": 0.39415299892425537, + "learning_rate": 2.5524e-05, + "loss": 0.0426, + "step": 8512 + }, + { + "epoch": 7.190033783783784, + "grad_norm": 0.3551677167415619, + "learning_rate": 2.5527e-05, + "loss": 0.027, + "step": 8513 + }, + { + "epoch": 7.190878378378378, + "grad_norm": 0.5311535596847534, + "learning_rate": 2.553e-05, + "loss": 0.0387, + "step": 8514 + }, + { + "epoch": 7.191722972972973, + "grad_norm": 0.5646861791610718, + "learning_rate": 2.5533e-05, + "loss": 0.03, + "step": 8515 + }, + { + "epoch": 7.1925675675675675, + "grad_norm": 0.43853116035461426, + "learning_rate": 2.5536e-05, + "loss": 0.0323, + "step": 8516 + }, + { + "epoch": 7.193412162162162, + "grad_norm": 0.30260998010635376, + "learning_rate": 2.5539e-05, + "loss": 0.0243, + "step": 8517 + }, + { + "epoch": 7.194256756756757, + "grad_norm": 0.5613541603088379, + "learning_rate": 2.5542e-05, + "loss": 0.0404, + "step": 8518 + }, + { + "epoch": 7.195101351351352, + "grad_norm": 0.5683395266532898, + "learning_rate": 2.5545e-05, + "loss": 0.0571, + "step": 8519 + }, + { + "epoch": 7.195945945945946, + "grad_norm": 0.3037087023258209, + "learning_rate": 2.5548e-05, + "loss": 0.024, + "step": 8520 + }, + { + "epoch": 7.19679054054054, + "grad_norm": 0.3768540620803833, + "learning_rate": 2.5551e-05, + "loss": 0.0274, + "step": 8521 + }, + { + "epoch": 7.197635135135135, + "grad_norm": 0.6306260824203491, + "learning_rate": 2.5554e-05, + "loss": 0.0515, + "step": 8522 + }, + { + "epoch": 7.19847972972973, + "grad_norm": 0.505100429058075, + "learning_rate": 2.5557e-05, + "loss": 0.0335, + "step": 8523 + }, + { + "epoch": 7.199324324324325, + "grad_norm": 0.5756714940071106, + "learning_rate": 2.556e-05, + "loss": 0.0453, + "step": 8524 + }, + { + "epoch": 7.200168918918919, + "grad_norm": 0.2683015465736389, + "learning_rate": 2.5563e-05, + "loss": 0.0202, + "step": 8525 + }, + { + "epoch": 7.201013513513513, + "grad_norm": 0.4096314013004303, + "learning_rate": 2.5566e-05, + "loss": 0.0328, + "step": 8526 + }, + { + "epoch": 7.201858108108108, + "grad_norm": 0.5098023414611816, + "learning_rate": 2.5569e-05, + "loss": 0.0372, + "step": 8527 + }, + { + "epoch": 7.202702702702703, + "grad_norm": 0.8285034894943237, + "learning_rate": 2.5572000000000002e-05, + "loss": 0.037, + "step": 8528 + }, + { + "epoch": 7.203547297297297, + "grad_norm": 0.3532913029193878, + "learning_rate": 2.5575e-05, + "loss": 0.0471, + "step": 8529 + }, + { + "epoch": 7.204391891891892, + "grad_norm": 0.6022900938987732, + "learning_rate": 2.5578e-05, + "loss": 0.0543, + "step": 8530 + }, + { + "epoch": 7.205236486486487, + "grad_norm": 0.4556272029876709, + "learning_rate": 2.5581e-05, + "loss": 0.0477, + "step": 8531 + }, + { + "epoch": 7.206081081081081, + "grad_norm": 0.6759253144264221, + "learning_rate": 2.5584e-05, + "loss": 0.0347, + "step": 8532 + }, + { + "epoch": 7.206925675675675, + "grad_norm": 0.46499571204185486, + "learning_rate": 2.5587e-05, + "loss": 0.0343, + "step": 8533 + }, + { + "epoch": 7.20777027027027, + "grad_norm": 0.5218018293380737, + "learning_rate": 2.559e-05, + "loss": 0.04, + "step": 8534 + }, + { + "epoch": 7.208614864864865, + "grad_norm": 0.7059751749038696, + "learning_rate": 2.5593e-05, + "loss": 0.0409, + "step": 8535 + }, + { + "epoch": 7.20945945945946, + "grad_norm": 1.534882664680481, + "learning_rate": 2.5596e-05, + "loss": 0.0737, + "step": 8536 + }, + { + "epoch": 7.210304054054054, + "grad_norm": 1.2319904565811157, + "learning_rate": 2.5599e-05, + "loss": 0.0618, + "step": 8537 + }, + { + "epoch": 7.211148648648648, + "grad_norm": 1.8113229274749756, + "learning_rate": 2.5602000000000003e-05, + "loss": 0.062, + "step": 8538 + }, + { + "epoch": 7.211993243243243, + "grad_norm": 0.9469133019447327, + "learning_rate": 2.5605000000000003e-05, + "loss": 0.2779, + "step": 8539 + }, + { + "epoch": 7.212837837837838, + "grad_norm": 0.800683856010437, + "learning_rate": 2.5608000000000003e-05, + "loss": 0.1561, + "step": 8540 + }, + { + "epoch": 7.2136824324324325, + "grad_norm": 0.7202900648117065, + "learning_rate": 2.5611000000000003e-05, + "loss": 0.1085, + "step": 8541 + }, + { + "epoch": 7.214527027027027, + "grad_norm": 0.4876973330974579, + "learning_rate": 2.5614000000000002e-05, + "loss": 0.0569, + "step": 8542 + }, + { + "epoch": 7.215371621621622, + "grad_norm": 0.5293607115745544, + "learning_rate": 2.5617e-05, + "loss": 0.0501, + "step": 8543 + }, + { + "epoch": 7.216216216216216, + "grad_norm": 0.5568744540214539, + "learning_rate": 2.562e-05, + "loss": 0.0791, + "step": 8544 + }, + { + "epoch": 7.2170608108108105, + "grad_norm": 0.4050944447517395, + "learning_rate": 2.5623e-05, + "loss": 0.0458, + "step": 8545 + }, + { + "epoch": 7.217905405405405, + "grad_norm": 0.47549527883529663, + "learning_rate": 2.5625999999999998e-05, + "loss": 0.0369, + "step": 8546 + }, + { + "epoch": 7.21875, + "grad_norm": 0.5973921418190002, + "learning_rate": 2.5628999999999998e-05, + "loss": 0.0549, + "step": 8547 + }, + { + "epoch": 7.219594594594595, + "grad_norm": 0.4639471173286438, + "learning_rate": 2.5632e-05, + "loss": 0.0353, + "step": 8548 + }, + { + "epoch": 7.2204391891891895, + "grad_norm": 0.5232431888580322, + "learning_rate": 2.5635e-05, + "loss": 0.0354, + "step": 8549 + }, + { + "epoch": 7.221283783783784, + "grad_norm": 0.4719361960887909, + "learning_rate": 2.5638e-05, + "loss": 0.0456, + "step": 8550 + }, + { + "epoch": 7.222128378378378, + "grad_norm": 0.26942315697669983, + "learning_rate": 2.5641e-05, + "loss": 0.0271, + "step": 8551 + }, + { + "epoch": 7.222972972972973, + "grad_norm": 0.37482258677482605, + "learning_rate": 2.5644e-05, + "loss": 0.03, + "step": 8552 + }, + { + "epoch": 7.2238175675675675, + "grad_norm": 0.4840698838233948, + "learning_rate": 2.5647e-05, + "loss": 0.0499, + "step": 8553 + }, + { + "epoch": 7.224662162162162, + "grad_norm": 0.533842921257019, + "learning_rate": 2.565e-05, + "loss": 0.0447, + "step": 8554 + }, + { + "epoch": 7.225506756756757, + "grad_norm": 0.3701755106449127, + "learning_rate": 2.5653e-05, + "loss": 0.0423, + "step": 8555 + }, + { + "epoch": 7.226351351351352, + "grad_norm": 0.6063529849052429, + "learning_rate": 2.5656e-05, + "loss": 0.0409, + "step": 8556 + }, + { + "epoch": 7.227195945945946, + "grad_norm": 0.3654711842536926, + "learning_rate": 2.5659e-05, + "loss": 0.0272, + "step": 8557 + }, + { + "epoch": 7.22804054054054, + "grad_norm": 0.5255247354507446, + "learning_rate": 2.5662000000000003e-05, + "loss": 0.0623, + "step": 8558 + }, + { + "epoch": 7.228885135135135, + "grad_norm": 0.6081040501594543, + "learning_rate": 2.5665000000000002e-05, + "loss": 0.0286, + "step": 8559 + }, + { + "epoch": 7.22972972972973, + "grad_norm": 0.3145562708377838, + "learning_rate": 2.5668000000000002e-05, + "loss": 0.0313, + "step": 8560 + }, + { + "epoch": 7.230574324324325, + "grad_norm": 0.450381338596344, + "learning_rate": 2.5671000000000002e-05, + "loss": 0.0339, + "step": 8561 + }, + { + "epoch": 7.231418918918919, + "grad_norm": 0.2907196879386902, + "learning_rate": 2.5674000000000002e-05, + "loss": 0.0242, + "step": 8562 + }, + { + "epoch": 7.232263513513513, + "grad_norm": 1.2510194778442383, + "learning_rate": 2.5677e-05, + "loss": 0.0324, + "step": 8563 + }, + { + "epoch": 7.233108108108108, + "grad_norm": 0.7666230797767639, + "learning_rate": 2.568e-05, + "loss": 0.0553, + "step": 8564 + }, + { + "epoch": 7.233952702702703, + "grad_norm": 0.35250377655029297, + "learning_rate": 2.5683e-05, + "loss": 0.0385, + "step": 8565 + }, + { + "epoch": 7.234797297297297, + "grad_norm": 0.4504818916320801, + "learning_rate": 2.5686e-05, + "loss": 0.0392, + "step": 8566 + }, + { + "epoch": 7.235641891891892, + "grad_norm": 0.3886551856994629, + "learning_rate": 2.5688999999999997e-05, + "loss": 0.0363, + "step": 8567 + }, + { + "epoch": 7.236486486486487, + "grad_norm": 1.3936587572097778, + "learning_rate": 2.5692e-05, + "loss": 0.0448, + "step": 8568 + }, + { + "epoch": 7.237331081081081, + "grad_norm": 0.34427130222320557, + "learning_rate": 2.5695e-05, + "loss": 0.0208, + "step": 8569 + }, + { + "epoch": 7.238175675675675, + "grad_norm": 0.6836726665496826, + "learning_rate": 2.5698e-05, + "loss": 0.0368, + "step": 8570 + }, + { + "epoch": 7.23902027027027, + "grad_norm": 0.46853944659233093, + "learning_rate": 2.5701e-05, + "loss": 0.0309, + "step": 8571 + }, + { + "epoch": 7.239864864864865, + "grad_norm": 0.29830658435821533, + "learning_rate": 2.5704e-05, + "loss": 0.026, + "step": 8572 + }, + { + "epoch": 7.24070945945946, + "grad_norm": 0.847935140132904, + "learning_rate": 2.5707e-05, + "loss": 0.0562, + "step": 8573 + }, + { + "epoch": 7.241554054054054, + "grad_norm": 0.3367640972137451, + "learning_rate": 2.571e-05, + "loss": 0.0297, + "step": 8574 + }, + { + "epoch": 7.242398648648648, + "grad_norm": 0.5794939398765564, + "learning_rate": 2.5713e-05, + "loss": 0.0463, + "step": 8575 + }, + { + "epoch": 7.243243243243243, + "grad_norm": 0.5206461548805237, + "learning_rate": 2.5716e-05, + "loss": 0.029, + "step": 8576 + }, + { + "epoch": 7.244087837837838, + "grad_norm": 0.5402809381484985, + "learning_rate": 2.5719e-05, + "loss": 0.0336, + "step": 8577 + }, + { + "epoch": 7.2449324324324325, + "grad_norm": 0.7943912148475647, + "learning_rate": 2.5722000000000002e-05, + "loss": 0.0488, + "step": 8578 + }, + { + "epoch": 7.245777027027027, + "grad_norm": 0.24932438135147095, + "learning_rate": 2.5725000000000002e-05, + "loss": 0.0253, + "step": 8579 + }, + { + "epoch": 7.246621621621622, + "grad_norm": 1.0628758668899536, + "learning_rate": 2.5728e-05, + "loss": 0.0518, + "step": 8580 + }, + { + "epoch": 7.247466216216216, + "grad_norm": 0.9450965523719788, + "learning_rate": 2.5731e-05, + "loss": 0.0587, + "step": 8581 + }, + { + "epoch": 7.2483108108108105, + "grad_norm": 0.5268612504005432, + "learning_rate": 2.5734e-05, + "loss": 0.0685, + "step": 8582 + }, + { + "epoch": 7.249155405405405, + "grad_norm": 0.6995041370391846, + "learning_rate": 2.5737e-05, + "loss": 0.0563, + "step": 8583 + }, + { + "epoch": 7.25, + "grad_norm": 0.4073774516582489, + "learning_rate": 2.574e-05, + "loss": 0.0386, + "step": 8584 + }, + { + "epoch": 7.250844594594595, + "grad_norm": 0.42945197224617004, + "learning_rate": 2.5743e-05, + "loss": 0.0334, + "step": 8585 + }, + { + "epoch": 7.2516891891891895, + "grad_norm": 0.584794819355011, + "learning_rate": 2.5746e-05, + "loss": 0.0531, + "step": 8586 + }, + { + "epoch": 7.252533783783784, + "grad_norm": 0.9067978262901306, + "learning_rate": 2.5749e-05, + "loss": 0.0869, + "step": 8587 + }, + { + "epoch": 7.253378378378378, + "grad_norm": 0.768437385559082, + "learning_rate": 2.5752000000000003e-05, + "loss": 0.0668, + "step": 8588 + }, + { + "epoch": 7.254222972972973, + "grad_norm": 1.6061158180236816, + "learning_rate": 2.5755000000000003e-05, + "loss": 0.4166, + "step": 8589 + }, + { + "epoch": 7.2550675675675675, + "grad_norm": 0.5052465200424194, + "learning_rate": 2.5758000000000003e-05, + "loss": 0.1511, + "step": 8590 + }, + { + "epoch": 7.255912162162162, + "grad_norm": 0.42295509576797485, + "learning_rate": 2.5761000000000003e-05, + "loss": 0.0756, + "step": 8591 + }, + { + "epoch": 7.256756756756757, + "grad_norm": 0.5117958188056946, + "learning_rate": 2.5764e-05, + "loss": 0.0614, + "step": 8592 + }, + { + "epoch": 7.257601351351352, + "grad_norm": 0.4267946779727936, + "learning_rate": 2.5767e-05, + "loss": 0.06, + "step": 8593 + }, + { + "epoch": 7.258445945945946, + "grad_norm": 0.40098392963409424, + "learning_rate": 2.577e-05, + "loss": 0.0385, + "step": 8594 + }, + { + "epoch": 7.25929054054054, + "grad_norm": 0.4249478578567505, + "learning_rate": 2.5773e-05, + "loss": 0.0437, + "step": 8595 + }, + { + "epoch": 7.260135135135135, + "grad_norm": 0.27191251516342163, + "learning_rate": 2.5776e-05, + "loss": 0.0302, + "step": 8596 + }, + { + "epoch": 7.26097972972973, + "grad_norm": 0.5626140832901001, + "learning_rate": 2.5779e-05, + "loss": 0.0444, + "step": 8597 + }, + { + "epoch": 7.261824324324325, + "grad_norm": 0.5263160467147827, + "learning_rate": 2.5782e-05, + "loss": 0.0505, + "step": 8598 + }, + { + "epoch": 7.262668918918919, + "grad_norm": 0.3674556612968445, + "learning_rate": 2.5785e-05, + "loss": 0.0384, + "step": 8599 + }, + { + "epoch": 7.263513513513513, + "grad_norm": 0.6449187994003296, + "learning_rate": 2.5788e-05, + "loss": 0.0448, + "step": 8600 + }, + { + "epoch": 7.264358108108108, + "grad_norm": 0.5586270093917847, + "learning_rate": 2.5791e-05, + "loss": 0.0328, + "step": 8601 + }, + { + "epoch": 7.265202702702703, + "grad_norm": 0.479469895362854, + "learning_rate": 2.5794e-05, + "loss": 0.0482, + "step": 8602 + }, + { + "epoch": 7.266047297297297, + "grad_norm": 0.35305944085121155, + "learning_rate": 2.5797e-05, + "loss": 0.0327, + "step": 8603 + }, + { + "epoch": 7.266891891891892, + "grad_norm": 0.42617860436439514, + "learning_rate": 2.58e-05, + "loss": 0.0318, + "step": 8604 + }, + { + "epoch": 7.267736486486487, + "grad_norm": 0.507222056388855, + "learning_rate": 2.5803e-05, + "loss": 0.0416, + "step": 8605 + }, + { + "epoch": 7.268581081081081, + "grad_norm": 0.5674110054969788, + "learning_rate": 2.5806e-05, + "loss": 0.0588, + "step": 8606 + }, + { + "epoch": 7.269425675675675, + "grad_norm": 0.2878146469593048, + "learning_rate": 2.5809e-05, + "loss": 0.0178, + "step": 8607 + }, + { + "epoch": 7.27027027027027, + "grad_norm": 0.34057557582855225, + "learning_rate": 2.5812000000000003e-05, + "loss": 0.0393, + "step": 8608 + }, + { + "epoch": 7.271114864864865, + "grad_norm": 0.4658903181552887, + "learning_rate": 2.5815000000000003e-05, + "loss": 0.0454, + "step": 8609 + }, + { + "epoch": 7.27195945945946, + "grad_norm": 0.3138697147369385, + "learning_rate": 2.5818000000000003e-05, + "loss": 0.0294, + "step": 8610 + }, + { + "epoch": 7.272804054054054, + "grad_norm": 0.38175714015960693, + "learning_rate": 2.5821000000000002e-05, + "loss": 0.041, + "step": 8611 + }, + { + "epoch": 7.273648648648648, + "grad_norm": 0.4187347888946533, + "learning_rate": 2.5824000000000002e-05, + "loss": 0.0513, + "step": 8612 + }, + { + "epoch": 7.274493243243243, + "grad_norm": 0.6072432994842529, + "learning_rate": 2.5827000000000002e-05, + "loss": 0.0767, + "step": 8613 + }, + { + "epoch": 7.275337837837838, + "grad_norm": 0.515201985836029, + "learning_rate": 2.5830000000000002e-05, + "loss": 0.0478, + "step": 8614 + }, + { + "epoch": 7.2761824324324325, + "grad_norm": 0.7071685791015625, + "learning_rate": 2.5833e-05, + "loss": 0.0404, + "step": 8615 + }, + { + "epoch": 7.277027027027027, + "grad_norm": 0.39766520261764526, + "learning_rate": 2.5835999999999998e-05, + "loss": 0.0344, + "step": 8616 + }, + { + "epoch": 7.277871621621622, + "grad_norm": 0.3867991864681244, + "learning_rate": 2.5838999999999998e-05, + "loss": 0.0354, + "step": 8617 + }, + { + "epoch": 7.278716216216216, + "grad_norm": 0.5186421871185303, + "learning_rate": 2.5842e-05, + "loss": 0.0386, + "step": 8618 + }, + { + "epoch": 7.2795608108108105, + "grad_norm": 0.562985360622406, + "learning_rate": 2.5845e-05, + "loss": 0.0415, + "step": 8619 + }, + { + "epoch": 7.280405405405405, + "grad_norm": 0.3351535499095917, + "learning_rate": 2.5848e-05, + "loss": 0.0387, + "step": 8620 + }, + { + "epoch": 7.28125, + "grad_norm": 0.4250340163707733, + "learning_rate": 2.5851e-05, + "loss": 0.0359, + "step": 8621 + }, + { + "epoch": 7.282094594594595, + "grad_norm": 0.6618567109107971, + "learning_rate": 2.5854e-05, + "loss": 0.039, + "step": 8622 + }, + { + "epoch": 7.2829391891891895, + "grad_norm": 0.5652826428413391, + "learning_rate": 2.5857e-05, + "loss": 0.0335, + "step": 8623 + }, + { + "epoch": 7.283783783783784, + "grad_norm": 0.3465772569179535, + "learning_rate": 2.586e-05, + "loss": 0.0268, + "step": 8624 + }, + { + "epoch": 7.284628378378378, + "grad_norm": 0.5968042612075806, + "learning_rate": 2.5863e-05, + "loss": 0.0565, + "step": 8625 + }, + { + "epoch": 7.285472972972973, + "grad_norm": 0.4769721031188965, + "learning_rate": 2.5866e-05, + "loss": 0.0333, + "step": 8626 + }, + { + "epoch": 7.2863175675675675, + "grad_norm": 0.34893089532852173, + "learning_rate": 2.5869e-05, + "loss": 0.0359, + "step": 8627 + }, + { + "epoch": 7.287162162162162, + "grad_norm": 0.4646770656108856, + "learning_rate": 2.5872000000000002e-05, + "loss": 0.0473, + "step": 8628 + }, + { + "epoch": 7.288006756756757, + "grad_norm": 0.9364349842071533, + "learning_rate": 2.5875000000000002e-05, + "loss": 0.0358, + "step": 8629 + }, + { + "epoch": 7.288851351351352, + "grad_norm": 0.4130035638809204, + "learning_rate": 2.5878000000000002e-05, + "loss": 0.0243, + "step": 8630 + }, + { + "epoch": 7.289695945945946, + "grad_norm": 0.5131937861442566, + "learning_rate": 2.5881000000000002e-05, + "loss": 0.0543, + "step": 8631 + }, + { + "epoch": 7.29054054054054, + "grad_norm": 0.5410783290863037, + "learning_rate": 2.5884e-05, + "loss": 0.0251, + "step": 8632 + }, + { + "epoch": 7.291385135135135, + "grad_norm": 0.5427870154380798, + "learning_rate": 2.5887e-05, + "loss": 0.0493, + "step": 8633 + }, + { + "epoch": 7.29222972972973, + "grad_norm": 0.8988054990768433, + "learning_rate": 2.589e-05, + "loss": 0.0589, + "step": 8634 + }, + { + "epoch": 7.293074324324325, + "grad_norm": 0.5338316559791565, + "learning_rate": 2.5893e-05, + "loss": 0.0571, + "step": 8635 + }, + { + "epoch": 7.293918918918919, + "grad_norm": 1.0917497873306274, + "learning_rate": 2.5896e-05, + "loss": 0.0861, + "step": 8636 + }, + { + "epoch": 7.294763513513513, + "grad_norm": 1.197060227394104, + "learning_rate": 2.5899e-05, + "loss": 0.1321, + "step": 8637 + }, + { + "epoch": 7.295608108108108, + "grad_norm": 0.7937166690826416, + "learning_rate": 2.5902e-05, + "loss": 0.069, + "step": 8638 + }, + { + "epoch": 7.296452702702703, + "grad_norm": 1.5845882892608643, + "learning_rate": 2.5905000000000004e-05, + "loss": 0.3512, + "step": 8639 + }, + { + "epoch": 7.297297297297297, + "grad_norm": 0.9585224986076355, + "learning_rate": 2.5908000000000003e-05, + "loss": 0.1987, + "step": 8640 + }, + { + "epoch": 7.298141891891892, + "grad_norm": 0.4205320179462433, + "learning_rate": 2.5911e-05, + "loss": 0.1154, + "step": 8641 + }, + { + "epoch": 7.298986486486487, + "grad_norm": 0.38306522369384766, + "learning_rate": 2.5914e-05, + "loss": 0.0653, + "step": 8642 + }, + { + "epoch": 7.299831081081081, + "grad_norm": 0.40715426206588745, + "learning_rate": 2.5917e-05, + "loss": 0.0506, + "step": 8643 + }, + { + "epoch": 7.300675675675675, + "grad_norm": 0.4549446702003479, + "learning_rate": 2.592e-05, + "loss": 0.0445, + "step": 8644 + }, + { + "epoch": 7.30152027027027, + "grad_norm": 0.4436320960521698, + "learning_rate": 2.5923e-05, + "loss": 0.037, + "step": 8645 + }, + { + "epoch": 7.302364864864865, + "grad_norm": 0.3271903395652771, + "learning_rate": 2.5926e-05, + "loss": 0.037, + "step": 8646 + }, + { + "epoch": 7.30320945945946, + "grad_norm": 0.37218421697616577, + "learning_rate": 2.5929e-05, + "loss": 0.0441, + "step": 8647 + }, + { + "epoch": 7.304054054054054, + "grad_norm": 0.6186562180519104, + "learning_rate": 2.5932e-05, + "loss": 0.0335, + "step": 8648 + }, + { + "epoch": 7.304898648648648, + "grad_norm": 0.3985997140407562, + "learning_rate": 2.5935e-05, + "loss": 0.0395, + "step": 8649 + }, + { + "epoch": 7.305743243243243, + "grad_norm": 0.48227986693382263, + "learning_rate": 2.5938e-05, + "loss": 0.0395, + "step": 8650 + }, + { + "epoch": 7.306587837837838, + "grad_norm": 0.4476497173309326, + "learning_rate": 2.5941e-05, + "loss": 0.0456, + "step": 8651 + }, + { + "epoch": 7.3074324324324325, + "grad_norm": 0.4074764847755432, + "learning_rate": 2.5944e-05, + "loss": 0.0321, + "step": 8652 + }, + { + "epoch": 7.308277027027027, + "grad_norm": 0.5161740183830261, + "learning_rate": 2.5947e-05, + "loss": 0.0372, + "step": 8653 + }, + { + "epoch": 7.309121621621622, + "grad_norm": 0.3511922061443329, + "learning_rate": 2.595e-05, + "loss": 0.031, + "step": 8654 + }, + { + "epoch": 7.309966216216216, + "grad_norm": 0.3920682668685913, + "learning_rate": 2.5953e-05, + "loss": 0.0341, + "step": 8655 + }, + { + "epoch": 7.3108108108108105, + "grad_norm": 0.5667632222175598, + "learning_rate": 2.5956e-05, + "loss": 0.0409, + "step": 8656 + }, + { + "epoch": 7.311655405405405, + "grad_norm": 0.3514721095561981, + "learning_rate": 2.5959e-05, + "loss": 0.0275, + "step": 8657 + }, + { + "epoch": 7.3125, + "grad_norm": 0.5375425219535828, + "learning_rate": 2.5962e-05, + "loss": 0.0413, + "step": 8658 + }, + { + "epoch": 7.313344594594595, + "grad_norm": 0.46534425020217896, + "learning_rate": 2.5965000000000003e-05, + "loss": 0.031, + "step": 8659 + }, + { + "epoch": 7.3141891891891895, + "grad_norm": 0.2927438020706177, + "learning_rate": 2.5968000000000003e-05, + "loss": 0.0386, + "step": 8660 + }, + { + "epoch": 7.315033783783784, + "grad_norm": 0.3845786154270172, + "learning_rate": 2.5971000000000003e-05, + "loss": 0.03, + "step": 8661 + }, + { + "epoch": 7.315878378378378, + "grad_norm": 0.396445095539093, + "learning_rate": 2.5974000000000002e-05, + "loss": 0.0302, + "step": 8662 + }, + { + "epoch": 7.316722972972973, + "grad_norm": 0.43079230189323425, + "learning_rate": 2.5977000000000002e-05, + "loss": 0.0303, + "step": 8663 + }, + { + "epoch": 7.3175675675675675, + "grad_norm": 0.5740931034088135, + "learning_rate": 2.5980000000000002e-05, + "loss": 0.0379, + "step": 8664 + }, + { + "epoch": 7.318412162162162, + "grad_norm": 0.2914179563522339, + "learning_rate": 2.5983000000000002e-05, + "loss": 0.032, + "step": 8665 + }, + { + "epoch": 7.319256756756757, + "grad_norm": 0.40606585144996643, + "learning_rate": 2.5985999999999998e-05, + "loss": 0.0314, + "step": 8666 + }, + { + "epoch": 7.320101351351352, + "grad_norm": 0.25999507308006287, + "learning_rate": 2.5988999999999998e-05, + "loss": 0.0154, + "step": 8667 + }, + { + "epoch": 7.320945945945946, + "grad_norm": 0.31808850169181824, + "learning_rate": 2.5991999999999998e-05, + "loss": 0.0259, + "step": 8668 + }, + { + "epoch": 7.32179054054054, + "grad_norm": 0.4533661901950836, + "learning_rate": 2.5995e-05, + "loss": 0.037, + "step": 8669 + }, + { + "epoch": 7.322635135135135, + "grad_norm": 0.4984133541584015, + "learning_rate": 2.5998e-05, + "loss": 0.0347, + "step": 8670 + }, + { + "epoch": 7.32347972972973, + "grad_norm": 0.4866260588169098, + "learning_rate": 2.6001e-05, + "loss": 0.0435, + "step": 8671 + }, + { + "epoch": 7.324324324324325, + "grad_norm": 0.9201352000236511, + "learning_rate": 2.6004e-05, + "loss": 0.046, + "step": 8672 + }, + { + "epoch": 7.325168918918919, + "grad_norm": 0.5781540870666504, + "learning_rate": 2.6007e-05, + "loss": 0.0338, + "step": 8673 + }, + { + "epoch": 7.326013513513513, + "grad_norm": 0.3367605209350586, + "learning_rate": 2.601e-05, + "loss": 0.028, + "step": 8674 + }, + { + "epoch": 7.326858108108108, + "grad_norm": 0.5143247842788696, + "learning_rate": 2.6013e-05, + "loss": 0.032, + "step": 8675 + }, + { + "epoch": 7.327702702702703, + "grad_norm": 0.5611680746078491, + "learning_rate": 2.6016e-05, + "loss": 0.0348, + "step": 8676 + }, + { + "epoch": 7.328547297297297, + "grad_norm": 0.6281130909919739, + "learning_rate": 2.6019e-05, + "loss": 0.0372, + "step": 8677 + }, + { + "epoch": 7.329391891891892, + "grad_norm": 0.5369094014167786, + "learning_rate": 2.6022e-05, + "loss": 0.0311, + "step": 8678 + }, + { + "epoch": 7.330236486486487, + "grad_norm": 0.4538416862487793, + "learning_rate": 2.6025000000000002e-05, + "loss": 0.0348, + "step": 8679 + }, + { + "epoch": 7.331081081081081, + "grad_norm": 0.7179861664772034, + "learning_rate": 2.6028000000000002e-05, + "loss": 0.032, + "step": 8680 + }, + { + "epoch": 7.331925675675675, + "grad_norm": 0.5055319666862488, + "learning_rate": 2.6031000000000002e-05, + "loss": 0.0333, + "step": 8681 + }, + { + "epoch": 7.33277027027027, + "grad_norm": 1.4941009283065796, + "learning_rate": 2.6034000000000002e-05, + "loss": 0.0501, + "step": 8682 + }, + { + "epoch": 7.333614864864865, + "grad_norm": 0.4732950031757355, + "learning_rate": 2.6037e-05, + "loss": 0.0276, + "step": 8683 + }, + { + "epoch": 7.33445945945946, + "grad_norm": 0.9671759009361267, + "learning_rate": 2.604e-05, + "loss": 0.0426, + "step": 8684 + }, + { + "epoch": 7.335304054054054, + "grad_norm": 1.0429844856262207, + "learning_rate": 2.6043e-05, + "loss": 0.0729, + "step": 8685 + }, + { + "epoch": 7.336148648648648, + "grad_norm": 0.4461454451084137, + "learning_rate": 2.6046e-05, + "loss": 0.027, + "step": 8686 + }, + { + "epoch": 7.336993243243243, + "grad_norm": 0.6704974174499512, + "learning_rate": 2.6049e-05, + "loss": 0.0373, + "step": 8687 + }, + { + "epoch": 7.337837837837838, + "grad_norm": 0.6631040573120117, + "learning_rate": 2.6052e-05, + "loss": 0.0532, + "step": 8688 + }, + { + "epoch": 7.3386824324324325, + "grad_norm": 1.0330389738082886, + "learning_rate": 2.6055000000000004e-05, + "loss": 0.2825, + "step": 8689 + }, + { + "epoch": 7.339527027027027, + "grad_norm": 0.7299720644950867, + "learning_rate": 2.6058e-05, + "loss": 0.17, + "step": 8690 + }, + { + "epoch": 7.340371621621622, + "grad_norm": 1.015002965927124, + "learning_rate": 2.6061e-05, + "loss": 0.1093, + "step": 8691 + }, + { + "epoch": 7.341216216216216, + "grad_norm": 0.3915115296840668, + "learning_rate": 2.6064e-05, + "loss": 0.0483, + "step": 8692 + }, + { + "epoch": 7.3420608108108105, + "grad_norm": 0.5515143871307373, + "learning_rate": 2.6067e-05, + "loss": 0.0862, + "step": 8693 + }, + { + "epoch": 7.342905405405405, + "grad_norm": 1.1004587411880493, + "learning_rate": 2.607e-05, + "loss": 0.0575, + "step": 8694 + }, + { + "epoch": 7.34375, + "grad_norm": 0.38967961072921753, + "learning_rate": 2.6073e-05, + "loss": 0.0375, + "step": 8695 + }, + { + "epoch": 7.344594594594595, + "grad_norm": 0.34241175651550293, + "learning_rate": 2.6076e-05, + "loss": 0.0389, + "step": 8696 + }, + { + "epoch": 7.3454391891891895, + "grad_norm": 0.4095127284526825, + "learning_rate": 2.6079e-05, + "loss": 0.0345, + "step": 8697 + }, + { + "epoch": 7.346283783783784, + "grad_norm": 0.408353716135025, + "learning_rate": 2.6082e-05, + "loss": 0.05, + "step": 8698 + }, + { + "epoch": 7.347128378378378, + "grad_norm": 0.34010133147239685, + "learning_rate": 2.6085000000000002e-05, + "loss": 0.0331, + "step": 8699 + }, + { + "epoch": 7.347972972972973, + "grad_norm": 0.41459378600120544, + "learning_rate": 2.6088e-05, + "loss": 0.0506, + "step": 8700 + }, + { + "epoch": 7.3488175675675675, + "grad_norm": 0.3765791058540344, + "learning_rate": 2.6091e-05, + "loss": 0.0368, + "step": 8701 + }, + { + "epoch": 7.349662162162162, + "grad_norm": 0.43193745613098145, + "learning_rate": 2.6094e-05, + "loss": 0.0508, + "step": 8702 + }, + { + "epoch": 7.350506756756757, + "grad_norm": 0.41706177592277527, + "learning_rate": 2.6097e-05, + "loss": 0.0326, + "step": 8703 + }, + { + "epoch": 7.351351351351352, + "grad_norm": 0.3396018147468567, + "learning_rate": 2.61e-05, + "loss": 0.0295, + "step": 8704 + }, + { + "epoch": 7.352195945945946, + "grad_norm": 0.6707274913787842, + "learning_rate": 2.6103e-05, + "loss": 0.0506, + "step": 8705 + }, + { + "epoch": 7.35304054054054, + "grad_norm": 0.6220424771308899, + "learning_rate": 2.6106e-05, + "loss": 0.0274, + "step": 8706 + }, + { + "epoch": 7.353885135135135, + "grad_norm": 0.39424869418144226, + "learning_rate": 2.6109e-05, + "loss": 0.0297, + "step": 8707 + }, + { + "epoch": 7.35472972972973, + "grad_norm": 0.4330022633075714, + "learning_rate": 2.6112e-05, + "loss": 0.0385, + "step": 8708 + }, + { + "epoch": 7.355574324324325, + "grad_norm": 0.45702749490737915, + "learning_rate": 2.6115000000000003e-05, + "loss": 0.0421, + "step": 8709 + }, + { + "epoch": 7.356418918918919, + "grad_norm": 0.7401296496391296, + "learning_rate": 2.6118000000000003e-05, + "loss": 0.0443, + "step": 8710 + }, + { + "epoch": 7.357263513513513, + "grad_norm": 0.5731768608093262, + "learning_rate": 2.6121000000000003e-05, + "loss": 0.0434, + "step": 8711 + }, + { + "epoch": 7.358108108108108, + "grad_norm": 0.7124797105789185, + "learning_rate": 2.6124000000000003e-05, + "loss": 0.0358, + "step": 8712 + }, + { + "epoch": 7.358952702702703, + "grad_norm": 0.7695537805557251, + "learning_rate": 2.6127000000000002e-05, + "loss": 0.0371, + "step": 8713 + }, + { + "epoch": 7.359797297297297, + "grad_norm": 0.3506016433238983, + "learning_rate": 2.6130000000000002e-05, + "loss": 0.0237, + "step": 8714 + }, + { + "epoch": 7.360641891891892, + "grad_norm": 1.4654065370559692, + "learning_rate": 2.6133e-05, + "loss": 0.054, + "step": 8715 + }, + { + "epoch": 7.361486486486487, + "grad_norm": 0.6220901012420654, + "learning_rate": 2.6136e-05, + "loss": 0.0441, + "step": 8716 + }, + { + "epoch": 7.362331081081081, + "grad_norm": 0.8654173612594604, + "learning_rate": 2.6138999999999998e-05, + "loss": 0.0316, + "step": 8717 + }, + { + "epoch": 7.363175675675675, + "grad_norm": 0.6739197969436646, + "learning_rate": 2.6141999999999998e-05, + "loss": 0.042, + "step": 8718 + }, + { + "epoch": 7.36402027027027, + "grad_norm": 0.36859166622161865, + "learning_rate": 2.6145e-05, + "loss": 0.0279, + "step": 8719 + }, + { + "epoch": 7.364864864864865, + "grad_norm": 0.4210512042045593, + "learning_rate": 2.6148e-05, + "loss": 0.0363, + "step": 8720 + }, + { + "epoch": 7.36570945945946, + "grad_norm": 0.6009230017662048, + "learning_rate": 2.6151e-05, + "loss": 0.0451, + "step": 8721 + }, + { + "epoch": 7.366554054054054, + "grad_norm": 0.7459216713905334, + "learning_rate": 2.6154e-05, + "loss": 0.0284, + "step": 8722 + }, + { + "epoch": 7.367398648648648, + "grad_norm": 0.6469944715499878, + "learning_rate": 2.6157e-05, + "loss": 0.043, + "step": 8723 + }, + { + "epoch": 7.368243243243243, + "grad_norm": 0.35342687368392944, + "learning_rate": 2.616e-05, + "loss": 0.0321, + "step": 8724 + }, + { + "epoch": 7.369087837837838, + "grad_norm": 0.465581476688385, + "learning_rate": 2.6163e-05, + "loss": 0.0434, + "step": 8725 + }, + { + "epoch": 7.3699324324324325, + "grad_norm": 0.604770302772522, + "learning_rate": 2.6166e-05, + "loss": 0.0387, + "step": 8726 + }, + { + "epoch": 7.370777027027027, + "grad_norm": 0.43512198328971863, + "learning_rate": 2.6169e-05, + "loss": 0.0396, + "step": 8727 + }, + { + "epoch": 7.371621621621622, + "grad_norm": 0.5767526030540466, + "learning_rate": 2.6172e-05, + "loss": 0.0377, + "step": 8728 + }, + { + "epoch": 7.372466216216216, + "grad_norm": 0.6290702819824219, + "learning_rate": 2.6175000000000003e-05, + "loss": 0.041, + "step": 8729 + }, + { + "epoch": 7.3733108108108105, + "grad_norm": 0.48331668972969055, + "learning_rate": 2.6178000000000002e-05, + "loss": 0.0364, + "step": 8730 + }, + { + "epoch": 7.374155405405405, + "grad_norm": 0.37729015946388245, + "learning_rate": 2.6181000000000002e-05, + "loss": 0.0395, + "step": 8731 + }, + { + "epoch": 7.375, + "grad_norm": 0.5642907619476318, + "learning_rate": 2.6184000000000002e-05, + "loss": 0.0324, + "step": 8732 + }, + { + "epoch": 7.375844594594595, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.6187000000000002e-05, + "loss": 0.0253, + "step": 8733 + }, + { + "epoch": 7.3766891891891895, + "grad_norm": 0.4670281708240509, + "learning_rate": 2.619e-05, + "loss": 0.0264, + "step": 8734 + }, + { + "epoch": 7.377533783783784, + "grad_norm": 0.6384817957878113, + "learning_rate": 2.6193e-05, + "loss": 0.0357, + "step": 8735 + }, + { + "epoch": 7.378378378378378, + "grad_norm": 0.5058684945106506, + "learning_rate": 2.6196e-05, + "loss": 0.0426, + "step": 8736 + }, + { + "epoch": 7.379222972972973, + "grad_norm": 1.1060138940811157, + "learning_rate": 2.6199e-05, + "loss": 0.0516, + "step": 8737 + }, + { + "epoch": 7.3800675675675675, + "grad_norm": 2.6437129974365234, + "learning_rate": 2.6202e-05, + "loss": 0.1272, + "step": 8738 + }, + { + "epoch": 7.380912162162162, + "grad_norm": 1.4078913927078247, + "learning_rate": 2.6205e-05, + "loss": 0.3738, + "step": 8739 + }, + { + "epoch": 7.381756756756757, + "grad_norm": 0.6549994945526123, + "learning_rate": 2.6208e-05, + "loss": 0.18, + "step": 8740 + }, + { + "epoch": 7.382601351351352, + "grad_norm": 0.3886873722076416, + "learning_rate": 2.6211e-05, + "loss": 0.0997, + "step": 8741 + }, + { + "epoch": 7.383445945945946, + "grad_norm": 0.5467851161956787, + "learning_rate": 2.6214e-05, + "loss": 0.0676, + "step": 8742 + }, + { + "epoch": 7.38429054054054, + "grad_norm": 0.39512017369270325, + "learning_rate": 2.6217e-05, + "loss": 0.0377, + "step": 8743 + }, + { + "epoch": 7.385135135135135, + "grad_norm": 0.4506242871284485, + "learning_rate": 2.622e-05, + "loss": 0.0511, + "step": 8744 + }, + { + "epoch": 7.38597972972973, + "grad_norm": 0.4816721975803375, + "learning_rate": 2.6223e-05, + "loss": 0.0501, + "step": 8745 + }, + { + "epoch": 7.386824324324325, + "grad_norm": 0.3809327483177185, + "learning_rate": 2.6226e-05, + "loss": 0.0403, + "step": 8746 + }, + { + "epoch": 7.387668918918919, + "grad_norm": 0.33181968331336975, + "learning_rate": 2.6229e-05, + "loss": 0.0409, + "step": 8747 + }, + { + "epoch": 7.388513513513513, + "grad_norm": 0.40859079360961914, + "learning_rate": 2.6232e-05, + "loss": 0.0424, + "step": 8748 + }, + { + "epoch": 7.389358108108108, + "grad_norm": 0.41955819725990295, + "learning_rate": 2.6235000000000002e-05, + "loss": 0.0492, + "step": 8749 + }, + { + "epoch": 7.390202702702703, + "grad_norm": 0.38046595454216003, + "learning_rate": 2.6238000000000002e-05, + "loss": 0.0259, + "step": 8750 + }, + { + "epoch": 7.391047297297297, + "grad_norm": 0.5138417482376099, + "learning_rate": 2.6241e-05, + "loss": 0.0637, + "step": 8751 + }, + { + "epoch": 7.391891891891892, + "grad_norm": 0.2777215540409088, + "learning_rate": 2.6244e-05, + "loss": 0.0287, + "step": 8752 + }, + { + "epoch": 7.392736486486487, + "grad_norm": 0.8079548478126526, + "learning_rate": 2.6247e-05, + "loss": 0.0415, + "step": 8753 + }, + { + "epoch": 7.393581081081081, + "grad_norm": 0.3695757985115051, + "learning_rate": 2.625e-05, + "loss": 0.0336, + "step": 8754 + }, + { + "epoch": 7.394425675675675, + "grad_norm": 0.5563624501228333, + "learning_rate": 2.6253e-05, + "loss": 0.0476, + "step": 8755 + }, + { + "epoch": 7.39527027027027, + "grad_norm": 0.25924429297447205, + "learning_rate": 2.6256e-05, + "loss": 0.0201, + "step": 8756 + }, + { + "epoch": 7.396114864864865, + "grad_norm": 0.4828863739967346, + "learning_rate": 2.6259e-05, + "loss": 0.0393, + "step": 8757 + }, + { + "epoch": 7.39695945945946, + "grad_norm": 0.5264036059379578, + "learning_rate": 2.6262e-05, + "loss": 0.0552, + "step": 8758 + }, + { + "epoch": 7.397804054054054, + "grad_norm": 0.39066243171691895, + "learning_rate": 2.6265e-05, + "loss": 0.0432, + "step": 8759 + }, + { + "epoch": 7.398648648648648, + "grad_norm": 1.4441510438919067, + "learning_rate": 2.6268000000000003e-05, + "loss": 0.0412, + "step": 8760 + }, + { + "epoch": 7.399493243243243, + "grad_norm": 0.46537500619888306, + "learning_rate": 2.6271000000000003e-05, + "loss": 0.049, + "step": 8761 + }, + { + "epoch": 7.400337837837838, + "grad_norm": 0.4778682589530945, + "learning_rate": 2.6274000000000003e-05, + "loss": 0.0484, + "step": 8762 + }, + { + "epoch": 7.4011824324324325, + "grad_norm": 0.3107921779155731, + "learning_rate": 2.6277000000000003e-05, + "loss": 0.0236, + "step": 8763 + }, + { + "epoch": 7.402027027027027, + "grad_norm": 0.5085343718528748, + "learning_rate": 2.628e-05, + "loss": 0.0537, + "step": 8764 + }, + { + "epoch": 7.402871621621622, + "grad_norm": 0.3882332146167755, + "learning_rate": 2.6283e-05, + "loss": 0.0376, + "step": 8765 + }, + { + "epoch": 7.403716216216216, + "grad_norm": 0.4844527244567871, + "learning_rate": 2.6286e-05, + "loss": 0.0245, + "step": 8766 + }, + { + "epoch": 7.4045608108108105, + "grad_norm": 0.3329801559448242, + "learning_rate": 2.6289e-05, + "loss": 0.0254, + "step": 8767 + }, + { + "epoch": 7.405405405405405, + "grad_norm": 2.582054615020752, + "learning_rate": 2.6292e-05, + "loss": 0.0556, + "step": 8768 + }, + { + "epoch": 7.40625, + "grad_norm": 0.3527508080005646, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0414, + "step": 8769 + }, + { + "epoch": 7.407094594594595, + "grad_norm": 0.5605509877204895, + "learning_rate": 2.6298e-05, + "loss": 0.0464, + "step": 8770 + }, + { + "epoch": 7.4079391891891895, + "grad_norm": 0.4373931288719177, + "learning_rate": 2.6301e-05, + "loss": 0.0343, + "step": 8771 + }, + { + "epoch": 7.408783783783784, + "grad_norm": 0.41874393820762634, + "learning_rate": 2.6304e-05, + "loss": 0.0513, + "step": 8772 + }, + { + "epoch": 7.409628378378378, + "grad_norm": 0.44314807653427124, + "learning_rate": 2.6307e-05, + "loss": 0.0422, + "step": 8773 + }, + { + "epoch": 7.410472972972973, + "grad_norm": 0.4996875822544098, + "learning_rate": 2.631e-05, + "loss": 0.0424, + "step": 8774 + }, + { + "epoch": 7.4113175675675675, + "grad_norm": 0.4497877359390259, + "learning_rate": 2.6313e-05, + "loss": 0.0421, + "step": 8775 + }, + { + "epoch": 7.412162162162162, + "grad_norm": 0.46756553649902344, + "learning_rate": 2.6316e-05, + "loss": 0.0447, + "step": 8776 + }, + { + "epoch": 7.413006756756757, + "grad_norm": 0.7099120616912842, + "learning_rate": 2.6319e-05, + "loss": 0.044, + "step": 8777 + }, + { + "epoch": 7.413851351351352, + "grad_norm": 0.4374513626098633, + "learning_rate": 2.6322e-05, + "loss": 0.0391, + "step": 8778 + }, + { + "epoch": 7.414695945945946, + "grad_norm": 0.5358356237411499, + "learning_rate": 2.6325e-05, + "loss": 0.0279, + "step": 8779 + }, + { + "epoch": 7.41554054054054, + "grad_norm": 0.4491637647151947, + "learning_rate": 2.6328000000000003e-05, + "loss": 0.029, + "step": 8780 + }, + { + "epoch": 7.416385135135135, + "grad_norm": 0.48646122217178345, + "learning_rate": 2.6331000000000003e-05, + "loss": 0.0407, + "step": 8781 + }, + { + "epoch": 7.41722972972973, + "grad_norm": 0.3898051679134369, + "learning_rate": 2.6334000000000002e-05, + "loss": 0.031, + "step": 8782 + }, + { + "epoch": 7.418074324324325, + "grad_norm": 0.7804967164993286, + "learning_rate": 2.6337000000000002e-05, + "loss": 0.0979, + "step": 8783 + }, + { + "epoch": 7.418918918918919, + "grad_norm": 0.5024483799934387, + "learning_rate": 2.6340000000000002e-05, + "loss": 0.0503, + "step": 8784 + }, + { + "epoch": 7.419763513513513, + "grad_norm": 0.3976469337940216, + "learning_rate": 2.6343000000000002e-05, + "loss": 0.0294, + "step": 8785 + }, + { + "epoch": 7.420608108108108, + "grad_norm": 0.4546374976634979, + "learning_rate": 2.6346e-05, + "loss": 0.0283, + "step": 8786 + }, + { + "epoch": 7.421452702702703, + "grad_norm": 0.6171902418136597, + "learning_rate": 2.6349e-05, + "loss": 0.0724, + "step": 8787 + }, + { + "epoch": 7.422297297297297, + "grad_norm": 0.6429029703140259, + "learning_rate": 2.6351999999999998e-05, + "loss": 0.0611, + "step": 8788 + }, + { + "epoch": 7.423141891891892, + "grad_norm": 0.9194183349609375, + "learning_rate": 2.6354999999999998e-05, + "loss": 0.2875, + "step": 8789 + }, + { + "epoch": 7.423986486486487, + "grad_norm": 1.0913002490997314, + "learning_rate": 2.6358e-05, + "loss": 0.1944, + "step": 8790 + }, + { + "epoch": 7.424831081081081, + "grad_norm": 0.4272562861442566, + "learning_rate": 2.6361e-05, + "loss": 0.0895, + "step": 8791 + }, + { + "epoch": 7.425675675675675, + "grad_norm": 0.502201497554779, + "learning_rate": 2.6364e-05, + "loss": 0.0921, + "step": 8792 + }, + { + "epoch": 7.42652027027027, + "grad_norm": 0.5437290072441101, + "learning_rate": 2.6367e-05, + "loss": 0.0769, + "step": 8793 + }, + { + "epoch": 7.427364864864865, + "grad_norm": 0.4713108539581299, + "learning_rate": 2.637e-05, + "loss": 0.0594, + "step": 8794 + }, + { + "epoch": 7.42820945945946, + "grad_norm": 0.5809654593467712, + "learning_rate": 2.6373e-05, + "loss": 0.0683, + "step": 8795 + }, + { + "epoch": 7.429054054054054, + "grad_norm": 0.40155383944511414, + "learning_rate": 2.6376e-05, + "loss": 0.0405, + "step": 8796 + }, + { + "epoch": 7.429898648648648, + "grad_norm": 0.4168226718902588, + "learning_rate": 2.6379e-05, + "loss": 0.0437, + "step": 8797 + }, + { + "epoch": 7.430743243243243, + "grad_norm": 0.34464311599731445, + "learning_rate": 2.6382e-05, + "loss": 0.0413, + "step": 8798 + }, + { + "epoch": 7.431587837837838, + "grad_norm": 0.29039254784584045, + "learning_rate": 2.6385e-05, + "loss": 0.0331, + "step": 8799 + }, + { + "epoch": 7.4324324324324325, + "grad_norm": 0.5163001418113708, + "learning_rate": 2.6388000000000002e-05, + "loss": 0.0502, + "step": 8800 + }, + { + "epoch": 7.433277027027027, + "grad_norm": 0.4428311288356781, + "learning_rate": 2.6391000000000002e-05, + "loss": 0.032, + "step": 8801 + }, + { + "epoch": 7.434121621621622, + "grad_norm": 0.5023691654205322, + "learning_rate": 2.6394000000000002e-05, + "loss": 0.0291, + "step": 8802 + }, + { + "epoch": 7.434966216216216, + "grad_norm": 0.36873722076416016, + "learning_rate": 2.6397e-05, + "loss": 0.0352, + "step": 8803 + }, + { + "epoch": 7.4358108108108105, + "grad_norm": 0.41836532950401306, + "learning_rate": 2.64e-05, + "loss": 0.0428, + "step": 8804 + }, + { + "epoch": 7.436655405405405, + "grad_norm": 0.5371358394622803, + "learning_rate": 2.6403e-05, + "loss": 0.0275, + "step": 8805 + }, + { + "epoch": 7.4375, + "grad_norm": 0.4698958396911621, + "learning_rate": 2.6406e-05, + "loss": 0.0556, + "step": 8806 + }, + { + "epoch": 7.438344594594595, + "grad_norm": 0.23763900995254517, + "learning_rate": 2.6409e-05, + "loss": 0.0268, + "step": 8807 + }, + { + "epoch": 7.4391891891891895, + "grad_norm": 0.35290464758872986, + "learning_rate": 2.6412e-05, + "loss": 0.0409, + "step": 8808 + }, + { + "epoch": 7.440033783783784, + "grad_norm": 0.3738830089569092, + "learning_rate": 2.6415e-05, + "loss": 0.0303, + "step": 8809 + }, + { + "epoch": 7.440878378378378, + "grad_norm": 0.33074089884757996, + "learning_rate": 2.6418000000000004e-05, + "loss": 0.0232, + "step": 8810 + }, + { + "epoch": 7.441722972972973, + "grad_norm": 0.4120955169200897, + "learning_rate": 2.6421000000000003e-05, + "loss": 0.0391, + "step": 8811 + }, + { + "epoch": 7.4425675675675675, + "grad_norm": 0.28442925214767456, + "learning_rate": 2.6424000000000003e-05, + "loss": 0.0205, + "step": 8812 + }, + { + "epoch": 7.443412162162162, + "grad_norm": 0.5197255611419678, + "learning_rate": 2.6427e-05, + "loss": 0.0377, + "step": 8813 + }, + { + "epoch": 7.444256756756757, + "grad_norm": 0.450625479221344, + "learning_rate": 2.643e-05, + "loss": 0.0404, + "step": 8814 + }, + { + "epoch": 7.445101351351352, + "grad_norm": 0.30095744132995605, + "learning_rate": 2.6433e-05, + "loss": 0.0297, + "step": 8815 + }, + { + "epoch": 7.445945945945946, + "grad_norm": 0.40153074264526367, + "learning_rate": 2.6436e-05, + "loss": 0.0372, + "step": 8816 + }, + { + "epoch": 7.44679054054054, + "grad_norm": 0.3927532136440277, + "learning_rate": 2.6439e-05, + "loss": 0.0266, + "step": 8817 + }, + { + "epoch": 7.447635135135135, + "grad_norm": 0.5410299301147461, + "learning_rate": 2.6442e-05, + "loss": 0.0588, + "step": 8818 + }, + { + "epoch": 7.44847972972973, + "grad_norm": 0.3913387656211853, + "learning_rate": 2.6445e-05, + "loss": 0.0387, + "step": 8819 + }, + { + "epoch": 7.449324324324325, + "grad_norm": 0.3402498960494995, + "learning_rate": 2.6448e-05, + "loss": 0.0285, + "step": 8820 + }, + { + "epoch": 7.450168918918919, + "grad_norm": 0.615804135799408, + "learning_rate": 2.6451e-05, + "loss": 0.0457, + "step": 8821 + }, + { + "epoch": 7.451013513513513, + "grad_norm": 0.3641479015350342, + "learning_rate": 2.6454e-05, + "loss": 0.0295, + "step": 8822 + }, + { + "epoch": 7.451858108108108, + "grad_norm": 0.4178515374660492, + "learning_rate": 2.6457e-05, + "loss": 0.0406, + "step": 8823 + }, + { + "epoch": 7.452702702702703, + "grad_norm": 0.55552738904953, + "learning_rate": 2.646e-05, + "loss": 0.0291, + "step": 8824 + }, + { + "epoch": 7.453547297297297, + "grad_norm": 0.38132739067077637, + "learning_rate": 2.6463e-05, + "loss": 0.0447, + "step": 8825 + }, + { + "epoch": 7.454391891891892, + "grad_norm": 0.4959137439727783, + "learning_rate": 2.6466e-05, + "loss": 0.0212, + "step": 8826 + }, + { + "epoch": 7.455236486486487, + "grad_norm": 0.448146790266037, + "learning_rate": 2.6469e-05, + "loss": 0.0196, + "step": 8827 + }, + { + "epoch": 7.456081081081081, + "grad_norm": 0.4842474162578583, + "learning_rate": 2.6472e-05, + "loss": 0.037, + "step": 8828 + }, + { + "epoch": 7.456925675675675, + "grad_norm": 0.3997977077960968, + "learning_rate": 2.6475e-05, + "loss": 0.0335, + "step": 8829 + }, + { + "epoch": 7.45777027027027, + "grad_norm": 0.437545508146286, + "learning_rate": 2.6478000000000003e-05, + "loss": 0.0332, + "step": 8830 + }, + { + "epoch": 7.458614864864865, + "grad_norm": 0.538192868232727, + "learning_rate": 2.6481000000000003e-05, + "loss": 0.0539, + "step": 8831 + }, + { + "epoch": 7.45945945945946, + "grad_norm": 0.4090082347393036, + "learning_rate": 2.6484000000000003e-05, + "loss": 0.0515, + "step": 8832 + }, + { + "epoch": 7.460304054054054, + "grad_norm": 0.4001328945159912, + "learning_rate": 2.6487000000000002e-05, + "loss": 0.0226, + "step": 8833 + }, + { + "epoch": 7.461148648648648, + "grad_norm": 0.9204752445220947, + "learning_rate": 2.6490000000000002e-05, + "loss": 0.0524, + "step": 8834 + }, + { + "epoch": 7.461993243243243, + "grad_norm": 0.5519177317619324, + "learning_rate": 2.6493000000000002e-05, + "loss": 0.0458, + "step": 8835 + }, + { + "epoch": 7.462837837837838, + "grad_norm": 0.7122607827186584, + "learning_rate": 2.6496000000000002e-05, + "loss": 0.0582, + "step": 8836 + }, + { + "epoch": 7.4636824324324325, + "grad_norm": 0.8498919606208801, + "learning_rate": 2.6499e-05, + "loss": 0.034, + "step": 8837 + }, + { + "epoch": 7.464527027027027, + "grad_norm": 0.7981672883033752, + "learning_rate": 2.6501999999999998e-05, + "loss": 0.0565, + "step": 8838 + }, + { + "epoch": 7.465371621621622, + "grad_norm": 1.1826926469802856, + "learning_rate": 2.6504999999999998e-05, + "loss": 0.3367, + "step": 8839 + }, + { + "epoch": 7.466216216216216, + "grad_norm": 0.6238749623298645, + "learning_rate": 2.6508e-05, + "loss": 0.174, + "step": 8840 + }, + { + "epoch": 7.4670608108108105, + "grad_norm": 0.5745618939399719, + "learning_rate": 2.6511e-05, + "loss": 0.13, + "step": 8841 + }, + { + "epoch": 7.467905405405405, + "grad_norm": 0.44495517015457153, + "learning_rate": 2.6514e-05, + "loss": 0.066, + "step": 8842 + }, + { + "epoch": 7.46875, + "grad_norm": 0.45129379630088806, + "learning_rate": 2.6517e-05, + "loss": 0.085, + "step": 8843 + }, + { + "epoch": 7.469594594594595, + "grad_norm": 0.4698414206504822, + "learning_rate": 2.652e-05, + "loss": 0.0603, + "step": 8844 + }, + { + "epoch": 7.4704391891891895, + "grad_norm": 0.33422303199768066, + "learning_rate": 2.6523e-05, + "loss": 0.0546, + "step": 8845 + }, + { + "epoch": 7.471283783783784, + "grad_norm": 0.5347591042518616, + "learning_rate": 2.6526e-05, + "loss": 0.0422, + "step": 8846 + }, + { + "epoch": 7.472128378378378, + "grad_norm": 0.27836287021636963, + "learning_rate": 2.6529e-05, + "loss": 0.0421, + "step": 8847 + }, + { + "epoch": 7.472972972972973, + "grad_norm": 0.4677610695362091, + "learning_rate": 2.6532e-05, + "loss": 0.039, + "step": 8848 + }, + { + "epoch": 7.4738175675675675, + "grad_norm": 0.4293942451477051, + "learning_rate": 2.6535e-05, + "loss": 0.0417, + "step": 8849 + }, + { + "epoch": 7.474662162162162, + "grad_norm": 0.37498149275779724, + "learning_rate": 2.6538000000000002e-05, + "loss": 0.036, + "step": 8850 + }, + { + "epoch": 7.475506756756757, + "grad_norm": 0.29321423172950745, + "learning_rate": 2.6541000000000002e-05, + "loss": 0.0319, + "step": 8851 + }, + { + "epoch": 7.476351351351352, + "grad_norm": 0.3066723942756653, + "learning_rate": 2.6544000000000002e-05, + "loss": 0.0357, + "step": 8852 + }, + { + "epoch": 7.477195945945946, + "grad_norm": 0.30355963110923767, + "learning_rate": 2.6547000000000002e-05, + "loss": 0.0326, + "step": 8853 + }, + { + "epoch": 7.47804054054054, + "grad_norm": 0.44471824169158936, + "learning_rate": 2.655e-05, + "loss": 0.0394, + "step": 8854 + }, + { + "epoch": 7.478885135135135, + "grad_norm": 0.4389614462852478, + "learning_rate": 2.6553e-05, + "loss": 0.0416, + "step": 8855 + }, + { + "epoch": 7.47972972972973, + "grad_norm": 0.406446635723114, + "learning_rate": 2.6556e-05, + "loss": 0.039, + "step": 8856 + }, + { + "epoch": 7.480574324324325, + "grad_norm": 0.3182470500469208, + "learning_rate": 2.6559e-05, + "loss": 0.0411, + "step": 8857 + }, + { + "epoch": 7.481418918918919, + "grad_norm": 0.5249274373054504, + "learning_rate": 2.6562e-05, + "loss": 0.032, + "step": 8858 + }, + { + "epoch": 7.482263513513513, + "grad_norm": 0.53228360414505, + "learning_rate": 2.6565e-05, + "loss": 0.0538, + "step": 8859 + }, + { + "epoch": 7.483108108108108, + "grad_norm": 0.4150148332118988, + "learning_rate": 2.6568000000000004e-05, + "loss": 0.034, + "step": 8860 + }, + { + "epoch": 7.483952702702703, + "grad_norm": 0.6283365488052368, + "learning_rate": 2.6571000000000004e-05, + "loss": 0.0411, + "step": 8861 + }, + { + "epoch": 7.484797297297297, + "grad_norm": 0.35300198197364807, + "learning_rate": 2.6574e-05, + "loss": 0.0255, + "step": 8862 + }, + { + "epoch": 7.485641891891892, + "grad_norm": 0.32709088921546936, + "learning_rate": 2.6577e-05, + "loss": 0.024, + "step": 8863 + }, + { + "epoch": 7.486486486486487, + "grad_norm": 0.46066227555274963, + "learning_rate": 2.658e-05, + "loss": 0.0455, + "step": 8864 + }, + { + "epoch": 7.487331081081081, + "grad_norm": 0.3819233477115631, + "learning_rate": 2.6583e-05, + "loss": 0.0406, + "step": 8865 + }, + { + "epoch": 7.488175675675675, + "grad_norm": 0.39936357736587524, + "learning_rate": 2.6586e-05, + "loss": 0.0345, + "step": 8866 + }, + { + "epoch": 7.48902027027027, + "grad_norm": 0.48296141624450684, + "learning_rate": 2.6589e-05, + "loss": 0.0234, + "step": 8867 + }, + { + "epoch": 7.489864864864865, + "grad_norm": 0.40915700793266296, + "learning_rate": 2.6592e-05, + "loss": 0.0349, + "step": 8868 + }, + { + "epoch": 7.49070945945946, + "grad_norm": 0.37172117829322815, + "learning_rate": 2.6595e-05, + "loss": 0.0311, + "step": 8869 + }, + { + "epoch": 7.491554054054054, + "grad_norm": 0.37738940119743347, + "learning_rate": 2.6598000000000002e-05, + "loss": 0.0347, + "step": 8870 + }, + { + "epoch": 7.492398648648648, + "grad_norm": 0.4592635929584503, + "learning_rate": 2.6601e-05, + "loss": 0.0255, + "step": 8871 + }, + { + "epoch": 7.493243243243243, + "grad_norm": 0.4334005117416382, + "learning_rate": 2.6604e-05, + "loss": 0.0205, + "step": 8872 + }, + { + "epoch": 7.494087837837838, + "grad_norm": 0.4990079700946808, + "learning_rate": 2.6607e-05, + "loss": 0.0355, + "step": 8873 + }, + { + "epoch": 7.4949324324324325, + "grad_norm": 0.5678676962852478, + "learning_rate": 2.661e-05, + "loss": 0.0635, + "step": 8874 + }, + { + "epoch": 7.495777027027027, + "grad_norm": 0.5755007863044739, + "learning_rate": 2.6613e-05, + "loss": 0.036, + "step": 8875 + }, + { + "epoch": 7.496621621621622, + "grad_norm": 0.5459357500076294, + "learning_rate": 2.6616e-05, + "loss": 0.0383, + "step": 8876 + }, + { + "epoch": 7.497466216216216, + "grad_norm": 0.4848540127277374, + "learning_rate": 2.6619e-05, + "loss": 0.0421, + "step": 8877 + }, + { + "epoch": 7.4983108108108105, + "grad_norm": 0.29438677430152893, + "learning_rate": 2.6622e-05, + "loss": 0.0292, + "step": 8878 + }, + { + "epoch": 7.499155405405405, + "grad_norm": 1.0072567462921143, + "learning_rate": 2.6625e-05, + "loss": 0.0555, + "step": 8879 + }, + { + "epoch": 7.5, + "grad_norm": 0.600717306137085, + "learning_rate": 2.6628e-05, + "loss": 0.0245, + "step": 8880 + }, + { + "epoch": 7.500844594594595, + "grad_norm": 0.4008028209209442, + "learning_rate": 2.6631000000000003e-05, + "loss": 0.0301, + "step": 8881 + }, + { + "epoch": 7.5016891891891895, + "grad_norm": 0.720203161239624, + "learning_rate": 2.6634000000000003e-05, + "loss": 0.0525, + "step": 8882 + }, + { + "epoch": 7.502533783783784, + "grad_norm": 1.3288378715515137, + "learning_rate": 2.6637000000000003e-05, + "loss": 0.1209, + "step": 8883 + }, + { + "epoch": 7.503378378378378, + "grad_norm": 0.6550973653793335, + "learning_rate": 2.6640000000000002e-05, + "loss": 0.0613, + "step": 8884 + }, + { + "epoch": 7.504222972972973, + "grad_norm": 0.4552750587463379, + "learning_rate": 2.6643000000000002e-05, + "loss": 0.0408, + "step": 8885 + }, + { + "epoch": 7.5050675675675675, + "grad_norm": 0.8188155889511108, + "learning_rate": 2.6646000000000002e-05, + "loss": 0.0575, + "step": 8886 + }, + { + "epoch": 7.505912162162162, + "grad_norm": 1.3791507482528687, + "learning_rate": 2.6649e-05, + "loss": 0.0964, + "step": 8887 + }, + { + "epoch": 7.506756756756757, + "grad_norm": 0.900925874710083, + "learning_rate": 2.6651999999999998e-05, + "loss": 0.0821, + "step": 8888 + }, + { + "epoch": 7.507601351351351, + "grad_norm": 1.7663081884384155, + "learning_rate": 2.6654999999999998e-05, + "loss": 0.3356, + "step": 8889 + }, + { + "epoch": 7.508445945945946, + "grad_norm": 0.9886401295661926, + "learning_rate": 2.6657999999999998e-05, + "loss": 0.1715, + "step": 8890 + }, + { + "epoch": 7.50929054054054, + "grad_norm": 0.47644174098968506, + "learning_rate": 2.6661e-05, + "loss": 0.1066, + "step": 8891 + }, + { + "epoch": 7.510135135135135, + "grad_norm": 0.5864638686180115, + "learning_rate": 2.6664e-05, + "loss": 0.0741, + "step": 8892 + }, + { + "epoch": 7.51097972972973, + "grad_norm": 0.49923211336135864, + "learning_rate": 2.6667e-05, + "loss": 0.0807, + "step": 8893 + }, + { + "epoch": 7.511824324324325, + "grad_norm": 0.5548279285430908, + "learning_rate": 2.667e-05, + "loss": 0.0589, + "step": 8894 + }, + { + "epoch": 7.512668918918919, + "grad_norm": 0.4097191393375397, + "learning_rate": 2.6673e-05, + "loss": 0.045, + "step": 8895 + }, + { + "epoch": 7.513513513513513, + "grad_norm": 0.35261502861976624, + "learning_rate": 2.6676e-05, + "loss": 0.0447, + "step": 8896 + }, + { + "epoch": 7.514358108108108, + "grad_norm": 0.3522523045539856, + "learning_rate": 2.6679e-05, + "loss": 0.0258, + "step": 8897 + }, + { + "epoch": 7.515202702702703, + "grad_norm": 0.3563631474971771, + "learning_rate": 2.6682e-05, + "loss": 0.0692, + "step": 8898 + }, + { + "epoch": 7.516047297297297, + "grad_norm": 0.4146444499492645, + "learning_rate": 2.6685e-05, + "loss": 0.0338, + "step": 8899 + }, + { + "epoch": 7.516891891891892, + "grad_norm": 0.35256725549697876, + "learning_rate": 2.6688e-05, + "loss": 0.0314, + "step": 8900 + }, + { + "epoch": 7.517736486486487, + "grad_norm": 0.484353244304657, + "learning_rate": 2.6691000000000002e-05, + "loss": 0.0356, + "step": 8901 + }, + { + "epoch": 7.518581081081081, + "grad_norm": 0.4202396273612976, + "learning_rate": 2.6694000000000002e-05, + "loss": 0.05, + "step": 8902 + }, + { + "epoch": 7.519425675675675, + "grad_norm": 0.37060433626174927, + "learning_rate": 2.6697000000000002e-05, + "loss": 0.0528, + "step": 8903 + }, + { + "epoch": 7.52027027027027, + "grad_norm": 0.3330853283405304, + "learning_rate": 2.6700000000000002e-05, + "loss": 0.0234, + "step": 8904 + }, + { + "epoch": 7.521114864864865, + "grad_norm": 0.3656148314476013, + "learning_rate": 2.6703e-05, + "loss": 0.0326, + "step": 8905 + }, + { + "epoch": 7.52195945945946, + "grad_norm": 0.48246482014656067, + "learning_rate": 2.6706e-05, + "loss": 0.0377, + "step": 8906 + }, + { + "epoch": 7.522804054054054, + "grad_norm": 0.25169071555137634, + "learning_rate": 2.6709e-05, + "loss": 0.0189, + "step": 8907 + }, + { + "epoch": 7.523648648648649, + "grad_norm": 0.33771494030952454, + "learning_rate": 2.6712e-05, + "loss": 0.0276, + "step": 8908 + }, + { + "epoch": 7.524493243243243, + "grad_norm": 0.45474129915237427, + "learning_rate": 2.6715e-05, + "loss": 0.0513, + "step": 8909 + }, + { + "epoch": 7.525337837837838, + "grad_norm": 0.5900114178657532, + "learning_rate": 2.6718e-05, + "loss": 0.0277, + "step": 8910 + }, + { + "epoch": 7.5261824324324325, + "grad_norm": 0.4738331735134125, + "learning_rate": 2.6721e-05, + "loss": 0.0362, + "step": 8911 + }, + { + "epoch": 7.527027027027027, + "grad_norm": 0.47516128420829773, + "learning_rate": 2.6724e-05, + "loss": 0.0341, + "step": 8912 + }, + { + "epoch": 7.527871621621622, + "grad_norm": 0.35215553641319275, + "learning_rate": 2.6727e-05, + "loss": 0.0359, + "step": 8913 + }, + { + "epoch": 7.528716216216216, + "grad_norm": 0.8759762048721313, + "learning_rate": 2.673e-05, + "loss": 0.0364, + "step": 8914 + }, + { + "epoch": 7.5295608108108105, + "grad_norm": 0.4805624186992645, + "learning_rate": 2.6733e-05, + "loss": 0.029, + "step": 8915 + }, + { + "epoch": 7.530405405405405, + "grad_norm": 0.3929317593574524, + "learning_rate": 2.6736e-05, + "loss": 0.0343, + "step": 8916 + }, + { + "epoch": 7.53125, + "grad_norm": 0.5097458362579346, + "learning_rate": 2.6739e-05, + "loss": 0.0336, + "step": 8917 + }, + { + "epoch": 7.532094594594595, + "grad_norm": 1.0942462682724, + "learning_rate": 2.6742e-05, + "loss": 0.0362, + "step": 8918 + }, + { + "epoch": 7.5329391891891895, + "grad_norm": 0.4275311231613159, + "learning_rate": 2.6745e-05, + "loss": 0.0384, + "step": 8919 + }, + { + "epoch": 7.533783783783784, + "grad_norm": 0.28879648447036743, + "learning_rate": 2.6748e-05, + "loss": 0.035, + "step": 8920 + }, + { + "epoch": 7.534628378378378, + "grad_norm": 0.5971543192863464, + "learning_rate": 2.6751000000000002e-05, + "loss": 0.0642, + "step": 8921 + }, + { + "epoch": 7.535472972972973, + "grad_norm": 0.7628942728042603, + "learning_rate": 2.6754e-05, + "loss": 0.0381, + "step": 8922 + }, + { + "epoch": 7.5363175675675675, + "grad_norm": 0.40288206934928894, + "learning_rate": 2.6757e-05, + "loss": 0.0456, + "step": 8923 + }, + { + "epoch": 7.537162162162162, + "grad_norm": 0.31114184856414795, + "learning_rate": 2.676e-05, + "loss": 0.0328, + "step": 8924 + }, + { + "epoch": 7.538006756756757, + "grad_norm": 0.5876600742340088, + "learning_rate": 2.6763e-05, + "loss": 0.0563, + "step": 8925 + }, + { + "epoch": 7.538851351351351, + "grad_norm": 0.659055233001709, + "learning_rate": 2.6766e-05, + "loss": 0.0327, + "step": 8926 + }, + { + "epoch": 7.539695945945946, + "grad_norm": 1.349340558052063, + "learning_rate": 2.6769e-05, + "loss": 0.0539, + "step": 8927 + }, + { + "epoch": 7.54054054054054, + "grad_norm": 0.5297999978065491, + "learning_rate": 2.6772e-05, + "loss": 0.0412, + "step": 8928 + }, + { + "epoch": 7.541385135135135, + "grad_norm": 0.3495180904865265, + "learning_rate": 2.6775e-05, + "loss": 0.0288, + "step": 8929 + }, + { + "epoch": 7.54222972972973, + "grad_norm": 0.439876914024353, + "learning_rate": 2.6778e-05, + "loss": 0.0449, + "step": 8930 + }, + { + "epoch": 7.543074324324325, + "grad_norm": 0.498200923204422, + "learning_rate": 2.6781000000000003e-05, + "loss": 0.0517, + "step": 8931 + }, + { + "epoch": 7.543918918918919, + "grad_norm": 0.43729639053344727, + "learning_rate": 2.6784000000000003e-05, + "loss": 0.0468, + "step": 8932 + }, + { + "epoch": 7.544763513513513, + "grad_norm": 0.704197883605957, + "learning_rate": 2.6787000000000003e-05, + "loss": 0.0483, + "step": 8933 + }, + { + "epoch": 7.545608108108108, + "grad_norm": 0.47199568152427673, + "learning_rate": 2.6790000000000003e-05, + "loss": 0.0421, + "step": 8934 + }, + { + "epoch": 7.546452702702703, + "grad_norm": 0.29080599546432495, + "learning_rate": 2.6793000000000002e-05, + "loss": 0.0267, + "step": 8935 + }, + { + "epoch": 7.547297297297297, + "grad_norm": 0.39861273765563965, + "learning_rate": 2.6796e-05, + "loss": 0.0418, + "step": 8936 + }, + { + "epoch": 7.548141891891892, + "grad_norm": 0.9965016841888428, + "learning_rate": 2.6799e-05, + "loss": 0.0796, + "step": 8937 + }, + { + "epoch": 7.548986486486487, + "grad_norm": 1.242696762084961, + "learning_rate": 2.6802e-05, + "loss": 0.1416, + "step": 8938 + }, + { + "epoch": 7.549831081081081, + "grad_norm": 2.374899387359619, + "learning_rate": 2.6805e-05, + "loss": 0.4431, + "step": 8939 + }, + { + "epoch": 7.550675675675675, + "grad_norm": 0.8875529170036316, + "learning_rate": 2.6807999999999998e-05, + "loss": 0.1819, + "step": 8940 + }, + { + "epoch": 7.55152027027027, + "grad_norm": 0.5657346248626709, + "learning_rate": 2.6811e-05, + "loss": 0.1313, + "step": 8941 + }, + { + "epoch": 7.552364864864865, + "grad_norm": 0.5455006957054138, + "learning_rate": 2.6814e-05, + "loss": 0.0917, + "step": 8942 + }, + { + "epoch": 7.55320945945946, + "grad_norm": 0.3903238773345947, + "learning_rate": 2.6817e-05, + "loss": 0.0465, + "step": 8943 + }, + { + "epoch": 7.554054054054054, + "grad_norm": 0.4984035789966583, + "learning_rate": 2.682e-05, + "loss": 0.041, + "step": 8944 + }, + { + "epoch": 7.554898648648649, + "grad_norm": 0.38938310742378235, + "learning_rate": 2.6823e-05, + "loss": 0.0539, + "step": 8945 + }, + { + "epoch": 7.555743243243243, + "grad_norm": 0.49352648854255676, + "learning_rate": 2.6826e-05, + "loss": 0.0513, + "step": 8946 + }, + { + "epoch": 7.556587837837838, + "grad_norm": 0.5751579403877258, + "learning_rate": 2.6829e-05, + "loss": 0.0624, + "step": 8947 + }, + { + "epoch": 7.5574324324324325, + "grad_norm": 0.49126574397087097, + "learning_rate": 2.6832e-05, + "loss": 0.0522, + "step": 8948 + }, + { + "epoch": 7.558277027027027, + "grad_norm": 0.36316585540771484, + "learning_rate": 2.6835e-05, + "loss": 0.0344, + "step": 8949 + }, + { + "epoch": 7.559121621621622, + "grad_norm": 0.30320557951927185, + "learning_rate": 2.6838e-05, + "loss": 0.0298, + "step": 8950 + }, + { + "epoch": 7.559966216216216, + "grad_norm": 0.43252938985824585, + "learning_rate": 2.6841000000000003e-05, + "loss": 0.0448, + "step": 8951 + }, + { + "epoch": 7.5608108108108105, + "grad_norm": 0.43090760707855225, + "learning_rate": 2.6844000000000003e-05, + "loss": 0.0384, + "step": 8952 + }, + { + "epoch": 7.561655405405405, + "grad_norm": 0.3008948266506195, + "learning_rate": 2.6847000000000002e-05, + "loss": 0.0239, + "step": 8953 + }, + { + "epoch": 7.5625, + "grad_norm": 0.5557988882064819, + "learning_rate": 2.6850000000000002e-05, + "loss": 0.0495, + "step": 8954 + }, + { + "epoch": 7.563344594594595, + "grad_norm": 0.3589186370372772, + "learning_rate": 2.6853000000000002e-05, + "loss": 0.0451, + "step": 8955 + }, + { + "epoch": 7.5641891891891895, + "grad_norm": 0.2696214020252228, + "learning_rate": 2.6856000000000002e-05, + "loss": 0.0267, + "step": 8956 + }, + { + "epoch": 7.565033783783784, + "grad_norm": 0.44103267788887024, + "learning_rate": 2.6859e-05, + "loss": 0.0241, + "step": 8957 + }, + { + "epoch": 7.565878378378378, + "grad_norm": 0.7248183488845825, + "learning_rate": 2.6862e-05, + "loss": 0.0332, + "step": 8958 + }, + { + "epoch": 7.566722972972973, + "grad_norm": 0.461327463388443, + "learning_rate": 2.6865e-05, + "loss": 0.0377, + "step": 8959 + }, + { + "epoch": 7.5675675675675675, + "grad_norm": 0.4423501491546631, + "learning_rate": 2.6867999999999998e-05, + "loss": 0.0415, + "step": 8960 + }, + { + "epoch": 7.568412162162162, + "grad_norm": 0.27797940373420715, + "learning_rate": 2.6871e-05, + "loss": 0.0314, + "step": 8961 + }, + { + "epoch": 7.569256756756757, + "grad_norm": 0.42894309759140015, + "learning_rate": 2.6874e-05, + "loss": 0.0252, + "step": 8962 + }, + { + "epoch": 7.570101351351351, + "grad_norm": 0.39743882417678833, + "learning_rate": 2.6877e-05, + "loss": 0.048, + "step": 8963 + }, + { + "epoch": 7.570945945945946, + "grad_norm": 0.43659335374832153, + "learning_rate": 2.688e-05, + "loss": 0.0301, + "step": 8964 + }, + { + "epoch": 7.57179054054054, + "grad_norm": 0.5097795724868774, + "learning_rate": 2.6883e-05, + "loss": 0.0506, + "step": 8965 + }, + { + "epoch": 7.572635135135135, + "grad_norm": 0.4739038944244385, + "learning_rate": 2.6886e-05, + "loss": 0.0363, + "step": 8966 + }, + { + "epoch": 7.57347972972973, + "grad_norm": 0.4363148808479309, + "learning_rate": 2.6889e-05, + "loss": 0.0216, + "step": 8967 + }, + { + "epoch": 7.574324324324325, + "grad_norm": 0.3528127670288086, + "learning_rate": 2.6892e-05, + "loss": 0.0284, + "step": 8968 + }, + { + "epoch": 7.575168918918919, + "grad_norm": 0.38927701115608215, + "learning_rate": 2.6895e-05, + "loss": 0.0286, + "step": 8969 + }, + { + "epoch": 7.576013513513513, + "grad_norm": 0.36193251609802246, + "learning_rate": 2.6898e-05, + "loss": 0.0342, + "step": 8970 + }, + { + "epoch": 7.576858108108108, + "grad_norm": 0.38044941425323486, + "learning_rate": 2.6901000000000002e-05, + "loss": 0.0318, + "step": 8971 + }, + { + "epoch": 7.577702702702703, + "grad_norm": 0.49363353848457336, + "learning_rate": 2.6904000000000002e-05, + "loss": 0.0295, + "step": 8972 + }, + { + "epoch": 7.578547297297297, + "grad_norm": 0.34875646233558655, + "learning_rate": 2.6907000000000002e-05, + "loss": 0.0337, + "step": 8973 + }, + { + "epoch": 7.579391891891892, + "grad_norm": 0.33369505405426025, + "learning_rate": 2.691e-05, + "loss": 0.0198, + "step": 8974 + }, + { + "epoch": 7.580236486486487, + "grad_norm": 0.5543609261512756, + "learning_rate": 2.6913e-05, + "loss": 0.0443, + "step": 8975 + }, + { + "epoch": 7.581081081081081, + "grad_norm": 1.1993385553359985, + "learning_rate": 2.6916e-05, + "loss": 0.066, + "step": 8976 + }, + { + "epoch": 7.581925675675675, + "grad_norm": 0.4185197651386261, + "learning_rate": 2.6919e-05, + "loss": 0.0352, + "step": 8977 + }, + { + "epoch": 7.58277027027027, + "grad_norm": 0.7511467337608337, + "learning_rate": 2.6922e-05, + "loss": 0.0403, + "step": 8978 + }, + { + "epoch": 7.583614864864865, + "grad_norm": 0.3987257182598114, + "learning_rate": 2.6925e-05, + "loss": 0.0256, + "step": 8979 + }, + { + "epoch": 7.58445945945946, + "grad_norm": 0.5832106471061707, + "learning_rate": 2.6928e-05, + "loss": 0.0385, + "step": 8980 + }, + { + "epoch": 7.585304054054054, + "grad_norm": 3.1346325874328613, + "learning_rate": 2.6931000000000004e-05, + "loss": 0.038, + "step": 8981 + }, + { + "epoch": 7.586148648648649, + "grad_norm": 0.8709176182746887, + "learning_rate": 2.6934000000000003e-05, + "loss": 0.0516, + "step": 8982 + }, + { + "epoch": 7.586993243243243, + "grad_norm": 0.5293030738830566, + "learning_rate": 2.6937000000000003e-05, + "loss": 0.0443, + "step": 8983 + }, + { + "epoch": 7.587837837837838, + "grad_norm": 0.6053323149681091, + "learning_rate": 2.6940000000000003e-05, + "loss": 0.0431, + "step": 8984 + }, + { + "epoch": 7.5886824324324325, + "grad_norm": 0.7405412197113037, + "learning_rate": 2.6943e-05, + "loss": 0.0479, + "step": 8985 + }, + { + "epoch": 7.589527027027027, + "grad_norm": 0.659496009349823, + "learning_rate": 2.6946e-05, + "loss": 0.05, + "step": 8986 + }, + { + "epoch": 7.590371621621622, + "grad_norm": 0.6512092351913452, + "learning_rate": 2.6949e-05, + "loss": 0.0682, + "step": 8987 + }, + { + "epoch": 7.591216216216216, + "grad_norm": 0.7198317050933838, + "learning_rate": 2.6952e-05, + "loss": 0.0806, + "step": 8988 + }, + { + "epoch": 7.5920608108108105, + "grad_norm": 2.3249902725219727, + "learning_rate": 2.6955e-05, + "loss": 0.3239, + "step": 8989 + }, + { + "epoch": 7.592905405405405, + "grad_norm": 0.7580312490463257, + "learning_rate": 2.6958e-05, + "loss": 0.2428, + "step": 8990 + }, + { + "epoch": 7.59375, + "grad_norm": 0.5371579527854919, + "learning_rate": 2.6961e-05, + "loss": 0.087, + "step": 8991 + }, + { + "epoch": 7.594594594594595, + "grad_norm": 0.3452083468437195, + "learning_rate": 2.6964e-05, + "loss": 0.0415, + "step": 8992 + }, + { + "epoch": 7.5954391891891895, + "grad_norm": 0.2312890887260437, + "learning_rate": 2.6967e-05, + "loss": 0.0324, + "step": 8993 + }, + { + "epoch": 7.596283783783784, + "grad_norm": 0.3597121238708496, + "learning_rate": 2.697e-05, + "loss": 0.0511, + "step": 8994 + }, + { + "epoch": 7.597128378378378, + "grad_norm": 0.51451176404953, + "learning_rate": 2.6973e-05, + "loss": 0.045, + "step": 8995 + }, + { + "epoch": 7.597972972972973, + "grad_norm": 0.31068161129951477, + "learning_rate": 2.6976e-05, + "loss": 0.0301, + "step": 8996 + }, + { + "epoch": 7.5988175675675675, + "grad_norm": 0.3836057186126709, + "learning_rate": 2.6979e-05, + "loss": 0.0278, + "step": 8997 + }, + { + "epoch": 7.599662162162162, + "grad_norm": 0.7379639148712158, + "learning_rate": 2.6982e-05, + "loss": 0.0354, + "step": 8998 + }, + { + "epoch": 7.600506756756757, + "grad_norm": 0.2938230037689209, + "learning_rate": 2.6985e-05, + "loss": 0.0469, + "step": 8999 + }, + { + "epoch": 7.601351351351351, + "grad_norm": 0.6291072368621826, + "learning_rate": 2.6988e-05, + "loss": 0.0516, + "step": 9000 + }, + { + "epoch": 7.601351351351351, + "eval_cer": 0.10055417235563409, + "eval_loss": 0.3242468237876892, + "eval_runtime": 8.3056, + "eval_samples_per_second": 12.161, + "eval_steps_per_second": 0.482, + "eval_wer": 0.2932098765432099, + "step": 9000 + }, + { + "epoch": 7.602195945945946, + "grad_norm": 0.47946393489837646, + "learning_rate": 2.6991000000000003e-05, + "loss": 0.0402, + "step": 9001 + }, + { + "epoch": 7.60304054054054, + "grad_norm": 0.3186628222465515, + "learning_rate": 2.6994000000000003e-05, + "loss": 0.0406, + "step": 9002 + }, + { + "epoch": 7.603885135135135, + "grad_norm": 0.3127364218235016, + "learning_rate": 2.6997000000000003e-05, + "loss": 0.0243, + "step": 9003 + }, + { + "epoch": 7.60472972972973, + "grad_norm": 0.5527377128601074, + "learning_rate": 2.7000000000000002e-05, + "loss": 0.0444, + "step": 9004 + }, + { + "epoch": 7.605574324324325, + "grad_norm": 0.3395150899887085, + "learning_rate": 2.7003000000000002e-05, + "loss": 0.0593, + "step": 9005 + }, + { + "epoch": 7.606418918918919, + "grad_norm": 0.3332461416721344, + "learning_rate": 2.7006000000000002e-05, + "loss": 0.0396, + "step": 9006 + }, + { + "epoch": 7.607263513513513, + "grad_norm": 0.3255160450935364, + "learning_rate": 2.7009000000000002e-05, + "loss": 0.025, + "step": 9007 + }, + { + "epoch": 7.608108108108108, + "grad_norm": 0.6578475832939148, + "learning_rate": 2.7012e-05, + "loss": 0.0415, + "step": 9008 + }, + { + "epoch": 7.608952702702703, + "grad_norm": 0.4706801474094391, + "learning_rate": 2.7015e-05, + "loss": 0.1211, + "step": 9009 + }, + { + "epoch": 7.609797297297297, + "grad_norm": 0.3796078562736511, + "learning_rate": 2.7017999999999998e-05, + "loss": 0.0455, + "step": 9010 + }, + { + "epoch": 7.610641891891892, + "grad_norm": 0.3995674252510071, + "learning_rate": 2.7020999999999998e-05, + "loss": 0.0395, + "step": 9011 + }, + { + "epoch": 7.611486486486487, + "grad_norm": 0.4852961301803589, + "learning_rate": 2.7024e-05, + "loss": 0.0391, + "step": 9012 + }, + { + "epoch": 7.612331081081081, + "grad_norm": 0.4626532196998596, + "learning_rate": 2.7027e-05, + "loss": 0.0439, + "step": 9013 + }, + { + "epoch": 7.613175675675675, + "grad_norm": 0.5158022046089172, + "learning_rate": 2.703e-05, + "loss": 0.043, + "step": 9014 + }, + { + "epoch": 7.61402027027027, + "grad_norm": 0.45480796694755554, + "learning_rate": 2.7033e-05, + "loss": 0.0521, + "step": 9015 + }, + { + "epoch": 7.614864864864865, + "grad_norm": 0.4163455367088318, + "learning_rate": 2.7036e-05, + "loss": 0.0497, + "step": 9016 + }, + { + "epoch": 7.61570945945946, + "grad_norm": 0.5744029879570007, + "learning_rate": 2.7039e-05, + "loss": 0.0363, + "step": 9017 + }, + { + "epoch": 7.616554054054054, + "grad_norm": 0.6152617931365967, + "learning_rate": 2.7042e-05, + "loss": 0.0434, + "step": 9018 + }, + { + "epoch": 7.617398648648649, + "grad_norm": 0.42694494128227234, + "learning_rate": 2.7045e-05, + "loss": 0.0413, + "step": 9019 + }, + { + "epoch": 7.618243243243243, + "grad_norm": 0.5256922245025635, + "learning_rate": 2.7048e-05, + "loss": 0.0456, + "step": 9020 + }, + { + "epoch": 7.619087837837838, + "grad_norm": 0.3528364598751068, + "learning_rate": 2.7051e-05, + "loss": 0.0336, + "step": 9021 + }, + { + "epoch": 7.6199324324324325, + "grad_norm": 0.5336401462554932, + "learning_rate": 2.7054000000000002e-05, + "loss": 0.031, + "step": 9022 + }, + { + "epoch": 7.620777027027027, + "grad_norm": 0.2757668197154999, + "learning_rate": 2.7057000000000002e-05, + "loss": 0.0233, + "step": 9023 + }, + { + "epoch": 7.621621621621622, + "grad_norm": 0.5681132674217224, + "learning_rate": 2.7060000000000002e-05, + "loss": 0.0461, + "step": 9024 + }, + { + "epoch": 7.622466216216216, + "grad_norm": 0.4155959188938141, + "learning_rate": 2.7063e-05, + "loss": 0.0478, + "step": 9025 + }, + { + "epoch": 7.6233108108108105, + "grad_norm": 0.5239137411117554, + "learning_rate": 2.7066e-05, + "loss": 0.0407, + "step": 9026 + }, + { + "epoch": 7.624155405405405, + "grad_norm": 0.4852501153945923, + "learning_rate": 2.7069e-05, + "loss": 0.032, + "step": 9027 + }, + { + "epoch": 7.625, + "grad_norm": 0.40730300545692444, + "learning_rate": 2.7072e-05, + "loss": 0.032, + "step": 9028 + }, + { + "epoch": 7.625844594594595, + "grad_norm": 0.4537937641143799, + "learning_rate": 2.7075e-05, + "loss": 0.0321, + "step": 9029 + }, + { + "epoch": 7.6266891891891895, + "grad_norm": 0.6548889875411987, + "learning_rate": 2.7078e-05, + "loss": 0.0325, + "step": 9030 + }, + { + "epoch": 7.627533783783784, + "grad_norm": 0.7048069834709167, + "learning_rate": 2.7081e-05, + "loss": 0.0649, + "step": 9031 + }, + { + "epoch": 7.628378378378378, + "grad_norm": 0.5572300553321838, + "learning_rate": 2.7084000000000004e-05, + "loss": 0.0518, + "step": 9032 + }, + { + "epoch": 7.629222972972973, + "grad_norm": 0.43249863386154175, + "learning_rate": 2.7087000000000003e-05, + "loss": 0.0434, + "step": 9033 + }, + { + "epoch": 7.6300675675675675, + "grad_norm": 0.768189013004303, + "learning_rate": 2.709e-05, + "loss": 0.0504, + "step": 9034 + }, + { + "epoch": 7.630912162162162, + "grad_norm": 0.8045995235443115, + "learning_rate": 2.7093e-05, + "loss": 0.0629, + "step": 9035 + }, + { + "epoch": 7.631756756756757, + "grad_norm": 0.4929402768611908, + "learning_rate": 2.7096e-05, + "loss": 0.0497, + "step": 9036 + }, + { + "epoch": 7.632601351351351, + "grad_norm": 0.6576979160308838, + "learning_rate": 2.7099e-05, + "loss": 0.0626, + "step": 9037 + }, + { + "epoch": 7.633445945945946, + "grad_norm": 0.7931728959083557, + "learning_rate": 2.7102e-05, + "loss": 0.0543, + "step": 9038 + }, + { + "epoch": 7.63429054054054, + "grad_norm": 1.5198975801467896, + "learning_rate": 2.7105e-05, + "loss": 0.3067, + "step": 9039 + }, + { + "epoch": 7.635135135135135, + "grad_norm": 0.9729706048965454, + "learning_rate": 2.7108e-05, + "loss": 0.1931, + "step": 9040 + }, + { + "epoch": 7.63597972972973, + "grad_norm": 0.4591344892978668, + "learning_rate": 2.7111e-05, + "loss": 0.0587, + "step": 9041 + }, + { + "epoch": 7.636824324324325, + "grad_norm": 0.4838632643222809, + "learning_rate": 2.7114e-05, + "loss": 0.0637, + "step": 9042 + }, + { + "epoch": 7.637668918918919, + "grad_norm": 0.46969038248062134, + "learning_rate": 2.7117e-05, + "loss": 0.048, + "step": 9043 + }, + { + "epoch": 7.638513513513513, + "grad_norm": 0.516925036907196, + "learning_rate": 2.712e-05, + "loss": 0.0499, + "step": 9044 + }, + { + "epoch": 7.639358108108108, + "grad_norm": 0.7645596265792847, + "learning_rate": 2.7123e-05, + "loss": 0.0458, + "step": 9045 + }, + { + "epoch": 7.640202702702703, + "grad_norm": 0.36742544174194336, + "learning_rate": 2.7126e-05, + "loss": 0.0403, + "step": 9046 + }, + { + "epoch": 7.641047297297297, + "grad_norm": 0.35303449630737305, + "learning_rate": 2.7129e-05, + "loss": 0.0442, + "step": 9047 + }, + { + "epoch": 7.641891891891892, + "grad_norm": 0.46987342834472656, + "learning_rate": 2.7132e-05, + "loss": 0.0484, + "step": 9048 + }, + { + "epoch": 7.642736486486487, + "grad_norm": 0.31629082560539246, + "learning_rate": 2.7135e-05, + "loss": 0.0263, + "step": 9049 + }, + { + "epoch": 7.643581081081081, + "grad_norm": 0.4751451313495636, + "learning_rate": 2.7138e-05, + "loss": 0.0397, + "step": 9050 + }, + { + "epoch": 7.644425675675675, + "grad_norm": 0.4552227258682251, + "learning_rate": 2.7141e-05, + "loss": 0.0356, + "step": 9051 + }, + { + "epoch": 7.64527027027027, + "grad_norm": 0.3743666112422943, + "learning_rate": 2.7144000000000003e-05, + "loss": 0.0466, + "step": 9052 + }, + { + "epoch": 7.646114864864865, + "grad_norm": 0.5503769516944885, + "learning_rate": 2.7147000000000003e-05, + "loss": 0.097, + "step": 9053 + }, + { + "epoch": 7.64695945945946, + "grad_norm": 0.43354156613349915, + "learning_rate": 2.7150000000000003e-05, + "loss": 0.0409, + "step": 9054 + }, + { + "epoch": 7.647804054054054, + "grad_norm": 0.36918243765830994, + "learning_rate": 2.7153000000000002e-05, + "loss": 0.039, + "step": 9055 + }, + { + "epoch": 7.648648648648649, + "grad_norm": 0.42583325505256653, + "learning_rate": 2.7156000000000002e-05, + "loss": 0.0439, + "step": 9056 + }, + { + "epoch": 7.649493243243243, + "grad_norm": 0.3259972631931305, + "learning_rate": 2.7159000000000002e-05, + "loss": 0.034, + "step": 9057 + }, + { + "epoch": 7.650337837837838, + "grad_norm": 0.5487648248672485, + "learning_rate": 2.7162000000000002e-05, + "loss": 0.0517, + "step": 9058 + }, + { + "epoch": 7.6511824324324325, + "grad_norm": 0.313600093126297, + "learning_rate": 2.7164999999999998e-05, + "loss": 0.0482, + "step": 9059 + }, + { + "epoch": 7.652027027027027, + "grad_norm": 0.3243439793586731, + "learning_rate": 2.7167999999999998e-05, + "loss": 0.0314, + "step": 9060 + }, + { + "epoch": 7.652871621621622, + "grad_norm": 0.5652727484703064, + "learning_rate": 2.7170999999999998e-05, + "loss": 0.0704, + "step": 9061 + }, + { + "epoch": 7.653716216216216, + "grad_norm": 0.5522606372833252, + "learning_rate": 2.7174e-05, + "loss": 0.0396, + "step": 9062 + }, + { + "epoch": 7.6545608108108105, + "grad_norm": 0.34812769293785095, + "learning_rate": 2.7177e-05, + "loss": 0.0251, + "step": 9063 + }, + { + "epoch": 7.655405405405405, + "grad_norm": 0.27948102355003357, + "learning_rate": 2.718e-05, + "loss": 0.0229, + "step": 9064 + }, + { + "epoch": 7.65625, + "grad_norm": 0.5502997040748596, + "learning_rate": 2.7183e-05, + "loss": 0.055, + "step": 9065 + }, + { + "epoch": 7.657094594594595, + "grad_norm": 0.4020352065563202, + "learning_rate": 2.7186e-05, + "loss": 0.0293, + "step": 9066 + }, + { + "epoch": 7.6579391891891895, + "grad_norm": 1.0213607549667358, + "learning_rate": 2.7189e-05, + "loss": 0.0436, + "step": 9067 + }, + { + "epoch": 7.658783783783784, + "grad_norm": 0.4079277217388153, + "learning_rate": 2.7192e-05, + "loss": 0.0397, + "step": 9068 + }, + { + "epoch": 7.659628378378378, + "grad_norm": 0.40060386061668396, + "learning_rate": 2.7195e-05, + "loss": 0.0296, + "step": 9069 + }, + { + "epoch": 7.660472972972973, + "grad_norm": 0.3912776708602905, + "learning_rate": 2.7198e-05, + "loss": 0.0317, + "step": 9070 + }, + { + "epoch": 7.6613175675675675, + "grad_norm": 0.33079516887664795, + "learning_rate": 2.7201e-05, + "loss": 0.0252, + "step": 9071 + }, + { + "epoch": 7.662162162162162, + "grad_norm": 0.5703173875808716, + "learning_rate": 2.7204000000000002e-05, + "loss": 0.0422, + "step": 9072 + }, + { + "epoch": 7.663006756756757, + "grad_norm": 0.4432836174964905, + "learning_rate": 2.7207000000000002e-05, + "loss": 0.0412, + "step": 9073 + }, + { + "epoch": 7.663851351351351, + "grad_norm": 0.40448200702667236, + "learning_rate": 2.7210000000000002e-05, + "loss": 0.068, + "step": 9074 + }, + { + "epoch": 7.664695945945946, + "grad_norm": 1.2279036045074463, + "learning_rate": 2.7213000000000002e-05, + "loss": 0.0313, + "step": 9075 + }, + { + "epoch": 7.66554054054054, + "grad_norm": 0.6125891208648682, + "learning_rate": 2.7216e-05, + "loss": 0.0485, + "step": 9076 + }, + { + "epoch": 7.666385135135135, + "grad_norm": 0.41184815764427185, + "learning_rate": 2.7219e-05, + "loss": 0.0466, + "step": 9077 + }, + { + "epoch": 7.66722972972973, + "grad_norm": 0.4756956398487091, + "learning_rate": 2.7222e-05, + "loss": 0.0222, + "step": 9078 + }, + { + "epoch": 7.668074324324325, + "grad_norm": 0.5790454149246216, + "learning_rate": 2.7225e-05, + "loss": 0.03, + "step": 9079 + }, + { + "epoch": 7.668918918918919, + "grad_norm": 4.132484436035156, + "learning_rate": 2.7228e-05, + "loss": 0.0591, + "step": 9080 + }, + { + "epoch": 7.669763513513513, + "grad_norm": 1.0362821817398071, + "learning_rate": 2.7231e-05, + "loss": 0.0373, + "step": 9081 + }, + { + "epoch": 7.670608108108108, + "grad_norm": 0.8364120125770569, + "learning_rate": 2.7234000000000004e-05, + "loss": 0.0481, + "step": 9082 + }, + { + "epoch": 7.671452702702703, + "grad_norm": 0.661207377910614, + "learning_rate": 2.7237e-05, + "loss": 0.0426, + "step": 9083 + }, + { + "epoch": 7.672297297297297, + "grad_norm": 0.49658262729644775, + "learning_rate": 2.724e-05, + "loss": 0.0468, + "step": 9084 + }, + { + "epoch": 7.673141891891892, + "grad_norm": 0.6967087388038635, + "learning_rate": 2.7243e-05, + "loss": 0.0475, + "step": 9085 + }, + { + "epoch": 7.673986486486487, + "grad_norm": 0.835150957107544, + "learning_rate": 2.7246e-05, + "loss": 0.0384, + "step": 9086 + }, + { + "epoch": 7.674831081081081, + "grad_norm": 0.5922648906707764, + "learning_rate": 2.7249e-05, + "loss": 0.0503, + "step": 9087 + }, + { + "epoch": 7.675675675675675, + "grad_norm": 0.8239454627037048, + "learning_rate": 2.7252e-05, + "loss": 0.0432, + "step": 9088 + }, + { + "epoch": 7.67652027027027, + "grad_norm": 1.1534043550491333, + "learning_rate": 2.7255e-05, + "loss": 0.2723, + "step": 9089 + }, + { + "epoch": 7.677364864864865, + "grad_norm": 0.7449226379394531, + "learning_rate": 2.7258e-05, + "loss": 0.1551, + "step": 9090 + }, + { + "epoch": 7.67820945945946, + "grad_norm": 0.38482755422592163, + "learning_rate": 2.7261e-05, + "loss": 0.0753, + "step": 9091 + }, + { + "epoch": 7.679054054054054, + "grad_norm": 0.44081562757492065, + "learning_rate": 2.7264000000000002e-05, + "loss": 0.051, + "step": 9092 + }, + { + "epoch": 7.679898648648649, + "grad_norm": 0.4116683304309845, + "learning_rate": 2.7267e-05, + "loss": 0.0414, + "step": 9093 + }, + { + "epoch": 7.680743243243243, + "grad_norm": 0.2872393727302551, + "learning_rate": 2.727e-05, + "loss": 0.0381, + "step": 9094 + }, + { + "epoch": 7.681587837837838, + "grad_norm": 0.512121856212616, + "learning_rate": 2.7273e-05, + "loss": 0.0505, + "step": 9095 + }, + { + "epoch": 7.6824324324324325, + "grad_norm": 0.33682751655578613, + "learning_rate": 2.7276e-05, + "loss": 0.0311, + "step": 9096 + }, + { + "epoch": 7.683277027027027, + "grad_norm": 0.39685508608818054, + "learning_rate": 2.7279e-05, + "loss": 0.0359, + "step": 9097 + }, + { + "epoch": 7.684121621621622, + "grad_norm": 0.32831016182899475, + "learning_rate": 2.7282e-05, + "loss": 0.0434, + "step": 9098 + }, + { + "epoch": 7.684966216216216, + "grad_norm": 0.6165780425071716, + "learning_rate": 2.7285e-05, + "loss": 0.0387, + "step": 9099 + }, + { + "epoch": 7.6858108108108105, + "grad_norm": 0.43933403491973877, + "learning_rate": 2.7288e-05, + "loss": 0.0372, + "step": 9100 + }, + { + "epoch": 7.686655405405405, + "grad_norm": 0.3857854902744293, + "learning_rate": 2.7291e-05, + "loss": 0.0542, + "step": 9101 + }, + { + "epoch": 7.6875, + "grad_norm": 0.3741030991077423, + "learning_rate": 2.7294000000000003e-05, + "loss": 0.0496, + "step": 9102 + }, + { + "epoch": 7.688344594594595, + "grad_norm": 0.397746205329895, + "learning_rate": 2.7297000000000003e-05, + "loss": 0.0452, + "step": 9103 + }, + { + "epoch": 7.6891891891891895, + "grad_norm": 0.5328171253204346, + "learning_rate": 2.7300000000000003e-05, + "loss": 0.0325, + "step": 9104 + }, + { + "epoch": 7.690033783783784, + "grad_norm": 0.3695710599422455, + "learning_rate": 2.7303000000000003e-05, + "loss": 0.0317, + "step": 9105 + }, + { + "epoch": 7.690878378378378, + "grad_norm": 0.574004590511322, + "learning_rate": 2.7306000000000002e-05, + "loss": 0.0395, + "step": 9106 + }, + { + "epoch": 7.691722972972973, + "grad_norm": 0.3234793245792389, + "learning_rate": 2.7309000000000002e-05, + "loss": 0.0308, + "step": 9107 + }, + { + "epoch": 7.6925675675675675, + "grad_norm": 0.40171128511428833, + "learning_rate": 2.7312e-05, + "loss": 0.039, + "step": 9108 + }, + { + "epoch": 7.693412162162162, + "grad_norm": 0.33643847703933716, + "learning_rate": 2.7315e-05, + "loss": 0.0306, + "step": 9109 + }, + { + "epoch": 7.694256756756757, + "grad_norm": 0.3337324261665344, + "learning_rate": 2.7318e-05, + "loss": 0.0436, + "step": 9110 + }, + { + "epoch": 7.695101351351351, + "grad_norm": 0.4240596294403076, + "learning_rate": 2.7320999999999998e-05, + "loss": 0.0353, + "step": 9111 + }, + { + "epoch": 7.695945945945946, + "grad_norm": 0.3020727038383484, + "learning_rate": 2.7324e-05, + "loss": 0.0342, + "step": 9112 + }, + { + "epoch": 7.69679054054054, + "grad_norm": 0.2892606556415558, + "learning_rate": 2.7327e-05, + "loss": 0.0293, + "step": 9113 + }, + { + "epoch": 7.697635135135135, + "grad_norm": 0.41324469447135925, + "learning_rate": 2.733e-05, + "loss": 0.0354, + "step": 9114 + }, + { + "epoch": 7.69847972972973, + "grad_norm": 0.4274594783782959, + "learning_rate": 2.7333e-05, + "loss": 0.0357, + "step": 9115 + }, + { + "epoch": 7.699324324324325, + "grad_norm": 0.487170934677124, + "learning_rate": 2.7336e-05, + "loss": 0.0403, + "step": 9116 + }, + { + "epoch": 7.700168918918919, + "grad_norm": 0.6243817210197449, + "learning_rate": 2.7339e-05, + "loss": 0.0682, + "step": 9117 + }, + { + "epoch": 7.701013513513513, + "grad_norm": 0.5337956547737122, + "learning_rate": 2.7342e-05, + "loss": 0.0454, + "step": 9118 + }, + { + "epoch": 7.701858108108108, + "grad_norm": 0.26538974046707153, + "learning_rate": 2.7345e-05, + "loss": 0.0216, + "step": 9119 + }, + { + "epoch": 7.702702702702703, + "grad_norm": 0.2899664044380188, + "learning_rate": 2.7348e-05, + "loss": 0.0209, + "step": 9120 + }, + { + "epoch": 7.703547297297297, + "grad_norm": 0.3763435482978821, + "learning_rate": 2.7351e-05, + "loss": 0.0367, + "step": 9121 + }, + { + "epoch": 7.704391891891892, + "grad_norm": 0.6279814839363098, + "learning_rate": 2.7354000000000003e-05, + "loss": 0.1191, + "step": 9122 + }, + { + "epoch": 7.705236486486487, + "grad_norm": 0.4190998375415802, + "learning_rate": 2.7357000000000003e-05, + "loss": 0.0362, + "step": 9123 + }, + { + "epoch": 7.706081081081081, + "grad_norm": 0.33559897541999817, + "learning_rate": 2.7360000000000002e-05, + "loss": 0.0223, + "step": 9124 + }, + { + "epoch": 7.706925675675675, + "grad_norm": 0.8959747552871704, + "learning_rate": 2.7363000000000002e-05, + "loss": 0.0571, + "step": 9125 + }, + { + "epoch": 7.70777027027027, + "grad_norm": 0.25757306814193726, + "learning_rate": 2.7366000000000002e-05, + "loss": 0.0207, + "step": 9126 + }, + { + "epoch": 7.708614864864865, + "grad_norm": 0.5149748921394348, + "learning_rate": 2.7369000000000002e-05, + "loss": 0.0487, + "step": 9127 + }, + { + "epoch": 7.70945945945946, + "grad_norm": 0.5578386783599854, + "learning_rate": 2.7372e-05, + "loss": 0.046, + "step": 9128 + }, + { + "epoch": 7.710304054054054, + "grad_norm": 0.6427662372589111, + "learning_rate": 2.7375e-05, + "loss": 0.0527, + "step": 9129 + }, + { + "epoch": 7.711148648648649, + "grad_norm": 0.6571086049079895, + "learning_rate": 2.7378e-05, + "loss": 0.0331, + "step": 9130 + }, + { + "epoch": 7.711993243243243, + "grad_norm": 0.510258674621582, + "learning_rate": 2.7381e-05, + "loss": 0.0495, + "step": 9131 + }, + { + "epoch": 7.712837837837838, + "grad_norm": 0.46619218587875366, + "learning_rate": 2.7383999999999997e-05, + "loss": 0.0577, + "step": 9132 + }, + { + "epoch": 7.7136824324324325, + "grad_norm": 0.3877384662628174, + "learning_rate": 2.7387e-05, + "loss": 0.0336, + "step": 9133 + }, + { + "epoch": 7.714527027027027, + "grad_norm": 0.8000475764274597, + "learning_rate": 2.739e-05, + "loss": 0.0402, + "step": 9134 + }, + { + "epoch": 7.715371621621622, + "grad_norm": 0.4497174322605133, + "learning_rate": 2.7393e-05, + "loss": 0.043, + "step": 9135 + }, + { + "epoch": 7.716216216216216, + "grad_norm": 0.7581391930580139, + "learning_rate": 2.7396e-05, + "loss": 0.0335, + "step": 9136 + }, + { + "epoch": 7.7170608108108105, + "grad_norm": 0.7827717661857605, + "learning_rate": 2.7399e-05, + "loss": 0.0675, + "step": 9137 + }, + { + "epoch": 7.717905405405405, + "grad_norm": 0.8160694241523743, + "learning_rate": 2.7402e-05, + "loss": 0.0815, + "step": 9138 + }, + { + "epoch": 7.71875, + "grad_norm": 1.2443760633468628, + "learning_rate": 2.7405e-05, + "loss": 0.3769, + "step": 9139 + }, + { + "epoch": 7.719594594594595, + "grad_norm": 0.6137304902076721, + "learning_rate": 2.7408e-05, + "loss": 0.1781, + "step": 9140 + }, + { + "epoch": 7.7204391891891895, + "grad_norm": 0.5829335451126099, + "learning_rate": 2.7411e-05, + "loss": 0.1334, + "step": 9141 + }, + { + "epoch": 7.721283783783784, + "grad_norm": 0.4312531054019928, + "learning_rate": 2.7414e-05, + "loss": 0.0543, + "step": 9142 + }, + { + "epoch": 7.722128378378378, + "grad_norm": 0.5582956075668335, + "learning_rate": 2.7417000000000002e-05, + "loss": 0.0513, + "step": 9143 + }, + { + "epoch": 7.722972972972973, + "grad_norm": 0.5118535757064819, + "learning_rate": 2.7420000000000002e-05, + "loss": 0.0515, + "step": 9144 + }, + { + "epoch": 7.7238175675675675, + "grad_norm": 0.49078071117401123, + "learning_rate": 2.7423e-05, + "loss": 0.0567, + "step": 9145 + }, + { + "epoch": 7.724662162162162, + "grad_norm": 0.3431398570537567, + "learning_rate": 2.7426e-05, + "loss": 0.0443, + "step": 9146 + }, + { + "epoch": 7.725506756756757, + "grad_norm": 0.5795283317565918, + "learning_rate": 2.7429e-05, + "loss": 0.0861, + "step": 9147 + }, + { + "epoch": 7.726351351351351, + "grad_norm": 0.36996427178382874, + "learning_rate": 2.7432e-05, + "loss": 0.0312, + "step": 9148 + }, + { + "epoch": 7.727195945945946, + "grad_norm": 0.33735430240631104, + "learning_rate": 2.7435e-05, + "loss": 0.0315, + "step": 9149 + }, + { + "epoch": 7.72804054054054, + "grad_norm": 0.3183879852294922, + "learning_rate": 2.7438e-05, + "loss": 0.0215, + "step": 9150 + }, + { + "epoch": 7.728885135135135, + "grad_norm": 0.4448535144329071, + "learning_rate": 2.7441e-05, + "loss": 0.041, + "step": 9151 + }, + { + "epoch": 7.72972972972973, + "grad_norm": 0.46790817379951477, + "learning_rate": 2.7444e-05, + "loss": 0.0315, + "step": 9152 + }, + { + "epoch": 7.730574324324325, + "grad_norm": 0.4399876296520233, + "learning_rate": 2.7447000000000003e-05, + "loss": 0.0469, + "step": 9153 + }, + { + "epoch": 7.731418918918919, + "grad_norm": 0.3366740942001343, + "learning_rate": 2.7450000000000003e-05, + "loss": 0.0207, + "step": 9154 + }, + { + "epoch": 7.732263513513513, + "grad_norm": 0.5323269367218018, + "learning_rate": 2.7453000000000003e-05, + "loss": 0.043, + "step": 9155 + }, + { + "epoch": 7.733108108108108, + "grad_norm": 0.3736453950405121, + "learning_rate": 2.7456000000000003e-05, + "loss": 0.0336, + "step": 9156 + }, + { + "epoch": 7.733952702702703, + "grad_norm": 0.32524794340133667, + "learning_rate": 2.7459e-05, + "loss": 0.035, + "step": 9157 + }, + { + "epoch": 7.734797297297297, + "grad_norm": 0.6349866390228271, + "learning_rate": 2.7462e-05, + "loss": 0.0351, + "step": 9158 + }, + { + "epoch": 7.735641891891892, + "grad_norm": 0.43626147508621216, + "learning_rate": 2.7465e-05, + "loss": 0.0405, + "step": 9159 + }, + { + "epoch": 7.736486486486487, + "grad_norm": 0.5152937769889832, + "learning_rate": 2.7468e-05, + "loss": 0.0467, + "step": 9160 + }, + { + "epoch": 7.737331081081081, + "grad_norm": 1.1154427528381348, + "learning_rate": 2.7471e-05, + "loss": 0.1442, + "step": 9161 + }, + { + "epoch": 7.738175675675675, + "grad_norm": 0.3353712260723114, + "learning_rate": 2.7473999999999998e-05, + "loss": 0.0324, + "step": 9162 + }, + { + "epoch": 7.73902027027027, + "grad_norm": 0.3255951404571533, + "learning_rate": 2.7477e-05, + "loss": 0.032, + "step": 9163 + }, + { + "epoch": 7.739864864864865, + "grad_norm": 0.679158091545105, + "learning_rate": 2.748e-05, + "loss": 0.0393, + "step": 9164 + }, + { + "epoch": 7.74070945945946, + "grad_norm": 0.38267001509666443, + "learning_rate": 2.7483e-05, + "loss": 0.0515, + "step": 9165 + }, + { + "epoch": 7.741554054054054, + "grad_norm": 0.3221381902694702, + "learning_rate": 2.7486e-05, + "loss": 0.0292, + "step": 9166 + }, + { + "epoch": 7.742398648648649, + "grad_norm": 0.28829294443130493, + "learning_rate": 2.7489e-05, + "loss": 0.0292, + "step": 9167 + }, + { + "epoch": 7.743243243243243, + "grad_norm": 0.40217432379722595, + "learning_rate": 2.7492e-05, + "loss": 0.0465, + "step": 9168 + }, + { + "epoch": 7.744087837837838, + "grad_norm": 0.27622008323669434, + "learning_rate": 2.7495e-05, + "loss": 0.0227, + "step": 9169 + }, + { + "epoch": 7.7449324324324325, + "grad_norm": 0.38839811086654663, + "learning_rate": 2.7498e-05, + "loss": 0.0396, + "step": 9170 + }, + { + "epoch": 7.745777027027027, + "grad_norm": 0.43641209602355957, + "learning_rate": 2.7501e-05, + "loss": 0.0401, + "step": 9171 + }, + { + "epoch": 7.746621621621622, + "grad_norm": 0.5499197840690613, + "learning_rate": 2.7504e-05, + "loss": 0.033, + "step": 9172 + }, + { + "epoch": 7.747466216216216, + "grad_norm": 0.983417272567749, + "learning_rate": 2.7507000000000003e-05, + "loss": 0.0883, + "step": 9173 + }, + { + "epoch": 7.7483108108108105, + "grad_norm": 0.4047367572784424, + "learning_rate": 2.7510000000000003e-05, + "loss": 0.0346, + "step": 9174 + }, + { + "epoch": 7.749155405405405, + "grad_norm": 0.5503190755844116, + "learning_rate": 2.7513000000000002e-05, + "loss": 0.0409, + "step": 9175 + }, + { + "epoch": 7.75, + "grad_norm": 0.395415723323822, + "learning_rate": 2.7516000000000002e-05, + "loss": 0.0401, + "step": 9176 + }, + { + "epoch": 7.750844594594595, + "grad_norm": 0.3154386878013611, + "learning_rate": 2.7519000000000002e-05, + "loss": 0.0248, + "step": 9177 + }, + { + "epoch": 7.7516891891891895, + "grad_norm": 0.4219930171966553, + "learning_rate": 2.7522000000000002e-05, + "loss": 0.0449, + "step": 9178 + }, + { + "epoch": 7.752533783783784, + "grad_norm": 0.44107291102409363, + "learning_rate": 2.7525e-05, + "loss": 0.0331, + "step": 9179 + }, + { + "epoch": 7.753378378378378, + "grad_norm": 0.7863014936447144, + "learning_rate": 2.7528e-05, + "loss": 0.037, + "step": 9180 + }, + { + "epoch": 7.754222972972973, + "grad_norm": 0.4458942413330078, + "learning_rate": 2.7531e-05, + "loss": 0.0348, + "step": 9181 + }, + { + "epoch": 7.7550675675675675, + "grad_norm": 0.7094752788543701, + "learning_rate": 2.7533999999999998e-05, + "loss": 0.0377, + "step": 9182 + }, + { + "epoch": 7.755912162162162, + "grad_norm": 0.614078938961029, + "learning_rate": 2.7537e-05, + "loss": 0.0351, + "step": 9183 + }, + { + "epoch": 7.756756756756757, + "grad_norm": 0.5325682759284973, + "learning_rate": 2.754e-05, + "loss": 0.0335, + "step": 9184 + }, + { + "epoch": 7.757601351351351, + "grad_norm": 0.6090453267097473, + "learning_rate": 2.7543e-05, + "loss": 0.0301, + "step": 9185 + }, + { + "epoch": 7.758445945945946, + "grad_norm": 1.0205435752868652, + "learning_rate": 2.7546e-05, + "loss": 0.05, + "step": 9186 + }, + { + "epoch": 7.75929054054054, + "grad_norm": 1.2059272527694702, + "learning_rate": 2.7549e-05, + "loss": 0.0684, + "step": 9187 + }, + { + "epoch": 7.760135135135135, + "grad_norm": 0.9239664673805237, + "learning_rate": 2.7552e-05, + "loss": 0.0597, + "step": 9188 + }, + { + "epoch": 7.76097972972973, + "grad_norm": 1.0393089056015015, + "learning_rate": 2.7555e-05, + "loss": 0.3266, + "step": 9189 + }, + { + "epoch": 7.761824324324325, + "grad_norm": 0.5734230279922485, + "learning_rate": 2.7558e-05, + "loss": 0.1643, + "step": 9190 + }, + { + "epoch": 7.762668918918919, + "grad_norm": 0.7796306610107422, + "learning_rate": 2.7561e-05, + "loss": 0.1002, + "step": 9191 + }, + { + "epoch": 7.763513513513513, + "grad_norm": 0.5876327157020569, + "learning_rate": 2.7564e-05, + "loss": 0.0636, + "step": 9192 + }, + { + "epoch": 7.764358108108108, + "grad_norm": 0.7370143532752991, + "learning_rate": 2.7567000000000002e-05, + "loss": 0.0795, + "step": 9193 + }, + { + "epoch": 7.765202702702703, + "grad_norm": 0.5703667998313904, + "learning_rate": 2.7570000000000002e-05, + "loss": 0.0478, + "step": 9194 + }, + { + "epoch": 7.766047297297297, + "grad_norm": 0.3099575340747833, + "learning_rate": 2.7573000000000002e-05, + "loss": 0.0218, + "step": 9195 + }, + { + "epoch": 7.766891891891892, + "grad_norm": 0.4126676321029663, + "learning_rate": 2.7576e-05, + "loss": 0.0451, + "step": 9196 + }, + { + "epoch": 7.767736486486487, + "grad_norm": 0.4356764256954193, + "learning_rate": 2.7579e-05, + "loss": 0.0632, + "step": 9197 + }, + { + "epoch": 7.768581081081081, + "grad_norm": 0.4856802523136139, + "learning_rate": 2.7582e-05, + "loss": 0.0585, + "step": 9198 + }, + { + "epoch": 7.769425675675675, + "grad_norm": 0.3059896230697632, + "learning_rate": 2.7585e-05, + "loss": 0.0411, + "step": 9199 + }, + { + "epoch": 7.77027027027027, + "grad_norm": 0.7258771061897278, + "learning_rate": 2.7588e-05, + "loss": 0.0412, + "step": 9200 + }, + { + "epoch": 7.771114864864865, + "grad_norm": 0.8277774453163147, + "learning_rate": 2.7591e-05, + "loss": 0.068, + "step": 9201 + }, + { + "epoch": 7.77195945945946, + "grad_norm": 0.5363995432853699, + "learning_rate": 2.7594e-05, + "loss": 0.0389, + "step": 9202 + }, + { + "epoch": 7.772804054054054, + "grad_norm": 0.43616345524787903, + "learning_rate": 2.7597000000000004e-05, + "loss": 0.0394, + "step": 9203 + }, + { + "epoch": 7.773648648648649, + "grad_norm": 0.43537962436676025, + "learning_rate": 2.7600000000000003e-05, + "loss": 0.0405, + "step": 9204 + }, + { + "epoch": 7.774493243243243, + "grad_norm": 0.46834051609039307, + "learning_rate": 2.7603000000000003e-05, + "loss": 0.0319, + "step": 9205 + }, + { + "epoch": 7.775337837837838, + "grad_norm": 0.46743208169937134, + "learning_rate": 2.7606e-05, + "loss": 0.0319, + "step": 9206 + }, + { + "epoch": 7.7761824324324325, + "grad_norm": 0.3366927206516266, + "learning_rate": 2.7609e-05, + "loss": 0.0358, + "step": 9207 + }, + { + "epoch": 7.777027027027027, + "grad_norm": 0.4973726272583008, + "learning_rate": 2.7612e-05, + "loss": 0.0375, + "step": 9208 + }, + { + "epoch": 7.777871621621622, + "grad_norm": 0.4730171859264374, + "learning_rate": 2.7615e-05, + "loss": 0.0458, + "step": 9209 + }, + { + "epoch": 7.778716216216216, + "grad_norm": 0.2903428375720978, + "learning_rate": 2.7618e-05, + "loss": 0.028, + "step": 9210 + }, + { + "epoch": 7.7795608108108105, + "grad_norm": 0.3635188639163971, + "learning_rate": 2.7621e-05, + "loss": 0.0363, + "step": 9211 + }, + { + "epoch": 7.780405405405405, + "grad_norm": 0.38373300433158875, + "learning_rate": 2.7624e-05, + "loss": 0.0358, + "step": 9212 + }, + { + "epoch": 7.78125, + "grad_norm": 0.40965959429740906, + "learning_rate": 2.7627e-05, + "loss": 0.0391, + "step": 9213 + }, + { + "epoch": 7.782094594594595, + "grad_norm": 0.5324976444244385, + "learning_rate": 2.763e-05, + "loss": 0.0379, + "step": 9214 + }, + { + "epoch": 7.7829391891891895, + "grad_norm": 0.4411926865577698, + "learning_rate": 2.7633e-05, + "loss": 0.0332, + "step": 9215 + }, + { + "epoch": 7.783783783783784, + "grad_norm": 0.43599528074264526, + "learning_rate": 2.7636e-05, + "loss": 0.0381, + "step": 9216 + }, + { + "epoch": 7.784628378378378, + "grad_norm": 0.3787018060684204, + "learning_rate": 2.7639e-05, + "loss": 0.0242, + "step": 9217 + }, + { + "epoch": 7.785472972972973, + "grad_norm": 0.4184933602809906, + "learning_rate": 2.7642e-05, + "loss": 0.0486, + "step": 9218 + }, + { + "epoch": 7.7863175675675675, + "grad_norm": 0.5363699197769165, + "learning_rate": 2.7645e-05, + "loss": 0.0399, + "step": 9219 + }, + { + "epoch": 7.787162162162162, + "grad_norm": 0.5706464052200317, + "learning_rate": 2.7648e-05, + "loss": 0.0598, + "step": 9220 + }, + { + "epoch": 7.788006756756757, + "grad_norm": 0.4627418518066406, + "learning_rate": 2.7651e-05, + "loss": 0.0338, + "step": 9221 + }, + { + "epoch": 7.788851351351351, + "grad_norm": 0.764654278755188, + "learning_rate": 2.7654e-05, + "loss": 0.044, + "step": 9222 + }, + { + "epoch": 7.789695945945946, + "grad_norm": 0.8838900327682495, + "learning_rate": 2.7657000000000003e-05, + "loss": 0.0663, + "step": 9223 + }, + { + "epoch": 7.79054054054054, + "grad_norm": 0.32218414545059204, + "learning_rate": 2.7660000000000003e-05, + "loss": 0.0242, + "step": 9224 + }, + { + "epoch": 7.791385135135135, + "grad_norm": 0.3361205756664276, + "learning_rate": 2.7663000000000003e-05, + "loss": 0.0354, + "step": 9225 + }, + { + "epoch": 7.79222972972973, + "grad_norm": 0.7042792439460754, + "learning_rate": 2.7666000000000002e-05, + "loss": 0.053, + "step": 9226 + }, + { + "epoch": 7.793074324324325, + "grad_norm": 0.38405898213386536, + "learning_rate": 2.7669000000000002e-05, + "loss": 0.0251, + "step": 9227 + }, + { + "epoch": 7.793918918918919, + "grad_norm": 0.4307830333709717, + "learning_rate": 2.7672000000000002e-05, + "loss": 0.0299, + "step": 9228 + }, + { + "epoch": 7.794763513513513, + "grad_norm": 0.35528257489204407, + "learning_rate": 2.7675000000000002e-05, + "loss": 0.0417, + "step": 9229 + }, + { + "epoch": 7.795608108108108, + "grad_norm": 0.7149633765220642, + "learning_rate": 2.7678e-05, + "loss": 0.0316, + "step": 9230 + }, + { + "epoch": 7.796452702702703, + "grad_norm": 0.4574403166770935, + "learning_rate": 2.7680999999999998e-05, + "loss": 0.0356, + "step": 9231 + }, + { + "epoch": 7.797297297297297, + "grad_norm": 0.40667760372161865, + "learning_rate": 2.7683999999999998e-05, + "loss": 0.0329, + "step": 9232 + }, + { + "epoch": 7.798141891891892, + "grad_norm": 0.5778651237487793, + "learning_rate": 2.7687e-05, + "loss": 0.0321, + "step": 9233 + }, + { + "epoch": 7.798986486486487, + "grad_norm": 0.46510031819343567, + "learning_rate": 2.769e-05, + "loss": 0.0442, + "step": 9234 + }, + { + "epoch": 7.799831081081081, + "grad_norm": 0.660209059715271, + "learning_rate": 2.7693e-05, + "loss": 0.0439, + "step": 9235 + }, + { + "epoch": 7.800675675675675, + "grad_norm": 0.5416033267974854, + "learning_rate": 2.7696e-05, + "loss": 0.04, + "step": 9236 + }, + { + "epoch": 7.80152027027027, + "grad_norm": 0.6057378649711609, + "learning_rate": 2.7699e-05, + "loss": 0.0709, + "step": 9237 + }, + { + "epoch": 7.802364864864865, + "grad_norm": 0.5793862342834473, + "learning_rate": 2.7702e-05, + "loss": 0.0688, + "step": 9238 + }, + { + "epoch": 7.80320945945946, + "grad_norm": 0.939458429813385, + "learning_rate": 2.7705e-05, + "loss": 0.2758, + "step": 9239 + }, + { + "epoch": 7.804054054054054, + "grad_norm": 0.6246511936187744, + "learning_rate": 2.7708e-05, + "loss": 0.2101, + "step": 9240 + }, + { + "epoch": 7.804898648648649, + "grad_norm": 0.541507363319397, + "learning_rate": 2.7711e-05, + "loss": 0.1243, + "step": 9241 + }, + { + "epoch": 7.805743243243243, + "grad_norm": 0.6448103189468384, + "learning_rate": 2.7714e-05, + "loss": 0.0855, + "step": 9242 + }, + { + "epoch": 7.806587837837838, + "grad_norm": 0.6149433255195618, + "learning_rate": 2.7717000000000002e-05, + "loss": 0.0463, + "step": 9243 + }, + { + "epoch": 7.8074324324324325, + "grad_norm": 0.384581595659256, + "learning_rate": 2.7720000000000002e-05, + "loss": 0.0379, + "step": 9244 + }, + { + "epoch": 7.808277027027027, + "grad_norm": 0.44582685828208923, + "learning_rate": 2.7723000000000002e-05, + "loss": 0.0367, + "step": 9245 + }, + { + "epoch": 7.809121621621622, + "grad_norm": 0.408128023147583, + "learning_rate": 2.7726000000000002e-05, + "loss": 0.0402, + "step": 9246 + }, + { + "epoch": 7.809966216216216, + "grad_norm": 0.3198162615299225, + "learning_rate": 2.7729e-05, + "loss": 0.0293, + "step": 9247 + }, + { + "epoch": 7.8108108108108105, + "grad_norm": 0.20606358349323273, + "learning_rate": 2.7732e-05, + "loss": 0.0205, + "step": 9248 + }, + { + "epoch": 7.811655405405405, + "grad_norm": 0.3395509123802185, + "learning_rate": 2.7735e-05, + "loss": 0.0434, + "step": 9249 + }, + { + "epoch": 7.8125, + "grad_norm": 0.5571742653846741, + "learning_rate": 2.7738e-05, + "loss": 0.0659, + "step": 9250 + }, + { + "epoch": 7.813344594594595, + "grad_norm": 0.40737810730934143, + "learning_rate": 2.7741e-05, + "loss": 0.0264, + "step": 9251 + }, + { + "epoch": 7.8141891891891895, + "grad_norm": 0.37451162934303284, + "learning_rate": 2.7744e-05, + "loss": 0.035, + "step": 9252 + }, + { + "epoch": 7.815033783783784, + "grad_norm": 0.3622531592845917, + "learning_rate": 2.7747000000000004e-05, + "loss": 0.0363, + "step": 9253 + }, + { + "epoch": 7.815878378378378, + "grad_norm": 0.2878440320491791, + "learning_rate": 2.7750000000000004e-05, + "loss": 0.0367, + "step": 9254 + }, + { + "epoch": 7.816722972972973, + "grad_norm": 0.41959241032600403, + "learning_rate": 2.7753e-05, + "loss": 0.0438, + "step": 9255 + }, + { + "epoch": 7.8175675675675675, + "grad_norm": 0.38035640120506287, + "learning_rate": 2.7756e-05, + "loss": 0.0363, + "step": 9256 + }, + { + "epoch": 7.818412162162162, + "grad_norm": 0.3320920467376709, + "learning_rate": 2.7759e-05, + "loss": 0.0354, + "step": 9257 + }, + { + "epoch": 7.819256756756757, + "grad_norm": 0.4620492160320282, + "learning_rate": 2.7762e-05, + "loss": 0.0295, + "step": 9258 + }, + { + "epoch": 7.820101351351351, + "grad_norm": 0.5286184549331665, + "learning_rate": 2.7765e-05, + "loss": 0.0685, + "step": 9259 + }, + { + "epoch": 7.820945945945946, + "grad_norm": 0.4082317352294922, + "learning_rate": 2.7768e-05, + "loss": 0.0402, + "step": 9260 + }, + { + "epoch": 7.82179054054054, + "grad_norm": 0.4969559907913208, + "learning_rate": 2.7771e-05, + "loss": 0.0345, + "step": 9261 + }, + { + "epoch": 7.822635135135135, + "grad_norm": 0.36426711082458496, + "learning_rate": 2.7774e-05, + "loss": 0.0486, + "step": 9262 + }, + { + "epoch": 7.82347972972973, + "grad_norm": 0.5015283226966858, + "learning_rate": 2.7777e-05, + "loss": 0.0377, + "step": 9263 + }, + { + "epoch": 7.824324324324325, + "grad_norm": 0.4561992287635803, + "learning_rate": 2.778e-05, + "loss": 0.0404, + "step": 9264 + }, + { + "epoch": 7.825168918918919, + "grad_norm": 0.3658468723297119, + "learning_rate": 2.7783e-05, + "loss": 0.0539, + "step": 9265 + }, + { + "epoch": 7.826013513513513, + "grad_norm": 0.25839799642562866, + "learning_rate": 2.7786e-05, + "loss": 0.0222, + "step": 9266 + }, + { + "epoch": 7.826858108108108, + "grad_norm": 0.4192708134651184, + "learning_rate": 2.7789e-05, + "loss": 0.0442, + "step": 9267 + }, + { + "epoch": 7.827702702702703, + "grad_norm": 0.366099089384079, + "learning_rate": 2.7792e-05, + "loss": 0.0327, + "step": 9268 + }, + { + "epoch": 7.828547297297297, + "grad_norm": 0.7890492081642151, + "learning_rate": 2.7795e-05, + "loss": 0.0414, + "step": 9269 + }, + { + "epoch": 7.829391891891892, + "grad_norm": 0.44537460803985596, + "learning_rate": 2.7798e-05, + "loss": 0.0437, + "step": 9270 + }, + { + "epoch": 7.830236486486487, + "grad_norm": 0.502359926700592, + "learning_rate": 2.7801e-05, + "loss": 0.0544, + "step": 9271 + }, + { + "epoch": 7.831081081081081, + "grad_norm": 0.2979062795639038, + "learning_rate": 2.7804e-05, + "loss": 0.0228, + "step": 9272 + }, + { + "epoch": 7.831925675675675, + "grad_norm": 0.4333314597606659, + "learning_rate": 2.7807e-05, + "loss": 0.0462, + "step": 9273 + }, + { + "epoch": 7.83277027027027, + "grad_norm": 0.46488621830940247, + "learning_rate": 2.7810000000000003e-05, + "loss": 0.0494, + "step": 9274 + }, + { + "epoch": 7.833614864864865, + "grad_norm": 0.4545307159423828, + "learning_rate": 2.7813000000000003e-05, + "loss": 0.042, + "step": 9275 + }, + { + "epoch": 7.83445945945946, + "grad_norm": 0.3498944342136383, + "learning_rate": 2.7816000000000003e-05, + "loss": 0.032, + "step": 9276 + }, + { + "epoch": 7.835304054054054, + "grad_norm": 0.429897665977478, + "learning_rate": 2.7819000000000002e-05, + "loss": 0.0343, + "step": 9277 + }, + { + "epoch": 7.836148648648649, + "grad_norm": 0.6995148658752441, + "learning_rate": 2.7822000000000002e-05, + "loss": 0.0376, + "step": 9278 + }, + { + "epoch": 7.836993243243243, + "grad_norm": 0.6449278593063354, + "learning_rate": 2.7825000000000002e-05, + "loss": 0.0404, + "step": 9279 + }, + { + "epoch": 7.837837837837838, + "grad_norm": 0.536142110824585, + "learning_rate": 2.7828e-05, + "loss": 0.0343, + "step": 9280 + }, + { + "epoch": 7.8386824324324325, + "grad_norm": 0.6557995676994324, + "learning_rate": 2.7831e-05, + "loss": 0.0379, + "step": 9281 + }, + { + "epoch": 7.839527027027027, + "grad_norm": 0.45582449436187744, + "learning_rate": 2.7833999999999998e-05, + "loss": 0.0355, + "step": 9282 + }, + { + "epoch": 7.840371621621622, + "grad_norm": 0.31869491934776306, + "learning_rate": 2.7836999999999998e-05, + "loss": 0.0265, + "step": 9283 + }, + { + "epoch": 7.841216216216216, + "grad_norm": 0.34204384684562683, + "learning_rate": 2.784e-05, + "loss": 0.04, + "step": 9284 + }, + { + "epoch": 7.8420608108108105, + "grad_norm": 0.6346569657325745, + "learning_rate": 2.7843e-05, + "loss": 0.0537, + "step": 9285 + }, + { + "epoch": 7.842905405405405, + "grad_norm": 0.5383906364440918, + "learning_rate": 2.7846e-05, + "loss": 0.0369, + "step": 9286 + }, + { + "epoch": 7.84375, + "grad_norm": 0.47314614057540894, + "learning_rate": 2.7849e-05, + "loss": 0.0435, + "step": 9287 + }, + { + "epoch": 7.844594594594595, + "grad_norm": 0.9389992356300354, + "learning_rate": 2.7852e-05, + "loss": 0.0752, + "step": 9288 + }, + { + "epoch": 7.8454391891891895, + "grad_norm": 1.7771269083023071, + "learning_rate": 2.7855e-05, + "loss": 0.3994, + "step": 9289 + }, + { + "epoch": 7.846283783783784, + "grad_norm": 0.7685503959655762, + "learning_rate": 2.7858e-05, + "loss": 0.2184, + "step": 9290 + }, + { + "epoch": 7.847128378378378, + "grad_norm": 0.5051020979881287, + "learning_rate": 2.7861e-05, + "loss": 0.086, + "step": 9291 + }, + { + "epoch": 7.847972972972973, + "grad_norm": 0.4898203909397125, + "learning_rate": 2.7864e-05, + "loss": 0.0715, + "step": 9292 + }, + { + "epoch": 7.8488175675675675, + "grad_norm": 0.4221985340118408, + "learning_rate": 2.7867e-05, + "loss": 0.0604, + "step": 9293 + }, + { + "epoch": 7.849662162162162, + "grad_norm": 0.3407154083251953, + "learning_rate": 2.7870000000000003e-05, + "loss": 0.0406, + "step": 9294 + }, + { + "epoch": 7.850506756756757, + "grad_norm": 0.3892662823200226, + "learning_rate": 2.7873000000000002e-05, + "loss": 0.0477, + "step": 9295 + }, + { + "epoch": 7.851351351351351, + "grad_norm": 0.4049908220767975, + "learning_rate": 2.7876000000000002e-05, + "loss": 0.0493, + "step": 9296 + }, + { + "epoch": 7.852195945945946, + "grad_norm": 0.40593206882476807, + "learning_rate": 2.7879000000000002e-05, + "loss": 0.0418, + "step": 9297 + }, + { + "epoch": 7.85304054054054, + "grad_norm": 0.4881313443183899, + "learning_rate": 2.7882000000000002e-05, + "loss": 0.0448, + "step": 9298 + }, + { + "epoch": 7.853885135135135, + "grad_norm": 0.3664951026439667, + "learning_rate": 2.7885e-05, + "loss": 0.037, + "step": 9299 + }, + { + "epoch": 7.85472972972973, + "grad_norm": 0.40053388476371765, + "learning_rate": 2.7888e-05, + "loss": 0.0364, + "step": 9300 + }, + { + "epoch": 7.855574324324325, + "grad_norm": 0.29461967945098877, + "learning_rate": 2.7891e-05, + "loss": 0.0271, + "step": 9301 + }, + { + "epoch": 7.856418918918919, + "grad_norm": 0.7477752566337585, + "learning_rate": 2.7894e-05, + "loss": 0.0763, + "step": 9302 + }, + { + "epoch": 7.857263513513513, + "grad_norm": 0.3457295596599579, + "learning_rate": 2.7897e-05, + "loss": 0.0287, + "step": 9303 + }, + { + "epoch": 7.858108108108108, + "grad_norm": 0.49803438782691956, + "learning_rate": 2.79e-05, + "loss": 0.0372, + "step": 9304 + }, + { + "epoch": 7.858952702702703, + "grad_norm": 0.6708433032035828, + "learning_rate": 2.7903e-05, + "loss": 0.0387, + "step": 9305 + }, + { + "epoch": 7.859797297297297, + "grad_norm": 0.46518760919570923, + "learning_rate": 2.7906e-05, + "loss": 0.035, + "step": 9306 + }, + { + "epoch": 7.860641891891892, + "grad_norm": 0.42564669251441956, + "learning_rate": 2.7909e-05, + "loss": 0.0562, + "step": 9307 + }, + { + "epoch": 7.861486486486487, + "grad_norm": 0.5964553356170654, + "learning_rate": 2.7912e-05, + "loss": 0.0426, + "step": 9308 + }, + { + "epoch": 7.862331081081081, + "grad_norm": 0.39590317010879517, + "learning_rate": 2.7915e-05, + "loss": 0.0351, + "step": 9309 + }, + { + "epoch": 7.863175675675675, + "grad_norm": 0.6326050162315369, + "learning_rate": 2.7918e-05, + "loss": 0.0561, + "step": 9310 + }, + { + "epoch": 7.86402027027027, + "grad_norm": 0.36786648631095886, + "learning_rate": 2.7921e-05, + "loss": 0.0353, + "step": 9311 + }, + { + "epoch": 7.864864864864865, + "grad_norm": 0.39927974343299866, + "learning_rate": 2.7924e-05, + "loss": 0.0411, + "step": 9312 + }, + { + "epoch": 7.86570945945946, + "grad_norm": 0.25518178939819336, + "learning_rate": 2.7927e-05, + "loss": 0.0298, + "step": 9313 + }, + { + "epoch": 7.866554054054054, + "grad_norm": 0.4918230175971985, + "learning_rate": 2.7930000000000002e-05, + "loss": 0.0374, + "step": 9314 + }, + { + "epoch": 7.867398648648649, + "grad_norm": 0.29182055592536926, + "learning_rate": 2.7933000000000002e-05, + "loss": 0.022, + "step": 9315 + }, + { + "epoch": 7.868243243243243, + "grad_norm": 0.5444723963737488, + "learning_rate": 2.7936e-05, + "loss": 0.0338, + "step": 9316 + }, + { + "epoch": 7.869087837837838, + "grad_norm": 0.3093208074569702, + "learning_rate": 2.7939e-05, + "loss": 0.0313, + "step": 9317 + }, + { + "epoch": 7.8699324324324325, + "grad_norm": 0.32039394974708557, + "learning_rate": 2.7942e-05, + "loss": 0.0464, + "step": 9318 + }, + { + "epoch": 7.870777027027027, + "grad_norm": 0.5408463478088379, + "learning_rate": 2.7945e-05, + "loss": 0.0347, + "step": 9319 + }, + { + "epoch": 7.871621621621622, + "grad_norm": 0.5481081604957581, + "learning_rate": 2.7948e-05, + "loss": 0.0443, + "step": 9320 + }, + { + "epoch": 7.872466216216216, + "grad_norm": 0.5263658165931702, + "learning_rate": 2.7951e-05, + "loss": 0.0339, + "step": 9321 + }, + { + "epoch": 7.8733108108108105, + "grad_norm": 0.434986412525177, + "learning_rate": 2.7954e-05, + "loss": 0.0303, + "step": 9322 + }, + { + "epoch": 7.874155405405405, + "grad_norm": 0.5876225829124451, + "learning_rate": 2.7957e-05, + "loss": 0.038, + "step": 9323 + }, + { + "epoch": 7.875, + "grad_norm": 0.6487839221954346, + "learning_rate": 2.7960000000000003e-05, + "loss": 0.039, + "step": 9324 + }, + { + "epoch": 7.875844594594595, + "grad_norm": 0.6946122646331787, + "learning_rate": 2.7963000000000003e-05, + "loss": 0.0802, + "step": 9325 + }, + { + "epoch": 7.8766891891891895, + "grad_norm": 0.47779470682144165, + "learning_rate": 2.7966000000000003e-05, + "loss": 0.0416, + "step": 9326 + }, + { + "epoch": 7.877533783783784, + "grad_norm": 0.28951218724250793, + "learning_rate": 2.7969000000000003e-05, + "loss": 0.0191, + "step": 9327 + }, + { + "epoch": 7.878378378378378, + "grad_norm": 0.6725450158119202, + "learning_rate": 2.7972000000000003e-05, + "loss": 0.0517, + "step": 9328 + }, + { + "epoch": 7.879222972972973, + "grad_norm": 0.5897263884544373, + "learning_rate": 2.7975e-05, + "loss": 0.035, + "step": 9329 + }, + { + "epoch": 7.8800675675675675, + "grad_norm": 0.7034357786178589, + "learning_rate": 2.7978e-05, + "loss": 0.0382, + "step": 9330 + }, + { + "epoch": 7.880912162162162, + "grad_norm": 0.4885276257991791, + "learning_rate": 2.7981e-05, + "loss": 0.048, + "step": 9331 + }, + { + "epoch": 7.881756756756757, + "grad_norm": 0.42865121364593506, + "learning_rate": 2.7984e-05, + "loss": 0.0231, + "step": 9332 + }, + { + "epoch": 7.882601351351351, + "grad_norm": 0.7931675314903259, + "learning_rate": 2.7986999999999998e-05, + "loss": 0.0644, + "step": 9333 + }, + { + "epoch": 7.883445945945946, + "grad_norm": 0.47027984261512756, + "learning_rate": 2.799e-05, + "loss": 0.0532, + "step": 9334 + }, + { + "epoch": 7.88429054054054, + "grad_norm": 0.8599826097488403, + "learning_rate": 2.7993e-05, + "loss": 0.0586, + "step": 9335 + }, + { + "epoch": 7.885135135135135, + "grad_norm": 0.4958348572254181, + "learning_rate": 2.7996e-05, + "loss": 0.045, + "step": 9336 + }, + { + "epoch": 7.88597972972973, + "grad_norm": 0.46920445561408997, + "learning_rate": 2.7999e-05, + "loss": 0.0579, + "step": 9337 + }, + { + "epoch": 7.886824324324325, + "grad_norm": 1.1038554906845093, + "learning_rate": 2.8002e-05, + "loss": 0.0871, + "step": 9338 + }, + { + "epoch": 7.887668918918919, + "grad_norm": 2.6774659156799316, + "learning_rate": 2.8005e-05, + "loss": 0.3477, + "step": 9339 + }, + { + "epoch": 7.888513513513513, + "grad_norm": 0.8349739909172058, + "learning_rate": 2.8008e-05, + "loss": 0.1974, + "step": 9340 + }, + { + "epoch": 7.889358108108108, + "grad_norm": 0.45283254981040955, + "learning_rate": 2.8011e-05, + "loss": 0.0827, + "step": 9341 + }, + { + "epoch": 7.890202702702703, + "grad_norm": 0.6575438976287842, + "learning_rate": 2.8014e-05, + "loss": 0.0687, + "step": 9342 + }, + { + "epoch": 7.891047297297297, + "grad_norm": 0.4018745720386505, + "learning_rate": 2.8017e-05, + "loss": 0.0638, + "step": 9343 + }, + { + "epoch": 7.891891891891892, + "grad_norm": 0.4872407019138336, + "learning_rate": 2.8020000000000003e-05, + "loss": 0.0391, + "step": 9344 + }, + { + "epoch": 7.892736486486487, + "grad_norm": 0.31244805455207825, + "learning_rate": 2.8023000000000003e-05, + "loss": 0.0347, + "step": 9345 + }, + { + "epoch": 7.893581081081081, + "grad_norm": 0.5363302826881409, + "learning_rate": 2.8026000000000002e-05, + "loss": 0.0321, + "step": 9346 + }, + { + "epoch": 7.894425675675675, + "grad_norm": 0.4387786090373993, + "learning_rate": 2.8029000000000002e-05, + "loss": 0.0424, + "step": 9347 + }, + { + "epoch": 7.89527027027027, + "grad_norm": 0.3501797318458557, + "learning_rate": 2.8032000000000002e-05, + "loss": 0.05, + "step": 9348 + }, + { + "epoch": 7.896114864864865, + "grad_norm": 0.5820484757423401, + "learning_rate": 2.8035000000000002e-05, + "loss": 0.0433, + "step": 9349 + }, + { + "epoch": 7.89695945945946, + "grad_norm": 0.3585938513278961, + "learning_rate": 2.8038e-05, + "loss": 0.0489, + "step": 9350 + }, + { + "epoch": 7.897804054054054, + "grad_norm": 0.2700430750846863, + "learning_rate": 2.8041e-05, + "loss": 0.0235, + "step": 9351 + }, + { + "epoch": 7.898648648648649, + "grad_norm": 0.3760232925415039, + "learning_rate": 2.8044e-05, + "loss": 0.0224, + "step": 9352 + }, + { + "epoch": 7.899493243243243, + "grad_norm": 0.37222108244895935, + "learning_rate": 2.8047e-05, + "loss": 0.0394, + "step": 9353 + }, + { + "epoch": 7.900337837837838, + "grad_norm": 0.3426555395126343, + "learning_rate": 2.805e-05, + "loss": 0.0308, + "step": 9354 + }, + { + "epoch": 7.9011824324324325, + "grad_norm": 0.40881749987602234, + "learning_rate": 2.8053e-05, + "loss": 0.0297, + "step": 9355 + }, + { + "epoch": 7.902027027027027, + "grad_norm": 0.3451775312423706, + "learning_rate": 2.8056e-05, + "loss": 0.046, + "step": 9356 + }, + { + "epoch": 7.902871621621622, + "grad_norm": 0.6227182149887085, + "learning_rate": 2.8059e-05, + "loss": 0.0466, + "step": 9357 + }, + { + "epoch": 7.903716216216216, + "grad_norm": 0.4288051426410675, + "learning_rate": 2.8062e-05, + "loss": 0.0404, + "step": 9358 + }, + { + "epoch": 7.9045608108108105, + "grad_norm": 0.5062849521636963, + "learning_rate": 2.8065e-05, + "loss": 0.0349, + "step": 9359 + }, + { + "epoch": 7.905405405405405, + "grad_norm": 0.3922819197177887, + "learning_rate": 2.8068e-05, + "loss": 0.0326, + "step": 9360 + }, + { + "epoch": 7.90625, + "grad_norm": 0.2713511288166046, + "learning_rate": 2.8071e-05, + "loss": 0.0243, + "step": 9361 + }, + { + "epoch": 7.907094594594595, + "grad_norm": 0.5109285712242126, + "learning_rate": 2.8074e-05, + "loss": 0.0354, + "step": 9362 + }, + { + "epoch": 7.9079391891891895, + "grad_norm": 0.39221781492233276, + "learning_rate": 2.8077e-05, + "loss": 0.0279, + "step": 9363 + }, + { + "epoch": 7.908783783783784, + "grad_norm": 0.344330370426178, + "learning_rate": 2.8080000000000002e-05, + "loss": 0.0228, + "step": 9364 + }, + { + "epoch": 7.909628378378378, + "grad_norm": 0.35214200615882874, + "learning_rate": 2.8083000000000002e-05, + "loss": 0.0427, + "step": 9365 + }, + { + "epoch": 7.910472972972973, + "grad_norm": 0.34910106658935547, + "learning_rate": 2.8086000000000002e-05, + "loss": 0.0307, + "step": 9366 + }, + { + "epoch": 7.9113175675675675, + "grad_norm": 0.3680610656738281, + "learning_rate": 2.8089e-05, + "loss": 0.0279, + "step": 9367 + }, + { + "epoch": 7.912162162162162, + "grad_norm": 0.26485058665275574, + "learning_rate": 2.8092e-05, + "loss": 0.0253, + "step": 9368 + }, + { + "epoch": 7.913006756756757, + "grad_norm": 0.40224117040634155, + "learning_rate": 2.8095e-05, + "loss": 0.0481, + "step": 9369 + }, + { + "epoch": 7.913851351351351, + "grad_norm": 0.3266649544239044, + "learning_rate": 2.8098e-05, + "loss": 0.0339, + "step": 9370 + }, + { + "epoch": 7.914695945945946, + "grad_norm": 1.8531941175460815, + "learning_rate": 2.8101e-05, + "loss": 0.0331, + "step": 9371 + }, + { + "epoch": 7.91554054054054, + "grad_norm": 0.5818595290184021, + "learning_rate": 2.8104e-05, + "loss": 0.0505, + "step": 9372 + }, + { + "epoch": 7.916385135135135, + "grad_norm": 0.4566586911678314, + "learning_rate": 2.8107e-05, + "loss": 0.0427, + "step": 9373 + }, + { + "epoch": 7.91722972972973, + "grad_norm": 0.4923684895038605, + "learning_rate": 2.8110000000000004e-05, + "loss": 0.0333, + "step": 9374 + }, + { + "epoch": 7.918074324324325, + "grad_norm": 0.3065045177936554, + "learning_rate": 2.8113000000000003e-05, + "loss": 0.0231, + "step": 9375 + }, + { + "epoch": 7.918918918918919, + "grad_norm": 0.37175726890563965, + "learning_rate": 2.8116000000000003e-05, + "loss": 0.0359, + "step": 9376 + }, + { + "epoch": 7.919763513513513, + "grad_norm": 0.47023630142211914, + "learning_rate": 2.8119000000000003e-05, + "loss": 0.0452, + "step": 9377 + }, + { + "epoch": 7.920608108108108, + "grad_norm": 0.8127476572990417, + "learning_rate": 2.8122e-05, + "loss": 0.0587, + "step": 9378 + }, + { + "epoch": 7.921452702702703, + "grad_norm": 0.5317134261131287, + "learning_rate": 2.8125e-05, + "loss": 0.0422, + "step": 9379 + }, + { + "epoch": 7.922297297297297, + "grad_norm": 0.5717044472694397, + "learning_rate": 2.8128e-05, + "loss": 0.0435, + "step": 9380 + }, + { + "epoch": 7.923141891891892, + "grad_norm": 0.4877568781375885, + "learning_rate": 2.8131e-05, + "loss": 0.0416, + "step": 9381 + }, + { + "epoch": 7.923986486486487, + "grad_norm": 0.4660654067993164, + "learning_rate": 2.8134e-05, + "loss": 0.0376, + "step": 9382 + }, + { + "epoch": 7.924831081081081, + "grad_norm": 0.41174447536468506, + "learning_rate": 2.8137e-05, + "loss": 0.0357, + "step": 9383 + }, + { + "epoch": 7.925675675675675, + "grad_norm": 0.6575204730033875, + "learning_rate": 2.8139999999999998e-05, + "loss": 0.0517, + "step": 9384 + }, + { + "epoch": 7.92652027027027, + "grad_norm": 0.41108405590057373, + "learning_rate": 2.8143e-05, + "loss": 0.0426, + "step": 9385 + }, + { + "epoch": 7.927364864864865, + "grad_norm": 0.876273512840271, + "learning_rate": 2.8146e-05, + "loss": 0.0595, + "step": 9386 + }, + { + "epoch": 7.92820945945946, + "grad_norm": 0.5389605760574341, + "learning_rate": 2.8149e-05, + "loss": 0.0572, + "step": 9387 + }, + { + "epoch": 7.929054054054054, + "grad_norm": 0.8308470249176025, + "learning_rate": 2.8152e-05, + "loss": 0.1139, + "step": 9388 + }, + { + "epoch": 7.929898648648649, + "grad_norm": 1.3757004737854004, + "learning_rate": 2.8155e-05, + "loss": 0.2859, + "step": 9389 + }, + { + "epoch": 7.930743243243243, + "grad_norm": 0.8965153694152832, + "learning_rate": 2.8158e-05, + "loss": 0.2726, + "step": 9390 + }, + { + "epoch": 7.931587837837838, + "grad_norm": 0.6075337529182434, + "learning_rate": 2.8161e-05, + "loss": 0.0915, + "step": 9391 + }, + { + "epoch": 7.9324324324324325, + "grad_norm": 0.32075756788253784, + "learning_rate": 2.8164e-05, + "loss": 0.0546, + "step": 9392 + }, + { + "epoch": 7.933277027027027, + "grad_norm": 0.43349936604499817, + "learning_rate": 2.8167e-05, + "loss": 0.0563, + "step": 9393 + }, + { + "epoch": 7.934121621621622, + "grad_norm": 0.5789790749549866, + "learning_rate": 2.817e-05, + "loss": 0.0824, + "step": 9394 + }, + { + "epoch": 7.934966216216216, + "grad_norm": 1.1660103797912598, + "learning_rate": 2.8173000000000003e-05, + "loss": 0.0637, + "step": 9395 + }, + { + "epoch": 7.9358108108108105, + "grad_norm": 0.3684857189655304, + "learning_rate": 2.8176000000000003e-05, + "loss": 0.0452, + "step": 9396 + }, + { + "epoch": 7.936655405405405, + "grad_norm": 0.30346909165382385, + "learning_rate": 2.8179000000000002e-05, + "loss": 0.0397, + "step": 9397 + }, + { + "epoch": 7.9375, + "grad_norm": 0.4142126441001892, + "learning_rate": 2.8182000000000002e-05, + "loss": 0.043, + "step": 9398 + }, + { + "epoch": 7.938344594594595, + "grad_norm": 0.30576157569885254, + "learning_rate": 2.8185000000000002e-05, + "loss": 0.0307, + "step": 9399 + }, + { + "epoch": 7.9391891891891895, + "grad_norm": 0.2906179428100586, + "learning_rate": 2.8188000000000002e-05, + "loss": 0.0326, + "step": 9400 + }, + { + "epoch": 7.940033783783784, + "grad_norm": 0.44985872507095337, + "learning_rate": 2.8191e-05, + "loss": 0.0394, + "step": 9401 + }, + { + "epoch": 7.940878378378378, + "grad_norm": 0.31426286697387695, + "learning_rate": 2.8194e-05, + "loss": 0.024, + "step": 9402 + }, + { + "epoch": 7.941722972972973, + "grad_norm": 0.33985960483551025, + "learning_rate": 2.8196999999999998e-05, + "loss": 0.0405, + "step": 9403 + }, + { + "epoch": 7.9425675675675675, + "grad_norm": 0.5262959599494934, + "learning_rate": 2.8199999999999998e-05, + "loss": 0.0319, + "step": 9404 + }, + { + "epoch": 7.943412162162162, + "grad_norm": 0.33755865693092346, + "learning_rate": 2.8203e-05, + "loss": 0.0385, + "step": 9405 + }, + { + "epoch": 7.944256756756757, + "grad_norm": 0.3756353557109833, + "learning_rate": 2.8206e-05, + "loss": 0.0255, + "step": 9406 + }, + { + "epoch": 7.945101351351351, + "grad_norm": 0.36007100343704224, + "learning_rate": 2.8209e-05, + "loss": 0.029, + "step": 9407 + }, + { + "epoch": 7.945945945945946, + "grad_norm": 0.47341275215148926, + "learning_rate": 2.8212e-05, + "loss": 0.0268, + "step": 9408 + }, + { + "epoch": 7.94679054054054, + "grad_norm": 0.42076990008354187, + "learning_rate": 2.8215e-05, + "loss": 0.0402, + "step": 9409 + }, + { + "epoch": 7.947635135135135, + "grad_norm": 0.38956621289253235, + "learning_rate": 2.8218e-05, + "loss": 0.0302, + "step": 9410 + }, + { + "epoch": 7.94847972972973, + "grad_norm": 0.3988093435764313, + "learning_rate": 2.8221e-05, + "loss": 0.0452, + "step": 9411 + }, + { + "epoch": 7.949324324324325, + "grad_norm": 0.5351442098617554, + "learning_rate": 2.8224e-05, + "loss": 0.0436, + "step": 9412 + }, + { + "epoch": 7.950168918918919, + "grad_norm": 0.5445466637611389, + "learning_rate": 2.8227e-05, + "loss": 0.0387, + "step": 9413 + }, + { + "epoch": 7.951013513513513, + "grad_norm": 0.5383580923080444, + "learning_rate": 2.823e-05, + "loss": 0.0449, + "step": 9414 + }, + { + "epoch": 7.951858108108108, + "grad_norm": 0.37851372361183167, + "learning_rate": 2.8233000000000002e-05, + "loss": 0.0377, + "step": 9415 + }, + { + "epoch": 7.952702702702703, + "grad_norm": 0.3706700801849365, + "learning_rate": 2.8236000000000002e-05, + "loss": 0.034, + "step": 9416 + }, + { + "epoch": 7.953547297297297, + "grad_norm": 0.4133135676383972, + "learning_rate": 2.8239000000000002e-05, + "loss": 0.0374, + "step": 9417 + }, + { + "epoch": 7.954391891891892, + "grad_norm": 0.6800333261489868, + "learning_rate": 2.8242e-05, + "loss": 0.0426, + "step": 9418 + }, + { + "epoch": 7.955236486486487, + "grad_norm": 0.4409661591053009, + "learning_rate": 2.8245e-05, + "loss": 0.0585, + "step": 9419 + }, + { + "epoch": 7.956081081081081, + "grad_norm": 0.45611444115638733, + "learning_rate": 2.8248e-05, + "loss": 0.0263, + "step": 9420 + }, + { + "epoch": 7.956925675675675, + "grad_norm": 0.6521497368812561, + "learning_rate": 2.8251e-05, + "loss": 0.0545, + "step": 9421 + }, + { + "epoch": 7.95777027027027, + "grad_norm": 0.5897963047027588, + "learning_rate": 2.8254e-05, + "loss": 0.0392, + "step": 9422 + }, + { + "epoch": 7.958614864864865, + "grad_norm": 0.36764177680015564, + "learning_rate": 2.8257e-05, + "loss": 0.0314, + "step": 9423 + }, + { + "epoch": 7.95945945945946, + "grad_norm": 0.4701443612575531, + "learning_rate": 2.826e-05, + "loss": 0.0359, + "step": 9424 + }, + { + "epoch": 7.960304054054054, + "grad_norm": 0.3158228397369385, + "learning_rate": 2.8263000000000004e-05, + "loss": 0.0274, + "step": 9425 + }, + { + "epoch": 7.961148648648649, + "grad_norm": 0.4384396970272064, + "learning_rate": 2.8266000000000003e-05, + "loss": 0.0305, + "step": 9426 + }, + { + "epoch": 7.961993243243243, + "grad_norm": 0.46546298265457153, + "learning_rate": 2.8269e-05, + "loss": 0.0459, + "step": 9427 + }, + { + "epoch": 7.962837837837838, + "grad_norm": 0.4065169095993042, + "learning_rate": 2.8272e-05, + "loss": 0.0293, + "step": 9428 + }, + { + "epoch": 7.9636824324324325, + "grad_norm": 0.5015691518783569, + "learning_rate": 2.8275e-05, + "loss": 0.0441, + "step": 9429 + }, + { + "epoch": 7.964527027027027, + "grad_norm": 0.5579302906990051, + "learning_rate": 2.8278e-05, + "loss": 0.0485, + "step": 9430 + }, + { + "epoch": 7.965371621621622, + "grad_norm": 0.7747341394424438, + "learning_rate": 2.8281e-05, + "loss": 0.0753, + "step": 9431 + }, + { + "epoch": 7.966216216216216, + "grad_norm": 0.4136282503604889, + "learning_rate": 2.8284e-05, + "loss": 0.0501, + "step": 9432 + }, + { + "epoch": 7.9670608108108105, + "grad_norm": 0.5279449820518494, + "learning_rate": 2.8287e-05, + "loss": 0.0537, + "step": 9433 + }, + { + "epoch": 7.967905405405405, + "grad_norm": 0.4033745527267456, + "learning_rate": 2.829e-05, + "loss": 0.0401, + "step": 9434 + }, + { + "epoch": 7.96875, + "grad_norm": 0.6987799406051636, + "learning_rate": 2.8293e-05, + "loss": 0.0404, + "step": 9435 + }, + { + "epoch": 7.969594594594595, + "grad_norm": 0.6268243193626404, + "learning_rate": 2.8296e-05, + "loss": 0.0541, + "step": 9436 + }, + { + "epoch": 7.9704391891891895, + "grad_norm": 0.8412638306617737, + "learning_rate": 2.8299e-05, + "loss": 0.0586, + "step": 9437 + }, + { + "epoch": 7.971283783783784, + "grad_norm": 0.8728684186935425, + "learning_rate": 2.8302e-05, + "loss": 0.0614, + "step": 9438 + }, + { + "epoch": 7.972128378378378, + "grad_norm": 0.8132959604263306, + "learning_rate": 2.8305e-05, + "loss": 0.2864, + "step": 9439 + }, + { + "epoch": 7.972972972972973, + "grad_norm": 0.6162016987800598, + "learning_rate": 2.8308e-05, + "loss": 0.1624, + "step": 9440 + }, + { + "epoch": 7.9738175675675675, + "grad_norm": 0.517288088798523, + "learning_rate": 2.8311e-05, + "loss": 0.0647, + "step": 9441 + }, + { + "epoch": 7.974662162162162, + "grad_norm": 0.43195584416389465, + "learning_rate": 2.8314e-05, + "loss": 0.047, + "step": 9442 + }, + { + "epoch": 7.975506756756757, + "grad_norm": 0.28635963797569275, + "learning_rate": 2.8317e-05, + "loss": 0.0495, + "step": 9443 + }, + { + "epoch": 7.976351351351351, + "grad_norm": 0.26164528727531433, + "learning_rate": 2.832e-05, + "loss": 0.0348, + "step": 9444 + }, + { + "epoch": 7.977195945945946, + "grad_norm": 0.3714352548122406, + "learning_rate": 2.8323000000000003e-05, + "loss": 0.044, + "step": 9445 + }, + { + "epoch": 7.97804054054054, + "grad_norm": 0.42057108879089355, + "learning_rate": 2.8326000000000003e-05, + "loss": 0.0492, + "step": 9446 + }, + { + "epoch": 7.978885135135135, + "grad_norm": 0.7492820024490356, + "learning_rate": 2.8329000000000003e-05, + "loss": 0.0446, + "step": 9447 + }, + { + "epoch": 7.97972972972973, + "grad_norm": 0.3230738937854767, + "learning_rate": 2.8332000000000002e-05, + "loss": 0.0295, + "step": 9448 + }, + { + "epoch": 7.980574324324325, + "grad_norm": 0.32142969965934753, + "learning_rate": 2.8335000000000002e-05, + "loss": 0.0263, + "step": 9449 + }, + { + "epoch": 7.981418918918919, + "grad_norm": 0.4818192720413208, + "learning_rate": 2.8338000000000002e-05, + "loss": 0.0417, + "step": 9450 + }, + { + "epoch": 7.982263513513513, + "grad_norm": 0.3831006586551666, + "learning_rate": 2.8341000000000002e-05, + "loss": 0.0288, + "step": 9451 + }, + { + "epoch": 7.983108108108108, + "grad_norm": 0.3407990038394928, + "learning_rate": 2.8344e-05, + "loss": 0.0342, + "step": 9452 + }, + { + "epoch": 7.983952702702703, + "grad_norm": 0.391888290643692, + "learning_rate": 2.8346999999999998e-05, + "loss": 0.0353, + "step": 9453 + }, + { + "epoch": 7.984797297297297, + "grad_norm": 0.5300914645195007, + "learning_rate": 2.8349999999999998e-05, + "loss": 0.0517, + "step": 9454 + }, + { + "epoch": 7.985641891891892, + "grad_norm": 0.5786417722702026, + "learning_rate": 2.8353e-05, + "loss": 0.0362, + "step": 9455 + }, + { + "epoch": 7.986486486486487, + "grad_norm": 0.3480454385280609, + "learning_rate": 2.8356e-05, + "loss": 0.0352, + "step": 9456 + }, + { + "epoch": 7.987331081081081, + "grad_norm": 0.42105957865715027, + "learning_rate": 2.8359e-05, + "loss": 0.0161, + "step": 9457 + }, + { + "epoch": 7.988175675675675, + "grad_norm": 0.5606430172920227, + "learning_rate": 2.8362e-05, + "loss": 0.0404, + "step": 9458 + }, + { + "epoch": 7.98902027027027, + "grad_norm": 0.45419764518737793, + "learning_rate": 2.8365e-05, + "loss": 0.0431, + "step": 9459 + }, + { + "epoch": 7.989864864864865, + "grad_norm": 0.36835846304893494, + "learning_rate": 2.8368e-05, + "loss": 0.0313, + "step": 9460 + }, + { + "epoch": 7.99070945945946, + "grad_norm": 0.3762608766555786, + "learning_rate": 2.8371e-05, + "loss": 0.0272, + "step": 9461 + }, + { + "epoch": 7.991554054054054, + "grad_norm": 0.4429268538951874, + "learning_rate": 2.8374e-05, + "loss": 0.0342, + "step": 9462 + }, + { + "epoch": 7.992398648648649, + "grad_norm": 0.4346221089363098, + "learning_rate": 2.8377e-05, + "loss": 0.0823, + "step": 9463 + }, + { + "epoch": 7.993243243243243, + "grad_norm": 0.22461210191249847, + "learning_rate": 2.838e-05, + "loss": 0.0272, + "step": 9464 + }, + { + "epoch": 7.994087837837838, + "grad_norm": 0.5194985270500183, + "learning_rate": 2.8383000000000003e-05, + "loss": 0.044, + "step": 9465 + }, + { + "epoch": 7.9949324324324325, + "grad_norm": 0.6661891937255859, + "learning_rate": 2.8386000000000002e-05, + "loss": 0.0571, + "step": 9466 + }, + { + "epoch": 7.995777027027027, + "grad_norm": 0.5138624906539917, + "learning_rate": 2.8389000000000002e-05, + "loss": 0.0385, + "step": 9467 + }, + { + "epoch": 7.996621621621622, + "grad_norm": 0.48124730587005615, + "learning_rate": 2.8392000000000002e-05, + "loss": 0.0412, + "step": 9468 + }, + { + "epoch": 7.997466216216216, + "grad_norm": 0.7642111778259277, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0411, + "step": 9469 + }, + { + "epoch": 7.9983108108108105, + "grad_norm": 0.3663571774959564, + "learning_rate": 2.8398e-05, + "loss": 0.0294, + "step": 9470 + }, + { + "epoch": 7.999155405405405, + "grad_norm": 0.5906552076339722, + "learning_rate": 2.8401e-05, + "loss": 0.0447, + "step": 9471 + }, + { + "epoch": 8.0, + "grad_norm": 1.3664604425430298, + "learning_rate": 2.8404e-05, + "loss": 0.0659, + "step": 9472 + }, + { + "epoch": 8.000844594594595, + "grad_norm": 0.8791149854660034, + "learning_rate": 2.8407e-05, + "loss": 0.2768, + "step": 9473 + }, + { + "epoch": 8.00168918918919, + "grad_norm": 0.6616712212562561, + "learning_rate": 2.841e-05, + "loss": 0.1989, + "step": 9474 + }, + { + "epoch": 8.002533783783784, + "grad_norm": 0.6193185448646545, + "learning_rate": 2.8413000000000004e-05, + "loss": 0.0982, + "step": 9475 + }, + { + "epoch": 8.003378378378379, + "grad_norm": 0.475092351436615, + "learning_rate": 2.8416e-05, + "loss": 0.0484, + "step": 9476 + }, + { + "epoch": 8.004222972972974, + "grad_norm": 0.3085997998714447, + "learning_rate": 2.8419e-05, + "loss": 0.0313, + "step": 9477 + }, + { + "epoch": 8.005067567567568, + "grad_norm": 0.3592433035373688, + "learning_rate": 2.8422e-05, + "loss": 0.0352, + "step": 9478 + }, + { + "epoch": 8.005912162162161, + "grad_norm": 0.3300180733203888, + "learning_rate": 2.8425e-05, + "loss": 0.031, + "step": 9479 + }, + { + "epoch": 8.006756756756756, + "grad_norm": 0.32832449674606323, + "learning_rate": 2.8428e-05, + "loss": 0.041, + "step": 9480 + }, + { + "epoch": 8.00760135135135, + "grad_norm": 0.34135979413986206, + "learning_rate": 2.8431e-05, + "loss": 0.0437, + "step": 9481 + }, + { + "epoch": 8.008445945945946, + "grad_norm": 0.23265370726585388, + "learning_rate": 2.8434e-05, + "loss": 0.0267, + "step": 9482 + }, + { + "epoch": 8.00929054054054, + "grad_norm": 0.3666171729564667, + "learning_rate": 2.8437e-05, + "loss": 0.0459, + "step": 9483 + }, + { + "epoch": 8.010135135135135, + "grad_norm": 0.3869459331035614, + "learning_rate": 2.844e-05, + "loss": 0.0446, + "step": 9484 + }, + { + "epoch": 8.01097972972973, + "grad_norm": 0.7293341755867004, + "learning_rate": 2.8443000000000002e-05, + "loss": 0.0327, + "step": 9485 + }, + { + "epoch": 8.011824324324325, + "grad_norm": 0.23629829287528992, + "learning_rate": 2.8446000000000002e-05, + "loss": 0.0265, + "step": 9486 + }, + { + "epoch": 8.01266891891892, + "grad_norm": 0.34962210059165955, + "learning_rate": 2.8449e-05, + "loss": 0.0392, + "step": 9487 + }, + { + "epoch": 8.013513513513514, + "grad_norm": 0.5118108987808228, + "learning_rate": 2.8452e-05, + "loss": 0.0309, + "step": 9488 + }, + { + "epoch": 8.014358108108109, + "grad_norm": 0.3321560025215149, + "learning_rate": 2.8455e-05, + "loss": 0.0379, + "step": 9489 + }, + { + "epoch": 8.015202702702704, + "grad_norm": 0.34539955854415894, + "learning_rate": 2.8458e-05, + "loss": 0.0279, + "step": 9490 + }, + { + "epoch": 8.016047297297296, + "grad_norm": 0.6038481593132019, + "learning_rate": 2.8461e-05, + "loss": 0.0299, + "step": 9491 + }, + { + "epoch": 8.016891891891891, + "grad_norm": 0.25933122634887695, + "learning_rate": 2.8464e-05, + "loss": 0.023, + "step": 9492 + }, + { + "epoch": 8.017736486486486, + "grad_norm": 0.25837451219558716, + "learning_rate": 2.8467e-05, + "loss": 0.0233, + "step": 9493 + }, + { + "epoch": 8.01858108108108, + "grad_norm": 0.31400397419929504, + "learning_rate": 2.847e-05, + "loss": 0.0397, + "step": 9494 + }, + { + "epoch": 8.019425675675675, + "grad_norm": 0.2741384208202362, + "learning_rate": 2.8473000000000003e-05, + "loss": 0.0271, + "step": 9495 + }, + { + "epoch": 8.02027027027027, + "grad_norm": 0.34391310811042786, + "learning_rate": 2.8476000000000003e-05, + "loss": 0.0271, + "step": 9496 + }, + { + "epoch": 8.021114864864865, + "grad_norm": 0.37374699115753174, + "learning_rate": 2.8479000000000003e-05, + "loss": 0.0285, + "step": 9497 + }, + { + "epoch": 8.02195945945946, + "grad_norm": 0.18867656588554382, + "learning_rate": 2.8482000000000003e-05, + "loss": 0.0216, + "step": 9498 + }, + { + "epoch": 8.022804054054054, + "grad_norm": 0.39624327421188354, + "learning_rate": 2.8485000000000003e-05, + "loss": 0.0374, + "step": 9499 + }, + { + "epoch": 8.02364864864865, + "grad_norm": 0.3120978772640228, + "learning_rate": 2.8488000000000002e-05, + "loss": 0.0199, + "step": 9500 + }, + { + "epoch": 8.024493243243244, + "grad_norm": 0.3768478333950043, + "learning_rate": 2.8491e-05, + "loss": 0.0246, + "step": 9501 + }, + { + "epoch": 8.025337837837839, + "grad_norm": 0.3587610423564911, + "learning_rate": 2.8494e-05, + "loss": 0.0268, + "step": 9502 + }, + { + "epoch": 8.026182432432432, + "grad_norm": 0.5014599561691284, + "learning_rate": 2.8497e-05, + "loss": 0.0222, + "step": 9503 + }, + { + "epoch": 8.027027027027026, + "grad_norm": 0.29312682151794434, + "learning_rate": 2.8499999999999998e-05, + "loss": 0.0287, + "step": 9504 + }, + { + "epoch": 8.027871621621621, + "grad_norm": 0.251787006855011, + "learning_rate": 2.8502999999999998e-05, + "loss": 0.0213, + "step": 9505 + }, + { + "epoch": 8.028716216216216, + "grad_norm": 0.3611963093280792, + "learning_rate": 2.8506e-05, + "loss": 0.0298, + "step": 9506 + }, + { + "epoch": 8.02956081081081, + "grad_norm": 0.31879323720932007, + "learning_rate": 2.8509e-05, + "loss": 0.0261, + "step": 9507 + }, + { + "epoch": 8.030405405405405, + "grad_norm": 0.28410646319389343, + "learning_rate": 2.8512e-05, + "loss": 0.0395, + "step": 9508 + }, + { + "epoch": 8.03125, + "grad_norm": 0.3172721564769745, + "learning_rate": 2.8515e-05, + "loss": 0.0303, + "step": 9509 + }, + { + "epoch": 8.032094594594595, + "grad_norm": 0.33835119009017944, + "learning_rate": 2.8518e-05, + "loss": 0.0282, + "step": 9510 + }, + { + "epoch": 8.03293918918919, + "grad_norm": 0.736020028591156, + "learning_rate": 2.8521e-05, + "loss": 0.0225, + "step": 9511 + }, + { + "epoch": 8.033783783783784, + "grad_norm": 0.8080905079841614, + "learning_rate": 2.8524e-05, + "loss": 0.035, + "step": 9512 + }, + { + "epoch": 8.034628378378379, + "grad_norm": 0.4127253293991089, + "learning_rate": 2.8527e-05, + "loss": 0.0386, + "step": 9513 + }, + { + "epoch": 8.035472972972974, + "grad_norm": 0.5426331162452698, + "learning_rate": 2.853e-05, + "loss": 0.0378, + "step": 9514 + }, + { + "epoch": 8.036317567567568, + "grad_norm": 0.390108197927475, + "learning_rate": 2.8533e-05, + "loss": 0.0235, + "step": 9515 + }, + { + "epoch": 8.037162162162161, + "grad_norm": 0.509853720664978, + "learning_rate": 2.8536000000000003e-05, + "loss": 0.0538, + "step": 9516 + }, + { + "epoch": 8.038006756756756, + "grad_norm": 0.3842853903770447, + "learning_rate": 2.8539000000000002e-05, + "loss": 0.0263, + "step": 9517 + }, + { + "epoch": 8.03885135135135, + "grad_norm": 0.4563670754432678, + "learning_rate": 2.8542000000000002e-05, + "loss": 0.0407, + "step": 9518 + }, + { + "epoch": 8.039695945945946, + "grad_norm": 0.5987266302108765, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0346, + "step": 9519 + }, + { + "epoch": 8.04054054054054, + "grad_norm": 0.4736505150794983, + "learning_rate": 2.8548000000000002e-05, + "loss": 0.0444, + "step": 9520 + }, + { + "epoch": 8.041385135135135, + "grad_norm": 0.683605432510376, + "learning_rate": 2.8551e-05, + "loss": 0.0573, + "step": 9521 + }, + { + "epoch": 8.04222972972973, + "grad_norm": 1.074741005897522, + "learning_rate": 2.8554e-05, + "loss": 0.0776, + "step": 9522 + }, + { + "epoch": 8.043074324324325, + "grad_norm": 0.8075774312019348, + "learning_rate": 2.8557e-05, + "loss": 0.2777, + "step": 9523 + }, + { + "epoch": 8.04391891891892, + "grad_norm": 0.5935079455375671, + "learning_rate": 2.856e-05, + "loss": 0.17, + "step": 9524 + }, + { + "epoch": 8.044763513513514, + "grad_norm": 0.7454742789268494, + "learning_rate": 2.8563e-05, + "loss": 0.0792, + "step": 9525 + }, + { + "epoch": 8.045608108108109, + "grad_norm": 0.3047192096710205, + "learning_rate": 2.8566e-05, + "loss": 0.0443, + "step": 9526 + }, + { + "epoch": 8.046452702702704, + "grad_norm": 0.3788641691207886, + "learning_rate": 2.8569e-05, + "loss": 0.0439, + "step": 9527 + }, + { + "epoch": 8.047297297297296, + "grad_norm": 0.354294091463089, + "learning_rate": 2.8572e-05, + "loss": 0.0335, + "step": 9528 + }, + { + "epoch": 8.048141891891891, + "grad_norm": 0.40487250685691833, + "learning_rate": 2.8575e-05, + "loss": 0.0379, + "step": 9529 + }, + { + "epoch": 8.048986486486486, + "grad_norm": 0.32733938097953796, + "learning_rate": 2.8578e-05, + "loss": 0.0317, + "step": 9530 + }, + { + "epoch": 8.04983108108108, + "grad_norm": 0.22714997828006744, + "learning_rate": 2.8581e-05, + "loss": 0.0267, + "step": 9531 + }, + { + "epoch": 8.050675675675675, + "grad_norm": 0.24106210470199585, + "learning_rate": 2.8584e-05, + "loss": 0.0271, + "step": 9532 + }, + { + "epoch": 8.05152027027027, + "grad_norm": 0.40162304043769836, + "learning_rate": 2.8587e-05, + "loss": 0.0479, + "step": 9533 + }, + { + "epoch": 8.052364864864865, + "grad_norm": 0.7485094666481018, + "learning_rate": 2.859e-05, + "loss": 0.0599, + "step": 9534 + }, + { + "epoch": 8.05320945945946, + "grad_norm": 0.29319319128990173, + "learning_rate": 2.8593e-05, + "loss": 0.0257, + "step": 9535 + }, + { + "epoch": 8.054054054054054, + "grad_norm": 0.3023678660392761, + "learning_rate": 2.8596000000000002e-05, + "loss": 0.0229, + "step": 9536 + }, + { + "epoch": 8.05489864864865, + "grad_norm": 0.251144677400589, + "learning_rate": 2.8599000000000002e-05, + "loss": 0.0248, + "step": 9537 + }, + { + "epoch": 8.055743243243244, + "grad_norm": 0.5693443417549133, + "learning_rate": 2.8602e-05, + "loss": 0.0396, + "step": 9538 + }, + { + "epoch": 8.056587837837839, + "grad_norm": 0.48247861862182617, + "learning_rate": 2.8605e-05, + "loss": 0.0307, + "step": 9539 + }, + { + "epoch": 8.057432432432432, + "grad_norm": 0.4132448434829712, + "learning_rate": 2.8608e-05, + "loss": 0.0369, + "step": 9540 + }, + { + "epoch": 8.058277027027026, + "grad_norm": 0.31900355219841003, + "learning_rate": 2.8611e-05, + "loss": 0.0318, + "step": 9541 + }, + { + "epoch": 8.059121621621621, + "grad_norm": 0.2314303070306778, + "learning_rate": 2.8614e-05, + "loss": 0.0275, + "step": 9542 + }, + { + "epoch": 8.059966216216216, + "grad_norm": 0.388744592666626, + "learning_rate": 2.8617e-05, + "loss": 0.0323, + "step": 9543 + }, + { + "epoch": 8.06081081081081, + "grad_norm": 0.46634504199028015, + "learning_rate": 2.862e-05, + "loss": 0.0262, + "step": 9544 + }, + { + "epoch": 8.061655405405405, + "grad_norm": 0.7968847751617432, + "learning_rate": 2.8623e-05, + "loss": 0.0313, + "step": 9545 + }, + { + "epoch": 8.0625, + "grad_norm": 0.3571992516517639, + "learning_rate": 2.8626000000000003e-05, + "loss": 0.0361, + "step": 9546 + }, + { + "epoch": 8.063344594594595, + "grad_norm": 0.25357523560523987, + "learning_rate": 2.8629000000000003e-05, + "loss": 0.0263, + "step": 9547 + }, + { + "epoch": 8.06418918918919, + "grad_norm": 0.38365888595581055, + "learning_rate": 2.8632000000000003e-05, + "loss": 0.0176, + "step": 9548 + }, + { + "epoch": 8.065033783783784, + "grad_norm": 0.5568618178367615, + "learning_rate": 2.8635000000000003e-05, + "loss": 0.041, + "step": 9549 + }, + { + "epoch": 8.065878378378379, + "grad_norm": 0.23921340703964233, + "learning_rate": 2.8638e-05, + "loss": 0.0231, + "step": 9550 + }, + { + "epoch": 8.066722972972974, + "grad_norm": 0.4306645095348358, + "learning_rate": 2.8641e-05, + "loss": 0.0282, + "step": 9551 + }, + { + "epoch": 8.067567567567568, + "grad_norm": 0.48820868134498596, + "learning_rate": 2.8644e-05, + "loss": 0.0393, + "step": 9552 + }, + { + "epoch": 8.068412162162161, + "grad_norm": 0.2359912097454071, + "learning_rate": 2.8647e-05, + "loss": 0.0205, + "step": 9553 + }, + { + "epoch": 8.069256756756756, + "grad_norm": 0.533497154712677, + "learning_rate": 2.865e-05, + "loss": 0.0365, + "step": 9554 + }, + { + "epoch": 8.07010135135135, + "grad_norm": 0.30799710750579834, + "learning_rate": 2.8652999999999998e-05, + "loss": 0.0246, + "step": 9555 + }, + { + "epoch": 8.070945945945946, + "grad_norm": 0.44823601841926575, + "learning_rate": 2.8656e-05, + "loss": 0.0302, + "step": 9556 + }, + { + "epoch": 8.07179054054054, + "grad_norm": 0.5838112235069275, + "learning_rate": 2.8659e-05, + "loss": 0.0578, + "step": 9557 + }, + { + "epoch": 8.072635135135135, + "grad_norm": 0.47322672605514526, + "learning_rate": 2.8662e-05, + "loss": 0.0414, + "step": 9558 + }, + { + "epoch": 8.07347972972973, + "grad_norm": 0.3967660367488861, + "learning_rate": 2.8665e-05, + "loss": 0.0304, + "step": 9559 + }, + { + "epoch": 8.074324324324325, + "grad_norm": 0.4158722460269928, + "learning_rate": 2.8668e-05, + "loss": 0.0653, + "step": 9560 + }, + { + "epoch": 8.07516891891892, + "grad_norm": 0.4273594617843628, + "learning_rate": 2.8671e-05, + "loss": 0.0304, + "step": 9561 + }, + { + "epoch": 8.076013513513514, + "grad_norm": 0.4968785345554352, + "learning_rate": 2.8674e-05, + "loss": 0.0356, + "step": 9562 + }, + { + "epoch": 8.076858108108109, + "grad_norm": 0.3963542878627777, + "learning_rate": 2.8677e-05, + "loss": 0.0285, + "step": 9563 + }, + { + "epoch": 8.077702702702704, + "grad_norm": 0.4797937870025635, + "learning_rate": 2.868e-05, + "loss": 0.0311, + "step": 9564 + }, + { + "epoch": 8.078547297297296, + "grad_norm": 1.59944748878479, + "learning_rate": 2.8683e-05, + "loss": 0.0619, + "step": 9565 + }, + { + "epoch": 8.079391891891891, + "grad_norm": 0.7486570477485657, + "learning_rate": 2.8686000000000003e-05, + "loss": 0.0272, + "step": 9566 + }, + { + "epoch": 8.080236486486486, + "grad_norm": 0.5656939148902893, + "learning_rate": 2.8689000000000003e-05, + "loss": 0.0316, + "step": 9567 + }, + { + "epoch": 8.08108108108108, + "grad_norm": 0.4482349753379822, + "learning_rate": 2.8692000000000002e-05, + "loss": 0.037, + "step": 9568 + }, + { + "epoch": 8.081925675675675, + "grad_norm": 0.44228214025497437, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0376, + "step": 9569 + }, + { + "epoch": 8.08277027027027, + "grad_norm": 0.4539628028869629, + "learning_rate": 2.8698000000000002e-05, + "loss": 0.0394, + "step": 9570 + }, + { + "epoch": 8.083614864864865, + "grad_norm": 0.38565880060195923, + "learning_rate": 2.8701000000000002e-05, + "loss": 0.031, + "step": 9571 + }, + { + "epoch": 8.08445945945946, + "grad_norm": 0.7057384252548218, + "learning_rate": 2.8704e-05, + "loss": 0.0444, + "step": 9572 + }, + { + "epoch": 8.085304054054054, + "grad_norm": 1.8619049787521362, + "learning_rate": 2.8707e-05, + "loss": 0.3097, + "step": 9573 + }, + { + "epoch": 8.08614864864865, + "grad_norm": 0.6305084824562073, + "learning_rate": 2.871e-05, + "loss": 0.1564, + "step": 9574 + }, + { + "epoch": 8.086993243243244, + "grad_norm": 0.37059029936790466, + "learning_rate": 2.8712999999999998e-05, + "loss": 0.0838, + "step": 9575 + }, + { + "epoch": 8.087837837837839, + "grad_norm": 0.4072357714176178, + "learning_rate": 2.8716e-05, + "loss": 0.0565, + "step": 9576 + }, + { + "epoch": 8.088682432432432, + "grad_norm": 0.4602712392807007, + "learning_rate": 2.8719e-05, + "loss": 0.0331, + "step": 9577 + }, + { + "epoch": 8.089527027027026, + "grad_norm": 0.44282227754592896, + "learning_rate": 2.8722e-05, + "loss": 0.0364, + "step": 9578 + }, + { + "epoch": 8.090371621621621, + "grad_norm": 0.33636119961738586, + "learning_rate": 2.8725e-05, + "loss": 0.0341, + "step": 9579 + }, + { + "epoch": 8.091216216216216, + "grad_norm": 0.3406696915626526, + "learning_rate": 2.8728e-05, + "loss": 0.0334, + "step": 9580 + }, + { + "epoch": 8.09206081081081, + "grad_norm": 0.36385810375213623, + "learning_rate": 2.8731e-05, + "loss": 0.0457, + "step": 9581 + }, + { + "epoch": 8.092905405405405, + "grad_norm": 0.45220890641212463, + "learning_rate": 2.8734e-05, + "loss": 0.0505, + "step": 9582 + }, + { + "epoch": 8.09375, + "grad_norm": 0.3362233638763428, + "learning_rate": 2.8737e-05, + "loss": 0.0307, + "step": 9583 + }, + { + "epoch": 8.094594594594595, + "grad_norm": 0.2630385160446167, + "learning_rate": 2.874e-05, + "loss": 0.0305, + "step": 9584 + }, + { + "epoch": 8.09543918918919, + "grad_norm": 0.43928560614585876, + "learning_rate": 2.8743e-05, + "loss": 0.0304, + "step": 9585 + }, + { + "epoch": 8.096283783783784, + "grad_norm": 0.2824574112892151, + "learning_rate": 2.8746000000000002e-05, + "loss": 0.0207, + "step": 9586 + }, + { + "epoch": 8.097128378378379, + "grad_norm": 0.329000860452652, + "learning_rate": 2.8749000000000002e-05, + "loss": 0.0276, + "step": 9587 + }, + { + "epoch": 8.097972972972974, + "grad_norm": 0.31800469756126404, + "learning_rate": 2.8752000000000002e-05, + "loss": 0.0262, + "step": 9588 + }, + { + "epoch": 8.098817567567568, + "grad_norm": 0.9112160205841064, + "learning_rate": 2.8755e-05, + "loss": 0.0324, + "step": 9589 + }, + { + "epoch": 8.099662162162161, + "grad_norm": 0.37127435207366943, + "learning_rate": 2.8758e-05, + "loss": 0.0355, + "step": 9590 + }, + { + "epoch": 8.100506756756756, + "grad_norm": 0.5977120399475098, + "learning_rate": 2.8761e-05, + "loss": 0.045, + "step": 9591 + }, + { + "epoch": 8.10135135135135, + "grad_norm": 0.515582263469696, + "learning_rate": 2.8764e-05, + "loss": 0.0449, + "step": 9592 + }, + { + "epoch": 8.102195945945946, + "grad_norm": 0.3743630349636078, + "learning_rate": 2.8767e-05, + "loss": 0.0308, + "step": 9593 + }, + { + "epoch": 8.10304054054054, + "grad_norm": 0.24974775314331055, + "learning_rate": 2.877e-05, + "loss": 0.0196, + "step": 9594 + }, + { + "epoch": 8.103885135135135, + "grad_norm": 0.5064236521720886, + "learning_rate": 2.8773e-05, + "loss": 0.0305, + "step": 9595 + }, + { + "epoch": 8.10472972972973, + "grad_norm": 0.33876779675483704, + "learning_rate": 2.8776000000000004e-05, + "loss": 0.0263, + "step": 9596 + }, + { + "epoch": 8.105574324324325, + "grad_norm": 0.3265869915485382, + "learning_rate": 2.8779000000000003e-05, + "loss": 0.0341, + "step": 9597 + }, + { + "epoch": 8.10641891891892, + "grad_norm": 0.4473784267902374, + "learning_rate": 2.8782000000000003e-05, + "loss": 0.0371, + "step": 9598 + }, + { + "epoch": 8.107263513513514, + "grad_norm": 0.5696384906768799, + "learning_rate": 2.8785e-05, + "loss": 0.0434, + "step": 9599 + }, + { + "epoch": 8.108108108108109, + "grad_norm": 0.49045777320861816, + "learning_rate": 2.8788e-05, + "loss": 0.0338, + "step": 9600 + }, + { + "epoch": 8.108952702702704, + "grad_norm": 0.5384039878845215, + "learning_rate": 2.8791e-05, + "loss": 0.031, + "step": 9601 + }, + { + "epoch": 8.109797297297296, + "grad_norm": 0.4253823459148407, + "learning_rate": 2.8794e-05, + "loss": 0.0569, + "step": 9602 + }, + { + "epoch": 8.110641891891891, + "grad_norm": 0.31200215220451355, + "learning_rate": 2.8797e-05, + "loss": 0.0238, + "step": 9603 + }, + { + "epoch": 8.111486486486486, + "grad_norm": 0.5391426086425781, + "learning_rate": 2.88e-05, + "loss": 0.0476, + "step": 9604 + }, + { + "epoch": 8.11233108108108, + "grad_norm": 0.3072352409362793, + "learning_rate": 2.8803e-05, + "loss": 0.0301, + "step": 9605 + }, + { + "epoch": 8.113175675675675, + "grad_norm": 0.4534766376018524, + "learning_rate": 2.8806e-05, + "loss": 0.0522, + "step": 9606 + }, + { + "epoch": 8.11402027027027, + "grad_norm": 0.4814991056919098, + "learning_rate": 2.8809e-05, + "loss": 0.0339, + "step": 9607 + }, + { + "epoch": 8.114864864864865, + "grad_norm": 0.4305620789527893, + "learning_rate": 2.8812e-05, + "loss": 0.0368, + "step": 9608 + }, + { + "epoch": 8.11570945945946, + "grad_norm": 0.48197999596595764, + "learning_rate": 2.8815e-05, + "loss": 0.0309, + "step": 9609 + }, + { + "epoch": 8.116554054054054, + "grad_norm": 0.4157969355583191, + "learning_rate": 2.8818e-05, + "loss": 0.0295, + "step": 9610 + }, + { + "epoch": 8.11739864864865, + "grad_norm": 0.41187620162963867, + "learning_rate": 2.8821e-05, + "loss": 0.049, + "step": 9611 + }, + { + "epoch": 8.118243243243244, + "grad_norm": 0.5130999684333801, + "learning_rate": 2.8824e-05, + "loss": 0.0317, + "step": 9612 + }, + { + "epoch": 8.119087837837839, + "grad_norm": 1.2414969205856323, + "learning_rate": 2.8827e-05, + "loss": 0.0329, + "step": 9613 + }, + { + "epoch": 8.119932432432432, + "grad_norm": 0.42406466603279114, + "learning_rate": 2.883e-05, + "loss": 0.0275, + "step": 9614 + }, + { + "epoch": 8.120777027027026, + "grad_norm": 0.6415259838104248, + "learning_rate": 2.8833e-05, + "loss": 0.0457, + "step": 9615 + }, + { + "epoch": 8.121621621621621, + "grad_norm": 0.8236889839172363, + "learning_rate": 2.8836000000000003e-05, + "loss": 0.0321, + "step": 9616 + }, + { + "epoch": 8.122466216216216, + "grad_norm": 0.7250033020973206, + "learning_rate": 2.8839000000000003e-05, + "loss": 0.037, + "step": 9617 + }, + { + "epoch": 8.12331081081081, + "grad_norm": 0.6466588377952576, + "learning_rate": 2.8842000000000003e-05, + "loss": 0.0485, + "step": 9618 + }, + { + "epoch": 8.124155405405405, + "grad_norm": 0.3344354033470154, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0206, + "step": 9619 + }, + { + "epoch": 8.125, + "grad_norm": 0.478120356798172, + "learning_rate": 2.8848000000000002e-05, + "loss": 0.0279, + "step": 9620 + }, + { + "epoch": 8.125844594594595, + "grad_norm": 0.5327579379081726, + "learning_rate": 2.8851000000000002e-05, + "loss": 0.0317, + "step": 9621 + }, + { + "epoch": 8.12668918918919, + "grad_norm": 1.7812031507492065, + "learning_rate": 2.8854000000000002e-05, + "loss": 0.0647, + "step": 9622 + }, + { + "epoch": 8.127533783783784, + "grad_norm": 0.774599552154541, + "learning_rate": 2.8857000000000002e-05, + "loss": 0.2586, + "step": 9623 + }, + { + "epoch": 8.128378378378379, + "grad_norm": 0.6869192719459534, + "learning_rate": 2.8859999999999998e-05, + "loss": 0.1611, + "step": 9624 + }, + { + "epoch": 8.129222972972974, + "grad_norm": 0.6024976968765259, + "learning_rate": 2.8862999999999998e-05, + "loss": 0.0857, + "step": 9625 + }, + { + "epoch": 8.130067567567568, + "grad_norm": 0.5607405304908752, + "learning_rate": 2.8866e-05, + "loss": 0.0387, + "step": 9626 + }, + { + "epoch": 8.130912162162161, + "grad_norm": 0.4115413427352905, + "learning_rate": 2.8869e-05, + "loss": 0.0352, + "step": 9627 + }, + { + "epoch": 8.131756756756756, + "grad_norm": 0.33118557929992676, + "learning_rate": 2.8872e-05, + "loss": 0.0284, + "step": 9628 + }, + { + "epoch": 8.13260135135135, + "grad_norm": 0.7538695931434631, + "learning_rate": 2.8875e-05, + "loss": 0.0527, + "step": 9629 + }, + { + "epoch": 8.133445945945946, + "grad_norm": 1.1809002161026, + "learning_rate": 2.8878e-05, + "loss": 0.0411, + "step": 9630 + }, + { + "epoch": 8.13429054054054, + "grad_norm": 0.27540621161460876, + "learning_rate": 2.8881e-05, + "loss": 0.0315, + "step": 9631 + }, + { + "epoch": 8.135135135135135, + "grad_norm": 0.27846062183380127, + "learning_rate": 2.8884e-05, + "loss": 0.0329, + "step": 9632 + }, + { + "epoch": 8.13597972972973, + "grad_norm": 0.3131551444530487, + "learning_rate": 2.8887e-05, + "loss": 0.0289, + "step": 9633 + }, + { + "epoch": 8.136824324324325, + "grad_norm": 0.4288376569747925, + "learning_rate": 2.889e-05, + "loss": 0.059, + "step": 9634 + }, + { + "epoch": 8.13766891891892, + "grad_norm": 0.3352106809616089, + "learning_rate": 2.8893e-05, + "loss": 0.0609, + "step": 9635 + }, + { + "epoch": 8.138513513513514, + "grad_norm": 0.3593440353870392, + "learning_rate": 2.8896e-05, + "loss": 0.0317, + "step": 9636 + }, + { + "epoch": 8.139358108108109, + "grad_norm": 0.3167284429073334, + "learning_rate": 2.8899000000000002e-05, + "loss": 0.0422, + "step": 9637 + }, + { + "epoch": 8.140202702702704, + "grad_norm": 0.45327842235565186, + "learning_rate": 2.8902000000000002e-05, + "loss": 0.0466, + "step": 9638 + }, + { + "epoch": 8.141047297297296, + "grad_norm": 0.30307814478874207, + "learning_rate": 2.8905000000000002e-05, + "loss": 0.0308, + "step": 9639 + }, + { + "epoch": 8.141891891891891, + "grad_norm": 0.5541331768035889, + "learning_rate": 2.8908000000000002e-05, + "loss": 0.0369, + "step": 9640 + }, + { + "epoch": 8.142736486486486, + "grad_norm": 0.44330981373786926, + "learning_rate": 2.8911e-05, + "loss": 0.0374, + "step": 9641 + }, + { + "epoch": 8.14358108108108, + "grad_norm": 0.30814680457115173, + "learning_rate": 2.8914e-05, + "loss": 0.0307, + "step": 9642 + }, + { + "epoch": 8.144425675675675, + "grad_norm": 0.4358811378479004, + "learning_rate": 2.8917e-05, + "loss": 0.0441, + "step": 9643 + }, + { + "epoch": 8.14527027027027, + "grad_norm": 0.29279324412345886, + "learning_rate": 2.892e-05, + "loss": 0.0193, + "step": 9644 + }, + { + "epoch": 8.146114864864865, + "grad_norm": 0.5346368551254272, + "learning_rate": 2.8923e-05, + "loss": 0.1173, + "step": 9645 + }, + { + "epoch": 8.14695945945946, + "grad_norm": 0.4217517673969269, + "learning_rate": 2.8926e-05, + "loss": 0.0644, + "step": 9646 + }, + { + "epoch": 8.147804054054054, + "grad_norm": 0.4642598032951355, + "learning_rate": 2.8929000000000004e-05, + "loss": 0.0345, + "step": 9647 + }, + { + "epoch": 8.14864864864865, + "grad_norm": 0.25268611311912537, + "learning_rate": 2.8932e-05, + "loss": 0.0305, + "step": 9648 + }, + { + "epoch": 8.149493243243244, + "grad_norm": 0.3869633972644806, + "learning_rate": 2.8935e-05, + "loss": 0.0341, + "step": 9649 + }, + { + "epoch": 8.150337837837839, + "grad_norm": 0.4043413996696472, + "learning_rate": 2.8938e-05, + "loss": 0.0229, + "step": 9650 + }, + { + "epoch": 8.151182432432432, + "grad_norm": 0.3529498875141144, + "learning_rate": 2.8941e-05, + "loss": 0.0224, + "step": 9651 + }, + { + "epoch": 8.152027027027026, + "grad_norm": 0.3330707848072052, + "learning_rate": 2.8944e-05, + "loss": 0.0273, + "step": 9652 + }, + { + "epoch": 8.152871621621621, + "grad_norm": 0.6893938779830933, + "learning_rate": 2.8947e-05, + "loss": 0.0314, + "step": 9653 + }, + { + "epoch": 8.153716216216216, + "grad_norm": 0.638129472732544, + "learning_rate": 2.895e-05, + "loss": 0.0307, + "step": 9654 + }, + { + "epoch": 8.15456081081081, + "grad_norm": 0.8833263516426086, + "learning_rate": 2.8953e-05, + "loss": 0.036, + "step": 9655 + }, + { + "epoch": 8.155405405405405, + "grad_norm": 0.33795595169067383, + "learning_rate": 2.8956e-05, + "loss": 0.0283, + "step": 9656 + }, + { + "epoch": 8.15625, + "grad_norm": 0.4432410001754761, + "learning_rate": 2.8959000000000002e-05, + "loss": 0.0431, + "step": 9657 + }, + { + "epoch": 8.157094594594595, + "grad_norm": 0.25405263900756836, + "learning_rate": 2.8962e-05, + "loss": 0.0207, + "step": 9658 + }, + { + "epoch": 8.15793918918919, + "grad_norm": 0.36517420411109924, + "learning_rate": 2.8965e-05, + "loss": 0.0256, + "step": 9659 + }, + { + "epoch": 8.158783783783784, + "grad_norm": 0.9603286981582642, + "learning_rate": 2.8968e-05, + "loss": 0.0283, + "step": 9660 + }, + { + "epoch": 8.159628378378379, + "grad_norm": 0.4652397036552429, + "learning_rate": 2.8971e-05, + "loss": 0.0391, + "step": 9661 + }, + { + "epoch": 8.160472972972974, + "grad_norm": 0.6097494959831238, + "learning_rate": 2.8974e-05, + "loss": 0.051, + "step": 9662 + }, + { + "epoch": 8.161317567567568, + "grad_norm": 0.3046312928199768, + "learning_rate": 2.8977e-05, + "loss": 0.0269, + "step": 9663 + }, + { + "epoch": 8.162162162162161, + "grad_norm": 0.576866090297699, + "learning_rate": 2.898e-05, + "loss": 0.0466, + "step": 9664 + }, + { + "epoch": 8.163006756756756, + "grad_norm": 0.6150354146957397, + "learning_rate": 2.8983e-05, + "loss": 0.0476, + "step": 9665 + }, + { + "epoch": 8.16385135135135, + "grad_norm": 0.527491569519043, + "learning_rate": 2.8986e-05, + "loss": 0.0531, + "step": 9666 + }, + { + "epoch": 8.164695945945946, + "grad_norm": 0.5584753751754761, + "learning_rate": 2.8989000000000003e-05, + "loss": 0.0335, + "step": 9667 + }, + { + "epoch": 8.16554054054054, + "grad_norm": 1.2520549297332764, + "learning_rate": 2.8992000000000003e-05, + "loss": 0.0494, + "step": 9668 + }, + { + "epoch": 8.166385135135135, + "grad_norm": 0.5749549865722656, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0457, + "step": 9669 + }, + { + "epoch": 8.16722972972973, + "grad_norm": 0.6586952805519104, + "learning_rate": 2.8998000000000003e-05, + "loss": 0.0485, + "step": 9670 + }, + { + "epoch": 8.168074324324325, + "grad_norm": 2.3980257511138916, + "learning_rate": 2.9001000000000002e-05, + "loss": 0.1765, + "step": 9671 + }, + { + "epoch": 8.16891891891892, + "grad_norm": 1.416041612625122, + "learning_rate": 2.9004000000000002e-05, + "loss": 0.1665, + "step": 9672 + }, + { + "epoch": 8.169763513513514, + "grad_norm": 0.9422590136528015, + "learning_rate": 2.9007e-05, + "loss": 0.2423, + "step": 9673 + }, + { + "epoch": 8.170608108108109, + "grad_norm": 0.6986570358276367, + "learning_rate": 2.901e-05, + "loss": 0.1613, + "step": 9674 + }, + { + "epoch": 8.171452702702704, + "grad_norm": 0.8692867755889893, + "learning_rate": 2.9012999999999998e-05, + "loss": 0.133, + "step": 9675 + }, + { + "epoch": 8.172297297297296, + "grad_norm": 0.40584662556648254, + "learning_rate": 2.9015999999999998e-05, + "loss": 0.0551, + "step": 9676 + }, + { + "epoch": 8.173141891891891, + "grad_norm": 0.5500714778900146, + "learning_rate": 2.9019e-05, + "loss": 0.057, + "step": 9677 + }, + { + "epoch": 8.173986486486486, + "grad_norm": 0.37664106488227844, + "learning_rate": 2.9022e-05, + "loss": 0.0471, + "step": 9678 + }, + { + "epoch": 8.17483108108108, + "grad_norm": 0.416106253862381, + "learning_rate": 2.9025e-05, + "loss": 0.0395, + "step": 9679 + }, + { + "epoch": 8.175675675675675, + "grad_norm": 0.2774165868759155, + "learning_rate": 2.9028e-05, + "loss": 0.0237, + "step": 9680 + }, + { + "epoch": 8.17652027027027, + "grad_norm": 0.41313543915748596, + "learning_rate": 2.9031e-05, + "loss": 0.043, + "step": 9681 + }, + { + "epoch": 8.177364864864865, + "grad_norm": 0.4816824197769165, + "learning_rate": 2.9034e-05, + "loss": 0.0486, + "step": 9682 + }, + { + "epoch": 8.17820945945946, + "grad_norm": 0.4110986590385437, + "learning_rate": 2.9037e-05, + "loss": 0.0523, + "step": 9683 + }, + { + "epoch": 8.179054054054054, + "grad_norm": 0.437182754278183, + "learning_rate": 2.904e-05, + "loss": 0.0391, + "step": 9684 + }, + { + "epoch": 8.17989864864865, + "grad_norm": 0.3705117702484131, + "learning_rate": 2.9043e-05, + "loss": 0.0277, + "step": 9685 + }, + { + "epoch": 8.180743243243244, + "grad_norm": 0.49730992317199707, + "learning_rate": 2.9046e-05, + "loss": 0.055, + "step": 9686 + }, + { + "epoch": 8.181587837837839, + "grad_norm": 0.3292195200920105, + "learning_rate": 2.9049000000000003e-05, + "loss": 0.0276, + "step": 9687 + }, + { + "epoch": 8.182432432432432, + "grad_norm": 0.2572038471698761, + "learning_rate": 2.9052000000000002e-05, + "loss": 0.0285, + "step": 9688 + }, + { + "epoch": 8.183277027027026, + "grad_norm": 0.3360028862953186, + "learning_rate": 2.9055000000000002e-05, + "loss": 0.039, + "step": 9689 + }, + { + "epoch": 8.184121621621621, + "grad_norm": 0.3949260115623474, + "learning_rate": 2.9058000000000002e-05, + "loss": 0.0268, + "step": 9690 + }, + { + "epoch": 8.184966216216216, + "grad_norm": 0.4732036590576172, + "learning_rate": 2.9061000000000002e-05, + "loss": 0.039, + "step": 9691 + }, + { + "epoch": 8.18581081081081, + "grad_norm": 0.4465363323688507, + "learning_rate": 2.9064e-05, + "loss": 0.0374, + "step": 9692 + }, + { + "epoch": 8.186655405405405, + "grad_norm": 0.41397786140441895, + "learning_rate": 2.9067e-05, + "loss": 0.0317, + "step": 9693 + }, + { + "epoch": 8.1875, + "grad_norm": 0.6161172986030579, + "learning_rate": 2.907e-05, + "loss": 0.0319, + "step": 9694 + }, + { + "epoch": 8.188344594594595, + "grad_norm": 0.41299542784690857, + "learning_rate": 2.9073e-05, + "loss": 0.0573, + "step": 9695 + }, + { + "epoch": 8.18918918918919, + "grad_norm": 0.28529372811317444, + "learning_rate": 2.9076e-05, + "loss": 0.0323, + "step": 9696 + }, + { + "epoch": 8.190033783783784, + "grad_norm": 0.4162665903568268, + "learning_rate": 2.9079e-05, + "loss": 0.0507, + "step": 9697 + }, + { + "epoch": 8.190878378378379, + "grad_norm": 0.4319911599159241, + "learning_rate": 2.9082e-05, + "loss": 0.0302, + "step": 9698 + }, + { + "epoch": 8.191722972972974, + "grad_norm": 0.5176273584365845, + "learning_rate": 2.9085e-05, + "loss": 0.0383, + "step": 9699 + }, + { + "epoch": 8.192567567567568, + "grad_norm": 0.31313514709472656, + "learning_rate": 2.9088e-05, + "loss": 0.0347, + "step": 9700 + }, + { + "epoch": 8.193412162162161, + "grad_norm": 0.32633939385414124, + "learning_rate": 2.9091e-05, + "loss": 0.0225, + "step": 9701 + }, + { + "epoch": 8.194256756756756, + "grad_norm": 0.32289111614227295, + "learning_rate": 2.9094e-05, + "loss": 0.0231, + "step": 9702 + }, + { + "epoch": 8.19510135135135, + "grad_norm": 0.4336199164390564, + "learning_rate": 2.9097e-05, + "loss": 0.0503, + "step": 9703 + }, + { + "epoch": 8.195945945945946, + "grad_norm": 0.46969136595726013, + "learning_rate": 2.91e-05, + "loss": 0.0437, + "step": 9704 + }, + { + "epoch": 8.19679054054054, + "grad_norm": 0.45433783531188965, + "learning_rate": 2.9103e-05, + "loss": 0.0479, + "step": 9705 + }, + { + "epoch": 8.197635135135135, + "grad_norm": 0.4223666787147522, + "learning_rate": 2.9106e-05, + "loss": 0.0367, + "step": 9706 + }, + { + "epoch": 8.19847972972973, + "grad_norm": 0.6665780544281006, + "learning_rate": 2.9109000000000002e-05, + "loss": 0.0353, + "step": 9707 + }, + { + "epoch": 8.199324324324325, + "grad_norm": 0.3781251907348633, + "learning_rate": 2.9112000000000002e-05, + "loss": 0.0275, + "step": 9708 + }, + { + "epoch": 8.20016891891892, + "grad_norm": 0.36531317234039307, + "learning_rate": 2.9115e-05, + "loss": 0.0337, + "step": 9709 + }, + { + "epoch": 8.201013513513514, + "grad_norm": 0.27690815925598145, + "learning_rate": 2.9118e-05, + "loss": 0.023, + "step": 9710 + }, + { + "epoch": 8.201858108108109, + "grad_norm": 0.4570677876472473, + "learning_rate": 2.9121e-05, + "loss": 0.0297, + "step": 9711 + }, + { + "epoch": 8.202702702702704, + "grad_norm": 0.5759666562080383, + "learning_rate": 2.9124e-05, + "loss": 0.0359, + "step": 9712 + }, + { + "epoch": 8.203547297297296, + "grad_norm": 0.6686625480651855, + "learning_rate": 2.9127e-05, + "loss": 0.0392, + "step": 9713 + }, + { + "epoch": 8.204391891891891, + "grad_norm": 0.3000946640968323, + "learning_rate": 2.913e-05, + "loss": 0.0321, + "step": 9714 + }, + { + "epoch": 8.205236486486486, + "grad_norm": 0.585986316204071, + "learning_rate": 2.9133e-05, + "loss": 0.037, + "step": 9715 + }, + { + "epoch": 8.20608108108108, + "grad_norm": 0.3357499837875366, + "learning_rate": 2.9136e-05, + "loss": 0.0341, + "step": 9716 + }, + { + "epoch": 8.206925675675675, + "grad_norm": 0.5039554238319397, + "learning_rate": 2.9139000000000003e-05, + "loss": 0.021, + "step": 9717 + }, + { + "epoch": 8.20777027027027, + "grad_norm": 1.2997440099716187, + "learning_rate": 2.9142000000000003e-05, + "loss": 0.0728, + "step": 9718 + }, + { + "epoch": 8.208614864864865, + "grad_norm": 0.8108497858047485, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0614, + "step": 9719 + }, + { + "epoch": 8.20945945945946, + "grad_norm": 0.904661238193512, + "learning_rate": 2.9148000000000003e-05, + "loss": 0.0397, + "step": 9720 + }, + { + "epoch": 8.210304054054054, + "grad_norm": 0.6079152226448059, + "learning_rate": 2.9151000000000003e-05, + "loss": 0.046, + "step": 9721 + }, + { + "epoch": 8.21114864864865, + "grad_norm": 0.8249703645706177, + "learning_rate": 2.9154e-05, + "loss": 0.0481, + "step": 9722 + }, + { + "epoch": 8.211993243243244, + "grad_norm": 0.8583188056945801, + "learning_rate": 2.9157e-05, + "loss": 0.2872, + "step": 9723 + }, + { + "epoch": 8.212837837837839, + "grad_norm": 0.6032621264457703, + "learning_rate": 2.916e-05, + "loss": 0.177, + "step": 9724 + }, + { + "epoch": 8.213682432432432, + "grad_norm": 1.3349742889404297, + "learning_rate": 2.9163e-05, + "loss": 0.1191, + "step": 9725 + }, + { + "epoch": 8.214527027027026, + "grad_norm": 0.3401716947555542, + "learning_rate": 2.9165999999999998e-05, + "loss": 0.0604, + "step": 9726 + }, + { + "epoch": 8.215371621621621, + "grad_norm": 0.36163315176963806, + "learning_rate": 2.9169e-05, + "loss": 0.0446, + "step": 9727 + }, + { + "epoch": 8.216216216216216, + "grad_norm": 0.37966933846473694, + "learning_rate": 2.9172e-05, + "loss": 0.0517, + "step": 9728 + }, + { + "epoch": 8.21706081081081, + "grad_norm": 0.8439114689826965, + "learning_rate": 2.9175e-05, + "loss": 0.0604, + "step": 9729 + }, + { + "epoch": 8.217905405405405, + "grad_norm": 0.4136158227920532, + "learning_rate": 2.9178e-05, + "loss": 0.0317, + "step": 9730 + }, + { + "epoch": 8.21875, + "grad_norm": 0.3385426700115204, + "learning_rate": 2.9181e-05, + "loss": 0.042, + "step": 9731 + }, + { + "epoch": 8.219594594594595, + "grad_norm": 0.45475801825523376, + "learning_rate": 2.9184e-05, + "loss": 0.0266, + "step": 9732 + }, + { + "epoch": 8.22043918918919, + "grad_norm": 0.33454012870788574, + "learning_rate": 2.9187e-05, + "loss": 0.0309, + "step": 9733 + }, + { + "epoch": 8.221283783783784, + "grad_norm": 0.3180294930934906, + "learning_rate": 2.919e-05, + "loss": 0.0406, + "step": 9734 + }, + { + "epoch": 8.222128378378379, + "grad_norm": 0.43576785922050476, + "learning_rate": 2.9193e-05, + "loss": 0.0311, + "step": 9735 + }, + { + "epoch": 8.222972972972974, + "grad_norm": 0.4488300681114197, + "learning_rate": 2.9196e-05, + "loss": 0.0424, + "step": 9736 + }, + { + "epoch": 8.223817567567568, + "grad_norm": 0.3022879660129547, + "learning_rate": 2.9199000000000003e-05, + "loss": 0.0311, + "step": 9737 + }, + { + "epoch": 8.224662162162161, + "grad_norm": 0.32453998923301697, + "learning_rate": 2.9202000000000003e-05, + "loss": 0.0239, + "step": 9738 + }, + { + "epoch": 8.225506756756756, + "grad_norm": 0.5356419682502747, + "learning_rate": 2.9205000000000002e-05, + "loss": 0.0331, + "step": 9739 + }, + { + "epoch": 8.22635135135135, + "grad_norm": 0.402974933385849, + "learning_rate": 2.9208000000000002e-05, + "loss": 0.0295, + "step": 9740 + }, + { + "epoch": 8.227195945945946, + "grad_norm": 0.3410854637622833, + "learning_rate": 2.9211000000000002e-05, + "loss": 0.0167, + "step": 9741 + }, + { + "epoch": 8.22804054054054, + "grad_norm": 0.459689199924469, + "learning_rate": 2.9214000000000002e-05, + "loss": 0.0381, + "step": 9742 + }, + { + "epoch": 8.228885135135135, + "grad_norm": 0.6475565433502197, + "learning_rate": 2.9217e-05, + "loss": 0.0275, + "step": 9743 + }, + { + "epoch": 8.22972972972973, + "grad_norm": 0.3378433585166931, + "learning_rate": 2.922e-05, + "loss": 0.0418, + "step": 9744 + }, + { + "epoch": 8.230574324324325, + "grad_norm": 0.470692902803421, + "learning_rate": 2.9223e-05, + "loss": 0.0415, + "step": 9745 + }, + { + "epoch": 8.23141891891892, + "grad_norm": 0.547380805015564, + "learning_rate": 2.9226e-05, + "loss": 0.0383, + "step": 9746 + }, + { + "epoch": 8.232263513513514, + "grad_norm": 0.4745738208293915, + "learning_rate": 2.9229e-05, + "loss": 0.0785, + "step": 9747 + }, + { + "epoch": 8.233108108108109, + "grad_norm": 0.7414456605911255, + "learning_rate": 2.9232e-05, + "loss": 0.0386, + "step": 9748 + }, + { + "epoch": 8.233952702702704, + "grad_norm": 0.5866261720657349, + "learning_rate": 2.9235e-05, + "loss": 0.0371, + "step": 9749 + }, + { + "epoch": 8.234797297297296, + "grad_norm": 0.3335975706577301, + "learning_rate": 2.9238e-05, + "loss": 0.0252, + "step": 9750 + }, + { + "epoch": 8.235641891891891, + "grad_norm": 0.3061463236808777, + "learning_rate": 2.9241e-05, + "loss": 0.0277, + "step": 9751 + }, + { + "epoch": 8.236486486486486, + "grad_norm": 0.6276715397834778, + "learning_rate": 2.9244e-05, + "loss": 0.0411, + "step": 9752 + }, + { + "epoch": 8.23733108108108, + "grad_norm": 0.3689601421356201, + "learning_rate": 2.9247e-05, + "loss": 0.0257, + "step": 9753 + }, + { + "epoch": 8.238175675675675, + "grad_norm": 0.40760278701782227, + "learning_rate": 2.925e-05, + "loss": 0.0277, + "step": 9754 + }, + { + "epoch": 8.23902027027027, + "grad_norm": 3.2779183387756348, + "learning_rate": 2.9253e-05, + "loss": 0.0775, + "step": 9755 + }, + { + "epoch": 8.239864864864865, + "grad_norm": 0.6416897177696228, + "learning_rate": 2.9256e-05, + "loss": 0.0383, + "step": 9756 + }, + { + "epoch": 8.24070945945946, + "grad_norm": 0.42753922939300537, + "learning_rate": 2.9259e-05, + "loss": 0.0321, + "step": 9757 + }, + { + "epoch": 8.241554054054054, + "grad_norm": 0.5417196750640869, + "learning_rate": 2.9262000000000002e-05, + "loss": 0.0225, + "step": 9758 + }, + { + "epoch": 8.24239864864865, + "grad_norm": 0.7041162252426147, + "learning_rate": 2.9265000000000002e-05, + "loss": 0.0959, + "step": 9759 + }, + { + "epoch": 8.243243243243244, + "grad_norm": 0.3922455608844757, + "learning_rate": 2.9268e-05, + "loss": 0.0286, + "step": 9760 + }, + { + "epoch": 8.244087837837839, + "grad_norm": 1.6996924877166748, + "learning_rate": 2.9271e-05, + "loss": 0.0521, + "step": 9761 + }, + { + "epoch": 8.244932432432432, + "grad_norm": 0.3609141707420349, + "learning_rate": 2.9274e-05, + "loss": 0.0409, + "step": 9762 + }, + { + "epoch": 8.245777027027026, + "grad_norm": 0.30127671360969543, + "learning_rate": 2.9277e-05, + "loss": 0.0262, + "step": 9763 + }, + { + "epoch": 8.246621621621621, + "grad_norm": 0.4844679534435272, + "learning_rate": 2.928e-05, + "loss": 0.0382, + "step": 9764 + }, + { + "epoch": 8.247466216216216, + "grad_norm": 0.4636600613594055, + "learning_rate": 2.9283e-05, + "loss": 0.0334, + "step": 9765 + }, + { + "epoch": 8.24831081081081, + "grad_norm": 0.39676451683044434, + "learning_rate": 2.9286e-05, + "loss": 0.0321, + "step": 9766 + }, + { + "epoch": 8.249155405405405, + "grad_norm": 0.670590341091156, + "learning_rate": 2.9289e-05, + "loss": 0.0386, + "step": 9767 + }, + { + "epoch": 8.25, + "grad_norm": 0.47093474864959717, + "learning_rate": 2.9292000000000003e-05, + "loss": 0.045, + "step": 9768 + }, + { + "epoch": 8.250844594594595, + "grad_norm": 0.8967562317848206, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0449, + "step": 9769 + }, + { + "epoch": 8.25168918918919, + "grad_norm": 0.4564213752746582, + "learning_rate": 2.9298000000000003e-05, + "loss": 0.037, + "step": 9770 + }, + { + "epoch": 8.252533783783784, + "grad_norm": 1.0561221837997437, + "learning_rate": 2.9301e-05, + "loss": 0.0557, + "step": 9771 + }, + { + "epoch": 8.253378378378379, + "grad_norm": 0.6937546730041504, + "learning_rate": 2.9304e-05, + "loss": 0.0628, + "step": 9772 + }, + { + "epoch": 8.254222972972974, + "grad_norm": 1.3350815773010254, + "learning_rate": 2.9307e-05, + "loss": 0.2774, + "step": 9773 + }, + { + "epoch": 8.255067567567568, + "grad_norm": 0.6993722319602966, + "learning_rate": 2.931e-05, + "loss": 0.2282, + "step": 9774 + }, + { + "epoch": 8.255912162162161, + "grad_norm": 0.7794854044914246, + "learning_rate": 2.9313e-05, + "loss": 0.099, + "step": 9775 + }, + { + "epoch": 8.256756756756756, + "grad_norm": 0.38608622550964355, + "learning_rate": 2.9316e-05, + "loss": 0.0405, + "step": 9776 + }, + { + "epoch": 8.25760135135135, + "grad_norm": 0.400599867105484, + "learning_rate": 2.9318999999999998e-05, + "loss": 0.0309, + "step": 9777 + }, + { + "epoch": 8.258445945945946, + "grad_norm": 0.4561619758605957, + "learning_rate": 2.9322e-05, + "loss": 0.051, + "step": 9778 + }, + { + "epoch": 8.25929054054054, + "grad_norm": 0.37245115637779236, + "learning_rate": 2.9325e-05, + "loss": 0.0347, + "step": 9779 + }, + { + "epoch": 8.260135135135135, + "grad_norm": 0.34977787733078003, + "learning_rate": 2.9328e-05, + "loss": 0.0304, + "step": 9780 + }, + { + "epoch": 8.26097972972973, + "grad_norm": 0.2999531030654907, + "learning_rate": 2.9331e-05, + "loss": 0.0343, + "step": 9781 + }, + { + "epoch": 8.261824324324325, + "grad_norm": 0.565119743347168, + "learning_rate": 2.9334e-05, + "loss": 0.04, + "step": 9782 + }, + { + "epoch": 8.26266891891892, + "grad_norm": 0.544462263584137, + "learning_rate": 2.9337e-05, + "loss": 0.0316, + "step": 9783 + }, + { + "epoch": 8.263513513513514, + "grad_norm": 0.2673066556453705, + "learning_rate": 2.934e-05, + "loss": 0.0252, + "step": 9784 + }, + { + "epoch": 8.264358108108109, + "grad_norm": 0.6138606667518616, + "learning_rate": 2.9343e-05, + "loss": 0.0366, + "step": 9785 + }, + { + "epoch": 8.265202702702704, + "grad_norm": 0.42304378747940063, + "learning_rate": 2.9346e-05, + "loss": 0.0235, + "step": 9786 + }, + { + "epoch": 8.266047297297296, + "grad_norm": 0.4394608438014984, + "learning_rate": 2.9349e-05, + "loss": 0.0344, + "step": 9787 + }, + { + "epoch": 8.266891891891891, + "grad_norm": 0.23241747915744781, + "learning_rate": 2.9352000000000003e-05, + "loss": 0.0151, + "step": 9788 + }, + { + "epoch": 8.267736486486486, + "grad_norm": 0.3068475127220154, + "learning_rate": 2.9355000000000003e-05, + "loss": 0.0378, + "step": 9789 + }, + { + "epoch": 8.26858108108108, + "grad_norm": 0.49834731221199036, + "learning_rate": 2.9358000000000003e-05, + "loss": 0.0292, + "step": 9790 + }, + { + "epoch": 8.269425675675675, + "grad_norm": 0.3850211501121521, + "learning_rate": 2.9361000000000002e-05, + "loss": 0.0296, + "step": 9791 + }, + { + "epoch": 8.27027027027027, + "grad_norm": 0.3472522497177124, + "learning_rate": 2.9364000000000002e-05, + "loss": 0.0337, + "step": 9792 + }, + { + "epoch": 8.271114864864865, + "grad_norm": 0.330323725938797, + "learning_rate": 2.9367000000000002e-05, + "loss": 0.022, + "step": 9793 + }, + { + "epoch": 8.27195945945946, + "grad_norm": 0.31690433621406555, + "learning_rate": 2.9370000000000002e-05, + "loss": 0.0311, + "step": 9794 + }, + { + "epoch": 8.272804054054054, + "grad_norm": 0.5096933245658875, + "learning_rate": 2.9373e-05, + "loss": 0.0423, + "step": 9795 + }, + { + "epoch": 8.27364864864865, + "grad_norm": 0.4114404022693634, + "learning_rate": 2.9375999999999998e-05, + "loss": 0.0299, + "step": 9796 + }, + { + "epoch": 8.274493243243244, + "grad_norm": 0.21966171264648438, + "learning_rate": 2.9378999999999998e-05, + "loss": 0.0163, + "step": 9797 + }, + { + "epoch": 8.275337837837839, + "grad_norm": 0.3663932681083679, + "learning_rate": 2.9382e-05, + "loss": 0.024, + "step": 9798 + }, + { + "epoch": 8.276182432432432, + "grad_norm": 0.3807730972766876, + "learning_rate": 2.9385e-05, + "loss": 0.0322, + "step": 9799 + }, + { + "epoch": 8.277027027027026, + "grad_norm": 0.28052371740341187, + "learning_rate": 2.9388e-05, + "loss": 0.0208, + "step": 9800 + }, + { + "epoch": 8.277871621621621, + "grad_norm": 0.4120447337627411, + "learning_rate": 2.9391e-05, + "loss": 0.0363, + "step": 9801 + }, + { + "epoch": 8.278716216216216, + "grad_norm": 0.46947023272514343, + "learning_rate": 2.9394e-05, + "loss": 0.0444, + "step": 9802 + }, + { + "epoch": 8.27956081081081, + "grad_norm": 0.34764334559440613, + "learning_rate": 2.9397e-05, + "loss": 0.0367, + "step": 9803 + }, + { + "epoch": 8.280405405405405, + "grad_norm": 0.44435837864875793, + "learning_rate": 2.94e-05, + "loss": 0.0491, + "step": 9804 + }, + { + "epoch": 8.28125, + "grad_norm": 0.5702604055404663, + "learning_rate": 2.9403e-05, + "loss": 0.0425, + "step": 9805 + }, + { + "epoch": 8.282094594594595, + "grad_norm": 0.5770161151885986, + "learning_rate": 2.9406e-05, + "loss": 0.0346, + "step": 9806 + }, + { + "epoch": 8.28293918918919, + "grad_norm": 0.254144549369812, + "learning_rate": 2.9409e-05, + "loss": 0.0174, + "step": 9807 + }, + { + "epoch": 8.283783783783784, + "grad_norm": 0.5328224897384644, + "learning_rate": 2.9412000000000002e-05, + "loss": 0.0896, + "step": 9808 + }, + { + "epoch": 8.284628378378379, + "grad_norm": 0.9048380255699158, + "learning_rate": 2.9415000000000002e-05, + "loss": 0.0562, + "step": 9809 + }, + { + "epoch": 8.285472972972974, + "grad_norm": 0.5539232492446899, + "learning_rate": 2.9418000000000002e-05, + "loss": 0.0239, + "step": 9810 + }, + { + "epoch": 8.286317567567568, + "grad_norm": 0.6666167378425598, + "learning_rate": 2.9421000000000002e-05, + "loss": 0.0422, + "step": 9811 + }, + { + "epoch": 8.287162162162161, + "grad_norm": 0.46793681383132935, + "learning_rate": 2.9424e-05, + "loss": 0.0339, + "step": 9812 + }, + { + "epoch": 8.288006756756756, + "grad_norm": 0.3879598379135132, + "learning_rate": 2.9427e-05, + "loss": 0.0269, + "step": 9813 + }, + { + "epoch": 8.28885135135135, + "grad_norm": 0.7699930667877197, + "learning_rate": 2.943e-05, + "loss": 0.0259, + "step": 9814 + }, + { + "epoch": 8.289695945945946, + "grad_norm": 0.6419495344161987, + "learning_rate": 2.9433e-05, + "loss": 0.0405, + "step": 9815 + }, + { + "epoch": 8.29054054054054, + "grad_norm": 0.6474230289459229, + "learning_rate": 2.9436e-05, + "loss": 0.0336, + "step": 9816 + }, + { + "epoch": 8.291385135135135, + "grad_norm": 0.5203482508659363, + "learning_rate": 2.9439e-05, + "loss": 0.0289, + "step": 9817 + }, + { + "epoch": 8.29222972972973, + "grad_norm": 0.42289024591445923, + "learning_rate": 2.9442000000000004e-05, + "loss": 0.0324, + "step": 9818 + }, + { + "epoch": 8.293074324324325, + "grad_norm": 0.5424103736877441, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.04, + "step": 9819 + }, + { + "epoch": 8.29391891891892, + "grad_norm": 0.6042732000350952, + "learning_rate": 2.9448e-05, + "loss": 0.0778, + "step": 9820 + }, + { + "epoch": 8.294763513513514, + "grad_norm": 0.6987577676773071, + "learning_rate": 2.9451e-05, + "loss": 0.0606, + "step": 9821 + }, + { + "epoch": 8.295608108108109, + "grad_norm": 1.42155122756958, + "learning_rate": 2.9454e-05, + "loss": 0.079, + "step": 9822 + }, + { + "epoch": 8.296452702702704, + "grad_norm": 1.3829692602157593, + "learning_rate": 2.9457e-05, + "loss": 0.343, + "step": 9823 + }, + { + "epoch": 8.297297297297296, + "grad_norm": 0.9336186051368713, + "learning_rate": 2.946e-05, + "loss": 0.2221, + "step": 9824 + }, + { + "epoch": 8.298141891891891, + "grad_norm": 0.42993393540382385, + "learning_rate": 2.9463e-05, + "loss": 0.0761, + "step": 9825 + }, + { + "epoch": 8.298986486486486, + "grad_norm": 0.40489184856414795, + "learning_rate": 2.9466e-05, + "loss": 0.0679, + "step": 9826 + }, + { + "epoch": 8.29983108108108, + "grad_norm": 0.3699130117893219, + "learning_rate": 2.9469e-05, + "loss": 0.0574, + "step": 9827 + }, + { + "epoch": 8.300675675675675, + "grad_norm": 0.4761160910129547, + "learning_rate": 2.9472000000000002e-05, + "loss": 0.0453, + "step": 9828 + }, + { + "epoch": 8.30152027027027, + "grad_norm": 0.3820236325263977, + "learning_rate": 2.9475e-05, + "loss": 0.0443, + "step": 9829 + }, + { + "epoch": 8.302364864864865, + "grad_norm": 0.32792556285858154, + "learning_rate": 2.9478e-05, + "loss": 0.0356, + "step": 9830 + }, + { + "epoch": 8.30320945945946, + "grad_norm": 0.5080179572105408, + "learning_rate": 2.9481e-05, + "loss": 0.06, + "step": 9831 + }, + { + "epoch": 8.304054054054054, + "grad_norm": 0.3684707283973694, + "learning_rate": 2.9484e-05, + "loss": 0.0427, + "step": 9832 + }, + { + "epoch": 8.30489864864865, + "grad_norm": 0.4055511951446533, + "learning_rate": 2.9487e-05, + "loss": 0.037, + "step": 9833 + }, + { + "epoch": 8.305743243243244, + "grad_norm": 0.47854748368263245, + "learning_rate": 2.949e-05, + "loss": 0.0419, + "step": 9834 + }, + { + "epoch": 8.306587837837839, + "grad_norm": 0.2538383901119232, + "learning_rate": 2.9493e-05, + "loss": 0.0194, + "step": 9835 + }, + { + "epoch": 8.307432432432432, + "grad_norm": 0.37971314787864685, + "learning_rate": 2.9496e-05, + "loss": 0.0377, + "step": 9836 + }, + { + "epoch": 8.308277027027026, + "grad_norm": 0.32507190108299255, + "learning_rate": 2.9499e-05, + "loss": 0.0424, + "step": 9837 + }, + { + "epoch": 8.309121621621621, + "grad_norm": 0.4851822257041931, + "learning_rate": 2.9502000000000003e-05, + "loss": 0.039, + "step": 9838 + }, + { + "epoch": 8.309966216216216, + "grad_norm": 0.5900083184242249, + "learning_rate": 2.9505000000000003e-05, + "loss": 0.0516, + "step": 9839 + }, + { + "epoch": 8.31081081081081, + "grad_norm": 0.382319837808609, + "learning_rate": 2.9508000000000003e-05, + "loss": 0.0345, + "step": 9840 + }, + { + "epoch": 8.311655405405405, + "grad_norm": 0.31102603673934937, + "learning_rate": 2.9511000000000003e-05, + "loss": 0.0189, + "step": 9841 + }, + { + "epoch": 8.3125, + "grad_norm": 0.2532280683517456, + "learning_rate": 2.9514000000000002e-05, + "loss": 0.0269, + "step": 9842 + }, + { + "epoch": 8.313344594594595, + "grad_norm": 0.5110490918159485, + "learning_rate": 2.9517000000000002e-05, + "loss": 0.0398, + "step": 9843 + }, + { + "epoch": 8.31418918918919, + "grad_norm": 0.2816256284713745, + "learning_rate": 2.9520000000000002e-05, + "loss": 0.0184, + "step": 9844 + }, + { + "epoch": 8.315033783783784, + "grad_norm": 0.46660786867141724, + "learning_rate": 2.9523e-05, + "loss": 0.045, + "step": 9845 + }, + { + "epoch": 8.315878378378379, + "grad_norm": 0.4987873136997223, + "learning_rate": 2.9525999999999998e-05, + "loss": 0.0431, + "step": 9846 + }, + { + "epoch": 8.316722972972974, + "grad_norm": 0.566863477230072, + "learning_rate": 2.9528999999999998e-05, + "loss": 0.0451, + "step": 9847 + }, + { + "epoch": 8.317567567567568, + "grad_norm": 0.3397541642189026, + "learning_rate": 2.9532e-05, + "loss": 0.0225, + "step": 9848 + }, + { + "epoch": 8.318412162162161, + "grad_norm": 0.3376161754131317, + "learning_rate": 2.9535e-05, + "loss": 0.0474, + "step": 9849 + }, + { + "epoch": 8.319256756756756, + "grad_norm": 0.4257998764514923, + "learning_rate": 2.9538e-05, + "loss": 0.0327, + "step": 9850 + }, + { + "epoch": 8.32010135135135, + "grad_norm": 0.37011510133743286, + "learning_rate": 2.9541e-05, + "loss": 0.0321, + "step": 9851 + }, + { + "epoch": 8.320945945945946, + "grad_norm": 0.2526995837688446, + "learning_rate": 2.9544e-05, + "loss": 0.0223, + "step": 9852 + }, + { + "epoch": 8.32179054054054, + "grad_norm": 0.4132858216762543, + "learning_rate": 2.9547e-05, + "loss": 0.0373, + "step": 9853 + }, + { + "epoch": 8.322635135135135, + "grad_norm": 0.4617381691932678, + "learning_rate": 2.955e-05, + "loss": 0.0442, + "step": 9854 + }, + { + "epoch": 8.32347972972973, + "grad_norm": 0.3236894905567169, + "learning_rate": 2.9553e-05, + "loss": 0.0389, + "step": 9855 + }, + { + "epoch": 8.324324324324325, + "grad_norm": 0.48247823119163513, + "learning_rate": 2.9556e-05, + "loss": 0.0437, + "step": 9856 + }, + { + "epoch": 8.32516891891892, + "grad_norm": 0.4016946852207184, + "learning_rate": 2.9559e-05, + "loss": 0.0407, + "step": 9857 + }, + { + "epoch": 8.326013513513514, + "grad_norm": 0.3417372703552246, + "learning_rate": 2.9562000000000003e-05, + "loss": 0.026, + "step": 9858 + }, + { + "epoch": 8.326858108108109, + "grad_norm": 0.39974483847618103, + "learning_rate": 2.9565000000000002e-05, + "loss": 0.0308, + "step": 9859 + }, + { + "epoch": 8.327702702702704, + "grad_norm": 0.47208577394485474, + "learning_rate": 2.9568000000000002e-05, + "loss": 0.034, + "step": 9860 + }, + { + "epoch": 8.328547297297296, + "grad_norm": 0.34635770320892334, + "learning_rate": 2.9571000000000002e-05, + "loss": 0.0327, + "step": 9861 + }, + { + "epoch": 8.329391891891891, + "grad_norm": 0.3472854793071747, + "learning_rate": 2.9574000000000002e-05, + "loss": 0.0358, + "step": 9862 + }, + { + "epoch": 8.330236486486486, + "grad_norm": 0.5433049201965332, + "learning_rate": 2.9577e-05, + "loss": 0.0236, + "step": 9863 + }, + { + "epoch": 8.33108108108108, + "grad_norm": 0.44350606203079224, + "learning_rate": 2.958e-05, + "loss": 0.0444, + "step": 9864 + }, + { + "epoch": 8.331925675675675, + "grad_norm": 0.456950306892395, + "learning_rate": 2.9583e-05, + "loss": 0.0389, + "step": 9865 + }, + { + "epoch": 8.33277027027027, + "grad_norm": 0.8517313599586487, + "learning_rate": 2.9586e-05, + "loss": 0.0409, + "step": 9866 + }, + { + "epoch": 8.333614864864865, + "grad_norm": 1.793155312538147, + "learning_rate": 2.9589e-05, + "loss": 0.0586, + "step": 9867 + }, + { + "epoch": 8.33445945945946, + "grad_norm": 0.5116609930992126, + "learning_rate": 2.9592000000000004e-05, + "loss": 0.0282, + "step": 9868 + }, + { + "epoch": 8.335304054054054, + "grad_norm": 0.39489880204200745, + "learning_rate": 2.9595e-05, + "loss": 0.0338, + "step": 9869 + }, + { + "epoch": 8.33614864864865, + "grad_norm": 0.4733010530471802, + "learning_rate": 2.9598e-05, + "loss": 0.0604, + "step": 9870 + }, + { + "epoch": 8.336993243243244, + "grad_norm": 0.8821116089820862, + "learning_rate": 2.9601e-05, + "loss": 0.0609, + "step": 9871 + }, + { + "epoch": 8.337837837837839, + "grad_norm": 0.9844779372215271, + "learning_rate": 2.9604e-05, + "loss": 0.058, + "step": 9872 + }, + { + "epoch": 8.338682432432432, + "grad_norm": 1.2828439474105835, + "learning_rate": 2.9607e-05, + "loss": 0.3205, + "step": 9873 + }, + { + "epoch": 8.339527027027026, + "grad_norm": 1.2625304460525513, + "learning_rate": 2.961e-05, + "loss": 0.2201, + "step": 9874 + }, + { + "epoch": 8.340371621621621, + "grad_norm": 0.4314119517803192, + "learning_rate": 2.9613e-05, + "loss": 0.088, + "step": 9875 + }, + { + "epoch": 8.341216216216216, + "grad_norm": 0.3751532733440399, + "learning_rate": 2.9616e-05, + "loss": 0.0358, + "step": 9876 + }, + { + "epoch": 8.34206081081081, + "grad_norm": 0.45490899682044983, + "learning_rate": 2.9619e-05, + "loss": 0.0635, + "step": 9877 + }, + { + "epoch": 8.342905405405405, + "grad_norm": 0.738350510597229, + "learning_rate": 2.9622000000000002e-05, + "loss": 0.0749, + "step": 9878 + }, + { + "epoch": 8.34375, + "grad_norm": 0.4444716274738312, + "learning_rate": 2.9625000000000002e-05, + "loss": 0.0584, + "step": 9879 + }, + { + "epoch": 8.344594594594595, + "grad_norm": 0.7976069450378418, + "learning_rate": 2.9628e-05, + "loss": 0.0385, + "step": 9880 + }, + { + "epoch": 8.34543918918919, + "grad_norm": 0.3304402828216553, + "learning_rate": 2.9631e-05, + "loss": 0.0287, + "step": 9881 + }, + { + "epoch": 8.346283783783784, + "grad_norm": 0.5108516216278076, + "learning_rate": 2.9634e-05, + "loss": 0.0411, + "step": 9882 + }, + { + "epoch": 8.347128378378379, + "grad_norm": 0.36118578910827637, + "learning_rate": 2.9637e-05, + "loss": 0.0296, + "step": 9883 + }, + { + "epoch": 8.347972972972974, + "grad_norm": 0.3087237775325775, + "learning_rate": 2.964e-05, + "loss": 0.0276, + "step": 9884 + }, + { + "epoch": 8.348817567567568, + "grad_norm": 0.4988625645637512, + "learning_rate": 2.9643e-05, + "loss": 0.0291, + "step": 9885 + }, + { + "epoch": 8.349662162162161, + "grad_norm": 0.34821751713752747, + "learning_rate": 2.9646e-05, + "loss": 0.043, + "step": 9886 + }, + { + "epoch": 8.350506756756756, + "grad_norm": 0.4055544137954712, + "learning_rate": 2.9649e-05, + "loss": 0.0332, + "step": 9887 + }, + { + "epoch": 8.35135135135135, + "grad_norm": 0.5708906650543213, + "learning_rate": 2.9652e-05, + "loss": 0.0274, + "step": 9888 + }, + { + "epoch": 8.352195945945946, + "grad_norm": 0.4687063694000244, + "learning_rate": 2.9655000000000003e-05, + "loss": 0.0483, + "step": 9889 + }, + { + "epoch": 8.35304054054054, + "grad_norm": 0.5566892027854919, + "learning_rate": 2.9658000000000003e-05, + "loss": 0.0368, + "step": 9890 + }, + { + "epoch": 8.353885135135135, + "grad_norm": 0.35155561566352844, + "learning_rate": 2.9661000000000003e-05, + "loss": 0.0356, + "step": 9891 + }, + { + "epoch": 8.35472972972973, + "grad_norm": 0.3713681399822235, + "learning_rate": 2.9664000000000003e-05, + "loss": 0.0297, + "step": 9892 + }, + { + "epoch": 8.355574324324325, + "grad_norm": 0.3116985261440277, + "learning_rate": 2.9667000000000002e-05, + "loss": 0.0289, + "step": 9893 + }, + { + "epoch": 8.35641891891892, + "grad_norm": 0.49059081077575684, + "learning_rate": 2.967e-05, + "loss": 0.0372, + "step": 9894 + }, + { + "epoch": 8.357263513513514, + "grad_norm": 0.5969400405883789, + "learning_rate": 2.9673e-05, + "loss": 0.0368, + "step": 9895 + }, + { + "epoch": 8.358108108108109, + "grad_norm": 0.41702762246131897, + "learning_rate": 2.9676e-05, + "loss": 0.0305, + "step": 9896 + }, + { + "epoch": 8.358952702702704, + "grad_norm": 0.4342298209667206, + "learning_rate": 2.9678999999999998e-05, + "loss": 0.0238, + "step": 9897 + }, + { + "epoch": 8.359797297297296, + "grad_norm": 0.32917293906211853, + "learning_rate": 2.9681999999999998e-05, + "loss": 0.0282, + "step": 9898 + }, + { + "epoch": 8.360641891891891, + "grad_norm": 0.27750328183174133, + "learning_rate": 2.9685e-05, + "loss": 0.0148, + "step": 9899 + }, + { + "epoch": 8.361486486486486, + "grad_norm": 0.2877489924430847, + "learning_rate": 2.9688e-05, + "loss": 0.0256, + "step": 9900 + }, + { + "epoch": 8.36233108108108, + "grad_norm": 0.45888036489486694, + "learning_rate": 2.9691e-05, + "loss": 0.0475, + "step": 9901 + }, + { + "epoch": 8.363175675675675, + "grad_norm": 0.405475378036499, + "learning_rate": 2.9694e-05, + "loss": 0.0424, + "step": 9902 + }, + { + "epoch": 8.36402027027027, + "grad_norm": 0.2703940272331238, + "learning_rate": 2.9697e-05, + "loss": 0.0277, + "step": 9903 + }, + { + "epoch": 8.364864864864865, + "grad_norm": 0.38473498821258545, + "learning_rate": 2.97e-05, + "loss": 0.0372, + "step": 9904 + }, + { + "epoch": 8.36570945945946, + "grad_norm": 0.46982747316360474, + "learning_rate": 2.9703e-05, + "loss": 0.019, + "step": 9905 + }, + { + "epoch": 8.366554054054054, + "grad_norm": 0.36023956537246704, + "learning_rate": 2.9706e-05, + "loss": 0.036, + "step": 9906 + }, + { + "epoch": 8.36739864864865, + "grad_norm": 0.3059995174407959, + "learning_rate": 2.9709e-05, + "loss": 0.0281, + "step": 9907 + }, + { + "epoch": 8.368243243243244, + "grad_norm": 0.48090067505836487, + "learning_rate": 2.9712e-05, + "loss": 0.0321, + "step": 9908 + }, + { + "epoch": 8.369087837837839, + "grad_norm": 0.421232134103775, + "learning_rate": 2.9715000000000003e-05, + "loss": 0.0289, + "step": 9909 + }, + { + "epoch": 8.369932432432432, + "grad_norm": 0.456407755613327, + "learning_rate": 2.9718000000000002e-05, + "loss": 0.0267, + "step": 9910 + }, + { + "epoch": 8.370777027027026, + "grad_norm": 0.3456100523471832, + "learning_rate": 2.9721000000000002e-05, + "loss": 0.0371, + "step": 9911 + }, + { + "epoch": 8.371621621621621, + "grad_norm": 0.5970425605773926, + "learning_rate": 2.9724000000000002e-05, + "loss": 0.0425, + "step": 9912 + }, + { + "epoch": 8.372466216216216, + "grad_norm": 0.4618177115917206, + "learning_rate": 2.9727000000000002e-05, + "loss": 0.0353, + "step": 9913 + }, + { + "epoch": 8.37331081081081, + "grad_norm": 0.5057112574577332, + "learning_rate": 2.973e-05, + "loss": 0.038, + "step": 9914 + }, + { + "epoch": 8.374155405405405, + "grad_norm": 0.6182736158370972, + "learning_rate": 2.9733e-05, + "loss": 0.0417, + "step": 9915 + }, + { + "epoch": 8.375, + "grad_norm": 0.9820671081542969, + "learning_rate": 2.9736e-05, + "loss": 0.0361, + "step": 9916 + }, + { + "epoch": 8.375844594594595, + "grad_norm": 0.40556037425994873, + "learning_rate": 2.9739e-05, + "loss": 0.0631, + "step": 9917 + }, + { + "epoch": 8.37668918918919, + "grad_norm": 0.5494636297225952, + "learning_rate": 2.9742e-05, + "loss": 0.0359, + "step": 9918 + }, + { + "epoch": 8.377533783783784, + "grad_norm": 0.8524896502494812, + "learning_rate": 2.9745e-05, + "loss": 0.0319, + "step": 9919 + }, + { + "epoch": 8.378378378378379, + "grad_norm": 0.5436961650848389, + "learning_rate": 2.9748e-05, + "loss": 0.0362, + "step": 9920 + }, + { + "epoch": 8.379222972972974, + "grad_norm": 0.4859200417995453, + "learning_rate": 2.9751e-05, + "loss": 0.0457, + "step": 9921 + }, + { + "epoch": 8.380067567567568, + "grad_norm": 0.5098329782485962, + "learning_rate": 2.9754e-05, + "loss": 0.0537, + "step": 9922 + }, + { + "epoch": 8.380912162162161, + "grad_norm": 1.8447802066802979, + "learning_rate": 2.9757e-05, + "loss": 0.297, + "step": 9923 + }, + { + "epoch": 8.381756756756756, + "grad_norm": 0.868167519569397, + "learning_rate": 2.976e-05, + "loss": 0.2564, + "step": 9924 + }, + { + "epoch": 8.38260135135135, + "grad_norm": 0.6080814003944397, + "learning_rate": 2.9763e-05, + "loss": 0.1251, + "step": 9925 + }, + { + "epoch": 8.383445945945946, + "grad_norm": 0.5755114555358887, + "learning_rate": 2.9766e-05, + "loss": 0.0479, + "step": 9926 + }, + { + "epoch": 8.38429054054054, + "grad_norm": 0.5954191088676453, + "learning_rate": 2.9769e-05, + "loss": 0.0762, + "step": 9927 + }, + { + "epoch": 8.385135135135135, + "grad_norm": 0.5781902074813843, + "learning_rate": 2.9772e-05, + "loss": 0.0637, + "step": 9928 + }, + { + "epoch": 8.38597972972973, + "grad_norm": 0.5552752614021301, + "learning_rate": 2.9775000000000002e-05, + "loss": 0.0476, + "step": 9929 + }, + { + "epoch": 8.386824324324325, + "grad_norm": 0.318167120218277, + "learning_rate": 2.9778000000000002e-05, + "loss": 0.0277, + "step": 9930 + }, + { + "epoch": 8.38766891891892, + "grad_norm": 0.33250561356544495, + "learning_rate": 2.9781e-05, + "loss": 0.0347, + "step": 9931 + }, + { + "epoch": 8.388513513513514, + "grad_norm": 0.5378555059432983, + "learning_rate": 2.9784e-05, + "loss": 0.0519, + "step": 9932 + }, + { + "epoch": 8.389358108108109, + "grad_norm": 0.2648445665836334, + "learning_rate": 2.9787e-05, + "loss": 0.0296, + "step": 9933 + }, + { + "epoch": 8.390202702702704, + "grad_norm": 0.5732834339141846, + "learning_rate": 2.979e-05, + "loss": 0.0309, + "step": 9934 + }, + { + "epoch": 8.391047297297296, + "grad_norm": 0.29077422618865967, + "learning_rate": 2.9793e-05, + "loss": 0.0302, + "step": 9935 + }, + { + "epoch": 8.391891891891891, + "grad_norm": 0.43192148208618164, + "learning_rate": 2.9796e-05, + "loss": 0.0335, + "step": 9936 + }, + { + "epoch": 8.392736486486486, + "grad_norm": 0.7144744992256165, + "learning_rate": 2.9799e-05, + "loss": 0.0432, + "step": 9937 + }, + { + "epoch": 8.39358108108108, + "grad_norm": 0.7301540970802307, + "learning_rate": 2.9802e-05, + "loss": 0.0409, + "step": 9938 + }, + { + "epoch": 8.394425675675675, + "grad_norm": 0.33485904335975647, + "learning_rate": 2.9805000000000003e-05, + "loss": 0.047, + "step": 9939 + }, + { + "epoch": 8.39527027027027, + "grad_norm": 0.6945831179618835, + "learning_rate": 2.9808000000000003e-05, + "loss": 0.0527, + "step": 9940 + }, + { + "epoch": 8.396114864864865, + "grad_norm": 0.3472124934196472, + "learning_rate": 2.9811000000000003e-05, + "loss": 0.0219, + "step": 9941 + }, + { + "epoch": 8.39695945945946, + "grad_norm": 0.5216554999351501, + "learning_rate": 2.9814000000000003e-05, + "loss": 0.1019, + "step": 9942 + }, + { + "epoch": 8.397804054054054, + "grad_norm": 0.22635796666145325, + "learning_rate": 2.9817e-05, + "loss": 0.0192, + "step": 9943 + }, + { + "epoch": 8.39864864864865, + "grad_norm": 0.4139545261859894, + "learning_rate": 2.982e-05, + "loss": 0.0273, + "step": 9944 + }, + { + "epoch": 8.399493243243244, + "grad_norm": 0.5960346460342407, + "learning_rate": 2.9823e-05, + "loss": 0.0388, + "step": 9945 + }, + { + "epoch": 8.400337837837839, + "grad_norm": 0.4914974868297577, + "learning_rate": 2.9826e-05, + "loss": 0.0273, + "step": 9946 + }, + { + "epoch": 8.401182432432432, + "grad_norm": 0.34172022342681885, + "learning_rate": 2.9829e-05, + "loss": 0.0298, + "step": 9947 + }, + { + "epoch": 8.402027027027026, + "grad_norm": 0.5172134637832642, + "learning_rate": 2.9831999999999998e-05, + "loss": 0.0498, + "step": 9948 + }, + { + "epoch": 8.402871621621621, + "grad_norm": 0.615259051322937, + "learning_rate": 2.9835e-05, + "loss": 0.0605, + "step": 9949 + }, + { + "epoch": 8.403716216216216, + "grad_norm": 0.3562181890010834, + "learning_rate": 2.9838e-05, + "loss": 0.0283, + "step": 9950 + }, + { + "epoch": 8.40456081081081, + "grad_norm": 0.7224397659301758, + "learning_rate": 2.9841e-05, + "loss": 0.0566, + "step": 9951 + }, + { + "epoch": 8.405405405405405, + "grad_norm": 0.3770376443862915, + "learning_rate": 2.9844e-05, + "loss": 0.0326, + "step": 9952 + }, + { + "epoch": 8.40625, + "grad_norm": 0.2907335162162781, + "learning_rate": 2.9847e-05, + "loss": 0.019, + "step": 9953 + }, + { + "epoch": 8.407094594594595, + "grad_norm": 0.32401004433631897, + "learning_rate": 2.985e-05, + "loss": 0.0282, + "step": 9954 + }, + { + "epoch": 8.40793918918919, + "grad_norm": 0.5498228073120117, + "learning_rate": 2.9853e-05, + "loss": 0.0484, + "step": 9955 + }, + { + "epoch": 8.408783783783784, + "grad_norm": 0.3940450847148895, + "learning_rate": 2.9856e-05, + "loss": 0.0187, + "step": 9956 + }, + { + "epoch": 8.409628378378379, + "grad_norm": 0.35273298621177673, + "learning_rate": 2.9859e-05, + "loss": 0.0386, + "step": 9957 + }, + { + "epoch": 8.410472972972974, + "grad_norm": 0.49174800515174866, + "learning_rate": 2.9862e-05, + "loss": 0.0311, + "step": 9958 + }, + { + "epoch": 8.411317567567568, + "grad_norm": 0.4309266209602356, + "learning_rate": 2.9865000000000003e-05, + "loss": 0.0326, + "step": 9959 + }, + { + "epoch": 8.412162162162161, + "grad_norm": 0.39159438014030457, + "learning_rate": 2.9868000000000003e-05, + "loss": 0.0384, + "step": 9960 + }, + { + "epoch": 8.413006756756756, + "grad_norm": 0.41820836067199707, + "learning_rate": 2.9871000000000003e-05, + "loss": 0.0527, + "step": 9961 + }, + { + "epoch": 8.41385135135135, + "grad_norm": 0.31428617238998413, + "learning_rate": 2.9874000000000002e-05, + "loss": 0.0343, + "step": 9962 + }, + { + "epoch": 8.414695945945946, + "grad_norm": 0.5072184205055237, + "learning_rate": 2.9877000000000002e-05, + "loss": 0.0357, + "step": 9963 + }, + { + "epoch": 8.41554054054054, + "grad_norm": 0.6466109752655029, + "learning_rate": 2.9880000000000002e-05, + "loss": 0.0348, + "step": 9964 + }, + { + "epoch": 8.416385135135135, + "grad_norm": 0.6989620327949524, + "learning_rate": 2.9883000000000002e-05, + "loss": 0.0399, + "step": 9965 + }, + { + "epoch": 8.41722972972973, + "grad_norm": 0.7565966248512268, + "learning_rate": 2.9886e-05, + "loss": 0.0508, + "step": 9966 + }, + { + "epoch": 8.418074324324325, + "grad_norm": 0.4179445803165436, + "learning_rate": 2.9889e-05, + "loss": 0.0319, + "step": 9967 + }, + { + "epoch": 8.41891891891892, + "grad_norm": 0.47687238454818726, + "learning_rate": 2.9891999999999998e-05, + "loss": 0.0479, + "step": 9968 + }, + { + "epoch": 8.419763513513514, + "grad_norm": 0.7367896437644958, + "learning_rate": 2.9895e-05, + "loss": 0.0492, + "step": 9969 + }, + { + "epoch": 8.420608108108109, + "grad_norm": 0.4416072368621826, + "learning_rate": 2.9898e-05, + "loss": 0.05, + "step": 9970 + }, + { + "epoch": 8.421452702702704, + "grad_norm": 0.5292587876319885, + "learning_rate": 2.9901e-05, + "loss": 0.0493, + "step": 9971 + }, + { + "epoch": 8.422297297297296, + "grad_norm": 0.8420377969741821, + "learning_rate": 2.9904e-05, + "loss": 0.0578, + "step": 9972 + }, + { + "epoch": 8.423141891891891, + "grad_norm": 3.32173752784729, + "learning_rate": 2.9907e-05, + "loss": 0.6277, + "step": 9973 + }, + { + "epoch": 8.423986486486486, + "grad_norm": 0.9559699296951294, + "learning_rate": 2.991e-05, + "loss": 0.2035, + "step": 9974 + }, + { + "epoch": 8.42483108108108, + "grad_norm": 0.5780783295631409, + "learning_rate": 2.9913e-05, + "loss": 0.081, + "step": 9975 + }, + { + "epoch": 8.425675675675675, + "grad_norm": 0.3865586221218109, + "learning_rate": 2.9916e-05, + "loss": 0.0603, + "step": 9976 + }, + { + "epoch": 8.42652027027027, + "grad_norm": 0.757068932056427, + "learning_rate": 2.9919e-05, + "loss": 0.0653, + "step": 9977 + }, + { + "epoch": 8.427364864864865, + "grad_norm": 0.4629226326942444, + "learning_rate": 2.9922e-05, + "loss": 0.0543, + "step": 9978 + }, + { + "epoch": 8.42820945945946, + "grad_norm": 0.6453736424446106, + "learning_rate": 2.9925000000000002e-05, + "loss": 0.0766, + "step": 9979 + }, + { + "epoch": 8.429054054054054, + "grad_norm": 0.40532171726226807, + "learning_rate": 2.9928000000000002e-05, + "loss": 0.0427, + "step": 9980 + }, + { + "epoch": 8.42989864864865, + "grad_norm": 0.3089383542537689, + "learning_rate": 2.9931000000000002e-05, + "loss": 0.033, + "step": 9981 + }, + { + "epoch": 8.430743243243244, + "grad_norm": 0.2836035490036011, + "learning_rate": 2.9934000000000002e-05, + "loss": 0.0391, + "step": 9982 + }, + { + "epoch": 8.431587837837839, + "grad_norm": 0.45389464497566223, + "learning_rate": 2.9937e-05, + "loss": 0.0266, + "step": 9983 + }, + { + "epoch": 8.432432432432432, + "grad_norm": 0.4883260130882263, + "learning_rate": 2.994e-05, + "loss": 0.0303, + "step": 9984 + }, + { + "epoch": 8.433277027027026, + "grad_norm": 0.330040842294693, + "learning_rate": 2.9943e-05, + "loss": 0.0266, + "step": 9985 + }, + { + "epoch": 8.434121621621621, + "grad_norm": 0.45535019040107727, + "learning_rate": 2.9946e-05, + "loss": 0.0455, + "step": 9986 + }, + { + "epoch": 8.434966216216216, + "grad_norm": 0.4122505187988281, + "learning_rate": 2.9949e-05, + "loss": 0.0357, + "step": 9987 + }, + { + "epoch": 8.43581081081081, + "grad_norm": 0.6890194416046143, + "learning_rate": 2.9952e-05, + "loss": 0.0667, + "step": 9988 + }, + { + "epoch": 8.436655405405405, + "grad_norm": 0.9110317826271057, + "learning_rate": 2.9955000000000004e-05, + "loss": 0.0348, + "step": 9989 + }, + { + "epoch": 8.4375, + "grad_norm": 0.38516300916671753, + "learning_rate": 2.9958000000000004e-05, + "loss": 0.0304, + "step": 9990 + }, + { + "epoch": 8.438344594594595, + "grad_norm": 0.28985869884490967, + "learning_rate": 2.9961000000000003e-05, + "loss": 0.0257, + "step": 9991 + }, + { + "epoch": 8.43918918918919, + "grad_norm": 0.5287166833877563, + "learning_rate": 2.9964e-05, + "loss": 0.0434, + "step": 9992 + }, + { + "epoch": 8.440033783783784, + "grad_norm": 0.4314271807670593, + "learning_rate": 2.9967e-05, + "loss": 0.0377, + "step": 9993 + }, + { + "epoch": 8.440878378378379, + "grad_norm": 0.2431717962026596, + "learning_rate": 2.997e-05, + "loss": 0.0225, + "step": 9994 + }, + { + "epoch": 8.441722972972974, + "grad_norm": 0.3028135299682617, + "learning_rate": 2.9973e-05, + "loss": 0.0291, + "step": 9995 + }, + { + "epoch": 8.442567567567568, + "grad_norm": 0.3239331543445587, + "learning_rate": 2.9976e-05, + "loss": 0.0351, + "step": 9996 + }, + { + "epoch": 8.443412162162161, + "grad_norm": 0.4729953408241272, + "learning_rate": 2.9979e-05, + "loss": 0.0417, + "step": 9997 + }, + { + "epoch": 8.444256756756756, + "grad_norm": 0.5430527329444885, + "learning_rate": 2.9982e-05, + "loss": 0.0384, + "step": 9998 + }, + { + "epoch": 8.44510135135135, + "grad_norm": 0.7789978981018066, + "learning_rate": 2.9985000000000002e-05, + "loss": 0.0575, + "step": 9999 + }, + { + "epoch": 8.445945945945946, + "grad_norm": 0.5578780174255371, + "learning_rate": 2.9988e-05, + "loss": 0.0362, + "step": 10000 + }, + { + "epoch": 8.445945945945946, + "eval_cer": 0.09669906031644045, + "eval_loss": 0.27601853013038635, + "eval_runtime": 8.3993, + "eval_samples_per_second": 12.025, + "eval_steps_per_second": 0.476, + "eval_wer": 0.304673721340388, + "step": 10000 + }, + { + "epoch": 8.44679054054054, + "grad_norm": 0.29332882165908813, + "learning_rate": 2.9991e-05, + "loss": 0.0187, + "step": 10001 + }, + { + "epoch": 8.447635135135135, + "grad_norm": 0.45710670948028564, + "learning_rate": 2.9994e-05, + "loss": 0.0384, + "step": 10002 + }, + { + "epoch": 8.44847972972973, + "grad_norm": 0.2227388620376587, + "learning_rate": 2.9997e-05, + "loss": 0.0181, + "step": 10003 + }, + { + "epoch": 8.449324324324325, + "grad_norm": 2.6589162349700928, + "learning_rate": 3e-05, + "loss": 0.0268, + "step": 10004 + }, + { + "epoch": 8.45016891891892, + "grad_norm": 0.5258316993713379, + "learning_rate": 2.999966666666667e-05, + "loss": 0.0381, + "step": 10005 + }, + { + "epoch": 8.451013513513514, + "grad_norm": 1.1929181814193726, + "learning_rate": 2.9999333333333333e-05, + "loss": 0.0785, + "step": 10006 + }, + { + "epoch": 8.451858108108109, + "grad_norm": 0.3378559947013855, + "learning_rate": 2.9999000000000002e-05, + "loss": 0.027, + "step": 10007 + }, + { + "epoch": 8.452702702702704, + "grad_norm": 0.543487012386322, + "learning_rate": 2.9998666666666668e-05, + "loss": 0.0282, + "step": 10008 + }, + { + "epoch": 8.453547297297296, + "grad_norm": 0.4818302392959595, + "learning_rate": 2.9998333333333334e-05, + "loss": 0.0571, + "step": 10009 + }, + { + "epoch": 8.454391891891891, + "grad_norm": 0.4315248429775238, + "learning_rate": 2.9998e-05, + "loss": 0.0436, + "step": 10010 + }, + { + "epoch": 8.455236486486486, + "grad_norm": 0.36648210883140564, + "learning_rate": 2.999766666666667e-05, + "loss": 0.0214, + "step": 10011 + }, + { + "epoch": 8.45608108108108, + "grad_norm": 0.6950953006744385, + "learning_rate": 2.999733333333333e-05, + "loss": 0.0291, + "step": 10012 + }, + { + "epoch": 8.456925675675675, + "grad_norm": 0.4919818341732025, + "learning_rate": 2.9997e-05, + "loss": 0.0502, + "step": 10013 + }, + { + "epoch": 8.45777027027027, + "grad_norm": 0.4790979325771332, + "learning_rate": 2.999666666666667e-05, + "loss": 0.0367, + "step": 10014 + }, + { + "epoch": 8.458614864864865, + "grad_norm": 0.44028621912002563, + "learning_rate": 2.9996333333333333e-05, + "loss": 0.0366, + "step": 10015 + }, + { + "epoch": 8.45945945945946, + "grad_norm": 0.8563729524612427, + "learning_rate": 2.9996000000000002e-05, + "loss": 0.0533, + "step": 10016 + }, + { + "epoch": 8.460304054054054, + "grad_norm": 0.44987863302230835, + "learning_rate": 2.9995666666666668e-05, + "loss": 0.041, + "step": 10017 + }, + { + "epoch": 8.46114864864865, + "grad_norm": 0.8020370006561279, + "learning_rate": 2.9995333333333334e-05, + "loss": 0.0653, + "step": 10018 + }, + { + "epoch": 8.461993243243244, + "grad_norm": 0.7784488201141357, + "learning_rate": 2.9995e-05, + "loss": 0.0559, + "step": 10019 + }, + { + "epoch": 8.462837837837839, + "grad_norm": 0.44123128056526184, + "learning_rate": 2.9994666666666666e-05, + "loss": 0.0308, + "step": 10020 + }, + { + "epoch": 8.463682432432432, + "grad_norm": 0.9377260804176331, + "learning_rate": 2.9994333333333335e-05, + "loss": 0.0551, + "step": 10021 + }, + { + "epoch": 8.464527027027026, + "grad_norm": 1.2517714500427246, + "learning_rate": 2.9994e-05, + "loss": 0.0681, + "step": 10022 + }, + { + "epoch": 8.465371621621621, + "grad_norm": 1.3638509511947632, + "learning_rate": 2.9993666666666667e-05, + "loss": 0.3108, + "step": 10023 + }, + { + "epoch": 8.466216216216216, + "grad_norm": 0.9941492080688477, + "learning_rate": 2.9993333333333333e-05, + "loss": 0.1693, + "step": 10024 + }, + { + "epoch": 8.46706081081081, + "grad_norm": 0.5072339773178101, + "learning_rate": 2.9993000000000002e-05, + "loss": 0.0795, + "step": 10025 + }, + { + "epoch": 8.467905405405405, + "grad_norm": 0.32843053340911865, + "learning_rate": 2.9992666666666665e-05, + "loss": 0.0392, + "step": 10026 + }, + { + "epoch": 8.46875, + "grad_norm": 0.4644387364387512, + "learning_rate": 2.9992333333333334e-05, + "loss": 0.0422, + "step": 10027 + }, + { + "epoch": 8.469594594594595, + "grad_norm": 0.5228708982467651, + "learning_rate": 2.9992e-05, + "loss": 0.0441, + "step": 10028 + }, + { + "epoch": 8.47043918918919, + "grad_norm": 0.4077681601047516, + "learning_rate": 2.9991666666666666e-05, + "loss": 0.0368, + "step": 10029 + }, + { + "epoch": 8.471283783783784, + "grad_norm": 1.0260398387908936, + "learning_rate": 2.9991333333333335e-05, + "loss": 0.0663, + "step": 10030 + }, + { + "epoch": 8.472128378378379, + "grad_norm": 0.5403810143470764, + "learning_rate": 2.9991e-05, + "loss": 0.0361, + "step": 10031 + }, + { + "epoch": 8.472972972972974, + "grad_norm": 0.43711063265800476, + "learning_rate": 2.9990666666666667e-05, + "loss": 0.0442, + "step": 10032 + }, + { + "epoch": 8.473817567567568, + "grad_norm": 0.3938285708427429, + "learning_rate": 2.9990333333333333e-05, + "loss": 0.0322, + "step": 10033 + }, + { + "epoch": 8.474662162162161, + "grad_norm": 0.44612589478492737, + "learning_rate": 2.9990000000000003e-05, + "loss": 0.0303, + "step": 10034 + }, + { + "epoch": 8.475506756756756, + "grad_norm": 0.6648224592208862, + "learning_rate": 2.9989666666666665e-05, + "loss": 0.045, + "step": 10035 + }, + { + "epoch": 8.47635135135135, + "grad_norm": 0.5270398855209351, + "learning_rate": 2.9989333333333334e-05, + "loss": 0.0306, + "step": 10036 + }, + { + "epoch": 8.477195945945946, + "grad_norm": 0.4946129322052002, + "learning_rate": 2.9989e-05, + "loss": 0.0411, + "step": 10037 + }, + { + "epoch": 8.47804054054054, + "grad_norm": 0.4367859363555908, + "learning_rate": 2.9988666666666666e-05, + "loss": 0.0305, + "step": 10038 + }, + { + "epoch": 8.478885135135135, + "grad_norm": 0.41608330607414246, + "learning_rate": 2.9988333333333336e-05, + "loss": 0.0376, + "step": 10039 + }, + { + "epoch": 8.47972972972973, + "grad_norm": 0.3225111663341522, + "learning_rate": 2.9988e-05, + "loss": 0.0287, + "step": 10040 + }, + { + "epoch": 8.480574324324325, + "grad_norm": 0.3487389385700226, + "learning_rate": 2.9987666666666667e-05, + "loss": 0.0349, + "step": 10041 + }, + { + "epoch": 8.48141891891892, + "grad_norm": 0.777411699295044, + "learning_rate": 2.9987333333333333e-05, + "loss": 0.0344, + "step": 10042 + }, + { + "epoch": 8.482263513513514, + "grad_norm": 0.5905370712280273, + "learning_rate": 2.9987000000000003e-05, + "loss": 0.0463, + "step": 10043 + }, + { + "epoch": 8.483108108108109, + "grad_norm": 0.33415651321411133, + "learning_rate": 2.9986666666666665e-05, + "loss": 0.0277, + "step": 10044 + }, + { + "epoch": 8.483952702702704, + "grad_norm": 0.6660674810409546, + "learning_rate": 2.9986333333333335e-05, + "loss": 0.0429, + "step": 10045 + }, + { + "epoch": 8.484797297297296, + "grad_norm": 0.2960694432258606, + "learning_rate": 2.9986000000000004e-05, + "loss": 0.0211, + "step": 10046 + }, + { + "epoch": 8.485641891891891, + "grad_norm": 0.3402048349380493, + "learning_rate": 2.9985666666666666e-05, + "loss": 0.0253, + "step": 10047 + }, + { + "epoch": 8.486486486486486, + "grad_norm": 0.3701155483722687, + "learning_rate": 2.9985333333333336e-05, + "loss": 0.0366, + "step": 10048 + }, + { + "epoch": 8.48733108108108, + "grad_norm": 0.4644334614276886, + "learning_rate": 2.9985000000000002e-05, + "loss": 0.037, + "step": 10049 + }, + { + "epoch": 8.488175675675675, + "grad_norm": 0.27485308051109314, + "learning_rate": 2.9984666666666668e-05, + "loss": 0.0262, + "step": 10050 + }, + { + "epoch": 8.48902027027027, + "grad_norm": 0.44015416502952576, + "learning_rate": 2.9984333333333334e-05, + "loss": 0.0302, + "step": 10051 + }, + { + "epoch": 8.489864864864865, + "grad_norm": 0.7137801647186279, + "learning_rate": 2.9984e-05, + "loss": 0.0244, + "step": 10052 + }, + { + "epoch": 8.49070945945946, + "grad_norm": 0.3609675467014313, + "learning_rate": 2.9983666666666665e-05, + "loss": 0.0255, + "step": 10053 + }, + { + "epoch": 8.491554054054054, + "grad_norm": 0.49382802844047546, + "learning_rate": 2.9983333333333335e-05, + "loss": 0.027, + "step": 10054 + }, + { + "epoch": 8.49239864864865, + "grad_norm": 0.3030758500099182, + "learning_rate": 2.9983e-05, + "loss": 0.0232, + "step": 10055 + }, + { + "epoch": 8.493243243243244, + "grad_norm": 0.3832799792289734, + "learning_rate": 2.9982666666666667e-05, + "loss": 0.0406, + "step": 10056 + }, + { + "epoch": 8.494087837837839, + "grad_norm": 0.7719296216964722, + "learning_rate": 2.9982333333333336e-05, + "loss": 0.0561, + "step": 10057 + }, + { + "epoch": 8.494932432432432, + "grad_norm": 0.4588877856731415, + "learning_rate": 2.9982e-05, + "loss": 0.0235, + "step": 10058 + }, + { + "epoch": 8.495777027027026, + "grad_norm": 0.4739333391189575, + "learning_rate": 2.9981666666666668e-05, + "loss": 0.0443, + "step": 10059 + }, + { + "epoch": 8.496621621621621, + "grad_norm": 0.3507082760334015, + "learning_rate": 2.9981333333333334e-05, + "loss": 0.026, + "step": 10060 + }, + { + "epoch": 8.497466216216216, + "grad_norm": 0.4642655551433563, + "learning_rate": 2.9981e-05, + "loss": 0.0484, + "step": 10061 + }, + { + "epoch": 8.49831081081081, + "grad_norm": 0.6044650673866272, + "learning_rate": 2.9980666666666666e-05, + "loss": 0.0459, + "step": 10062 + }, + { + "epoch": 8.499155405405405, + "grad_norm": 1.1170768737792969, + "learning_rate": 2.9980333333333335e-05, + "loss": 0.0268, + "step": 10063 + }, + { + "epoch": 8.5, + "grad_norm": 1.0055484771728516, + "learning_rate": 2.998e-05, + "loss": 0.0326, + "step": 10064 + }, + { + "epoch": 8.500844594594595, + "grad_norm": 0.5718240141868591, + "learning_rate": 2.9979666666666667e-05, + "loss": 0.0555, + "step": 10065 + }, + { + "epoch": 8.50168918918919, + "grad_norm": 0.6536818146705627, + "learning_rate": 2.9979333333333336e-05, + "loss": 0.0859, + "step": 10066 + }, + { + "epoch": 8.502533783783784, + "grad_norm": 0.41386690735816956, + "learning_rate": 2.9979e-05, + "loss": 0.0313, + "step": 10067 + }, + { + "epoch": 8.503378378378379, + "grad_norm": 0.5251802206039429, + "learning_rate": 2.9978666666666668e-05, + "loss": 0.0375, + "step": 10068 + }, + { + "epoch": 8.504222972972974, + "grad_norm": 1.7093147039413452, + "learning_rate": 2.9978333333333334e-05, + "loss": 0.0446, + "step": 10069 + }, + { + "epoch": 8.505067567567568, + "grad_norm": 1.7715219259262085, + "learning_rate": 2.9978e-05, + "loss": 0.0606, + "step": 10070 + }, + { + "epoch": 8.505912162162161, + "grad_norm": 1.918428659439087, + "learning_rate": 2.997766666666667e-05, + "loss": 0.0682, + "step": 10071 + }, + { + "epoch": 8.506756756756756, + "grad_norm": 1.5705554485321045, + "learning_rate": 2.9977333333333335e-05, + "loss": 0.0962, + "step": 10072 + }, + { + "epoch": 8.50760135135135, + "grad_norm": 0.8980259895324707, + "learning_rate": 2.9977e-05, + "loss": 0.3202, + "step": 10073 + }, + { + "epoch": 8.508445945945946, + "grad_norm": 0.5488793849945068, + "learning_rate": 2.9976666666666667e-05, + "loss": 0.1573, + "step": 10074 + }, + { + "epoch": 8.50929054054054, + "grad_norm": 0.46585339307785034, + "learning_rate": 2.9976333333333336e-05, + "loss": 0.0986, + "step": 10075 + }, + { + "epoch": 8.510135135135135, + "grad_norm": 0.7133052945137024, + "learning_rate": 2.9976e-05, + "loss": 0.0637, + "step": 10076 + }, + { + "epoch": 8.51097972972973, + "grad_norm": 0.5516826510429382, + "learning_rate": 2.9975666666666668e-05, + "loss": 0.0536, + "step": 10077 + }, + { + "epoch": 8.511824324324325, + "grad_norm": 0.4303953945636749, + "learning_rate": 2.9975333333333334e-05, + "loss": 0.0481, + "step": 10078 + }, + { + "epoch": 8.51266891891892, + "grad_norm": 0.2995903491973877, + "learning_rate": 2.9975e-05, + "loss": 0.0262, + "step": 10079 + }, + { + "epoch": 8.513513513513514, + "grad_norm": 0.4061271548271179, + "learning_rate": 2.997466666666667e-05, + "loss": 0.0413, + "step": 10080 + }, + { + "epoch": 8.514358108108109, + "grad_norm": 0.37996241450309753, + "learning_rate": 2.9974333333333332e-05, + "loss": 0.0415, + "step": 10081 + }, + { + "epoch": 8.515202702702704, + "grad_norm": 0.6056141257286072, + "learning_rate": 2.9974e-05, + "loss": 0.0775, + "step": 10082 + }, + { + "epoch": 8.516047297297296, + "grad_norm": 0.3057810068130493, + "learning_rate": 2.9973666666666667e-05, + "loss": 0.0408, + "step": 10083 + }, + { + "epoch": 8.516891891891891, + "grad_norm": 0.40614140033721924, + "learning_rate": 2.9973333333333333e-05, + "loss": 0.0487, + "step": 10084 + }, + { + "epoch": 8.517736486486486, + "grad_norm": 0.4901714324951172, + "learning_rate": 2.9973e-05, + "loss": 0.0333, + "step": 10085 + }, + { + "epoch": 8.51858108108108, + "grad_norm": 0.40271201729774475, + "learning_rate": 2.997266666666667e-05, + "loss": 0.0367, + "step": 10086 + }, + { + "epoch": 8.519425675675675, + "grad_norm": 0.4629638195037842, + "learning_rate": 2.997233333333333e-05, + "loss": 0.0377, + "step": 10087 + }, + { + "epoch": 8.52027027027027, + "grad_norm": 0.4157496988773346, + "learning_rate": 2.9972e-05, + "loss": 0.0279, + "step": 10088 + }, + { + "epoch": 8.521114864864865, + "grad_norm": 0.28606757521629333, + "learning_rate": 2.997166666666667e-05, + "loss": 0.0299, + "step": 10089 + }, + { + "epoch": 8.52195945945946, + "grad_norm": 0.29977989196777344, + "learning_rate": 2.9971333333333332e-05, + "loss": 0.0249, + "step": 10090 + }, + { + "epoch": 8.522804054054054, + "grad_norm": 0.349982887506485, + "learning_rate": 2.9971e-05, + "loss": 0.0285, + "step": 10091 + }, + { + "epoch": 8.52364864864865, + "grad_norm": 0.3606393039226532, + "learning_rate": 2.9970666666666667e-05, + "loss": 0.0377, + "step": 10092 + }, + { + "epoch": 8.524493243243244, + "grad_norm": 0.35181206464767456, + "learning_rate": 2.9970333333333333e-05, + "loss": 0.0248, + "step": 10093 + }, + { + "epoch": 8.525337837837839, + "grad_norm": 0.44088080525398254, + "learning_rate": 2.997e-05, + "loss": 0.0292, + "step": 10094 + }, + { + "epoch": 8.526182432432432, + "grad_norm": 0.40281912684440613, + "learning_rate": 2.996966666666667e-05, + "loss": 0.0257, + "step": 10095 + }, + { + "epoch": 8.527027027027026, + "grad_norm": 0.4377671480178833, + "learning_rate": 2.9969333333333335e-05, + "loss": 0.0431, + "step": 10096 + }, + { + "epoch": 8.527871621621621, + "grad_norm": 0.4256654679775238, + "learning_rate": 2.9969e-05, + "loss": 0.0265, + "step": 10097 + }, + { + "epoch": 8.528716216216216, + "grad_norm": 0.39295461773872375, + "learning_rate": 2.996866666666667e-05, + "loss": 0.0267, + "step": 10098 + }, + { + "epoch": 8.52956081081081, + "grad_norm": 0.502119779586792, + "learning_rate": 2.9968333333333332e-05, + "loss": 0.0221, + "step": 10099 + }, + { + "epoch": 8.530405405405405, + "grad_norm": 0.48187315464019775, + "learning_rate": 2.9968000000000002e-05, + "loss": 0.0297, + "step": 10100 + }, + { + "epoch": 8.53125, + "grad_norm": 0.4808312654495239, + "learning_rate": 2.9967666666666668e-05, + "loss": 0.023, + "step": 10101 + }, + { + "epoch": 8.532094594594595, + "grad_norm": 0.3158288896083832, + "learning_rate": 2.9967333333333334e-05, + "loss": 0.0182, + "step": 10102 + }, + { + "epoch": 8.53293918918919, + "grad_norm": 0.3643815517425537, + "learning_rate": 2.9967e-05, + "loss": 0.0186, + "step": 10103 + }, + { + "epoch": 8.533783783783784, + "grad_norm": 0.5563183426856995, + "learning_rate": 2.996666666666667e-05, + "loss": 0.0326, + "step": 10104 + }, + { + "epoch": 8.534628378378379, + "grad_norm": 0.8903801441192627, + "learning_rate": 2.9966333333333335e-05, + "loss": 0.0344, + "step": 10105 + }, + { + "epoch": 8.535472972972974, + "grad_norm": 0.5360203385353088, + "learning_rate": 2.9966e-05, + "loss": 0.0459, + "step": 10106 + }, + { + "epoch": 8.536317567567568, + "grad_norm": 0.6547129154205322, + "learning_rate": 2.996566666666667e-05, + "loss": 0.0565, + "step": 10107 + }, + { + "epoch": 8.537162162162161, + "grad_norm": 0.787948727607727, + "learning_rate": 2.9965333333333333e-05, + "loss": 0.0607, + "step": 10108 + }, + { + "epoch": 8.538006756756756, + "grad_norm": 0.41397953033447266, + "learning_rate": 2.9965000000000002e-05, + "loss": 0.0357, + "step": 10109 + }, + { + "epoch": 8.53885135135135, + "grad_norm": 0.39068418741226196, + "learning_rate": 2.9964666666666664e-05, + "loss": 0.028, + "step": 10110 + }, + { + "epoch": 8.539695945945946, + "grad_norm": 0.38312965631484985, + "learning_rate": 2.9964333333333334e-05, + "loss": 0.0345, + "step": 10111 + }, + { + "epoch": 8.54054054054054, + "grad_norm": 0.5153236389160156, + "learning_rate": 2.9964e-05, + "loss": 0.0431, + "step": 10112 + }, + { + "epoch": 8.541385135135135, + "grad_norm": 0.5143445134162903, + "learning_rate": 2.9963666666666666e-05, + "loss": 0.0318, + "step": 10113 + }, + { + "epoch": 8.54222972972973, + "grad_norm": 0.9833884239196777, + "learning_rate": 2.9963333333333335e-05, + "loss": 0.0537, + "step": 10114 + }, + { + "epoch": 8.543074324324325, + "grad_norm": 0.48594650626182556, + "learning_rate": 2.9963e-05, + "loss": 0.0325, + "step": 10115 + }, + { + "epoch": 8.54391891891892, + "grad_norm": 0.6472426056861877, + "learning_rate": 2.9962666666666667e-05, + "loss": 0.0339, + "step": 10116 + }, + { + "epoch": 8.544763513513514, + "grad_norm": 0.483673632144928, + "learning_rate": 2.9962333333333333e-05, + "loss": 0.034, + "step": 10117 + }, + { + "epoch": 8.545608108108109, + "grad_norm": 0.8780347108840942, + "learning_rate": 2.9962000000000002e-05, + "loss": 0.0409, + "step": 10118 + }, + { + "epoch": 8.546452702702704, + "grad_norm": 0.3526596426963806, + "learning_rate": 2.9961666666666665e-05, + "loss": 0.0244, + "step": 10119 + }, + { + "epoch": 8.547297297297296, + "grad_norm": 0.718930184841156, + "learning_rate": 2.9961333333333334e-05, + "loss": 0.0406, + "step": 10120 + }, + { + "epoch": 8.548141891891891, + "grad_norm": 1.0988421440124512, + "learning_rate": 2.9961000000000003e-05, + "loss": 0.0903, + "step": 10121 + }, + { + "epoch": 8.548986486486486, + "grad_norm": 1.2824286222457886, + "learning_rate": 2.9960666666666666e-05, + "loss": 0.0779, + "step": 10122 + }, + { + "epoch": 8.54983108108108, + "grad_norm": 1.1330013275146484, + "learning_rate": 2.9960333333333335e-05, + "loss": 0.3403, + "step": 10123 + }, + { + "epoch": 8.550675675675675, + "grad_norm": 0.5669223666191101, + "learning_rate": 2.996e-05, + "loss": 0.2085, + "step": 10124 + }, + { + "epoch": 8.55152027027027, + "grad_norm": 0.6123809814453125, + "learning_rate": 2.9959666666666667e-05, + "loss": 0.0874, + "step": 10125 + }, + { + "epoch": 8.552364864864865, + "grad_norm": 0.5533405542373657, + "learning_rate": 2.9959333333333333e-05, + "loss": 0.0586, + "step": 10126 + }, + { + "epoch": 8.55320945945946, + "grad_norm": 0.5059563517570496, + "learning_rate": 2.9959000000000002e-05, + "loss": 0.0474, + "step": 10127 + }, + { + "epoch": 8.554054054054054, + "grad_norm": 0.510195255279541, + "learning_rate": 2.9958666666666665e-05, + "loss": 0.0534, + "step": 10128 + }, + { + "epoch": 8.55489864864865, + "grad_norm": 0.5479398965835571, + "learning_rate": 2.9958333333333334e-05, + "loss": 0.0367, + "step": 10129 + }, + { + "epoch": 8.555743243243244, + "grad_norm": 0.41482460498809814, + "learning_rate": 2.9958000000000004e-05, + "loss": 0.0371, + "step": 10130 + }, + { + "epoch": 8.556587837837839, + "grad_norm": 0.4511183798313141, + "learning_rate": 2.9957666666666666e-05, + "loss": 0.0479, + "step": 10131 + }, + { + "epoch": 8.557432432432432, + "grad_norm": 0.3520013093948364, + "learning_rate": 2.9957333333333335e-05, + "loss": 0.0468, + "step": 10132 + }, + { + "epoch": 8.558277027027026, + "grad_norm": 0.5488730072975159, + "learning_rate": 2.9957e-05, + "loss": 0.0331, + "step": 10133 + }, + { + "epoch": 8.559121621621621, + "grad_norm": 0.5760298371315002, + "learning_rate": 2.9956666666666667e-05, + "loss": 0.0419, + "step": 10134 + }, + { + "epoch": 8.559966216216216, + "grad_norm": 0.4816664755344391, + "learning_rate": 2.9956333333333333e-05, + "loss": 0.0356, + "step": 10135 + }, + { + "epoch": 8.56081081081081, + "grad_norm": 0.49606311321258545, + "learning_rate": 2.9956000000000003e-05, + "loss": 0.0762, + "step": 10136 + }, + { + "epoch": 8.561655405405405, + "grad_norm": 0.38945192098617554, + "learning_rate": 2.9955666666666665e-05, + "loss": 0.0418, + "step": 10137 + }, + { + "epoch": 8.5625, + "grad_norm": 0.47477859258651733, + "learning_rate": 2.9955333333333334e-05, + "loss": 0.061, + "step": 10138 + }, + { + "epoch": 8.563344594594595, + "grad_norm": 0.28347864747047424, + "learning_rate": 2.9955000000000004e-05, + "loss": 0.0257, + "step": 10139 + }, + { + "epoch": 8.56418918918919, + "grad_norm": 0.7143532037734985, + "learning_rate": 2.9954666666666666e-05, + "loss": 0.037, + "step": 10140 + }, + { + "epoch": 8.565033783783784, + "grad_norm": 0.3507101535797119, + "learning_rate": 2.9954333333333336e-05, + "loss": 0.02, + "step": 10141 + }, + { + "epoch": 8.565878378378379, + "grad_norm": 0.21388953924179077, + "learning_rate": 2.9953999999999998e-05, + "loss": 0.0244, + "step": 10142 + }, + { + "epoch": 8.566722972972974, + "grad_norm": 0.9635578989982605, + "learning_rate": 2.9953666666666667e-05, + "loss": 0.0425, + "step": 10143 + }, + { + "epoch": 8.567567567567568, + "grad_norm": 1.2973726987838745, + "learning_rate": 2.9953333333333333e-05, + "loss": 0.0435, + "step": 10144 + }, + { + "epoch": 8.568412162162161, + "grad_norm": 0.41147828102111816, + "learning_rate": 2.9953e-05, + "loss": 0.0309, + "step": 10145 + }, + { + "epoch": 8.569256756756756, + "grad_norm": 0.2567380964756012, + "learning_rate": 2.995266666666667e-05, + "loss": 0.0206, + "step": 10146 + }, + { + "epoch": 8.57010135135135, + "grad_norm": 0.4565578103065491, + "learning_rate": 2.9952333333333335e-05, + "loss": 0.0444, + "step": 10147 + }, + { + "epoch": 8.570945945945946, + "grad_norm": 0.42262905836105347, + "learning_rate": 2.9952e-05, + "loss": 0.0438, + "step": 10148 + }, + { + "epoch": 8.57179054054054, + "grad_norm": 0.4803393483161926, + "learning_rate": 2.9951666666666666e-05, + "loss": 0.0464, + "step": 10149 + }, + { + "epoch": 8.572635135135135, + "grad_norm": 0.7815998196601868, + "learning_rate": 2.9951333333333336e-05, + "loss": 0.0406, + "step": 10150 + }, + { + "epoch": 8.57347972972973, + "grad_norm": 0.34384697675704956, + "learning_rate": 2.9951e-05, + "loss": 0.0259, + "step": 10151 + }, + { + "epoch": 8.574324324324325, + "grad_norm": 0.3154940903186798, + "learning_rate": 2.9950666666666668e-05, + "loss": 0.0325, + "step": 10152 + }, + { + "epoch": 8.57516891891892, + "grad_norm": 0.4723922908306122, + "learning_rate": 2.9950333333333334e-05, + "loss": 0.0463, + "step": 10153 + }, + { + "epoch": 8.576013513513514, + "grad_norm": 0.5947287678718567, + "learning_rate": 2.995e-05, + "loss": 0.0554, + "step": 10154 + }, + { + "epoch": 8.576858108108109, + "grad_norm": 0.36807143688201904, + "learning_rate": 2.994966666666667e-05, + "loss": 0.0383, + "step": 10155 + }, + { + "epoch": 8.577702702702704, + "grad_norm": 0.3118795156478882, + "learning_rate": 2.9949333333333335e-05, + "loss": 0.0347, + "step": 10156 + }, + { + "epoch": 8.578547297297296, + "grad_norm": 0.27517420053482056, + "learning_rate": 2.9949e-05, + "loss": 0.0246, + "step": 10157 + }, + { + "epoch": 8.579391891891891, + "grad_norm": 0.41746577620506287, + "learning_rate": 2.9948666666666667e-05, + "loss": 0.0305, + "step": 10158 + }, + { + "epoch": 8.580236486486486, + "grad_norm": 0.25652649998664856, + "learning_rate": 2.9948333333333336e-05, + "loss": 0.0204, + "step": 10159 + }, + { + "epoch": 8.58108108108108, + "grad_norm": 0.33634331822395325, + "learning_rate": 2.9948e-05, + "loss": 0.0254, + "step": 10160 + }, + { + "epoch": 8.581925675675675, + "grad_norm": 0.6129982471466064, + "learning_rate": 2.9947666666666668e-05, + "loss": 0.0359, + "step": 10161 + }, + { + "epoch": 8.58277027027027, + "grad_norm": 0.4663855731487274, + "learning_rate": 2.9947333333333334e-05, + "loss": 0.0273, + "step": 10162 + }, + { + "epoch": 8.583614864864865, + "grad_norm": 0.3526194393634796, + "learning_rate": 2.9947e-05, + "loss": 0.0263, + "step": 10163 + }, + { + "epoch": 8.58445945945946, + "grad_norm": 0.4049128293991089, + "learning_rate": 2.994666666666667e-05, + "loss": 0.0269, + "step": 10164 + }, + { + "epoch": 8.585304054054054, + "grad_norm": 0.6291057467460632, + "learning_rate": 2.9946333333333335e-05, + "loss": 0.0273, + "step": 10165 + }, + { + "epoch": 8.58614864864865, + "grad_norm": 0.7264106273651123, + "learning_rate": 2.9946e-05, + "loss": 0.0346, + "step": 10166 + }, + { + "epoch": 8.586993243243244, + "grad_norm": 0.7709944844245911, + "learning_rate": 2.9945666666666667e-05, + "loss": 0.0515, + "step": 10167 + }, + { + "epoch": 8.587837837837839, + "grad_norm": 0.48391827940940857, + "learning_rate": 2.9945333333333336e-05, + "loss": 0.0359, + "step": 10168 + }, + { + "epoch": 8.588682432432432, + "grad_norm": 0.5605697631835938, + "learning_rate": 2.9945e-05, + "loss": 0.0504, + "step": 10169 + }, + { + "epoch": 8.589527027027026, + "grad_norm": 0.8199106454849243, + "learning_rate": 2.9944666666666668e-05, + "loss": 0.0769, + "step": 10170 + }, + { + "epoch": 8.590371621621621, + "grad_norm": 0.7370114326477051, + "learning_rate": 2.9944333333333334e-05, + "loss": 0.1066, + "step": 10171 + }, + { + "epoch": 8.591216216216216, + "grad_norm": 0.8665898442268372, + "learning_rate": 2.9944e-05, + "loss": 0.0778, + "step": 10172 + }, + { + "epoch": 8.59206081081081, + "grad_norm": 1.2559864521026611, + "learning_rate": 2.994366666666667e-05, + "loss": 0.3156, + "step": 10173 + }, + { + "epoch": 8.592905405405405, + "grad_norm": 0.7417587637901306, + "learning_rate": 2.9943333333333332e-05, + "loss": 0.1975, + "step": 10174 + }, + { + "epoch": 8.59375, + "grad_norm": 0.41965529322624207, + "learning_rate": 2.9943e-05, + "loss": 0.1107, + "step": 10175 + }, + { + "epoch": 8.594594594594595, + "grad_norm": 0.38509225845336914, + "learning_rate": 2.9942666666666667e-05, + "loss": 0.0602, + "step": 10176 + }, + { + "epoch": 8.59543918918919, + "grad_norm": 0.37399330735206604, + "learning_rate": 2.9942333333333333e-05, + "loss": 0.055, + "step": 10177 + }, + { + "epoch": 8.596283783783784, + "grad_norm": 0.5941076874732971, + "learning_rate": 2.9942e-05, + "loss": 0.0684, + "step": 10178 + }, + { + "epoch": 8.597128378378379, + "grad_norm": 0.5148724317550659, + "learning_rate": 2.9941666666666668e-05, + "loss": 0.0552, + "step": 10179 + }, + { + "epoch": 8.597972972972974, + "grad_norm": 0.32706332206726074, + "learning_rate": 2.9941333333333334e-05, + "loss": 0.0411, + "step": 10180 + }, + { + "epoch": 8.598817567567568, + "grad_norm": 0.39390265941619873, + "learning_rate": 2.9941e-05, + "loss": 0.0375, + "step": 10181 + }, + { + "epoch": 8.599662162162161, + "grad_norm": 0.35751867294311523, + "learning_rate": 2.994066666666667e-05, + "loss": 0.033, + "step": 10182 + }, + { + "epoch": 8.600506756756756, + "grad_norm": 0.32115888595581055, + "learning_rate": 2.9940333333333332e-05, + "loss": 0.0268, + "step": 10183 + }, + { + "epoch": 8.60135135135135, + "grad_norm": 0.3034404516220093, + "learning_rate": 2.994e-05, + "loss": 0.0263, + "step": 10184 + }, + { + "epoch": 8.602195945945946, + "grad_norm": 0.7044649720191956, + "learning_rate": 2.9939666666666667e-05, + "loss": 0.0603, + "step": 10185 + }, + { + "epoch": 8.60304054054054, + "grad_norm": 0.40339478850364685, + "learning_rate": 2.9939333333333333e-05, + "loss": 0.0404, + "step": 10186 + }, + { + "epoch": 8.603885135135135, + "grad_norm": 0.5868296027183533, + "learning_rate": 2.9939e-05, + "loss": 0.0521, + "step": 10187 + }, + { + "epoch": 8.60472972972973, + "grad_norm": 0.32559576630592346, + "learning_rate": 2.993866666666667e-05, + "loss": 0.0245, + "step": 10188 + }, + { + "epoch": 8.605574324324325, + "grad_norm": 0.5952154994010925, + "learning_rate": 2.9938333333333334e-05, + "loss": 0.0428, + "step": 10189 + }, + { + "epoch": 8.60641891891892, + "grad_norm": 0.306674987077713, + "learning_rate": 2.9938e-05, + "loss": 0.0344, + "step": 10190 + }, + { + "epoch": 8.607263513513514, + "grad_norm": 0.38755112886428833, + "learning_rate": 2.993766666666667e-05, + "loss": 0.0378, + "step": 10191 + }, + { + "epoch": 8.608108108108109, + "grad_norm": 0.43185165524482727, + "learning_rate": 2.9937333333333332e-05, + "loss": 0.039, + "step": 10192 + }, + { + "epoch": 8.608952702702704, + "grad_norm": 0.4858570396900177, + "learning_rate": 2.9937e-05, + "loss": 0.0437, + "step": 10193 + }, + { + "epoch": 8.609797297297296, + "grad_norm": 0.6638676524162292, + "learning_rate": 2.9936666666666667e-05, + "loss": 0.041, + "step": 10194 + }, + { + "epoch": 8.610641891891891, + "grad_norm": 0.48580604791641235, + "learning_rate": 2.9936333333333333e-05, + "loss": 0.0218, + "step": 10195 + }, + { + "epoch": 8.611486486486486, + "grad_norm": 0.6757107973098755, + "learning_rate": 2.9936000000000003e-05, + "loss": 0.0394, + "step": 10196 + }, + { + "epoch": 8.61233108108108, + "grad_norm": 0.39729467034339905, + "learning_rate": 2.993566666666667e-05, + "loss": 0.0307, + "step": 10197 + }, + { + "epoch": 8.613175675675675, + "grad_norm": 0.4106625020503998, + "learning_rate": 2.9935333333333335e-05, + "loss": 0.0413, + "step": 10198 + }, + { + "epoch": 8.61402027027027, + "grad_norm": 0.40303218364715576, + "learning_rate": 2.9935e-05, + "loss": 0.0316, + "step": 10199 + }, + { + "epoch": 8.614864864864865, + "grad_norm": 0.3036094009876251, + "learning_rate": 2.993466666666667e-05, + "loss": 0.0265, + "step": 10200 + }, + { + "epoch": 8.61570945945946, + "grad_norm": 0.5123484134674072, + "learning_rate": 2.9934333333333332e-05, + "loss": 0.0387, + "step": 10201 + }, + { + "epoch": 8.616554054054054, + "grad_norm": 1.201342225074768, + "learning_rate": 2.9934000000000002e-05, + "loss": 0.0643, + "step": 10202 + }, + { + "epoch": 8.61739864864865, + "grad_norm": 0.3625534772872925, + "learning_rate": 2.9933666666666664e-05, + "loss": 0.0365, + "step": 10203 + }, + { + "epoch": 8.618243243243244, + "grad_norm": 0.44614675641059875, + "learning_rate": 2.9933333333333334e-05, + "loss": 0.0392, + "step": 10204 + }, + { + "epoch": 8.619087837837839, + "grad_norm": 0.36334651708602905, + "learning_rate": 2.9933000000000003e-05, + "loss": 0.0318, + "step": 10205 + }, + { + "epoch": 8.619932432432432, + "grad_norm": 0.4135650396347046, + "learning_rate": 2.9932666666666665e-05, + "loss": 0.0367, + "step": 10206 + }, + { + "epoch": 8.620777027027026, + "grad_norm": 0.7069926857948303, + "learning_rate": 2.9932333333333335e-05, + "loss": 0.0557, + "step": 10207 + }, + { + "epoch": 8.621621621621621, + "grad_norm": 0.3726401627063751, + "learning_rate": 2.9932e-05, + "loss": 0.0249, + "step": 10208 + }, + { + "epoch": 8.622466216216216, + "grad_norm": 0.456214040517807, + "learning_rate": 2.9931666666666667e-05, + "loss": 0.053, + "step": 10209 + }, + { + "epoch": 8.62331081081081, + "grad_norm": 0.47007182240486145, + "learning_rate": 2.9931333333333333e-05, + "loss": 0.0346, + "step": 10210 + }, + { + "epoch": 8.624155405405405, + "grad_norm": 0.6197255253791809, + "learning_rate": 2.9931000000000002e-05, + "loss": 0.0343, + "step": 10211 + }, + { + "epoch": 8.625, + "grad_norm": 0.4076075851917267, + "learning_rate": 2.9930666666666668e-05, + "loss": 0.0433, + "step": 10212 + }, + { + "epoch": 8.625844594594595, + "grad_norm": 0.4121657609939575, + "learning_rate": 2.9930333333333334e-05, + "loss": 0.0386, + "step": 10213 + }, + { + "epoch": 8.62668918918919, + "grad_norm": 0.37592247128486633, + "learning_rate": 2.9930000000000003e-05, + "loss": 0.0293, + "step": 10214 + }, + { + "epoch": 8.627533783783784, + "grad_norm": 0.4247586727142334, + "learning_rate": 2.9929666666666666e-05, + "loss": 0.0365, + "step": 10215 + }, + { + "epoch": 8.628378378378379, + "grad_norm": 0.6570274233818054, + "learning_rate": 2.9929333333333335e-05, + "loss": 0.0678, + "step": 10216 + }, + { + "epoch": 8.629222972972974, + "grad_norm": 0.3602847456932068, + "learning_rate": 2.9929e-05, + "loss": 0.0352, + "step": 10217 + }, + { + "epoch": 8.630067567567568, + "grad_norm": 0.3925158977508545, + "learning_rate": 2.9928666666666667e-05, + "loss": 0.0479, + "step": 10218 + }, + { + "epoch": 8.630912162162161, + "grad_norm": 0.7977339029312134, + "learning_rate": 2.9928333333333333e-05, + "loss": 0.0537, + "step": 10219 + }, + { + "epoch": 8.631756756756756, + "grad_norm": 0.8683387041091919, + "learning_rate": 2.9928000000000002e-05, + "loss": 0.0648, + "step": 10220 + }, + { + "epoch": 8.63260135135135, + "grad_norm": 0.7475566267967224, + "learning_rate": 2.9927666666666668e-05, + "loss": 0.0557, + "step": 10221 + }, + { + "epoch": 8.633445945945946, + "grad_norm": 0.7283245921134949, + "learning_rate": 2.9927333333333334e-05, + "loss": 0.0467, + "step": 10222 + }, + { + "epoch": 8.63429054054054, + "grad_norm": 0.7992246747016907, + "learning_rate": 2.9927000000000003e-05, + "loss": 0.2963, + "step": 10223 + }, + { + "epoch": 8.635135135135135, + "grad_norm": 0.5144175291061401, + "learning_rate": 2.9926666666666666e-05, + "loss": 0.1864, + "step": 10224 + }, + { + "epoch": 8.63597972972973, + "grad_norm": 0.4843386709690094, + "learning_rate": 2.9926333333333335e-05, + "loss": 0.1129, + "step": 10225 + }, + { + "epoch": 8.636824324324325, + "grad_norm": 0.44167056679725647, + "learning_rate": 2.9926e-05, + "loss": 0.0863, + "step": 10226 + }, + { + "epoch": 8.63766891891892, + "grad_norm": 0.464246541261673, + "learning_rate": 2.9925666666666667e-05, + "loss": 0.0682, + "step": 10227 + }, + { + "epoch": 8.638513513513514, + "grad_norm": 0.5725159049034119, + "learning_rate": 2.9925333333333333e-05, + "loss": 0.0586, + "step": 10228 + }, + { + "epoch": 8.639358108108109, + "grad_norm": 0.4052066504955292, + "learning_rate": 2.9925000000000002e-05, + "loss": 0.0492, + "step": 10229 + }, + { + "epoch": 8.640202702702704, + "grad_norm": 0.28704607486724854, + "learning_rate": 2.9924666666666668e-05, + "loss": 0.032, + "step": 10230 + }, + { + "epoch": 8.641047297297296, + "grad_norm": 0.40047305822372437, + "learning_rate": 2.9924333333333334e-05, + "loss": 0.0378, + "step": 10231 + }, + { + "epoch": 8.641891891891891, + "grad_norm": 0.3537921905517578, + "learning_rate": 2.9924e-05, + "loss": 0.0464, + "step": 10232 + }, + { + "epoch": 8.642736486486486, + "grad_norm": 0.3461790382862091, + "learning_rate": 2.9923666666666666e-05, + "loss": 0.0347, + "step": 10233 + }, + { + "epoch": 8.64358108108108, + "grad_norm": 0.34986916184425354, + "learning_rate": 2.9923333333333335e-05, + "loss": 0.0417, + "step": 10234 + }, + { + "epoch": 8.644425675675675, + "grad_norm": 0.369067519903183, + "learning_rate": 2.9922999999999998e-05, + "loss": 0.0295, + "step": 10235 + }, + { + "epoch": 8.64527027027027, + "grad_norm": 0.22358718514442444, + "learning_rate": 2.9922666666666667e-05, + "loss": 0.021, + "step": 10236 + }, + { + "epoch": 8.646114864864865, + "grad_norm": 0.37398889660835266, + "learning_rate": 2.9922333333333333e-05, + "loss": 0.0518, + "step": 10237 + }, + { + "epoch": 8.64695945945946, + "grad_norm": 0.44247952103614807, + "learning_rate": 2.9922e-05, + "loss": 0.0408, + "step": 10238 + }, + { + "epoch": 8.647804054054054, + "grad_norm": 0.3084888458251953, + "learning_rate": 2.992166666666667e-05, + "loss": 0.0303, + "step": 10239 + }, + { + "epoch": 8.64864864864865, + "grad_norm": 0.34925562143325806, + "learning_rate": 2.9921333333333334e-05, + "loss": 0.0321, + "step": 10240 + }, + { + "epoch": 8.649493243243244, + "grad_norm": 0.42019692063331604, + "learning_rate": 2.9921e-05, + "loss": 0.0407, + "step": 10241 + }, + { + "epoch": 8.650337837837839, + "grad_norm": 0.44859620928764343, + "learning_rate": 2.9920666666666666e-05, + "loss": 0.035, + "step": 10242 + }, + { + "epoch": 8.651182432432432, + "grad_norm": 0.47769293189048767, + "learning_rate": 2.9920333333333336e-05, + "loss": 0.049, + "step": 10243 + }, + { + "epoch": 8.652027027027026, + "grad_norm": 0.8659946322441101, + "learning_rate": 2.9919999999999998e-05, + "loss": 0.0239, + "step": 10244 + }, + { + "epoch": 8.652871621621621, + "grad_norm": 0.47897204756736755, + "learning_rate": 2.9919666666666667e-05, + "loss": 0.0346, + "step": 10245 + }, + { + "epoch": 8.653716216216216, + "grad_norm": 0.6358274817466736, + "learning_rate": 2.9919333333333337e-05, + "loss": 0.0305, + "step": 10246 + }, + { + "epoch": 8.65456081081081, + "grad_norm": 0.3694375157356262, + "learning_rate": 2.9919e-05, + "loss": 0.0295, + "step": 10247 + }, + { + "epoch": 8.655405405405405, + "grad_norm": 0.45633673667907715, + "learning_rate": 2.991866666666667e-05, + "loss": 0.0248, + "step": 10248 + }, + { + "epoch": 8.65625, + "grad_norm": 0.3785530626773834, + "learning_rate": 2.9918333333333335e-05, + "loss": 0.0464, + "step": 10249 + }, + { + "epoch": 8.657094594594595, + "grad_norm": 0.4049910604953766, + "learning_rate": 2.9918e-05, + "loss": 0.0361, + "step": 10250 + }, + { + "epoch": 8.65793918918919, + "grad_norm": 0.4138730466365814, + "learning_rate": 2.9917666666666666e-05, + "loss": 0.0479, + "step": 10251 + }, + { + "epoch": 8.658783783783784, + "grad_norm": 0.4052395522594452, + "learning_rate": 2.9917333333333336e-05, + "loss": 0.0214, + "step": 10252 + }, + { + "epoch": 8.659628378378379, + "grad_norm": 0.39883649349212646, + "learning_rate": 2.9917e-05, + "loss": 0.0271, + "step": 10253 + }, + { + "epoch": 8.660472972972974, + "grad_norm": 0.4853348135948181, + "learning_rate": 2.9916666666666668e-05, + "loss": 0.0399, + "step": 10254 + }, + { + "epoch": 8.661317567567568, + "grad_norm": 0.3468082547187805, + "learning_rate": 2.9916333333333337e-05, + "loss": 0.04, + "step": 10255 + }, + { + "epoch": 8.662162162162161, + "grad_norm": 0.4348318874835968, + "learning_rate": 2.9916e-05, + "loss": 0.0343, + "step": 10256 + }, + { + "epoch": 8.663006756756756, + "grad_norm": 0.6557223200798035, + "learning_rate": 2.991566666666667e-05, + "loss": 0.0409, + "step": 10257 + }, + { + "epoch": 8.66385135135135, + "grad_norm": 0.5775102972984314, + "learning_rate": 2.9915333333333335e-05, + "loss": 0.0485, + "step": 10258 + }, + { + "epoch": 8.664695945945946, + "grad_norm": 0.39404863119125366, + "learning_rate": 2.9915e-05, + "loss": 0.0265, + "step": 10259 + }, + { + "epoch": 8.66554054054054, + "grad_norm": 0.5083813667297363, + "learning_rate": 2.9914666666666667e-05, + "loss": 0.0317, + "step": 10260 + }, + { + "epoch": 8.666385135135135, + "grad_norm": 0.44900137186050415, + "learning_rate": 2.9914333333333336e-05, + "loss": 0.0423, + "step": 10261 + }, + { + "epoch": 8.66722972972973, + "grad_norm": 0.39685168862342834, + "learning_rate": 2.9914000000000002e-05, + "loss": 0.0438, + "step": 10262 + }, + { + "epoch": 8.668074324324325, + "grad_norm": 0.5459296107292175, + "learning_rate": 2.9913666666666668e-05, + "loss": 0.035, + "step": 10263 + }, + { + "epoch": 8.66891891891892, + "grad_norm": 0.4948076903820038, + "learning_rate": 2.9913333333333334e-05, + "loss": 0.0331, + "step": 10264 + }, + { + "epoch": 8.669763513513514, + "grad_norm": 0.54075026512146, + "learning_rate": 2.9913e-05, + "loss": 0.0421, + "step": 10265 + }, + { + "epoch": 8.670608108108109, + "grad_norm": 0.6643034219741821, + "learning_rate": 2.991266666666667e-05, + "loss": 0.0471, + "step": 10266 + }, + { + "epoch": 8.671452702702704, + "grad_norm": 0.3709309697151184, + "learning_rate": 2.991233333333333e-05, + "loss": 0.0364, + "step": 10267 + }, + { + "epoch": 8.672297297297296, + "grad_norm": 0.7900141477584839, + "learning_rate": 2.9912e-05, + "loss": 0.0536, + "step": 10268 + }, + { + "epoch": 8.673141891891891, + "grad_norm": 0.6114487051963806, + "learning_rate": 2.9911666666666667e-05, + "loss": 0.0733, + "step": 10269 + }, + { + "epoch": 8.673986486486486, + "grad_norm": 0.411606103181839, + "learning_rate": 2.9911333333333333e-05, + "loss": 0.0404, + "step": 10270 + }, + { + "epoch": 8.67483108108108, + "grad_norm": 1.250777006149292, + "learning_rate": 2.9911000000000002e-05, + "loss": 0.0744, + "step": 10271 + }, + { + "epoch": 8.675675675675675, + "grad_norm": 1.02536141872406, + "learning_rate": 2.9910666666666668e-05, + "loss": 0.0653, + "step": 10272 + }, + { + "epoch": 8.67652027027027, + "grad_norm": 0.9328819513320923, + "learning_rate": 2.9910333333333334e-05, + "loss": 0.2555, + "step": 10273 + }, + { + "epoch": 8.677364864864865, + "grad_norm": 0.6589503288269043, + "learning_rate": 2.991e-05, + "loss": 0.167, + "step": 10274 + }, + { + "epoch": 8.67820945945946, + "grad_norm": 0.6051692366600037, + "learning_rate": 2.990966666666667e-05, + "loss": 0.0872, + "step": 10275 + }, + { + "epoch": 8.679054054054054, + "grad_norm": 0.43322324752807617, + "learning_rate": 2.9909333333333332e-05, + "loss": 0.0721, + "step": 10276 + }, + { + "epoch": 8.67989864864865, + "grad_norm": 0.39164310693740845, + "learning_rate": 2.9909e-05, + "loss": 0.0666, + "step": 10277 + }, + { + "epoch": 8.680743243243244, + "grad_norm": 0.6401993036270142, + "learning_rate": 2.9908666666666667e-05, + "loss": 0.0534, + "step": 10278 + }, + { + "epoch": 8.681587837837839, + "grad_norm": 0.5741422176361084, + "learning_rate": 2.9908333333333333e-05, + "loss": 0.0406, + "step": 10279 + }, + { + "epoch": 8.682432432432432, + "grad_norm": 0.41071873903274536, + "learning_rate": 2.9908000000000002e-05, + "loss": 0.0331, + "step": 10280 + }, + { + "epoch": 8.683277027027026, + "grad_norm": 0.5167336463928223, + "learning_rate": 2.9907666666666668e-05, + "loss": 0.0487, + "step": 10281 + }, + { + "epoch": 8.684121621621621, + "grad_norm": 0.2658461928367615, + "learning_rate": 2.9907333333333334e-05, + "loss": 0.0283, + "step": 10282 + }, + { + "epoch": 8.684966216216216, + "grad_norm": 0.2857321500778198, + "learning_rate": 2.9907e-05, + "loss": 0.0333, + "step": 10283 + }, + { + "epoch": 8.68581081081081, + "grad_norm": 0.3424749970436096, + "learning_rate": 2.990666666666667e-05, + "loss": 0.0342, + "step": 10284 + }, + { + "epoch": 8.686655405405405, + "grad_norm": 0.39741477370262146, + "learning_rate": 2.9906333333333332e-05, + "loss": 0.0412, + "step": 10285 + }, + { + "epoch": 8.6875, + "grad_norm": 0.28553974628448486, + "learning_rate": 2.9906e-05, + "loss": 0.0286, + "step": 10286 + }, + { + "epoch": 8.688344594594595, + "grad_norm": 0.36239755153656006, + "learning_rate": 2.9905666666666667e-05, + "loss": 0.0413, + "step": 10287 + }, + { + "epoch": 8.68918918918919, + "grad_norm": 0.2672688663005829, + "learning_rate": 2.9905333333333333e-05, + "loss": 0.0352, + "step": 10288 + }, + { + "epoch": 8.690033783783784, + "grad_norm": 0.3510688543319702, + "learning_rate": 2.9905000000000003e-05, + "loss": 0.0343, + "step": 10289 + }, + { + "epoch": 8.690878378378379, + "grad_norm": 0.32639339566230774, + "learning_rate": 2.990466666666667e-05, + "loss": 0.027, + "step": 10290 + }, + { + "epoch": 8.691722972972974, + "grad_norm": 0.5682740807533264, + "learning_rate": 2.9904333333333334e-05, + "loss": 0.0308, + "step": 10291 + }, + { + "epoch": 8.692567567567568, + "grad_norm": 0.23320339620113373, + "learning_rate": 2.9904e-05, + "loss": 0.0179, + "step": 10292 + }, + { + "epoch": 8.693412162162161, + "grad_norm": 0.4285568594932556, + "learning_rate": 2.9903666666666666e-05, + "loss": 0.0401, + "step": 10293 + }, + { + "epoch": 8.694256756756756, + "grad_norm": 0.5003628134727478, + "learning_rate": 2.9903333333333332e-05, + "loss": 0.0347, + "step": 10294 + }, + { + "epoch": 8.69510135135135, + "grad_norm": 0.5100786089897156, + "learning_rate": 2.9903e-05, + "loss": 0.0172, + "step": 10295 + }, + { + "epoch": 8.695945945945946, + "grad_norm": 0.5013890266418457, + "learning_rate": 2.9902666666666667e-05, + "loss": 0.0306, + "step": 10296 + }, + { + "epoch": 8.69679054054054, + "grad_norm": 0.37049636244773865, + "learning_rate": 2.9902333333333333e-05, + "loss": 0.0374, + "step": 10297 + }, + { + "epoch": 8.697635135135135, + "grad_norm": 0.3338242471218109, + "learning_rate": 2.9902000000000003e-05, + "loss": 0.0238, + "step": 10298 + }, + { + "epoch": 8.69847972972973, + "grad_norm": 0.6282346844673157, + "learning_rate": 2.9901666666666665e-05, + "loss": 0.0366, + "step": 10299 + }, + { + "epoch": 8.699324324324325, + "grad_norm": 0.44355276226997375, + "learning_rate": 2.9901333333333335e-05, + "loss": 0.0304, + "step": 10300 + }, + { + "epoch": 8.70016891891892, + "grad_norm": 0.4991931915283203, + "learning_rate": 2.9901e-05, + "loss": 0.0337, + "step": 10301 + }, + { + "epoch": 8.701013513513514, + "grad_norm": 0.5835630297660828, + "learning_rate": 2.9900666666666666e-05, + "loss": 0.0315, + "step": 10302 + }, + { + "epoch": 8.701858108108109, + "grad_norm": 0.5570759177207947, + "learning_rate": 2.9900333333333332e-05, + "loss": 0.0354, + "step": 10303 + }, + { + "epoch": 8.702702702702704, + "grad_norm": 0.3082534670829773, + "learning_rate": 2.9900000000000002e-05, + "loss": 0.0245, + "step": 10304 + }, + { + "epoch": 8.703547297297296, + "grad_norm": 0.5562629103660583, + "learning_rate": 2.9899666666666668e-05, + "loss": 0.0509, + "step": 10305 + }, + { + "epoch": 8.704391891891891, + "grad_norm": 0.3394651412963867, + "learning_rate": 2.9899333333333334e-05, + "loss": 0.0284, + "step": 10306 + }, + { + "epoch": 8.705236486486486, + "grad_norm": 0.4441518187522888, + "learning_rate": 2.9899000000000003e-05, + "loss": 0.041, + "step": 10307 + }, + { + "epoch": 8.70608108108108, + "grad_norm": 0.3049052059650421, + "learning_rate": 2.9898666666666665e-05, + "loss": 0.0247, + "step": 10308 + }, + { + "epoch": 8.706925675675675, + "grad_norm": 0.938850462436676, + "learning_rate": 2.9898333333333335e-05, + "loss": 0.0372, + "step": 10309 + }, + { + "epoch": 8.70777027027027, + "grad_norm": 0.37319841980934143, + "learning_rate": 2.9898e-05, + "loss": 0.0314, + "step": 10310 + }, + { + "epoch": 8.708614864864865, + "grad_norm": 0.3801862299442291, + "learning_rate": 2.9897666666666667e-05, + "loss": 0.0388, + "step": 10311 + }, + { + "epoch": 8.70945945945946, + "grad_norm": 0.42370113730430603, + "learning_rate": 2.9897333333333336e-05, + "loss": 0.0344, + "step": 10312 + }, + { + "epoch": 8.710304054054054, + "grad_norm": 0.7294020652770996, + "learning_rate": 2.9897000000000002e-05, + "loss": 0.0571, + "step": 10313 + }, + { + "epoch": 8.71114864864865, + "grad_norm": 0.3521493077278137, + "learning_rate": 2.9896666666666668e-05, + "loss": 0.0302, + "step": 10314 + }, + { + "epoch": 8.711993243243244, + "grad_norm": 0.6813415288925171, + "learning_rate": 2.9896333333333334e-05, + "loss": 0.0466, + "step": 10315 + }, + { + "epoch": 8.712837837837839, + "grad_norm": 0.5063830614089966, + "learning_rate": 2.9896000000000003e-05, + "loss": 0.0371, + "step": 10316 + }, + { + "epoch": 8.713682432432432, + "grad_norm": 0.4293546676635742, + "learning_rate": 2.9895666666666666e-05, + "loss": 0.0351, + "step": 10317 + }, + { + "epoch": 8.714527027027026, + "grad_norm": 0.3893905580043793, + "learning_rate": 2.9895333333333335e-05, + "loss": 0.0274, + "step": 10318 + }, + { + "epoch": 8.715371621621621, + "grad_norm": 0.5537062883377075, + "learning_rate": 2.9895e-05, + "loss": 0.035, + "step": 10319 + }, + { + "epoch": 8.716216216216216, + "grad_norm": 0.3091246783733368, + "learning_rate": 2.9894666666666667e-05, + "loss": 0.0283, + "step": 10320 + }, + { + "epoch": 8.71706081081081, + "grad_norm": 0.5389294028282166, + "learning_rate": 2.9894333333333336e-05, + "loss": 0.047, + "step": 10321 + }, + { + "epoch": 8.717905405405405, + "grad_norm": 1.0262999534606934, + "learning_rate": 2.9894e-05, + "loss": 0.0716, + "step": 10322 + }, + { + "epoch": 8.71875, + "grad_norm": 0.9892295002937317, + "learning_rate": 2.9893666666666668e-05, + "loss": 0.3028, + "step": 10323 + }, + { + "epoch": 8.719594594594595, + "grad_norm": 0.5884274244308472, + "learning_rate": 2.9893333333333334e-05, + "loss": 0.1166, + "step": 10324 + }, + { + "epoch": 8.72043918918919, + "grad_norm": 0.7941684126853943, + "learning_rate": 2.9893e-05, + "loss": 0.126, + "step": 10325 + }, + { + "epoch": 8.721283783783784, + "grad_norm": 0.3010696470737457, + "learning_rate": 2.9892666666666666e-05, + "loss": 0.0439, + "step": 10326 + }, + { + "epoch": 8.722128378378379, + "grad_norm": 0.29132530093193054, + "learning_rate": 2.9892333333333335e-05, + "loss": 0.0376, + "step": 10327 + }, + { + "epoch": 8.722972972972974, + "grad_norm": 0.3162004351615906, + "learning_rate": 2.9891999999999998e-05, + "loss": 0.0247, + "step": 10328 + }, + { + "epoch": 8.723817567567568, + "grad_norm": 0.370413213968277, + "learning_rate": 2.9891666666666667e-05, + "loss": 0.0472, + "step": 10329 + }, + { + "epoch": 8.724662162162161, + "grad_norm": 0.4455658793449402, + "learning_rate": 2.9891333333333336e-05, + "loss": 0.0356, + "step": 10330 + }, + { + "epoch": 8.725506756756756, + "grad_norm": 0.2848832309246063, + "learning_rate": 2.9891e-05, + "loss": 0.031, + "step": 10331 + }, + { + "epoch": 8.72635135135135, + "grad_norm": 0.3504176437854767, + "learning_rate": 2.9890666666666668e-05, + "loss": 0.0342, + "step": 10332 + }, + { + "epoch": 8.727195945945946, + "grad_norm": 0.45834049582481384, + "learning_rate": 2.9890333333333334e-05, + "loss": 0.0518, + "step": 10333 + }, + { + "epoch": 8.72804054054054, + "grad_norm": 0.3452169597148895, + "learning_rate": 2.989e-05, + "loss": 0.0302, + "step": 10334 + }, + { + "epoch": 8.728885135135135, + "grad_norm": 0.5031566619873047, + "learning_rate": 2.9889666666666666e-05, + "loss": 0.0448, + "step": 10335 + }, + { + "epoch": 8.72972972972973, + "grad_norm": 0.9200527667999268, + "learning_rate": 2.9889333333333335e-05, + "loss": 0.0568, + "step": 10336 + }, + { + "epoch": 8.730574324324325, + "grad_norm": 0.3184449374675751, + "learning_rate": 2.9889e-05, + "loss": 0.0387, + "step": 10337 + }, + { + "epoch": 8.73141891891892, + "grad_norm": 0.4162719249725342, + "learning_rate": 2.9888666666666667e-05, + "loss": 0.0447, + "step": 10338 + }, + { + "epoch": 8.732263513513514, + "grad_norm": 0.5374095439910889, + "learning_rate": 2.9888333333333337e-05, + "loss": 0.0473, + "step": 10339 + }, + { + "epoch": 8.733108108108109, + "grad_norm": 0.24874922633171082, + "learning_rate": 2.9888e-05, + "loss": 0.0278, + "step": 10340 + }, + { + "epoch": 8.733952702702704, + "grad_norm": 0.4453246593475342, + "learning_rate": 2.988766666666667e-05, + "loss": 0.0431, + "step": 10341 + }, + { + "epoch": 8.734797297297296, + "grad_norm": 0.333804726600647, + "learning_rate": 2.9887333333333334e-05, + "loss": 0.0413, + "step": 10342 + }, + { + "epoch": 8.735641891891891, + "grad_norm": 0.31283700466156006, + "learning_rate": 2.9887e-05, + "loss": 0.0368, + "step": 10343 + }, + { + "epoch": 8.736486486486486, + "grad_norm": 0.43168529868125916, + "learning_rate": 2.9886666666666666e-05, + "loss": 0.0403, + "step": 10344 + }, + { + "epoch": 8.73733108108108, + "grad_norm": 0.3076123297214508, + "learning_rate": 2.9886333333333336e-05, + "loss": 0.0188, + "step": 10345 + }, + { + "epoch": 8.738175675675675, + "grad_norm": 0.34453511238098145, + "learning_rate": 2.9886e-05, + "loss": 0.0373, + "step": 10346 + }, + { + "epoch": 8.73902027027027, + "grad_norm": 0.3154539167881012, + "learning_rate": 2.9885666666666667e-05, + "loss": 0.0283, + "step": 10347 + }, + { + "epoch": 8.739864864864865, + "grad_norm": 0.4347982406616211, + "learning_rate": 2.9885333333333337e-05, + "loss": 0.0447, + "step": 10348 + }, + { + "epoch": 8.74070945945946, + "grad_norm": 0.43185001611709595, + "learning_rate": 2.9885e-05, + "loss": 0.0361, + "step": 10349 + }, + { + "epoch": 8.741554054054054, + "grad_norm": 0.48760658502578735, + "learning_rate": 2.988466666666667e-05, + "loss": 0.0346, + "step": 10350 + }, + { + "epoch": 8.74239864864865, + "grad_norm": 1.260246992111206, + "learning_rate": 2.9884333333333335e-05, + "loss": 0.0271, + "step": 10351 + }, + { + "epoch": 8.743243243243244, + "grad_norm": 0.28046441078186035, + "learning_rate": 2.9884e-05, + "loss": 0.0227, + "step": 10352 + }, + { + "epoch": 8.744087837837839, + "grad_norm": 0.35758116841316223, + "learning_rate": 2.9883666666666666e-05, + "loss": 0.0221, + "step": 10353 + }, + { + "epoch": 8.744932432432432, + "grad_norm": 0.3389449715614319, + "learning_rate": 2.9883333333333332e-05, + "loss": 0.0203, + "step": 10354 + }, + { + "epoch": 8.745777027027026, + "grad_norm": 0.6159526109695435, + "learning_rate": 2.9883000000000002e-05, + "loss": 0.0385, + "step": 10355 + }, + { + "epoch": 8.746621621621621, + "grad_norm": 0.28183290362358093, + "learning_rate": 2.9882666666666668e-05, + "loss": 0.0213, + "step": 10356 + }, + { + "epoch": 8.747466216216216, + "grad_norm": 0.6719575524330139, + "learning_rate": 2.9882333333333334e-05, + "loss": 0.0523, + "step": 10357 + }, + { + "epoch": 8.74831081081081, + "grad_norm": 0.43035048246383667, + "learning_rate": 2.9882e-05, + "loss": 0.0432, + "step": 10358 + }, + { + "epoch": 8.749155405405405, + "grad_norm": 0.5363022685050964, + "learning_rate": 2.988166666666667e-05, + "loss": 0.048, + "step": 10359 + }, + { + "epoch": 8.75, + "grad_norm": 0.3953740894794464, + "learning_rate": 2.988133333333333e-05, + "loss": 0.0359, + "step": 10360 + }, + { + "epoch": 8.750844594594595, + "grad_norm": 0.4714615046977997, + "learning_rate": 2.9881e-05, + "loss": 0.0346, + "step": 10361 + }, + { + "epoch": 8.75168918918919, + "grad_norm": 0.6065964698791504, + "learning_rate": 2.988066666666667e-05, + "loss": 0.0401, + "step": 10362 + }, + { + "epoch": 8.752533783783784, + "grad_norm": 0.48903870582580566, + "learning_rate": 2.9880333333333333e-05, + "loss": 0.0327, + "step": 10363 + }, + { + "epoch": 8.753378378378379, + "grad_norm": 0.5677671432495117, + "learning_rate": 2.9880000000000002e-05, + "loss": 0.0346, + "step": 10364 + }, + { + "epoch": 8.754222972972974, + "grad_norm": 0.4310459792613983, + "learning_rate": 2.9879666666666668e-05, + "loss": 0.0255, + "step": 10365 + }, + { + "epoch": 8.755067567567568, + "grad_norm": 0.3992469608783722, + "learning_rate": 2.9879333333333334e-05, + "loss": 0.0308, + "step": 10366 + }, + { + "epoch": 8.755912162162161, + "grad_norm": 0.4151046574115753, + "learning_rate": 2.9879e-05, + "loss": 0.0328, + "step": 10367 + }, + { + "epoch": 8.756756756756756, + "grad_norm": 0.4690706133842468, + "learning_rate": 2.987866666666667e-05, + "loss": 0.0468, + "step": 10368 + }, + { + "epoch": 8.75760135135135, + "grad_norm": 0.48011329770088196, + "learning_rate": 2.987833333333333e-05, + "loss": 0.0473, + "step": 10369 + }, + { + "epoch": 8.758445945945946, + "grad_norm": 0.8340473175048828, + "learning_rate": 2.9878e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 8.75929054054054, + "grad_norm": 0.7271409630775452, + "learning_rate": 2.987766666666667e-05, + "loss": 0.0586, + "step": 10371 + }, + { + "epoch": 8.760135135135135, + "grad_norm": 0.8699197173118591, + "learning_rate": 2.9877333333333333e-05, + "loss": 0.0672, + "step": 10372 + }, + { + "epoch": 8.76097972972973, + "grad_norm": 0.8298888206481934, + "learning_rate": 2.9877000000000002e-05, + "loss": 0.3062, + "step": 10373 + }, + { + "epoch": 8.761824324324325, + "grad_norm": 0.7788753509521484, + "learning_rate": 2.9876666666666668e-05, + "loss": 0.1608, + "step": 10374 + }, + { + "epoch": 8.76266891891892, + "grad_norm": 0.40989306569099426, + "learning_rate": 2.9876333333333334e-05, + "loss": 0.081, + "step": 10375 + }, + { + "epoch": 8.763513513513514, + "grad_norm": 0.4702301621437073, + "learning_rate": 2.9876e-05, + "loss": 0.0508, + "step": 10376 + }, + { + "epoch": 8.764358108108109, + "grad_norm": 0.6272776126861572, + "learning_rate": 2.987566666666667e-05, + "loss": 0.052, + "step": 10377 + }, + { + "epoch": 8.765202702702704, + "grad_norm": 0.5454742312431335, + "learning_rate": 2.9875333333333332e-05, + "loss": 0.0447, + "step": 10378 + }, + { + "epoch": 8.766047297297296, + "grad_norm": 0.4605020582675934, + "learning_rate": 2.9875e-05, + "loss": 0.0464, + "step": 10379 + }, + { + "epoch": 8.766891891891891, + "grad_norm": 0.2880735695362091, + "learning_rate": 2.987466666666667e-05, + "loss": 0.0408, + "step": 10380 + }, + { + "epoch": 8.767736486486486, + "grad_norm": 0.3455023467540741, + "learning_rate": 2.9874333333333333e-05, + "loss": 0.0336, + "step": 10381 + }, + { + "epoch": 8.76858108108108, + "grad_norm": 0.530762791633606, + "learning_rate": 2.9874000000000002e-05, + "loss": 0.0453, + "step": 10382 + }, + { + "epoch": 8.769425675675675, + "grad_norm": 0.40522879362106323, + "learning_rate": 2.9873666666666665e-05, + "loss": 0.0465, + "step": 10383 + }, + { + "epoch": 8.77027027027027, + "grad_norm": 0.36667922139167786, + "learning_rate": 2.9873333333333334e-05, + "loss": 0.0269, + "step": 10384 + }, + { + "epoch": 8.771114864864865, + "grad_norm": 0.3255351483821869, + "learning_rate": 2.9873e-05, + "loss": 0.031, + "step": 10385 + }, + { + "epoch": 8.77195945945946, + "grad_norm": 0.4896239638328552, + "learning_rate": 2.9872666666666666e-05, + "loss": 0.0509, + "step": 10386 + }, + { + "epoch": 8.772804054054054, + "grad_norm": 0.23119904100894928, + "learning_rate": 2.9872333333333335e-05, + "loss": 0.0273, + "step": 10387 + }, + { + "epoch": 8.77364864864865, + "grad_norm": 0.5297448635101318, + "learning_rate": 2.9872e-05, + "loss": 0.0304, + "step": 10388 + }, + { + "epoch": 8.774493243243244, + "grad_norm": 0.23841506242752075, + "learning_rate": 2.9871666666666667e-05, + "loss": 0.0283, + "step": 10389 + }, + { + "epoch": 8.775337837837839, + "grad_norm": 0.32499074935913086, + "learning_rate": 2.9871333333333333e-05, + "loss": 0.0254, + "step": 10390 + }, + { + "epoch": 8.776182432432432, + "grad_norm": 0.567251980304718, + "learning_rate": 2.9871000000000003e-05, + "loss": 0.0344, + "step": 10391 + }, + { + "epoch": 8.777027027027026, + "grad_norm": 0.6479577422142029, + "learning_rate": 2.9870666666666665e-05, + "loss": 0.0351, + "step": 10392 + }, + { + "epoch": 8.777871621621621, + "grad_norm": 0.7805019021034241, + "learning_rate": 2.9870333333333334e-05, + "loss": 0.0272, + "step": 10393 + }, + { + "epoch": 8.778716216216216, + "grad_norm": 0.3487522006034851, + "learning_rate": 2.987e-05, + "loss": 0.0228, + "step": 10394 + }, + { + "epoch": 8.77956081081081, + "grad_norm": 0.6188486218452454, + "learning_rate": 2.9869666666666666e-05, + "loss": 0.0429, + "step": 10395 + }, + { + "epoch": 8.780405405405405, + "grad_norm": 0.40825390815734863, + "learning_rate": 2.9869333333333336e-05, + "loss": 0.0351, + "step": 10396 + }, + { + "epoch": 8.78125, + "grad_norm": 0.4788919985294342, + "learning_rate": 2.9869e-05, + "loss": 0.0334, + "step": 10397 + }, + { + "epoch": 8.782094594594595, + "grad_norm": 0.24637579917907715, + "learning_rate": 2.9868666666666667e-05, + "loss": 0.0225, + "step": 10398 + }, + { + "epoch": 8.78293918918919, + "grad_norm": 0.4701840877532959, + "learning_rate": 2.9868333333333333e-05, + "loss": 0.0346, + "step": 10399 + }, + { + "epoch": 8.783783783783784, + "grad_norm": 0.4717649221420288, + "learning_rate": 2.9868000000000003e-05, + "loss": 0.0371, + "step": 10400 + }, + { + "epoch": 8.784628378378379, + "grad_norm": 0.508380651473999, + "learning_rate": 2.9867666666666665e-05, + "loss": 0.0424, + "step": 10401 + }, + { + "epoch": 8.785472972972974, + "grad_norm": 0.39527204632759094, + "learning_rate": 2.9867333333333335e-05, + "loss": 0.025, + "step": 10402 + }, + { + "epoch": 8.786317567567568, + "grad_norm": 0.4182630181312561, + "learning_rate": 2.9867e-05, + "loss": 0.0428, + "step": 10403 + }, + { + "epoch": 8.787162162162161, + "grad_norm": 0.4790171682834625, + "learning_rate": 2.9866666666666666e-05, + "loss": 0.0265, + "step": 10404 + }, + { + "epoch": 8.788006756756756, + "grad_norm": 0.35827088356018066, + "learning_rate": 2.9866333333333336e-05, + "loss": 0.031, + "step": 10405 + }, + { + "epoch": 8.78885135135135, + "grad_norm": 0.38431471586227417, + "learning_rate": 2.9866000000000002e-05, + "loss": 0.0328, + "step": 10406 + }, + { + "epoch": 8.789695945945946, + "grad_norm": 0.48376160860061646, + "learning_rate": 2.9865666666666668e-05, + "loss": 0.0273, + "step": 10407 + }, + { + "epoch": 8.79054054054054, + "grad_norm": 0.3727709650993347, + "learning_rate": 2.9865333333333334e-05, + "loss": 0.0327, + "step": 10408 + }, + { + "epoch": 8.791385135135135, + "grad_norm": 0.5176593065261841, + "learning_rate": 2.9865000000000003e-05, + "loss": 0.0388, + "step": 10409 + }, + { + "epoch": 8.79222972972973, + "grad_norm": 0.2543307840824127, + "learning_rate": 2.9864666666666665e-05, + "loss": 0.0204, + "step": 10410 + }, + { + "epoch": 8.793074324324325, + "grad_norm": 0.6594468355178833, + "learning_rate": 2.9864333333333335e-05, + "loss": 0.0488, + "step": 10411 + }, + { + "epoch": 8.79391891891892, + "grad_norm": 0.43105077743530273, + "learning_rate": 2.9864000000000004e-05, + "loss": 0.064, + "step": 10412 + }, + { + "epoch": 8.794763513513514, + "grad_norm": 0.6629084348678589, + "learning_rate": 2.9863666666666667e-05, + "loss": 0.0451, + "step": 10413 + }, + { + "epoch": 8.795608108108109, + "grad_norm": 0.4075204133987427, + "learning_rate": 2.9863333333333336e-05, + "loss": 0.0335, + "step": 10414 + }, + { + "epoch": 8.796452702702704, + "grad_norm": 0.9507907629013062, + "learning_rate": 2.9863e-05, + "loss": 0.0291, + "step": 10415 + }, + { + "epoch": 8.797297297297296, + "grad_norm": 0.33569222688674927, + "learning_rate": 2.9862666666666668e-05, + "loss": 0.0326, + "step": 10416 + }, + { + "epoch": 8.798141891891891, + "grad_norm": 0.5025061368942261, + "learning_rate": 2.9862333333333334e-05, + "loss": 0.0426, + "step": 10417 + }, + { + "epoch": 8.798986486486486, + "grad_norm": 0.4705736041069031, + "learning_rate": 2.9862e-05, + "loss": 0.0451, + "step": 10418 + }, + { + "epoch": 8.79983108108108, + "grad_norm": 0.7827404737472534, + "learning_rate": 2.9861666666666666e-05, + "loss": 0.0525, + "step": 10419 + }, + { + "epoch": 8.800675675675675, + "grad_norm": 0.5876433849334717, + "learning_rate": 2.9861333333333335e-05, + "loss": 0.054, + "step": 10420 + }, + { + "epoch": 8.80152027027027, + "grad_norm": 0.8260504603385925, + "learning_rate": 2.9861e-05, + "loss": 0.0372, + "step": 10421 + }, + { + "epoch": 8.802364864864865, + "grad_norm": 0.9055311679840088, + "learning_rate": 2.9860666666666667e-05, + "loss": 0.0521, + "step": 10422 + }, + { + "epoch": 8.80320945945946, + "grad_norm": 0.6776443719863892, + "learning_rate": 2.9860333333333336e-05, + "loss": 0.2331, + "step": 10423 + }, + { + "epoch": 8.804054054054054, + "grad_norm": 0.898996114730835, + "learning_rate": 2.986e-05, + "loss": 0.1589, + "step": 10424 + }, + { + "epoch": 8.80489864864865, + "grad_norm": 0.6576224565505981, + "learning_rate": 2.9859666666666668e-05, + "loss": 0.0844, + "step": 10425 + }, + { + "epoch": 8.805743243243244, + "grad_norm": 0.3642970025539398, + "learning_rate": 2.9859333333333334e-05, + "loss": 0.0532, + "step": 10426 + }, + { + "epoch": 8.806587837837839, + "grad_norm": 0.29299288988113403, + "learning_rate": 2.9859e-05, + "loss": 0.0487, + "step": 10427 + }, + { + "epoch": 8.807432432432432, + "grad_norm": 0.3808072507381439, + "learning_rate": 2.9858666666666666e-05, + "loss": 0.0335, + "step": 10428 + }, + { + "epoch": 8.808277027027026, + "grad_norm": 0.4202885031700134, + "learning_rate": 2.9858333333333335e-05, + "loss": 0.0591, + "step": 10429 + }, + { + "epoch": 8.809121621621621, + "grad_norm": 0.4049687087535858, + "learning_rate": 2.9858e-05, + "loss": 0.0395, + "step": 10430 + }, + { + "epoch": 8.809966216216216, + "grad_norm": 0.6581936478614807, + "learning_rate": 2.9857666666666667e-05, + "loss": 0.0569, + "step": 10431 + }, + { + "epoch": 8.81081081081081, + "grad_norm": 0.3260175287723541, + "learning_rate": 2.9857333333333336e-05, + "loss": 0.0381, + "step": 10432 + }, + { + "epoch": 8.811655405405405, + "grad_norm": 0.37167876958847046, + "learning_rate": 2.9857e-05, + "loss": 0.0288, + "step": 10433 + }, + { + "epoch": 8.8125, + "grad_norm": 0.4089118540287018, + "learning_rate": 2.9856666666666668e-05, + "loss": 0.0389, + "step": 10434 + }, + { + "epoch": 8.813344594594595, + "grad_norm": 0.39999818801879883, + "learning_rate": 2.9856333333333334e-05, + "loss": 0.0347, + "step": 10435 + }, + { + "epoch": 8.81418918918919, + "grad_norm": 0.28729546070098877, + "learning_rate": 2.9856e-05, + "loss": 0.0308, + "step": 10436 + }, + { + "epoch": 8.815033783783784, + "grad_norm": 0.21724559366703033, + "learning_rate": 2.985566666666667e-05, + "loss": 0.0294, + "step": 10437 + }, + { + "epoch": 8.815878378378379, + "grad_norm": 0.5473465919494629, + "learning_rate": 2.9855333333333335e-05, + "loss": 0.0341, + "step": 10438 + }, + { + "epoch": 8.816722972972974, + "grad_norm": 0.3430584967136383, + "learning_rate": 2.9855e-05, + "loss": 0.034, + "step": 10439 + }, + { + "epoch": 8.817567567567568, + "grad_norm": 0.3984736502170563, + "learning_rate": 2.9854666666666667e-05, + "loss": 0.0304, + "step": 10440 + }, + { + "epoch": 8.818412162162161, + "grad_norm": 0.45838016271591187, + "learning_rate": 2.9854333333333337e-05, + "loss": 0.0462, + "step": 10441 + }, + { + "epoch": 8.819256756756756, + "grad_norm": 0.730785071849823, + "learning_rate": 2.9854e-05, + "loss": 0.1142, + "step": 10442 + }, + { + "epoch": 8.82010135135135, + "grad_norm": 0.3495108187198639, + "learning_rate": 2.985366666666667e-05, + "loss": 0.0411, + "step": 10443 + }, + { + "epoch": 8.820945945945946, + "grad_norm": 0.42796966433525085, + "learning_rate": 2.985333333333333e-05, + "loss": 0.0313, + "step": 10444 + }, + { + "epoch": 8.82179054054054, + "grad_norm": 0.34789153933525085, + "learning_rate": 2.9853e-05, + "loss": 0.0307, + "step": 10445 + }, + { + "epoch": 8.822635135135135, + "grad_norm": 0.5627415776252747, + "learning_rate": 2.985266666666667e-05, + "loss": 0.0347, + "step": 10446 + }, + { + "epoch": 8.82347972972973, + "grad_norm": 0.41764846444129944, + "learning_rate": 2.9852333333333332e-05, + "loss": 0.0325, + "step": 10447 + }, + { + "epoch": 8.824324324324325, + "grad_norm": 0.46140140295028687, + "learning_rate": 2.9852e-05, + "loss": 0.0323, + "step": 10448 + }, + { + "epoch": 8.82516891891892, + "grad_norm": 0.32104960083961487, + "learning_rate": 2.9851666666666667e-05, + "loss": 0.0375, + "step": 10449 + }, + { + "epoch": 8.826013513513514, + "grad_norm": 0.6575955152511597, + "learning_rate": 2.9851333333333333e-05, + "loss": 0.0483, + "step": 10450 + }, + { + "epoch": 8.826858108108109, + "grad_norm": 0.3779786229133606, + "learning_rate": 2.9851e-05, + "loss": 0.0273, + "step": 10451 + }, + { + "epoch": 8.827702702702704, + "grad_norm": 0.33272045850753784, + "learning_rate": 2.985066666666667e-05, + "loss": 0.0318, + "step": 10452 + }, + { + "epoch": 8.828547297297296, + "grad_norm": 0.31696173548698425, + "learning_rate": 2.985033333333333e-05, + "loss": 0.0196, + "step": 10453 + }, + { + "epoch": 8.829391891891891, + "grad_norm": 0.6247479915618896, + "learning_rate": 2.985e-05, + "loss": 0.0558, + "step": 10454 + }, + { + "epoch": 8.830236486486486, + "grad_norm": 0.6075249910354614, + "learning_rate": 2.984966666666667e-05, + "loss": 0.041, + "step": 10455 + }, + { + "epoch": 8.83108108108108, + "grad_norm": 0.3685269355773926, + "learning_rate": 2.9849333333333332e-05, + "loss": 0.0273, + "step": 10456 + }, + { + "epoch": 8.831925675675675, + "grad_norm": 0.38344788551330566, + "learning_rate": 2.9849000000000002e-05, + "loss": 0.0348, + "step": 10457 + }, + { + "epoch": 8.83277027027027, + "grad_norm": 0.35215458273887634, + "learning_rate": 2.9848666666666668e-05, + "loss": 0.0293, + "step": 10458 + }, + { + "epoch": 8.833614864864865, + "grad_norm": 0.41046467423439026, + "learning_rate": 2.9848333333333334e-05, + "loss": 0.0284, + "step": 10459 + }, + { + "epoch": 8.83445945945946, + "grad_norm": 0.395139217376709, + "learning_rate": 2.9848e-05, + "loss": 0.0307, + "step": 10460 + }, + { + "epoch": 8.835304054054054, + "grad_norm": 0.47123444080352783, + "learning_rate": 2.984766666666667e-05, + "loss": 0.0315, + "step": 10461 + }, + { + "epoch": 8.83614864864865, + "grad_norm": 0.6984919905662537, + "learning_rate": 2.9847333333333335e-05, + "loss": 0.0266, + "step": 10462 + }, + { + "epoch": 8.836993243243244, + "grad_norm": 0.45944178104400635, + "learning_rate": 2.9847e-05, + "loss": 0.0481, + "step": 10463 + }, + { + "epoch": 8.837837837837839, + "grad_norm": 0.7353988885879517, + "learning_rate": 2.984666666666667e-05, + "loss": 0.0347, + "step": 10464 + }, + { + "epoch": 8.838682432432432, + "grad_norm": 0.5188436508178711, + "learning_rate": 2.9846333333333333e-05, + "loss": 0.043, + "step": 10465 + }, + { + "epoch": 8.839527027027026, + "grad_norm": 0.6088624596595764, + "learning_rate": 2.9846000000000002e-05, + "loss": 0.0416, + "step": 10466 + }, + { + "epoch": 8.840371621621621, + "grad_norm": 0.4118756949901581, + "learning_rate": 2.9845666666666668e-05, + "loss": 0.0234, + "step": 10467 + }, + { + "epoch": 8.841216216216216, + "grad_norm": 0.3312223553657532, + "learning_rate": 2.9845333333333334e-05, + "loss": 0.029, + "step": 10468 + }, + { + "epoch": 8.84206081081081, + "grad_norm": 0.40767863392829895, + "learning_rate": 2.9845e-05, + "loss": 0.0397, + "step": 10469 + }, + { + "epoch": 8.842905405405405, + "grad_norm": 0.8356384634971619, + "learning_rate": 2.984466666666667e-05, + "loss": 0.0528, + "step": 10470 + }, + { + "epoch": 8.84375, + "grad_norm": 0.8301467895507812, + "learning_rate": 2.9844333333333335e-05, + "loss": 0.0373, + "step": 10471 + }, + { + "epoch": 8.844594594594595, + "grad_norm": 1.4877049922943115, + "learning_rate": 2.9844e-05, + "loss": 0.0529, + "step": 10472 + }, + { + "epoch": 8.84543918918919, + "grad_norm": 0.7513725161552429, + "learning_rate": 2.9843666666666667e-05, + "loss": 0.2765, + "step": 10473 + }, + { + "epoch": 8.846283783783784, + "grad_norm": 0.7343279123306274, + "learning_rate": 2.9843333333333333e-05, + "loss": 0.1675, + "step": 10474 + }, + { + "epoch": 8.847128378378379, + "grad_norm": 0.9880204796791077, + "learning_rate": 2.9843000000000002e-05, + "loss": 0.0784, + "step": 10475 + }, + { + "epoch": 8.847972972972974, + "grad_norm": 0.5076517462730408, + "learning_rate": 2.9842666666666665e-05, + "loss": 0.0586, + "step": 10476 + }, + { + "epoch": 8.848817567567568, + "grad_norm": 0.42531198263168335, + "learning_rate": 2.9842333333333334e-05, + "loss": 0.0489, + "step": 10477 + }, + { + "epoch": 8.849662162162161, + "grad_norm": 0.7126221656799316, + "learning_rate": 2.9842e-05, + "loss": 0.0532, + "step": 10478 + }, + { + "epoch": 8.850506756756756, + "grad_norm": 0.3610564172267914, + "learning_rate": 2.9841666666666666e-05, + "loss": 0.0365, + "step": 10479 + }, + { + "epoch": 8.85135135135135, + "grad_norm": 0.2728837728500366, + "learning_rate": 2.9841333333333335e-05, + "loss": 0.0407, + "step": 10480 + }, + { + "epoch": 8.852195945945946, + "grad_norm": 0.5315036773681641, + "learning_rate": 2.9841e-05, + "loss": 0.0446, + "step": 10481 + }, + { + "epoch": 8.85304054054054, + "grad_norm": 0.5597589015960693, + "learning_rate": 2.9840666666666667e-05, + "loss": 0.0356, + "step": 10482 + }, + { + "epoch": 8.853885135135135, + "grad_norm": 0.3041195571422577, + "learning_rate": 2.9840333333333333e-05, + "loss": 0.0356, + "step": 10483 + }, + { + "epoch": 8.85472972972973, + "grad_norm": 0.43142205476760864, + "learning_rate": 2.9840000000000002e-05, + "loss": 0.0388, + "step": 10484 + }, + { + "epoch": 8.855574324324325, + "grad_norm": 0.32980477809906006, + "learning_rate": 2.9839666666666665e-05, + "loss": 0.0284, + "step": 10485 + }, + { + "epoch": 8.85641891891892, + "grad_norm": 0.4334642291069031, + "learning_rate": 2.9839333333333334e-05, + "loss": 0.0451, + "step": 10486 + }, + { + "epoch": 8.857263513513514, + "grad_norm": 0.31453198194503784, + "learning_rate": 2.9839000000000003e-05, + "loss": 0.0323, + "step": 10487 + }, + { + "epoch": 8.858108108108109, + "grad_norm": 0.39217543601989746, + "learning_rate": 2.9838666666666666e-05, + "loss": 0.0133, + "step": 10488 + }, + { + "epoch": 8.858952702702704, + "grad_norm": 0.3232496678829193, + "learning_rate": 2.9838333333333335e-05, + "loss": 0.0281, + "step": 10489 + }, + { + "epoch": 8.859797297297296, + "grad_norm": 0.30026575922966003, + "learning_rate": 2.9838e-05, + "loss": 0.0295, + "step": 10490 + }, + { + "epoch": 8.860641891891891, + "grad_norm": 0.35529664158821106, + "learning_rate": 2.9837666666666667e-05, + "loss": 0.0292, + "step": 10491 + }, + { + "epoch": 8.861486486486486, + "grad_norm": 0.484145849943161, + "learning_rate": 2.9837333333333333e-05, + "loss": 0.0434, + "step": 10492 + }, + { + "epoch": 8.86233108108108, + "grad_norm": 0.5022293925285339, + "learning_rate": 2.9837000000000002e-05, + "loss": 0.0409, + "step": 10493 + }, + { + "epoch": 8.863175675675675, + "grad_norm": 0.5644683837890625, + "learning_rate": 2.9836666666666665e-05, + "loss": 0.045, + "step": 10494 + }, + { + "epoch": 8.86402027027027, + "grad_norm": 0.39028453826904297, + "learning_rate": 2.9836333333333334e-05, + "loss": 0.0465, + "step": 10495 + }, + { + "epoch": 8.864864864864865, + "grad_norm": 0.3980507254600525, + "learning_rate": 2.9836000000000004e-05, + "loss": 0.0291, + "step": 10496 + }, + { + "epoch": 8.86570945945946, + "grad_norm": 0.3545221984386444, + "learning_rate": 2.9835666666666666e-05, + "loss": 0.0237, + "step": 10497 + }, + { + "epoch": 8.866554054054054, + "grad_norm": 0.6034770607948303, + "learning_rate": 2.9835333333333336e-05, + "loss": 0.0517, + "step": 10498 + }, + { + "epoch": 8.86739864864865, + "grad_norm": 0.46435829997062683, + "learning_rate": 2.9835e-05, + "loss": 0.0482, + "step": 10499 + }, + { + "epoch": 8.868243243243244, + "grad_norm": 0.26872649788856506, + "learning_rate": 2.9834666666666667e-05, + "loss": 0.0199, + "step": 10500 + }, + { + "epoch": 8.869087837837839, + "grad_norm": 0.25738033652305603, + "learning_rate": 2.9834333333333333e-05, + "loss": 0.0269, + "step": 10501 + }, + { + "epoch": 8.869932432432432, + "grad_norm": 0.3131464421749115, + "learning_rate": 2.9834000000000003e-05, + "loss": 0.0301, + "step": 10502 + }, + { + "epoch": 8.870777027027026, + "grad_norm": 0.39936426281929016, + "learning_rate": 2.9833666666666665e-05, + "loss": 0.0534, + "step": 10503 + }, + { + "epoch": 8.871621621621621, + "grad_norm": 0.360369473695755, + "learning_rate": 2.9833333333333335e-05, + "loss": 0.0201, + "step": 10504 + }, + { + "epoch": 8.872466216216216, + "grad_norm": 0.366769403219223, + "learning_rate": 2.9833e-05, + "loss": 0.0415, + "step": 10505 + }, + { + "epoch": 8.87331081081081, + "grad_norm": 0.3377207815647125, + "learning_rate": 2.9832666666666666e-05, + "loss": 0.0205, + "step": 10506 + }, + { + "epoch": 8.874155405405405, + "grad_norm": 0.462761789560318, + "learning_rate": 2.9832333333333336e-05, + "loss": 0.0448, + "step": 10507 + }, + { + "epoch": 8.875, + "grad_norm": 0.64863520860672, + "learning_rate": 2.9831999999999998e-05, + "loss": 0.05, + "step": 10508 + }, + { + "epoch": 8.875844594594595, + "grad_norm": 0.3847024142742157, + "learning_rate": 2.9831666666666668e-05, + "loss": 0.0241, + "step": 10509 + }, + { + "epoch": 8.87668918918919, + "grad_norm": 0.523971438407898, + "learning_rate": 2.9831333333333334e-05, + "loss": 0.0458, + "step": 10510 + }, + { + "epoch": 8.877533783783784, + "grad_norm": 0.3643254339694977, + "learning_rate": 2.9831e-05, + "loss": 0.0444, + "step": 10511 + }, + { + "epoch": 8.878378378378379, + "grad_norm": 1.457175374031067, + "learning_rate": 2.983066666666667e-05, + "loss": 0.0413, + "step": 10512 + }, + { + "epoch": 8.879222972972974, + "grad_norm": 0.3381975591182709, + "learning_rate": 2.9830333333333335e-05, + "loss": 0.0256, + "step": 10513 + }, + { + "epoch": 8.880067567567568, + "grad_norm": 0.5923339128494263, + "learning_rate": 2.983e-05, + "loss": 0.0275, + "step": 10514 + }, + { + "epoch": 8.880912162162161, + "grad_norm": 0.6500594019889832, + "learning_rate": 2.9829666666666667e-05, + "loss": 0.0354, + "step": 10515 + }, + { + "epoch": 8.881756756756756, + "grad_norm": 0.38600531220436096, + "learning_rate": 2.9829333333333336e-05, + "loss": 0.0276, + "step": 10516 + }, + { + "epoch": 8.88260135135135, + "grad_norm": 0.6631804704666138, + "learning_rate": 2.9829e-05, + "loss": 0.0366, + "step": 10517 + }, + { + "epoch": 8.883445945945946, + "grad_norm": 1.0820732116699219, + "learning_rate": 2.9828666666666668e-05, + "loss": 0.0414, + "step": 10518 + }, + { + "epoch": 8.88429054054054, + "grad_norm": 0.4040422737598419, + "learning_rate": 2.9828333333333334e-05, + "loss": 0.0294, + "step": 10519 + }, + { + "epoch": 8.885135135135135, + "grad_norm": 1.2479981184005737, + "learning_rate": 2.9828e-05, + "loss": 0.0415, + "step": 10520 + }, + { + "epoch": 8.88597972972973, + "grad_norm": 1.104303240776062, + "learning_rate": 2.982766666666667e-05, + "loss": 0.0491, + "step": 10521 + }, + { + "epoch": 8.886824324324325, + "grad_norm": 1.5847930908203125, + "learning_rate": 2.9827333333333335e-05, + "loss": 0.0624, + "step": 10522 + }, + { + "epoch": 8.88766891891892, + "grad_norm": 0.8528720736503601, + "learning_rate": 2.9827e-05, + "loss": 0.3451, + "step": 10523 + }, + { + "epoch": 8.888513513513514, + "grad_norm": 0.6238109469413757, + "learning_rate": 2.9826666666666667e-05, + "loss": 0.1665, + "step": 10524 + }, + { + "epoch": 8.889358108108109, + "grad_norm": 0.5073246359825134, + "learning_rate": 2.9826333333333336e-05, + "loss": 0.1155, + "step": 10525 + }, + { + "epoch": 8.890202702702704, + "grad_norm": 0.43824467062950134, + "learning_rate": 2.9826e-05, + "loss": 0.0637, + "step": 10526 + }, + { + "epoch": 8.891047297297296, + "grad_norm": 0.5354230403900146, + "learning_rate": 2.9825666666666668e-05, + "loss": 0.0697, + "step": 10527 + }, + { + "epoch": 8.891891891891891, + "grad_norm": 0.7506477236747742, + "learning_rate": 2.9825333333333334e-05, + "loss": 0.0616, + "step": 10528 + }, + { + "epoch": 8.892736486486486, + "grad_norm": 0.5205955505371094, + "learning_rate": 2.9825e-05, + "loss": 0.0429, + "step": 10529 + }, + { + "epoch": 8.89358108108108, + "grad_norm": 0.34622493386268616, + "learning_rate": 2.982466666666667e-05, + "loss": 0.036, + "step": 10530 + }, + { + "epoch": 8.894425675675675, + "grad_norm": 0.38695815205574036, + "learning_rate": 2.9824333333333335e-05, + "loss": 0.0408, + "step": 10531 + }, + { + "epoch": 8.89527027027027, + "grad_norm": 0.20136959850788116, + "learning_rate": 2.9824e-05, + "loss": 0.0256, + "step": 10532 + }, + { + "epoch": 8.896114864864865, + "grad_norm": 0.39088061451911926, + "learning_rate": 2.9823666666666667e-05, + "loss": 0.0257, + "step": 10533 + }, + { + "epoch": 8.89695945945946, + "grad_norm": 0.26487046480178833, + "learning_rate": 2.9823333333333333e-05, + "loss": 0.0277, + "step": 10534 + }, + { + "epoch": 8.897804054054054, + "grad_norm": 0.3015342950820923, + "learning_rate": 2.9823e-05, + "loss": 0.0367, + "step": 10535 + }, + { + "epoch": 8.89864864864865, + "grad_norm": 0.2646124064922333, + "learning_rate": 2.9822666666666668e-05, + "loss": 0.028, + "step": 10536 + }, + { + "epoch": 8.899493243243244, + "grad_norm": 0.49205946922302246, + "learning_rate": 2.9822333333333334e-05, + "loss": 0.0364, + "step": 10537 + }, + { + "epoch": 8.900337837837839, + "grad_norm": 0.3060421943664551, + "learning_rate": 2.9822e-05, + "loss": 0.0288, + "step": 10538 + }, + { + "epoch": 8.901182432432432, + "grad_norm": 0.40165409445762634, + "learning_rate": 2.982166666666667e-05, + "loss": 0.041, + "step": 10539 + }, + { + "epoch": 8.902027027027026, + "grad_norm": 0.4965469241142273, + "learning_rate": 2.9821333333333332e-05, + "loss": 0.0719, + "step": 10540 + }, + { + "epoch": 8.902871621621621, + "grad_norm": 0.47253602743148804, + "learning_rate": 2.9821e-05, + "loss": 0.0452, + "step": 10541 + }, + { + "epoch": 8.903716216216216, + "grad_norm": 0.47846469283103943, + "learning_rate": 2.9820666666666667e-05, + "loss": 0.0287, + "step": 10542 + }, + { + "epoch": 8.90456081081081, + "grad_norm": 0.3312705457210541, + "learning_rate": 2.9820333333333333e-05, + "loss": 0.0308, + "step": 10543 + }, + { + "epoch": 8.905405405405405, + "grad_norm": 0.42878207564353943, + "learning_rate": 2.982e-05, + "loss": 0.0515, + "step": 10544 + }, + { + "epoch": 8.90625, + "grad_norm": 0.372369647026062, + "learning_rate": 2.981966666666667e-05, + "loss": 0.0423, + "step": 10545 + }, + { + "epoch": 8.907094594594595, + "grad_norm": 0.2872657775878906, + "learning_rate": 2.9819333333333334e-05, + "loss": 0.0231, + "step": 10546 + }, + { + "epoch": 8.90793918918919, + "grad_norm": 0.42472702264785767, + "learning_rate": 2.9819e-05, + "loss": 0.0308, + "step": 10547 + }, + { + "epoch": 8.908783783783784, + "grad_norm": 0.4363534152507782, + "learning_rate": 2.981866666666667e-05, + "loss": 0.036, + "step": 10548 + }, + { + "epoch": 8.909628378378379, + "grad_norm": 0.1470242589712143, + "learning_rate": 2.9818333333333332e-05, + "loss": 0.0115, + "step": 10549 + }, + { + "epoch": 8.910472972972974, + "grad_norm": 0.4321247935295105, + "learning_rate": 2.9818e-05, + "loss": 0.0343, + "step": 10550 + }, + { + "epoch": 8.911317567567568, + "grad_norm": 0.3218495547771454, + "learning_rate": 2.9817666666666667e-05, + "loss": 0.0293, + "step": 10551 + }, + { + "epoch": 8.912162162162161, + "grad_norm": 0.5195683836936951, + "learning_rate": 2.9817333333333333e-05, + "loss": 0.0485, + "step": 10552 + }, + { + "epoch": 8.913006756756756, + "grad_norm": 0.36471137404441833, + "learning_rate": 2.9817e-05, + "loss": 0.0423, + "step": 10553 + }, + { + "epoch": 8.91385135135135, + "grad_norm": 0.36873891949653625, + "learning_rate": 2.981666666666667e-05, + "loss": 0.0485, + "step": 10554 + }, + { + "epoch": 8.914695945945946, + "grad_norm": 0.44486555457115173, + "learning_rate": 2.9816333333333335e-05, + "loss": 0.0336, + "step": 10555 + }, + { + "epoch": 8.91554054054054, + "grad_norm": 0.38353946805000305, + "learning_rate": 2.9816e-05, + "loss": 0.0276, + "step": 10556 + }, + { + "epoch": 8.916385135135135, + "grad_norm": 0.39127662777900696, + "learning_rate": 2.981566666666667e-05, + "loss": 0.0397, + "step": 10557 + }, + { + "epoch": 8.91722972972973, + "grad_norm": 0.4485887289047241, + "learning_rate": 2.9815333333333332e-05, + "loss": 0.0302, + "step": 10558 + }, + { + "epoch": 8.918074324324325, + "grad_norm": 0.7621645927429199, + "learning_rate": 2.9815e-05, + "loss": 0.0522, + "step": 10559 + }, + { + "epoch": 8.91891891891892, + "grad_norm": 0.2683061361312866, + "learning_rate": 2.9814666666666668e-05, + "loss": 0.0196, + "step": 10560 + }, + { + "epoch": 8.919763513513514, + "grad_norm": 0.36001136898994446, + "learning_rate": 2.9814333333333334e-05, + "loss": 0.0335, + "step": 10561 + }, + { + "epoch": 8.920608108108109, + "grad_norm": 0.4626258909702301, + "learning_rate": 2.9814000000000003e-05, + "loss": 0.0498, + "step": 10562 + }, + { + "epoch": 8.921452702702704, + "grad_norm": 0.6284313201904297, + "learning_rate": 2.981366666666667e-05, + "loss": 0.0373, + "step": 10563 + }, + { + "epoch": 8.922297297297296, + "grad_norm": 0.3293011784553528, + "learning_rate": 2.9813333333333335e-05, + "loss": 0.0305, + "step": 10564 + }, + { + "epoch": 8.923141891891891, + "grad_norm": 0.5116462111473083, + "learning_rate": 2.9813e-05, + "loss": 0.0442, + "step": 10565 + }, + { + "epoch": 8.923986486486486, + "grad_norm": 0.40468186140060425, + "learning_rate": 2.9812666666666667e-05, + "loss": 0.041, + "step": 10566 + }, + { + "epoch": 8.92483108108108, + "grad_norm": 0.45144370198249817, + "learning_rate": 2.9812333333333333e-05, + "loss": 0.0405, + "step": 10567 + }, + { + "epoch": 8.925675675675675, + "grad_norm": 0.5100306272506714, + "learning_rate": 2.9812000000000002e-05, + "loss": 0.0408, + "step": 10568 + }, + { + "epoch": 8.92652027027027, + "grad_norm": 0.6501714587211609, + "learning_rate": 2.9811666666666664e-05, + "loss": 0.0248, + "step": 10569 + }, + { + "epoch": 8.927364864864865, + "grad_norm": 0.4448523223400116, + "learning_rate": 2.9811333333333334e-05, + "loss": 0.0319, + "step": 10570 + }, + { + "epoch": 8.92820945945946, + "grad_norm": 0.5574675798416138, + "learning_rate": 2.9811000000000003e-05, + "loss": 0.0622, + "step": 10571 + }, + { + "epoch": 8.929054054054054, + "grad_norm": 0.7120972871780396, + "learning_rate": 2.9810666666666666e-05, + "loss": 0.0397, + "step": 10572 + }, + { + "epoch": 8.92989864864865, + "grad_norm": 1.0347962379455566, + "learning_rate": 2.9810333333333335e-05, + "loss": 0.2807, + "step": 10573 + }, + { + "epoch": 8.930743243243244, + "grad_norm": 0.6245193481445312, + "learning_rate": 2.981e-05, + "loss": 0.1446, + "step": 10574 + }, + { + "epoch": 8.931587837837839, + "grad_norm": 0.5803101062774658, + "learning_rate": 2.9809666666666667e-05, + "loss": 0.0906, + "step": 10575 + }, + { + "epoch": 8.932432432432432, + "grad_norm": 0.5339227318763733, + "learning_rate": 2.9809333333333333e-05, + "loss": 0.0518, + "step": 10576 + }, + { + "epoch": 8.933277027027026, + "grad_norm": 0.40816232562065125, + "learning_rate": 2.9809000000000002e-05, + "loss": 0.0424, + "step": 10577 + }, + { + "epoch": 8.934121621621621, + "grad_norm": 0.44952327013015747, + "learning_rate": 2.9808666666666665e-05, + "loss": 0.0565, + "step": 10578 + }, + { + "epoch": 8.934966216216216, + "grad_norm": 0.3679756820201874, + "learning_rate": 2.9808333333333334e-05, + "loss": 0.0459, + "step": 10579 + }, + { + "epoch": 8.93581081081081, + "grad_norm": 0.5534399151802063, + "learning_rate": 2.9808000000000003e-05, + "loss": 0.0386, + "step": 10580 + }, + { + "epoch": 8.936655405405405, + "grad_norm": 0.4371817111968994, + "learning_rate": 2.9807666666666666e-05, + "loss": 0.0327, + "step": 10581 + }, + { + "epoch": 8.9375, + "grad_norm": 0.4549400210380554, + "learning_rate": 2.9807333333333335e-05, + "loss": 0.0382, + "step": 10582 + }, + { + "epoch": 8.938344594594595, + "grad_norm": 0.39835092425346375, + "learning_rate": 2.9807e-05, + "loss": 0.0428, + "step": 10583 + }, + { + "epoch": 8.93918918918919, + "grad_norm": 0.2784932553768158, + "learning_rate": 2.9806666666666667e-05, + "loss": 0.0255, + "step": 10584 + }, + { + "epoch": 8.940033783783784, + "grad_norm": 0.3055592477321625, + "learning_rate": 2.9806333333333333e-05, + "loss": 0.0409, + "step": 10585 + }, + { + "epoch": 8.940878378378379, + "grad_norm": 0.38137149810791016, + "learning_rate": 2.9806000000000002e-05, + "loss": 0.0344, + "step": 10586 + }, + { + "epoch": 8.941722972972974, + "grad_norm": 0.4535558521747589, + "learning_rate": 2.9805666666666668e-05, + "loss": 0.0597, + "step": 10587 + }, + { + "epoch": 8.942567567567568, + "grad_norm": 0.39827898144721985, + "learning_rate": 2.9805333333333334e-05, + "loss": 0.0386, + "step": 10588 + }, + { + "epoch": 8.943412162162161, + "grad_norm": 0.7018169164657593, + "learning_rate": 2.9805000000000003e-05, + "loss": 0.0319, + "step": 10589 + }, + { + "epoch": 8.944256756756756, + "grad_norm": 0.349080890417099, + "learning_rate": 2.9804666666666666e-05, + "loss": 0.0283, + "step": 10590 + }, + { + "epoch": 8.94510135135135, + "grad_norm": 0.42001697421073914, + "learning_rate": 2.9804333333333335e-05, + "loss": 0.0278, + "step": 10591 + }, + { + "epoch": 8.945945945945946, + "grad_norm": 0.5278732180595398, + "learning_rate": 2.9804e-05, + "loss": 0.0314, + "step": 10592 + }, + { + "epoch": 8.94679054054054, + "grad_norm": 0.4700312614440918, + "learning_rate": 2.9803666666666667e-05, + "loss": 0.0318, + "step": 10593 + }, + { + "epoch": 8.947635135135135, + "grad_norm": 0.4385684132575989, + "learning_rate": 2.9803333333333333e-05, + "loss": 0.042, + "step": 10594 + }, + { + "epoch": 8.94847972972973, + "grad_norm": 0.3093216121196747, + "learning_rate": 2.9803e-05, + "loss": 0.0287, + "step": 10595 + }, + { + "epoch": 8.949324324324325, + "grad_norm": 0.36315006017684937, + "learning_rate": 2.980266666666667e-05, + "loss": 0.0401, + "step": 10596 + }, + { + "epoch": 8.95016891891892, + "grad_norm": 0.3755655884742737, + "learning_rate": 2.9802333333333334e-05, + "loss": 0.0288, + "step": 10597 + }, + { + "epoch": 8.951013513513514, + "grad_norm": 0.5911944508552551, + "learning_rate": 2.9802e-05, + "loss": 0.027, + "step": 10598 + }, + { + "epoch": 8.951858108108109, + "grad_norm": 0.23865263164043427, + "learning_rate": 2.9801666666666666e-05, + "loss": 0.0235, + "step": 10599 + }, + { + "epoch": 8.952702702702704, + "grad_norm": 0.43175840377807617, + "learning_rate": 2.9801333333333336e-05, + "loss": 0.0228, + "step": 10600 + }, + { + "epoch": 8.953547297297296, + "grad_norm": 0.41897156834602356, + "learning_rate": 2.9800999999999998e-05, + "loss": 0.0302, + "step": 10601 + }, + { + "epoch": 8.954391891891891, + "grad_norm": 0.7362703084945679, + "learning_rate": 2.9800666666666667e-05, + "loss": 0.0562, + "step": 10602 + }, + { + "epoch": 8.955236486486486, + "grad_norm": 0.5139796733856201, + "learning_rate": 2.9800333333333333e-05, + "loss": 0.0234, + "step": 10603 + }, + { + "epoch": 8.95608108108108, + "grad_norm": 0.49643903970718384, + "learning_rate": 2.98e-05, + "loss": 0.0396, + "step": 10604 + }, + { + "epoch": 8.956925675675675, + "grad_norm": 0.4343770742416382, + "learning_rate": 2.979966666666667e-05, + "loss": 0.0313, + "step": 10605 + }, + { + "epoch": 8.95777027027027, + "grad_norm": 0.46806228160858154, + "learning_rate": 2.9799333333333335e-05, + "loss": 0.0416, + "step": 10606 + }, + { + "epoch": 8.958614864864865, + "grad_norm": 0.48358359932899475, + "learning_rate": 2.9799e-05, + "loss": 0.0325, + "step": 10607 + }, + { + "epoch": 8.95945945945946, + "grad_norm": 0.5481849312782288, + "learning_rate": 2.9798666666666666e-05, + "loss": 0.0588, + "step": 10608 + }, + { + "epoch": 8.960304054054054, + "grad_norm": 0.43365678191185, + "learning_rate": 2.9798333333333336e-05, + "loss": 0.0302, + "step": 10609 + }, + { + "epoch": 8.96114864864865, + "grad_norm": 0.3537215292453766, + "learning_rate": 2.9797999999999998e-05, + "loss": 0.0273, + "step": 10610 + }, + { + "epoch": 8.961993243243244, + "grad_norm": 0.6789443492889404, + "learning_rate": 2.9797666666666668e-05, + "loss": 0.0388, + "step": 10611 + }, + { + "epoch": 8.962837837837839, + "grad_norm": 0.5852158069610596, + "learning_rate": 2.9797333333333337e-05, + "loss": 0.045, + "step": 10612 + }, + { + "epoch": 8.963682432432432, + "grad_norm": 0.47535642981529236, + "learning_rate": 2.9797e-05, + "loss": 0.0482, + "step": 10613 + }, + { + "epoch": 8.964527027027026, + "grad_norm": 0.5940541625022888, + "learning_rate": 2.979666666666667e-05, + "loss": 0.0441, + "step": 10614 + }, + { + "epoch": 8.965371621621621, + "grad_norm": 0.4849696457386017, + "learning_rate": 2.9796333333333335e-05, + "loss": 0.0482, + "step": 10615 + }, + { + "epoch": 8.966216216216216, + "grad_norm": 0.7003757953643799, + "learning_rate": 2.9796e-05, + "loss": 0.0412, + "step": 10616 + }, + { + "epoch": 8.96706081081081, + "grad_norm": 0.5499294400215149, + "learning_rate": 2.9795666666666667e-05, + "loss": 0.0299, + "step": 10617 + }, + { + "epoch": 8.967905405405405, + "grad_norm": 1.2881736755371094, + "learning_rate": 2.9795333333333336e-05, + "loss": 0.058, + "step": 10618 + }, + { + "epoch": 8.96875, + "grad_norm": 0.7997590899467468, + "learning_rate": 2.9795e-05, + "loss": 0.0393, + "step": 10619 + }, + { + "epoch": 8.969594594594595, + "grad_norm": 4.760157108306885, + "learning_rate": 2.9794666666666668e-05, + "loss": 0.1093, + "step": 10620 + }, + { + "epoch": 8.97043918918919, + "grad_norm": 0.9021480679512024, + "learning_rate": 2.9794333333333337e-05, + "loss": 0.0566, + "step": 10621 + }, + { + "epoch": 8.971283783783784, + "grad_norm": 0.7959802746772766, + "learning_rate": 2.9794e-05, + "loss": 0.0964, + "step": 10622 + }, + { + "epoch": 8.972128378378379, + "grad_norm": 0.9060487151145935, + "learning_rate": 2.979366666666667e-05, + "loss": 0.2159, + "step": 10623 + }, + { + "epoch": 8.972972972972974, + "grad_norm": 0.37051451206207275, + "learning_rate": 2.979333333333333e-05, + "loss": 0.1018, + "step": 10624 + }, + { + "epoch": 8.973817567567568, + "grad_norm": 0.4916180372238159, + "learning_rate": 2.9793e-05, + "loss": 0.06, + "step": 10625 + }, + { + "epoch": 8.974662162162161, + "grad_norm": 0.2992907166481018, + "learning_rate": 2.9792666666666667e-05, + "loss": 0.0291, + "step": 10626 + }, + { + "epoch": 8.975506756756756, + "grad_norm": 0.36702051758766174, + "learning_rate": 2.9792333333333333e-05, + "loss": 0.032, + "step": 10627 + }, + { + "epoch": 8.97635135135135, + "grad_norm": 0.4882814288139343, + "learning_rate": 2.9792e-05, + "loss": 0.0261, + "step": 10628 + }, + { + "epoch": 8.977195945945946, + "grad_norm": 0.25889670848846436, + "learning_rate": 2.9791666666666668e-05, + "loss": 0.035, + "step": 10629 + }, + { + "epoch": 8.97804054054054, + "grad_norm": 0.42303794622421265, + "learning_rate": 2.9791333333333334e-05, + "loss": 0.0306, + "step": 10630 + }, + { + "epoch": 8.978885135135135, + "grad_norm": 0.37629950046539307, + "learning_rate": 2.9791e-05, + "loss": 0.0403, + "step": 10631 + }, + { + "epoch": 8.97972972972973, + "grad_norm": 0.3138103783130646, + "learning_rate": 2.979066666666667e-05, + "loss": 0.0293, + "step": 10632 + }, + { + "epoch": 8.980574324324325, + "grad_norm": 0.4122527837753296, + "learning_rate": 2.9790333333333332e-05, + "loss": 0.0405, + "step": 10633 + }, + { + "epoch": 8.98141891891892, + "grad_norm": 0.4711419641971588, + "learning_rate": 2.979e-05, + "loss": 0.0398, + "step": 10634 + }, + { + "epoch": 8.982263513513514, + "grad_norm": 0.4508354365825653, + "learning_rate": 2.9789666666666667e-05, + "loss": 0.0514, + "step": 10635 + }, + { + "epoch": 8.983108108108109, + "grad_norm": 0.5015857815742493, + "learning_rate": 2.9789333333333333e-05, + "loss": 0.0332, + "step": 10636 + }, + { + "epoch": 8.983952702702704, + "grad_norm": 0.7338237762451172, + "learning_rate": 2.9789000000000002e-05, + "loss": 0.0388, + "step": 10637 + }, + { + "epoch": 8.984797297297296, + "grad_norm": 0.4817555546760559, + "learning_rate": 2.9788666666666668e-05, + "loss": 0.0467, + "step": 10638 + }, + { + "epoch": 8.985641891891891, + "grad_norm": 0.3091513216495514, + "learning_rate": 2.9788333333333334e-05, + "loss": 0.0287, + "step": 10639 + }, + { + "epoch": 8.986486486486486, + "grad_norm": 0.45689550042152405, + "learning_rate": 2.9788e-05, + "loss": 0.0291, + "step": 10640 + }, + { + "epoch": 8.98733108108108, + "grad_norm": 0.5821790099143982, + "learning_rate": 2.978766666666667e-05, + "loss": 0.0316, + "step": 10641 + }, + { + "epoch": 8.988175675675675, + "grad_norm": 0.3651062846183777, + "learning_rate": 2.9787333333333332e-05, + "loss": 0.0342, + "step": 10642 + }, + { + "epoch": 8.98902027027027, + "grad_norm": 0.38100457191467285, + "learning_rate": 2.9787e-05, + "loss": 0.0299, + "step": 10643 + }, + { + "epoch": 8.989864864864865, + "grad_norm": 0.4164263606071472, + "learning_rate": 2.9786666666666667e-05, + "loss": 0.0332, + "step": 10644 + }, + { + "epoch": 8.99070945945946, + "grad_norm": 0.6623162627220154, + "learning_rate": 2.9786333333333333e-05, + "loss": 0.051, + "step": 10645 + }, + { + "epoch": 8.991554054054054, + "grad_norm": 0.37227511405944824, + "learning_rate": 2.9786000000000002e-05, + "loss": 0.0242, + "step": 10646 + }, + { + "epoch": 8.99239864864865, + "grad_norm": 0.3485133647918701, + "learning_rate": 2.978566666666667e-05, + "loss": 0.0309, + "step": 10647 + }, + { + "epoch": 8.993243243243244, + "grad_norm": 0.3227030038833618, + "learning_rate": 2.9785333333333334e-05, + "loss": 0.0282, + "step": 10648 + }, + { + "epoch": 8.994087837837839, + "grad_norm": 0.46683046221733093, + "learning_rate": 2.9785e-05, + "loss": 0.0417, + "step": 10649 + }, + { + "epoch": 8.994932432432432, + "grad_norm": 0.5926070213317871, + "learning_rate": 2.978466666666667e-05, + "loss": 0.0376, + "step": 10650 + }, + { + "epoch": 8.995777027027026, + "grad_norm": 0.3708416223526001, + "learning_rate": 2.9784333333333332e-05, + "loss": 0.0351, + "step": 10651 + }, + { + "epoch": 8.996621621621621, + "grad_norm": 1.7224725484848022, + "learning_rate": 2.9784e-05, + "loss": 0.044, + "step": 10652 + }, + { + "epoch": 8.997466216216216, + "grad_norm": 0.5113392472267151, + "learning_rate": 2.9783666666666667e-05, + "loss": 0.0514, + "step": 10653 + }, + { + "epoch": 8.99831081081081, + "grad_norm": 0.5488256216049194, + "learning_rate": 2.9783333333333333e-05, + "loss": 0.0354, + "step": 10654 + }, + { + "epoch": 8.999155405405405, + "grad_norm": 0.7616792321205139, + "learning_rate": 2.9783000000000003e-05, + "loss": 0.0513, + "step": 10655 + }, + { + "epoch": 9.0, + "grad_norm": 0.8146515488624573, + "learning_rate": 2.9782666666666665e-05, + "loss": 0.0658, + "step": 10656 + }, + { + "epoch": 9.000844594594595, + "grad_norm": 0.827131986618042, + "learning_rate": 2.9782333333333335e-05, + "loss": 0.2478, + "step": 10657 + }, + { + "epoch": 9.00168918918919, + "grad_norm": 0.8404821157455444, + "learning_rate": 2.9782e-05, + "loss": 0.2492, + "step": 10658 + }, + { + "epoch": 9.002533783783784, + "grad_norm": 0.557927668094635, + "learning_rate": 2.9781666666666666e-05, + "loss": 0.0855, + "step": 10659 + }, + { + "epoch": 9.003378378378379, + "grad_norm": 0.4351741075515747, + "learning_rate": 2.9781333333333332e-05, + "loss": 0.042, + "step": 10660 + }, + { + "epoch": 9.004222972972974, + "grad_norm": 0.6085464954376221, + "learning_rate": 2.9781e-05, + "loss": 0.068, + "step": 10661 + }, + { + "epoch": 9.005067567567568, + "grad_norm": 0.34418368339538574, + "learning_rate": 2.9780666666666668e-05, + "loss": 0.0465, + "step": 10662 + }, + { + "epoch": 9.005912162162161, + "grad_norm": 0.3830600678920746, + "learning_rate": 2.9780333333333334e-05, + "loss": 0.0295, + "step": 10663 + }, + { + "epoch": 9.006756756756756, + "grad_norm": 0.29228475689888, + "learning_rate": 2.9780000000000003e-05, + "loss": 0.0286, + "step": 10664 + }, + { + "epoch": 9.00760135135135, + "grad_norm": 0.331356942653656, + "learning_rate": 2.9779666666666665e-05, + "loss": 0.0342, + "step": 10665 + }, + { + "epoch": 9.008445945945946, + "grad_norm": 0.271052747964859, + "learning_rate": 2.9779333333333335e-05, + "loss": 0.0241, + "step": 10666 + }, + { + "epoch": 9.00929054054054, + "grad_norm": 0.5255920886993408, + "learning_rate": 2.9779e-05, + "loss": 0.0364, + "step": 10667 + }, + { + "epoch": 9.010135135135135, + "grad_norm": 0.3357698321342468, + "learning_rate": 2.9778666666666667e-05, + "loss": 0.027, + "step": 10668 + }, + { + "epoch": 9.01097972972973, + "grad_norm": 0.2518639862537384, + "learning_rate": 2.9778333333333333e-05, + "loss": 0.0244, + "step": 10669 + }, + { + "epoch": 9.011824324324325, + "grad_norm": 0.33089691400527954, + "learning_rate": 2.9778000000000002e-05, + "loss": 0.0412, + "step": 10670 + }, + { + "epoch": 9.01266891891892, + "grad_norm": 0.3473902940750122, + "learning_rate": 2.9777666666666668e-05, + "loss": 0.0182, + "step": 10671 + }, + { + "epoch": 9.013513513513514, + "grad_norm": 0.4950304925441742, + "learning_rate": 2.9777333333333334e-05, + "loss": 0.0346, + "step": 10672 + }, + { + "epoch": 9.014358108108109, + "grad_norm": 0.3429747521877289, + "learning_rate": 2.9777000000000003e-05, + "loss": 0.0323, + "step": 10673 + }, + { + "epoch": 9.015202702702704, + "grad_norm": 0.42804449796676636, + "learning_rate": 2.9776666666666666e-05, + "loss": 0.0251, + "step": 10674 + }, + { + "epoch": 9.016047297297296, + "grad_norm": 0.4894997477531433, + "learning_rate": 2.9776333333333335e-05, + "loss": 0.0362, + "step": 10675 + }, + { + "epoch": 9.016891891891891, + "grad_norm": 0.40078938007354736, + "learning_rate": 2.9776e-05, + "loss": 0.0246, + "step": 10676 + }, + { + "epoch": 9.017736486486486, + "grad_norm": 0.5254570841789246, + "learning_rate": 2.9775666666666667e-05, + "loss": 0.0348, + "step": 10677 + }, + { + "epoch": 9.01858108108108, + "grad_norm": 0.5020730495452881, + "learning_rate": 2.9775333333333333e-05, + "loss": 0.0516, + "step": 10678 + }, + { + "epoch": 9.019425675675675, + "grad_norm": 0.27517032623291016, + "learning_rate": 2.9775000000000002e-05, + "loss": 0.0262, + "step": 10679 + }, + { + "epoch": 9.02027027027027, + "grad_norm": 0.7732148766517639, + "learning_rate": 2.9774666666666668e-05, + "loss": 0.02, + "step": 10680 + }, + { + "epoch": 9.021114864864865, + "grad_norm": 0.23505640029907227, + "learning_rate": 2.9774333333333334e-05, + "loss": 0.0161, + "step": 10681 + }, + { + "epoch": 9.02195945945946, + "grad_norm": 0.5302535891532898, + "learning_rate": 2.9774000000000003e-05, + "loss": 0.0351, + "step": 10682 + }, + { + "epoch": 9.022804054054054, + "grad_norm": 0.25438791513442993, + "learning_rate": 2.9773666666666666e-05, + "loss": 0.0174, + "step": 10683 + }, + { + "epoch": 9.02364864864865, + "grad_norm": 0.322693794965744, + "learning_rate": 2.9773333333333335e-05, + "loss": 0.0207, + "step": 10684 + }, + { + "epoch": 9.024493243243244, + "grad_norm": 0.38612353801727295, + "learning_rate": 2.9772999999999998e-05, + "loss": 0.0265, + "step": 10685 + }, + { + "epoch": 9.025337837837839, + "grad_norm": 0.5278806686401367, + "learning_rate": 2.9772666666666667e-05, + "loss": 0.0319, + "step": 10686 + }, + { + "epoch": 9.026182432432432, + "grad_norm": 0.29624342918395996, + "learning_rate": 2.9772333333333336e-05, + "loss": 0.0249, + "step": 10687 + }, + { + "epoch": 9.027027027027026, + "grad_norm": 0.4114421308040619, + "learning_rate": 2.9772e-05, + "loss": 0.0304, + "step": 10688 + }, + { + "epoch": 9.027871621621621, + "grad_norm": 0.31752631068229675, + "learning_rate": 2.9771666666666668e-05, + "loss": 0.0286, + "step": 10689 + }, + { + "epoch": 9.028716216216216, + "grad_norm": 0.23769307136535645, + "learning_rate": 2.9771333333333334e-05, + "loss": 0.0192, + "step": 10690 + }, + { + "epoch": 9.02956081081081, + "grad_norm": 0.7237430810928345, + "learning_rate": 2.9771e-05, + "loss": 0.0531, + "step": 10691 + }, + { + "epoch": 9.030405405405405, + "grad_norm": 0.3258844017982483, + "learning_rate": 2.9770666666666666e-05, + "loss": 0.0138, + "step": 10692 + }, + { + "epoch": 9.03125, + "grad_norm": 0.4931457042694092, + "learning_rate": 2.9770333333333335e-05, + "loss": 0.0426, + "step": 10693 + }, + { + "epoch": 9.032094594594595, + "grad_norm": 0.3510929048061371, + "learning_rate": 2.9769999999999998e-05, + "loss": 0.0255, + "step": 10694 + }, + { + "epoch": 9.03293918918919, + "grad_norm": 0.6722108721733093, + "learning_rate": 2.9769666666666667e-05, + "loss": 0.0252, + "step": 10695 + }, + { + "epoch": 9.033783783783784, + "grad_norm": 0.23422327637672424, + "learning_rate": 2.9769333333333337e-05, + "loss": 0.0141, + "step": 10696 + }, + { + "epoch": 9.034628378378379, + "grad_norm": 0.33057278394699097, + "learning_rate": 2.9769e-05, + "loss": 0.0188, + "step": 10697 + }, + { + "epoch": 9.035472972972974, + "grad_norm": 0.5441006422042847, + "learning_rate": 2.976866666666667e-05, + "loss": 0.0338, + "step": 10698 + }, + { + "epoch": 9.036317567567568, + "grad_norm": 0.5260345339775085, + "learning_rate": 2.9768333333333334e-05, + "loss": 0.0488, + "step": 10699 + }, + { + "epoch": 9.037162162162161, + "grad_norm": 0.39407992362976074, + "learning_rate": 2.9768e-05, + "loss": 0.0202, + "step": 10700 + }, + { + "epoch": 9.038006756756756, + "grad_norm": 0.6732671856880188, + "learning_rate": 2.9767666666666666e-05, + "loss": 0.0282, + "step": 10701 + }, + { + "epoch": 9.03885135135135, + "grad_norm": 1.0085967779159546, + "learning_rate": 2.9767333333333336e-05, + "loss": 0.0332, + "step": 10702 + }, + { + "epoch": 9.039695945945946, + "grad_norm": 0.38970109820365906, + "learning_rate": 2.9766999999999998e-05, + "loss": 0.0331, + "step": 10703 + }, + { + "epoch": 9.04054054054054, + "grad_norm": 0.2893628776073456, + "learning_rate": 2.9766666666666667e-05, + "loss": 0.0254, + "step": 10704 + }, + { + "epoch": 9.041385135135135, + "grad_norm": 0.7534977197647095, + "learning_rate": 2.9766333333333337e-05, + "loss": 0.0682, + "step": 10705 + }, + { + "epoch": 9.04222972972973, + "grad_norm": 0.5668459534645081, + "learning_rate": 2.9766e-05, + "loss": 0.0278, + "step": 10706 + }, + { + "epoch": 9.043074324324325, + "grad_norm": 2.3452610969543457, + "learning_rate": 2.976566666666667e-05, + "loss": 0.3256, + "step": 10707 + }, + { + "epoch": 9.04391891891892, + "grad_norm": 0.6756065487861633, + "learning_rate": 2.9765333333333335e-05, + "loss": 0.1495, + "step": 10708 + }, + { + "epoch": 9.044763513513514, + "grad_norm": 0.4850168526172638, + "learning_rate": 2.9765e-05, + "loss": 0.0772, + "step": 10709 + }, + { + "epoch": 9.045608108108109, + "grad_norm": 0.3907698690891266, + "learning_rate": 2.9764666666666666e-05, + "loss": 0.0243, + "step": 10710 + }, + { + "epoch": 9.046452702702704, + "grad_norm": 0.5026588439941406, + "learning_rate": 2.9764333333333336e-05, + "loss": 0.0508, + "step": 10711 + }, + { + "epoch": 9.047297297297296, + "grad_norm": 0.326725035905838, + "learning_rate": 2.9764e-05, + "loss": 0.0345, + "step": 10712 + }, + { + "epoch": 9.048141891891891, + "grad_norm": 0.4328555464744568, + "learning_rate": 2.9763666666666668e-05, + "loss": 0.0247, + "step": 10713 + }, + { + "epoch": 9.048986486486486, + "grad_norm": 0.35064831376075745, + "learning_rate": 2.9763333333333337e-05, + "loss": 0.0321, + "step": 10714 + }, + { + "epoch": 9.04983108108108, + "grad_norm": 0.5934526920318604, + "learning_rate": 2.9763e-05, + "loss": 0.0424, + "step": 10715 + }, + { + "epoch": 9.050675675675675, + "grad_norm": 0.314237505197525, + "learning_rate": 2.976266666666667e-05, + "loss": 0.0224, + "step": 10716 + }, + { + "epoch": 9.05152027027027, + "grad_norm": 0.4842422306537628, + "learning_rate": 2.976233333333333e-05, + "loss": 0.027, + "step": 10717 + }, + { + "epoch": 9.052364864864865, + "grad_norm": 0.26052096486091614, + "learning_rate": 2.9762e-05, + "loss": 0.0222, + "step": 10718 + }, + { + "epoch": 9.05320945945946, + "grad_norm": 0.41783297061920166, + "learning_rate": 2.9761666666666667e-05, + "loss": 0.022, + "step": 10719 + }, + { + "epoch": 9.054054054054054, + "grad_norm": 0.295421838760376, + "learning_rate": 2.9761333333333333e-05, + "loss": 0.0219, + "step": 10720 + }, + { + "epoch": 9.05489864864865, + "grad_norm": 0.4121617376804352, + "learning_rate": 2.9761000000000002e-05, + "loss": 0.0348, + "step": 10721 + }, + { + "epoch": 9.055743243243244, + "grad_norm": 0.48467305302619934, + "learning_rate": 2.9760666666666668e-05, + "loss": 0.0212, + "step": 10722 + }, + { + "epoch": 9.056587837837839, + "grad_norm": 0.3448594808578491, + "learning_rate": 2.9760333333333334e-05, + "loss": 0.021, + "step": 10723 + }, + { + "epoch": 9.057432432432432, + "grad_norm": 0.20183952152729034, + "learning_rate": 2.976e-05, + "loss": 0.0245, + "step": 10724 + }, + { + "epoch": 9.058277027027026, + "grad_norm": 0.45988699793815613, + "learning_rate": 2.975966666666667e-05, + "loss": 0.0261, + "step": 10725 + }, + { + "epoch": 9.059121621621621, + "grad_norm": 0.2943425476551056, + "learning_rate": 2.975933333333333e-05, + "loss": 0.0293, + "step": 10726 + }, + { + "epoch": 9.059966216216216, + "grad_norm": 0.40142491459846497, + "learning_rate": 2.9759e-05, + "loss": 0.0283, + "step": 10727 + }, + { + "epoch": 9.06081081081081, + "grad_norm": 0.4853915572166443, + "learning_rate": 2.9758666666666667e-05, + "loss": 0.0355, + "step": 10728 + }, + { + "epoch": 9.061655405405405, + "grad_norm": 0.5366557240486145, + "learning_rate": 2.9758333333333333e-05, + "loss": 0.0279, + "step": 10729 + }, + { + "epoch": 9.0625, + "grad_norm": 0.3104571998119354, + "learning_rate": 2.9758000000000002e-05, + "loss": 0.0144, + "step": 10730 + }, + { + "epoch": 9.063344594594595, + "grad_norm": 0.3592892289161682, + "learning_rate": 2.9757666666666668e-05, + "loss": 0.0291, + "step": 10731 + }, + { + "epoch": 9.06418918918919, + "grad_norm": 0.3479751646518707, + "learning_rate": 2.9757333333333334e-05, + "loss": 0.0319, + "step": 10732 + }, + { + "epoch": 9.065033783783784, + "grad_norm": 0.5809136033058167, + "learning_rate": 2.9757e-05, + "loss": 0.0305, + "step": 10733 + }, + { + "epoch": 9.065878378378379, + "grad_norm": 0.4362649619579315, + "learning_rate": 2.975666666666667e-05, + "loss": 0.0273, + "step": 10734 + }, + { + "epoch": 9.066722972972974, + "grad_norm": 0.4011545777320862, + "learning_rate": 2.9756333333333332e-05, + "loss": 0.0303, + "step": 10735 + }, + { + "epoch": 9.067567567567568, + "grad_norm": 0.28148314356803894, + "learning_rate": 2.9756e-05, + "loss": 0.0243, + "step": 10736 + }, + { + "epoch": 9.068412162162161, + "grad_norm": 0.48114216327667236, + "learning_rate": 2.975566666666667e-05, + "loss": 0.0269, + "step": 10737 + }, + { + "epoch": 9.069256756756756, + "grad_norm": 0.7114810347557068, + "learning_rate": 2.9755333333333333e-05, + "loss": 0.0563, + "step": 10738 + }, + { + "epoch": 9.07010135135135, + "grad_norm": 0.34570205211639404, + "learning_rate": 2.9755000000000002e-05, + "loss": 0.0319, + "step": 10739 + }, + { + "epoch": 9.070945945945946, + "grad_norm": 1.4457391500473022, + "learning_rate": 2.9754666666666668e-05, + "loss": 0.1415, + "step": 10740 + }, + { + "epoch": 9.07179054054054, + "grad_norm": 0.7047023177146912, + "learning_rate": 2.9754333333333334e-05, + "loss": 0.0928, + "step": 10741 + }, + { + "epoch": 9.072635135135135, + "grad_norm": 0.2450011521577835, + "learning_rate": 2.9754e-05, + "loss": 0.0175, + "step": 10742 + }, + { + "epoch": 9.07347972972973, + "grad_norm": 0.41010525822639465, + "learning_rate": 2.975366666666667e-05, + "loss": 0.0295, + "step": 10743 + }, + { + "epoch": 9.074324324324325, + "grad_norm": 0.4646199941635132, + "learning_rate": 2.9753333333333332e-05, + "loss": 0.0384, + "step": 10744 + }, + { + "epoch": 9.07516891891892, + "grad_norm": 0.3880273401737213, + "learning_rate": 2.9753e-05, + "loss": 0.0317, + "step": 10745 + }, + { + "epoch": 9.076013513513514, + "grad_norm": 0.49249252676963806, + "learning_rate": 2.9752666666666667e-05, + "loss": 0.0376, + "step": 10746 + }, + { + "epoch": 9.076858108108109, + "grad_norm": 0.26128482818603516, + "learning_rate": 2.9752333333333333e-05, + "loss": 0.0217, + "step": 10747 + }, + { + "epoch": 9.077702702702704, + "grad_norm": 0.49629437923431396, + "learning_rate": 2.9752000000000002e-05, + "loss": 0.0268, + "step": 10748 + }, + { + "epoch": 9.078547297297296, + "grad_norm": 0.29315027594566345, + "learning_rate": 2.9751666666666665e-05, + "loss": 0.0243, + "step": 10749 + }, + { + "epoch": 9.079391891891891, + "grad_norm": 0.6028382182121277, + "learning_rate": 2.9751333333333334e-05, + "loss": 0.0297, + "step": 10750 + }, + { + "epoch": 9.080236486486486, + "grad_norm": 1.0490577220916748, + "learning_rate": 2.9751e-05, + "loss": 0.0393, + "step": 10751 + }, + { + "epoch": 9.08108108108108, + "grad_norm": 0.5138123035430908, + "learning_rate": 2.9750666666666666e-05, + "loss": 0.0375, + "step": 10752 + }, + { + "epoch": 9.081925675675675, + "grad_norm": 1.011077880859375, + "learning_rate": 2.9750333333333332e-05, + "loss": 0.0517, + "step": 10753 + }, + { + "epoch": 9.08277027027027, + "grad_norm": 0.8093308210372925, + "learning_rate": 2.975e-05, + "loss": 0.073, + "step": 10754 + }, + { + "epoch": 9.083614864864865, + "grad_norm": 0.6616553068161011, + "learning_rate": 2.9749666666666667e-05, + "loss": 0.0447, + "step": 10755 + }, + { + "epoch": 9.08445945945946, + "grad_norm": 2.0457513332366943, + "learning_rate": 2.9749333333333333e-05, + "loss": 0.0888, + "step": 10756 + }, + { + "epoch": 9.085304054054054, + "grad_norm": 2.079265594482422, + "learning_rate": 2.9749000000000003e-05, + "loss": 0.3806, + "step": 10757 + }, + { + "epoch": 9.08614864864865, + "grad_norm": 0.9040412902832031, + "learning_rate": 2.9748666666666665e-05, + "loss": 0.1482, + "step": 10758 + }, + { + "epoch": 9.086993243243244, + "grad_norm": 0.5432671904563904, + "learning_rate": 2.9748333333333335e-05, + "loss": 0.1165, + "step": 10759 + }, + { + "epoch": 9.087837837837839, + "grad_norm": 0.4990927278995514, + "learning_rate": 2.9748e-05, + "loss": 0.0584, + "step": 10760 + }, + { + "epoch": 9.088682432432432, + "grad_norm": 0.5709635019302368, + "learning_rate": 2.9747666666666666e-05, + "loss": 0.0645, + "step": 10761 + }, + { + "epoch": 9.089527027027026, + "grad_norm": 0.38175204396247864, + "learning_rate": 2.9747333333333336e-05, + "loss": 0.0411, + "step": 10762 + }, + { + "epoch": 9.090371621621621, + "grad_norm": 0.287569135427475, + "learning_rate": 2.9747e-05, + "loss": 0.0269, + "step": 10763 + }, + { + "epoch": 9.091216216216216, + "grad_norm": 0.42658618092536926, + "learning_rate": 2.9746666666666668e-05, + "loss": 0.0451, + "step": 10764 + }, + { + "epoch": 9.09206081081081, + "grad_norm": 0.28371256589889526, + "learning_rate": 2.9746333333333334e-05, + "loss": 0.0297, + "step": 10765 + }, + { + "epoch": 9.092905405405405, + "grad_norm": 0.37969720363616943, + "learning_rate": 2.9746000000000003e-05, + "loss": 0.0352, + "step": 10766 + }, + { + "epoch": 9.09375, + "grad_norm": 0.26810869574546814, + "learning_rate": 2.9745666666666665e-05, + "loss": 0.0206, + "step": 10767 + }, + { + "epoch": 9.094594594594595, + "grad_norm": 0.27247998118400574, + "learning_rate": 2.9745333333333335e-05, + "loss": 0.0316, + "step": 10768 + }, + { + "epoch": 9.09543918918919, + "grad_norm": 0.4040432274341583, + "learning_rate": 2.9745e-05, + "loss": 0.0325, + "step": 10769 + }, + { + "epoch": 9.096283783783784, + "grad_norm": 0.20945225656032562, + "learning_rate": 2.9744666666666667e-05, + "loss": 0.0243, + "step": 10770 + }, + { + "epoch": 9.097128378378379, + "grad_norm": 0.2557338774204254, + "learning_rate": 2.9744333333333336e-05, + "loss": 0.0181, + "step": 10771 + }, + { + "epoch": 9.097972972972974, + "grad_norm": 0.30611279606819153, + "learning_rate": 2.9744000000000002e-05, + "loss": 0.0361, + "step": 10772 + }, + { + "epoch": 9.098817567567568, + "grad_norm": 0.3846624493598938, + "learning_rate": 2.9743666666666668e-05, + "loss": 0.037, + "step": 10773 + }, + { + "epoch": 9.099662162162161, + "grad_norm": 0.5016037821769714, + "learning_rate": 2.9743333333333334e-05, + "loss": 0.0434, + "step": 10774 + }, + { + "epoch": 9.100506756756756, + "grad_norm": 0.31864580512046814, + "learning_rate": 2.9743000000000003e-05, + "loss": 0.0257, + "step": 10775 + }, + { + "epoch": 9.10135135135135, + "grad_norm": 0.3588368892669678, + "learning_rate": 2.9742666666666666e-05, + "loss": 0.0357, + "step": 10776 + }, + { + "epoch": 9.102195945945946, + "grad_norm": 0.4650505483150482, + "learning_rate": 2.9742333333333335e-05, + "loss": 0.0305, + "step": 10777 + }, + { + "epoch": 9.10304054054054, + "grad_norm": 0.338365763425827, + "learning_rate": 2.9742e-05, + "loss": 0.0383, + "step": 10778 + }, + { + "epoch": 9.103885135135135, + "grad_norm": 0.8015097975730896, + "learning_rate": 2.9741666666666667e-05, + "loss": 0.0347, + "step": 10779 + }, + { + "epoch": 9.10472972972973, + "grad_norm": 0.28853079676628113, + "learning_rate": 2.9741333333333336e-05, + "loss": 0.0229, + "step": 10780 + }, + { + "epoch": 9.105574324324325, + "grad_norm": 0.42726603150367737, + "learning_rate": 2.9741e-05, + "loss": 0.0345, + "step": 10781 + }, + { + "epoch": 9.10641891891892, + "grad_norm": 0.3581509292125702, + "learning_rate": 2.9740666666666668e-05, + "loss": 0.0311, + "step": 10782 + }, + { + "epoch": 9.107263513513514, + "grad_norm": 0.2628098428249359, + "learning_rate": 2.9740333333333334e-05, + "loss": 0.0211, + "step": 10783 + }, + { + "epoch": 9.108108108108109, + "grad_norm": 0.5857098698616028, + "learning_rate": 2.974e-05, + "loss": 0.0382, + "step": 10784 + }, + { + "epoch": 9.108952702702704, + "grad_norm": 0.29706084728240967, + "learning_rate": 2.9739666666666666e-05, + "loss": 0.032, + "step": 10785 + }, + { + "epoch": 9.109797297297296, + "grad_norm": 0.6261327862739563, + "learning_rate": 2.9739333333333335e-05, + "loss": 0.026, + "step": 10786 + }, + { + "epoch": 9.110641891891891, + "grad_norm": 0.2654208838939667, + "learning_rate": 2.9739e-05, + "loss": 0.023, + "step": 10787 + }, + { + "epoch": 9.111486486486486, + "grad_norm": 0.7504265904426575, + "learning_rate": 2.9738666666666667e-05, + "loss": 0.0387, + "step": 10788 + }, + { + "epoch": 9.11233108108108, + "grad_norm": 0.32404136657714844, + "learning_rate": 2.9738333333333336e-05, + "loss": 0.0262, + "step": 10789 + }, + { + "epoch": 9.113175675675675, + "grad_norm": 0.3698839545249939, + "learning_rate": 2.9738e-05, + "loss": 0.0286, + "step": 10790 + }, + { + "epoch": 9.11402027027027, + "grad_norm": 0.36844027042388916, + "learning_rate": 2.9737666666666668e-05, + "loss": 0.0287, + "step": 10791 + }, + { + "epoch": 9.114864864864865, + "grad_norm": 0.30116620659828186, + "learning_rate": 2.9737333333333334e-05, + "loss": 0.0206, + "step": 10792 + }, + { + "epoch": 9.11570945945946, + "grad_norm": 0.5169225335121155, + "learning_rate": 2.9737e-05, + "loss": 0.0339, + "step": 10793 + }, + { + "epoch": 9.116554054054054, + "grad_norm": 0.3931833505630493, + "learning_rate": 2.9736666666666666e-05, + "loss": 0.0317, + "step": 10794 + }, + { + "epoch": 9.11739864864865, + "grad_norm": 0.34961649775505066, + "learning_rate": 2.9736333333333335e-05, + "loss": 0.0266, + "step": 10795 + }, + { + "epoch": 9.118243243243244, + "grad_norm": 0.4953862428665161, + "learning_rate": 2.9736e-05, + "loss": 0.054, + "step": 10796 + }, + { + "epoch": 9.119087837837839, + "grad_norm": 0.5340989232063293, + "learning_rate": 2.9735666666666667e-05, + "loss": 0.0326, + "step": 10797 + }, + { + "epoch": 9.119932432432432, + "grad_norm": 0.27687889337539673, + "learning_rate": 2.9735333333333337e-05, + "loss": 0.0218, + "step": 10798 + }, + { + "epoch": 9.120777027027026, + "grad_norm": 0.3759852349758148, + "learning_rate": 2.9735e-05, + "loss": 0.0314, + "step": 10799 + }, + { + "epoch": 9.121621621621621, + "grad_norm": 0.9333899021148682, + "learning_rate": 2.973466666666667e-05, + "loss": 0.0454, + "step": 10800 + }, + { + "epoch": 9.122466216216216, + "grad_norm": 0.2347603440284729, + "learning_rate": 2.9734333333333334e-05, + "loss": 0.0208, + "step": 10801 + }, + { + "epoch": 9.12331081081081, + "grad_norm": 0.5688439011573792, + "learning_rate": 2.9734e-05, + "loss": 0.0359, + "step": 10802 + }, + { + "epoch": 9.124155405405405, + "grad_norm": 0.9395857453346252, + "learning_rate": 2.9733666666666666e-05, + "loss": 0.0253, + "step": 10803 + }, + { + "epoch": 9.125, + "grad_norm": 0.48535051941871643, + "learning_rate": 2.9733333333333336e-05, + "loss": 0.0392, + "step": 10804 + }, + { + "epoch": 9.125844594594595, + "grad_norm": 2.1096248626708984, + "learning_rate": 2.9733e-05, + "loss": 0.0561, + "step": 10805 + }, + { + "epoch": 9.12668918918919, + "grad_norm": 0.808618426322937, + "learning_rate": 2.9732666666666667e-05, + "loss": 0.0427, + "step": 10806 + }, + { + "epoch": 9.127533783783784, + "grad_norm": 0.8745734691619873, + "learning_rate": 2.9732333333333333e-05, + "loss": 0.2587, + "step": 10807 + }, + { + "epoch": 9.128378378378379, + "grad_norm": 0.6593582034111023, + "learning_rate": 2.9732e-05, + "loss": 0.1856, + "step": 10808 + }, + { + "epoch": 9.129222972972974, + "grad_norm": 0.5848621129989624, + "learning_rate": 2.973166666666667e-05, + "loss": 0.1063, + "step": 10809 + }, + { + "epoch": 9.130067567567568, + "grad_norm": 0.3423236906528473, + "learning_rate": 2.973133333333333e-05, + "loss": 0.0543, + "step": 10810 + }, + { + "epoch": 9.130912162162161, + "grad_norm": 0.4674443304538727, + "learning_rate": 2.9731e-05, + "loss": 0.0376, + "step": 10811 + }, + { + "epoch": 9.131756756756756, + "grad_norm": 0.27869197726249695, + "learning_rate": 2.973066666666667e-05, + "loss": 0.0387, + "step": 10812 + }, + { + "epoch": 9.13260135135135, + "grad_norm": 0.35202261805534363, + "learning_rate": 2.9730333333333332e-05, + "loss": 0.042, + "step": 10813 + }, + { + "epoch": 9.133445945945946, + "grad_norm": 0.5612779855728149, + "learning_rate": 2.973e-05, + "loss": 0.0428, + "step": 10814 + }, + { + "epoch": 9.13429054054054, + "grad_norm": 0.32788342237472534, + "learning_rate": 2.9729666666666668e-05, + "loss": 0.0366, + "step": 10815 + }, + { + "epoch": 9.135135135135135, + "grad_norm": 0.6067360639572144, + "learning_rate": 2.9729333333333334e-05, + "loss": 0.0473, + "step": 10816 + }, + { + "epoch": 9.13597972972973, + "grad_norm": 0.4725850522518158, + "learning_rate": 2.9729e-05, + "loss": 0.0393, + "step": 10817 + }, + { + "epoch": 9.136824324324325, + "grad_norm": 0.4751059412956238, + "learning_rate": 2.972866666666667e-05, + "loss": 0.0324, + "step": 10818 + }, + { + "epoch": 9.13766891891892, + "grad_norm": 0.41220295429229736, + "learning_rate": 2.972833333333333e-05, + "loss": 0.0356, + "step": 10819 + }, + { + "epoch": 9.138513513513514, + "grad_norm": 0.23270882666110992, + "learning_rate": 2.9728e-05, + "loss": 0.0301, + "step": 10820 + }, + { + "epoch": 9.139358108108109, + "grad_norm": 0.7802748084068298, + "learning_rate": 2.972766666666667e-05, + "loss": 0.0563, + "step": 10821 + }, + { + "epoch": 9.140202702702704, + "grad_norm": 0.716010332107544, + "learning_rate": 2.9727333333333333e-05, + "loss": 0.0334, + "step": 10822 + }, + { + "epoch": 9.141047297297296, + "grad_norm": 0.38803258538246155, + "learning_rate": 2.9727000000000002e-05, + "loss": 0.0398, + "step": 10823 + }, + { + "epoch": 9.141891891891891, + "grad_norm": 0.35399961471557617, + "learning_rate": 2.9726666666666668e-05, + "loss": 0.039, + "step": 10824 + }, + { + "epoch": 9.142736486486486, + "grad_norm": 0.3485441207885742, + "learning_rate": 2.9726333333333334e-05, + "loss": 0.0311, + "step": 10825 + }, + { + "epoch": 9.14358108108108, + "grad_norm": 0.2634943723678589, + "learning_rate": 2.9726e-05, + "loss": 0.0242, + "step": 10826 + }, + { + "epoch": 9.144425675675675, + "grad_norm": 0.35129448771476746, + "learning_rate": 2.972566666666667e-05, + "loss": 0.0795, + "step": 10827 + }, + { + "epoch": 9.14527027027027, + "grad_norm": 0.27949604392051697, + "learning_rate": 2.9725333333333335e-05, + "loss": 0.0165, + "step": 10828 + }, + { + "epoch": 9.146114864864865, + "grad_norm": 0.34740525484085083, + "learning_rate": 2.9725e-05, + "loss": 0.0513, + "step": 10829 + }, + { + "epoch": 9.14695945945946, + "grad_norm": 0.2725411355495453, + "learning_rate": 2.972466666666667e-05, + "loss": 0.0317, + "step": 10830 + }, + { + "epoch": 9.147804054054054, + "grad_norm": 0.6252082586288452, + "learning_rate": 2.9724333333333333e-05, + "loss": 0.0436, + "step": 10831 + }, + { + "epoch": 9.14864864864865, + "grad_norm": 0.29694226384162903, + "learning_rate": 2.9724000000000002e-05, + "loss": 0.0455, + "step": 10832 + }, + { + "epoch": 9.149493243243244, + "grad_norm": 0.2243078649044037, + "learning_rate": 2.9723666666666668e-05, + "loss": 0.0256, + "step": 10833 + }, + { + "epoch": 9.150337837837839, + "grad_norm": 0.3519865870475769, + "learning_rate": 2.9723333333333334e-05, + "loss": 0.0316, + "step": 10834 + }, + { + "epoch": 9.151182432432432, + "grad_norm": 0.3789595067501068, + "learning_rate": 2.9723e-05, + "loss": 0.0253, + "step": 10835 + }, + { + "epoch": 9.152027027027026, + "grad_norm": 0.2799396216869354, + "learning_rate": 2.9722666666666666e-05, + "loss": 0.0296, + "step": 10836 + }, + { + "epoch": 9.152871621621621, + "grad_norm": 0.4650908410549164, + "learning_rate": 2.9722333333333335e-05, + "loss": 0.0352, + "step": 10837 + }, + { + "epoch": 9.153716216216216, + "grad_norm": 0.1725734919309616, + "learning_rate": 2.9722e-05, + "loss": 0.0154, + "step": 10838 + }, + { + "epoch": 9.15456081081081, + "grad_norm": 0.3881082236766815, + "learning_rate": 2.9721666666666667e-05, + "loss": 0.027, + "step": 10839 + }, + { + "epoch": 9.155405405405405, + "grad_norm": 0.6588100790977478, + "learning_rate": 2.9721333333333333e-05, + "loss": 0.0303, + "step": 10840 + }, + { + "epoch": 9.15625, + "grad_norm": 0.16981816291809082, + "learning_rate": 2.9721000000000002e-05, + "loss": 0.0117, + "step": 10841 + }, + { + "epoch": 9.157094594594595, + "grad_norm": 0.20978416502475739, + "learning_rate": 2.9720666666666665e-05, + "loss": 0.018, + "step": 10842 + }, + { + "epoch": 9.15793918918919, + "grad_norm": 0.7485329508781433, + "learning_rate": 2.9720333333333334e-05, + "loss": 0.0494, + "step": 10843 + }, + { + "epoch": 9.158783783783784, + "grad_norm": 1.036661148071289, + "learning_rate": 2.972e-05, + "loss": 0.0307, + "step": 10844 + }, + { + "epoch": 9.159628378378379, + "grad_norm": 0.40236371755599976, + "learning_rate": 2.9719666666666666e-05, + "loss": 0.0253, + "step": 10845 + }, + { + "epoch": 9.160472972972974, + "grad_norm": 0.24362505972385406, + "learning_rate": 2.9719333333333335e-05, + "loss": 0.0207, + "step": 10846 + }, + { + "epoch": 9.161317567567568, + "grad_norm": 0.2911420166492462, + "learning_rate": 2.9719e-05, + "loss": 0.0224, + "step": 10847 + }, + { + "epoch": 9.162162162162161, + "grad_norm": 0.5717020630836487, + "learning_rate": 2.9718666666666667e-05, + "loss": 0.0318, + "step": 10848 + }, + { + "epoch": 9.163006756756756, + "grad_norm": 0.5250759124755859, + "learning_rate": 2.9718333333333333e-05, + "loss": 0.0402, + "step": 10849 + }, + { + "epoch": 9.16385135135135, + "grad_norm": 0.3119019865989685, + "learning_rate": 2.9718000000000002e-05, + "loss": 0.0191, + "step": 10850 + }, + { + "epoch": 9.164695945945946, + "grad_norm": 0.3624788820743561, + "learning_rate": 2.9717666666666665e-05, + "loss": 0.0304, + "step": 10851 + }, + { + "epoch": 9.16554054054054, + "grad_norm": 0.7378984689712524, + "learning_rate": 2.9717333333333334e-05, + "loss": 0.0404, + "step": 10852 + }, + { + "epoch": 9.166385135135135, + "grad_norm": 0.7374562621116638, + "learning_rate": 2.9717e-05, + "loss": 0.0528, + "step": 10853 + }, + { + "epoch": 9.16722972972973, + "grad_norm": 1.7545111179351807, + "learning_rate": 2.9716666666666666e-05, + "loss": 0.0479, + "step": 10854 + }, + { + "epoch": 9.168074324324325, + "grad_norm": 1.8828704357147217, + "learning_rate": 2.9716333333333336e-05, + "loss": 0.056, + "step": 10855 + }, + { + "epoch": 9.16891891891892, + "grad_norm": 0.8676890134811401, + "learning_rate": 2.9716e-05, + "loss": 0.1087, + "step": 10856 + }, + { + "epoch": 9.169763513513514, + "grad_norm": 0.7534636855125427, + "learning_rate": 2.9715666666666667e-05, + "loss": 0.2341, + "step": 10857 + }, + { + "epoch": 9.170608108108109, + "grad_norm": 0.5206340551376343, + "learning_rate": 2.9715333333333333e-05, + "loss": 0.1469, + "step": 10858 + }, + { + "epoch": 9.171452702702704, + "grad_norm": 0.5623742341995239, + "learning_rate": 2.9715000000000003e-05, + "loss": 0.1021, + "step": 10859 + }, + { + "epoch": 9.172297297297296, + "grad_norm": 0.688697099685669, + "learning_rate": 2.9714666666666665e-05, + "loss": 0.0556, + "step": 10860 + }, + { + "epoch": 9.173141891891891, + "grad_norm": 0.6284759640693665, + "learning_rate": 2.9714333333333335e-05, + "loss": 0.0512, + "step": 10861 + }, + { + "epoch": 9.173986486486486, + "grad_norm": 0.40805649757385254, + "learning_rate": 2.9714000000000004e-05, + "loss": 0.0369, + "step": 10862 + }, + { + "epoch": 9.17483108108108, + "grad_norm": 0.27320268750190735, + "learning_rate": 2.9713666666666666e-05, + "loss": 0.0305, + "step": 10863 + }, + { + "epoch": 9.175675675675675, + "grad_norm": 0.37530091404914856, + "learning_rate": 2.9713333333333336e-05, + "loss": 0.0308, + "step": 10864 + }, + { + "epoch": 9.17652027027027, + "grad_norm": 0.2290542870759964, + "learning_rate": 2.9713e-05, + "loss": 0.0201, + "step": 10865 + }, + { + "epoch": 9.177364864864865, + "grad_norm": 0.389905720949173, + "learning_rate": 2.9712666666666668e-05, + "loss": 0.0333, + "step": 10866 + }, + { + "epoch": 9.17820945945946, + "grad_norm": 0.3553386926651001, + "learning_rate": 2.9712333333333334e-05, + "loss": 0.0261, + "step": 10867 + }, + { + "epoch": 9.179054054054054, + "grad_norm": 0.30695489048957825, + "learning_rate": 2.9712e-05, + "loss": 0.0197, + "step": 10868 + }, + { + "epoch": 9.17989864864865, + "grad_norm": 0.5333396792411804, + "learning_rate": 2.9711666666666665e-05, + "loss": 0.0392, + "step": 10869 + }, + { + "epoch": 9.180743243243244, + "grad_norm": 0.47488200664520264, + "learning_rate": 2.9711333333333335e-05, + "loss": 0.0742, + "step": 10870 + }, + { + "epoch": 9.181587837837839, + "grad_norm": 0.2787235379219055, + "learning_rate": 2.9711e-05, + "loss": 0.0221, + "step": 10871 + }, + { + "epoch": 9.182432432432432, + "grad_norm": 0.41993820667266846, + "learning_rate": 2.9710666666666667e-05, + "loss": 0.0363, + "step": 10872 + }, + { + "epoch": 9.183277027027026, + "grad_norm": 0.4449954032897949, + "learning_rate": 2.9710333333333336e-05, + "loss": 0.0394, + "step": 10873 + }, + { + "epoch": 9.184121621621621, + "grad_norm": 0.3338054120540619, + "learning_rate": 2.971e-05, + "loss": 0.0311, + "step": 10874 + }, + { + "epoch": 9.184966216216216, + "grad_norm": 0.6008748412132263, + "learning_rate": 2.9709666666666668e-05, + "loss": 0.0402, + "step": 10875 + }, + { + "epoch": 9.18581081081081, + "grad_norm": 0.5060599446296692, + "learning_rate": 2.9709333333333334e-05, + "loss": 0.0372, + "step": 10876 + }, + { + "epoch": 9.186655405405405, + "grad_norm": 0.6230486631393433, + "learning_rate": 2.9709e-05, + "loss": 0.0432, + "step": 10877 + }, + { + "epoch": 9.1875, + "grad_norm": 0.35284167528152466, + "learning_rate": 2.970866666666667e-05, + "loss": 0.0359, + "step": 10878 + }, + { + "epoch": 9.188344594594595, + "grad_norm": 0.3481127917766571, + "learning_rate": 2.9708333333333335e-05, + "loss": 0.0259, + "step": 10879 + }, + { + "epoch": 9.18918918918919, + "grad_norm": 0.46403768658638, + "learning_rate": 2.9708e-05, + "loss": 0.0702, + "step": 10880 + }, + { + "epoch": 9.190033783783784, + "grad_norm": 0.3763406276702881, + "learning_rate": 2.9707666666666667e-05, + "loss": 0.0279, + "step": 10881 + }, + { + "epoch": 9.190878378378379, + "grad_norm": 0.26850906014442444, + "learning_rate": 2.9707333333333336e-05, + "loss": 0.0256, + "step": 10882 + }, + { + "epoch": 9.191722972972974, + "grad_norm": 0.22012406587600708, + "learning_rate": 2.9707e-05, + "loss": 0.0279, + "step": 10883 + }, + { + "epoch": 9.192567567567568, + "grad_norm": 0.24399475753307343, + "learning_rate": 2.9706666666666668e-05, + "loss": 0.0202, + "step": 10884 + }, + { + "epoch": 9.193412162162161, + "grad_norm": 0.27495917677879333, + "learning_rate": 2.9706333333333334e-05, + "loss": 0.0217, + "step": 10885 + }, + { + "epoch": 9.194256756756756, + "grad_norm": 0.3966186046600342, + "learning_rate": 2.9706e-05, + "loss": 0.0269, + "step": 10886 + }, + { + "epoch": 9.19510135135135, + "grad_norm": 0.23030497133731842, + "learning_rate": 2.970566666666667e-05, + "loss": 0.0159, + "step": 10887 + }, + { + "epoch": 9.195945945945946, + "grad_norm": 0.41154715418815613, + "learning_rate": 2.9705333333333335e-05, + "loss": 0.046, + "step": 10888 + }, + { + "epoch": 9.19679054054054, + "grad_norm": 0.36227405071258545, + "learning_rate": 2.9705e-05, + "loss": 0.0334, + "step": 10889 + }, + { + "epoch": 9.197635135135135, + "grad_norm": 0.5283817648887634, + "learning_rate": 2.9704666666666667e-05, + "loss": 0.0348, + "step": 10890 + }, + { + "epoch": 9.19847972972973, + "grad_norm": 0.3242928981781006, + "learning_rate": 2.9704333333333336e-05, + "loss": 0.0271, + "step": 10891 + }, + { + "epoch": 9.199324324324325, + "grad_norm": 0.25663578510284424, + "learning_rate": 2.9704e-05, + "loss": 0.027, + "step": 10892 + }, + { + "epoch": 9.20016891891892, + "grad_norm": 0.37971732020378113, + "learning_rate": 2.9703666666666668e-05, + "loss": 0.0349, + "step": 10893 + }, + { + "epoch": 9.201013513513514, + "grad_norm": 0.3541983962059021, + "learning_rate": 2.9703333333333334e-05, + "loss": 0.0237, + "step": 10894 + }, + { + "epoch": 9.201858108108109, + "grad_norm": 0.3213193714618683, + "learning_rate": 2.9703e-05, + "loss": 0.0264, + "step": 10895 + }, + { + "epoch": 9.202702702702704, + "grad_norm": 0.29612302780151367, + "learning_rate": 2.970266666666667e-05, + "loss": 0.0225, + "step": 10896 + }, + { + "epoch": 9.203547297297296, + "grad_norm": 0.3376896381378174, + "learning_rate": 2.9702333333333332e-05, + "loss": 0.0268, + "step": 10897 + }, + { + "epoch": 9.204391891891891, + "grad_norm": 0.43843257427215576, + "learning_rate": 2.9702e-05, + "loss": 0.0392, + "step": 10898 + }, + { + "epoch": 9.205236486486486, + "grad_norm": 0.4906822144985199, + "learning_rate": 2.9701666666666667e-05, + "loss": 0.0355, + "step": 10899 + }, + { + "epoch": 9.20608108108108, + "grad_norm": 0.5027570724487305, + "learning_rate": 2.9701333333333333e-05, + "loss": 0.0344, + "step": 10900 + }, + { + "epoch": 9.206925675675675, + "grad_norm": 0.3426749110221863, + "learning_rate": 2.9701e-05, + "loss": 0.0174, + "step": 10901 + }, + { + "epoch": 9.20777027027027, + "grad_norm": 0.4736900329589844, + "learning_rate": 2.970066666666667e-05, + "loss": 0.0367, + "step": 10902 + }, + { + "epoch": 9.208614864864865, + "grad_norm": 0.834811806678772, + "learning_rate": 2.9700333333333334e-05, + "loss": 0.0505, + "step": 10903 + }, + { + "epoch": 9.20945945945946, + "grad_norm": 0.46560633182525635, + "learning_rate": 2.97e-05, + "loss": 0.0332, + "step": 10904 + }, + { + "epoch": 9.210304054054054, + "grad_norm": 1.6600408554077148, + "learning_rate": 2.969966666666667e-05, + "loss": 0.1321, + "step": 10905 + }, + { + "epoch": 9.21114864864865, + "grad_norm": 0.7073683142662048, + "learning_rate": 2.9699333333333332e-05, + "loss": 0.0359, + "step": 10906 + }, + { + "epoch": 9.211993243243244, + "grad_norm": 0.9668869972229004, + "learning_rate": 2.9699e-05, + "loss": 0.3056, + "step": 10907 + }, + { + "epoch": 9.212837837837839, + "grad_norm": 1.399893045425415, + "learning_rate": 2.9698666666666667e-05, + "loss": 0.1977, + "step": 10908 + }, + { + "epoch": 9.213682432432432, + "grad_norm": 0.48222506046295166, + "learning_rate": 2.9698333333333333e-05, + "loss": 0.0965, + "step": 10909 + }, + { + "epoch": 9.214527027027026, + "grad_norm": 0.2672390043735504, + "learning_rate": 2.9698e-05, + "loss": 0.0588, + "step": 10910 + }, + { + "epoch": 9.215371621621621, + "grad_norm": 0.4761359989643097, + "learning_rate": 2.969766666666667e-05, + "loss": 0.0485, + "step": 10911 + }, + { + "epoch": 9.216216216216216, + "grad_norm": 0.3391360640525818, + "learning_rate": 2.9697333333333335e-05, + "loss": 0.0314, + "step": 10912 + }, + { + "epoch": 9.21706081081081, + "grad_norm": 0.42176902294158936, + "learning_rate": 2.9697e-05, + "loss": 0.0428, + "step": 10913 + }, + { + "epoch": 9.217905405405405, + "grad_norm": 0.3882176876068115, + "learning_rate": 2.969666666666667e-05, + "loss": 0.0323, + "step": 10914 + }, + { + "epoch": 9.21875, + "grad_norm": 0.38121476769447327, + "learning_rate": 2.9696333333333332e-05, + "loss": 0.0344, + "step": 10915 + }, + { + "epoch": 9.219594594594595, + "grad_norm": 0.6861129999160767, + "learning_rate": 2.9696e-05, + "loss": 0.0285, + "step": 10916 + }, + { + "epoch": 9.22043918918919, + "grad_norm": 0.5186518430709839, + "learning_rate": 2.9695666666666668e-05, + "loss": 0.0376, + "step": 10917 + }, + { + "epoch": 9.221283783783784, + "grad_norm": 0.3390999436378479, + "learning_rate": 2.9695333333333334e-05, + "loss": 0.0308, + "step": 10918 + }, + { + "epoch": 9.222128378378379, + "grad_norm": 0.28980034589767456, + "learning_rate": 2.9695e-05, + "loss": 0.0246, + "step": 10919 + }, + { + "epoch": 9.222972972972974, + "grad_norm": 0.29470592737197876, + "learning_rate": 2.969466666666667e-05, + "loss": 0.0242, + "step": 10920 + }, + { + "epoch": 9.223817567567568, + "grad_norm": 0.470750093460083, + "learning_rate": 2.9694333333333335e-05, + "loss": 0.0369, + "step": 10921 + }, + { + "epoch": 9.224662162162161, + "grad_norm": 0.21512526273727417, + "learning_rate": 2.9694e-05, + "loss": 0.0156, + "step": 10922 + }, + { + "epoch": 9.225506756756756, + "grad_norm": 0.24548247456550598, + "learning_rate": 2.969366666666667e-05, + "loss": 0.0207, + "step": 10923 + }, + { + "epoch": 9.22635135135135, + "grad_norm": 0.2988262474536896, + "learning_rate": 2.9693333333333333e-05, + "loss": 0.0201, + "step": 10924 + }, + { + "epoch": 9.227195945945946, + "grad_norm": 0.3857201337814331, + "learning_rate": 2.9693000000000002e-05, + "loss": 0.0266, + "step": 10925 + }, + { + "epoch": 9.22804054054054, + "grad_norm": 0.3151756227016449, + "learning_rate": 2.9692666666666668e-05, + "loss": 0.0347, + "step": 10926 + }, + { + "epoch": 9.228885135135135, + "grad_norm": 0.46917885541915894, + "learning_rate": 2.9692333333333334e-05, + "loss": 0.0432, + "step": 10927 + }, + { + "epoch": 9.22972972972973, + "grad_norm": 0.5184805989265442, + "learning_rate": 2.9692000000000003e-05, + "loss": 0.0208, + "step": 10928 + }, + { + "epoch": 9.230574324324325, + "grad_norm": 0.22664448618888855, + "learning_rate": 2.9691666666666666e-05, + "loss": 0.0197, + "step": 10929 + }, + { + "epoch": 9.23141891891892, + "grad_norm": 0.5383444428443909, + "learning_rate": 2.9691333333333335e-05, + "loss": 0.063, + "step": 10930 + }, + { + "epoch": 9.232263513513514, + "grad_norm": 0.29889243841171265, + "learning_rate": 2.9691e-05, + "loss": 0.0368, + "step": 10931 + }, + { + "epoch": 9.233108108108109, + "grad_norm": 0.3752661645412445, + "learning_rate": 2.9690666666666667e-05, + "loss": 0.0303, + "step": 10932 + }, + { + "epoch": 9.233952702702704, + "grad_norm": 0.4357648193836212, + "learning_rate": 2.9690333333333333e-05, + "loss": 0.035, + "step": 10933 + }, + { + "epoch": 9.234797297297296, + "grad_norm": 0.4168435037136078, + "learning_rate": 2.9690000000000002e-05, + "loss": 0.0293, + "step": 10934 + }, + { + "epoch": 9.235641891891891, + "grad_norm": 0.3224618136882782, + "learning_rate": 2.9689666666666665e-05, + "loss": 0.0289, + "step": 10935 + }, + { + "epoch": 9.236486486486486, + "grad_norm": 0.34966444969177246, + "learning_rate": 2.9689333333333334e-05, + "loss": 0.0338, + "step": 10936 + }, + { + "epoch": 9.23733108108108, + "grad_norm": 0.4755205512046814, + "learning_rate": 2.9689000000000003e-05, + "loss": 0.0479, + "step": 10937 + }, + { + "epoch": 9.238175675675675, + "grad_norm": 0.2731526494026184, + "learning_rate": 2.9688666666666666e-05, + "loss": 0.0296, + "step": 10938 + }, + { + "epoch": 9.23902027027027, + "grad_norm": 0.21455763280391693, + "learning_rate": 2.9688333333333335e-05, + "loss": 0.0249, + "step": 10939 + }, + { + "epoch": 9.239864864864865, + "grad_norm": 0.3356986343860626, + "learning_rate": 2.9688e-05, + "loss": 0.0278, + "step": 10940 + }, + { + "epoch": 9.24070945945946, + "grad_norm": 0.2545987069606781, + "learning_rate": 2.9687666666666667e-05, + "loss": 0.0221, + "step": 10941 + }, + { + "epoch": 9.241554054054054, + "grad_norm": 0.40413907170295715, + "learning_rate": 2.9687333333333333e-05, + "loss": 0.0266, + "step": 10942 + }, + { + "epoch": 9.24239864864865, + "grad_norm": 0.35348644852638245, + "learning_rate": 2.9687000000000002e-05, + "loss": 0.0276, + "step": 10943 + }, + { + "epoch": 9.243243243243244, + "grad_norm": 1.6649906635284424, + "learning_rate": 2.9686666666666665e-05, + "loss": 0.0383, + "step": 10944 + }, + { + "epoch": 9.244087837837839, + "grad_norm": 0.36535435914993286, + "learning_rate": 2.9686333333333334e-05, + "loss": 0.0384, + "step": 10945 + }, + { + "epoch": 9.244932432432432, + "grad_norm": 0.37122249603271484, + "learning_rate": 2.9686000000000003e-05, + "loss": 0.0229, + "step": 10946 + }, + { + "epoch": 9.245777027027026, + "grad_norm": 0.47144803404808044, + "learning_rate": 2.9685666666666666e-05, + "loss": 0.0286, + "step": 10947 + }, + { + "epoch": 9.246621621621621, + "grad_norm": 0.34946271777153015, + "learning_rate": 2.9685333333333335e-05, + "loss": 0.028, + "step": 10948 + }, + { + "epoch": 9.247466216216216, + "grad_norm": 0.9527696371078491, + "learning_rate": 2.9685e-05, + "loss": 0.0363, + "step": 10949 + }, + { + "epoch": 9.24831081081081, + "grad_norm": 0.32067397236824036, + "learning_rate": 2.9684666666666667e-05, + "loss": 0.0233, + "step": 10950 + }, + { + "epoch": 9.249155405405405, + "grad_norm": 0.5504553318023682, + "learning_rate": 2.9684333333333333e-05, + "loss": 0.0392, + "step": 10951 + }, + { + "epoch": 9.25, + "grad_norm": 0.2180592119693756, + "learning_rate": 2.9684000000000002e-05, + "loss": 0.0191, + "step": 10952 + }, + { + "epoch": 9.250844594594595, + "grad_norm": 0.9173136353492737, + "learning_rate": 2.968366666666667e-05, + "loss": 0.0604, + "step": 10953 + }, + { + "epoch": 9.25168918918919, + "grad_norm": 0.548980176448822, + "learning_rate": 2.9683333333333334e-05, + "loss": 0.0292, + "step": 10954 + }, + { + "epoch": 9.252533783783784, + "grad_norm": 0.5278878211975098, + "learning_rate": 2.9683000000000004e-05, + "loss": 0.0605, + "step": 10955 + }, + { + "epoch": 9.253378378378379, + "grad_norm": 0.9152034521102905, + "learning_rate": 2.9682666666666666e-05, + "loss": 0.079, + "step": 10956 + }, + { + "epoch": 9.254222972972974, + "grad_norm": 0.9028202891349792, + "learning_rate": 2.9682333333333335e-05, + "loss": 0.2381, + "step": 10957 + }, + { + "epoch": 9.255067567567568, + "grad_norm": 0.6272192597389221, + "learning_rate": 2.9681999999999998e-05, + "loss": 0.1385, + "step": 10958 + }, + { + "epoch": 9.255912162162161, + "grad_norm": 0.4134725332260132, + "learning_rate": 2.9681666666666667e-05, + "loss": 0.072, + "step": 10959 + }, + { + "epoch": 9.256756756756756, + "grad_norm": 0.3764966130256653, + "learning_rate": 2.9681333333333333e-05, + "loss": 0.0557, + "step": 10960 + }, + { + "epoch": 9.25760135135135, + "grad_norm": 0.5029194355010986, + "learning_rate": 2.9681e-05, + "loss": 0.0348, + "step": 10961 + }, + { + "epoch": 9.258445945945946, + "grad_norm": 0.4803997576236725, + "learning_rate": 2.968066666666667e-05, + "loss": 0.0594, + "step": 10962 + }, + { + "epoch": 9.25929054054054, + "grad_norm": 0.3119674324989319, + "learning_rate": 2.9680333333333334e-05, + "loss": 0.0303, + "step": 10963 + }, + { + "epoch": 9.260135135135135, + "grad_norm": 0.3788740932941437, + "learning_rate": 2.968e-05, + "loss": 0.0577, + "step": 10964 + }, + { + "epoch": 9.26097972972973, + "grad_norm": 0.2620469629764557, + "learning_rate": 2.9679666666666666e-05, + "loss": 0.0312, + "step": 10965 + }, + { + "epoch": 9.261824324324325, + "grad_norm": 0.28563880920410156, + "learning_rate": 2.9679333333333336e-05, + "loss": 0.0313, + "step": 10966 + }, + { + "epoch": 9.26266891891892, + "grad_norm": 0.21546588838100433, + "learning_rate": 2.9678999999999998e-05, + "loss": 0.0199, + "step": 10967 + }, + { + "epoch": 9.263513513513514, + "grad_norm": 0.35715335607528687, + "learning_rate": 2.9678666666666668e-05, + "loss": 0.0248, + "step": 10968 + }, + { + "epoch": 9.264358108108109, + "grad_norm": 0.27275899052619934, + "learning_rate": 2.9678333333333334e-05, + "loss": 0.0306, + "step": 10969 + }, + { + "epoch": 9.265202702702704, + "grad_norm": 0.597184956073761, + "learning_rate": 2.9678e-05, + "loss": 0.0396, + "step": 10970 + }, + { + "epoch": 9.266047297297296, + "grad_norm": 0.40979963541030884, + "learning_rate": 2.967766666666667e-05, + "loss": 0.0305, + "step": 10971 + }, + { + "epoch": 9.266891891891891, + "grad_norm": 0.1714208573102951, + "learning_rate": 2.9677333333333335e-05, + "loss": 0.0115, + "step": 10972 + }, + { + "epoch": 9.267736486486486, + "grad_norm": 0.2750939726829529, + "learning_rate": 2.9677e-05, + "loss": 0.0283, + "step": 10973 + }, + { + "epoch": 9.26858108108108, + "grad_norm": 0.3796765208244324, + "learning_rate": 2.9676666666666667e-05, + "loss": 0.0391, + "step": 10974 + }, + { + "epoch": 9.269425675675675, + "grad_norm": 0.6056897640228271, + "learning_rate": 2.9676333333333336e-05, + "loss": 0.0279, + "step": 10975 + }, + { + "epoch": 9.27027027027027, + "grad_norm": 0.20404545962810516, + "learning_rate": 2.9676e-05, + "loss": 0.0251, + "step": 10976 + }, + { + "epoch": 9.271114864864865, + "grad_norm": 0.4056433141231537, + "learning_rate": 2.9675666666666668e-05, + "loss": 0.0348, + "step": 10977 + }, + { + "epoch": 9.27195945945946, + "grad_norm": 0.5783957242965698, + "learning_rate": 2.9675333333333337e-05, + "loss": 0.0235, + "step": 10978 + }, + { + "epoch": 9.272804054054054, + "grad_norm": 0.5819535851478577, + "learning_rate": 2.9675e-05, + "loss": 0.098, + "step": 10979 + }, + { + "epoch": 9.27364864864865, + "grad_norm": 0.47789350152015686, + "learning_rate": 2.967466666666667e-05, + "loss": 0.02, + "step": 10980 + }, + { + "epoch": 9.274493243243244, + "grad_norm": 0.2864607572555542, + "learning_rate": 2.9674333333333335e-05, + "loss": 0.0241, + "step": 10981 + }, + { + "epoch": 9.275337837837839, + "grad_norm": 0.27490368485450745, + "learning_rate": 2.9674e-05, + "loss": 0.0272, + "step": 10982 + }, + { + "epoch": 9.276182432432432, + "grad_norm": 0.34629595279693604, + "learning_rate": 2.9673666666666667e-05, + "loss": 0.038, + "step": 10983 + }, + { + "epoch": 9.277027027027026, + "grad_norm": 0.30977070331573486, + "learning_rate": 2.9673333333333336e-05, + "loss": 0.0252, + "step": 10984 + }, + { + "epoch": 9.277871621621621, + "grad_norm": 0.598230242729187, + "learning_rate": 2.9673e-05, + "loss": 0.0302, + "step": 10985 + }, + { + "epoch": 9.278716216216216, + "grad_norm": 0.6328708529472351, + "learning_rate": 2.9672666666666668e-05, + "loss": 0.02, + "step": 10986 + }, + { + "epoch": 9.27956081081081, + "grad_norm": 0.25146710872650146, + "learning_rate": 2.9672333333333334e-05, + "loss": 0.0213, + "step": 10987 + }, + { + "epoch": 9.280405405405405, + "grad_norm": 0.2701118290424347, + "learning_rate": 2.9672e-05, + "loss": 0.0303, + "step": 10988 + }, + { + "epoch": 9.28125, + "grad_norm": 0.23531103134155273, + "learning_rate": 2.967166666666667e-05, + "loss": 0.0187, + "step": 10989 + }, + { + "epoch": 9.282094594594595, + "grad_norm": 0.3980681896209717, + "learning_rate": 2.9671333333333332e-05, + "loss": 0.0274, + "step": 10990 + }, + { + "epoch": 9.28293918918919, + "grad_norm": 0.37917065620422363, + "learning_rate": 2.9671e-05, + "loss": 0.0238, + "step": 10991 + }, + { + "epoch": 9.283783783783784, + "grad_norm": 0.3214472234249115, + "learning_rate": 2.9670666666666667e-05, + "loss": 0.0237, + "step": 10992 + }, + { + "epoch": 9.284628378378379, + "grad_norm": 0.5046930909156799, + "learning_rate": 2.9670333333333333e-05, + "loss": 0.0291, + "step": 10993 + }, + { + "epoch": 9.285472972972974, + "grad_norm": 0.41611310839653015, + "learning_rate": 2.967e-05, + "loss": 0.0329, + "step": 10994 + }, + { + "epoch": 9.286317567567568, + "grad_norm": 0.29383325576782227, + "learning_rate": 2.9669666666666668e-05, + "loss": 0.0274, + "step": 10995 + }, + { + "epoch": 9.287162162162161, + "grad_norm": 0.5015666484832764, + "learning_rate": 2.9669333333333334e-05, + "loss": 0.0318, + "step": 10996 + }, + { + "epoch": 9.288006756756756, + "grad_norm": 0.41708844900131226, + "learning_rate": 2.9669e-05, + "loss": 0.0235, + "step": 10997 + }, + { + "epoch": 9.28885135135135, + "grad_norm": 0.5886309146881104, + "learning_rate": 2.966866666666667e-05, + "loss": 0.0306, + "step": 10998 + }, + { + "epoch": 9.289695945945946, + "grad_norm": 0.5202608108520508, + "learning_rate": 2.9668333333333332e-05, + "loss": 0.0379, + "step": 10999 + }, + { + "epoch": 9.29054054054054, + "grad_norm": 0.3067750930786133, + "learning_rate": 2.9668e-05, + "loss": 0.0253, + "step": 11000 + }, + { + "epoch": 9.29054054054054, + "eval_cer": 0.09083607742350013, + "eval_loss": 0.27267903089523315, + "eval_runtime": 8.5585, + "eval_samples_per_second": 11.801, + "eval_steps_per_second": 0.467, + "eval_wer": 0.27821869488536155, + "step": 11000 + }, + { + "epoch": 9.291385135135135, + "grad_norm": 0.44835326075553894, + "learning_rate": 2.9667666666666667e-05, + "loss": 0.0242, + "step": 11001 + }, + { + "epoch": 9.29222972972973, + "grad_norm": 0.5185922384262085, + "learning_rate": 2.9667333333333333e-05, + "loss": 0.0386, + "step": 11002 + }, + { + "epoch": 9.293074324324325, + "grad_norm": 0.26466047763824463, + "learning_rate": 2.9667000000000002e-05, + "loss": 0.0176, + "step": 11003 + }, + { + "epoch": 9.29391891891892, + "grad_norm": 0.46872594952583313, + "learning_rate": 2.966666666666667e-05, + "loss": 0.0326, + "step": 11004 + }, + { + "epoch": 9.294763513513514, + "grad_norm": 0.4886261820793152, + "learning_rate": 2.9666333333333334e-05, + "loss": 0.0318, + "step": 11005 + }, + { + "epoch": 9.295608108108109, + "grad_norm": 0.5405016541481018, + "learning_rate": 2.9666e-05, + "loss": 0.0328, + "step": 11006 + }, + { + "epoch": 9.296452702702704, + "grad_norm": 0.9693438410758972, + "learning_rate": 2.966566666666667e-05, + "loss": 0.2918, + "step": 11007 + }, + { + "epoch": 9.297297297297296, + "grad_norm": 0.6634128093719482, + "learning_rate": 2.9665333333333332e-05, + "loss": 0.1769, + "step": 11008 + }, + { + "epoch": 9.298141891891891, + "grad_norm": 0.3588038980960846, + "learning_rate": 2.9665e-05, + "loss": 0.0656, + "step": 11009 + }, + { + "epoch": 9.298986486486486, + "grad_norm": 0.3043780028820038, + "learning_rate": 2.9664666666666667e-05, + "loss": 0.0471, + "step": 11010 + }, + { + "epoch": 9.29983108108108, + "grad_norm": 0.2315264344215393, + "learning_rate": 2.9664333333333333e-05, + "loss": 0.0228, + "step": 11011 + }, + { + "epoch": 9.300675675675675, + "grad_norm": 0.32198864221572876, + "learning_rate": 2.9664000000000003e-05, + "loss": 0.0491, + "step": 11012 + }, + { + "epoch": 9.30152027027027, + "grad_norm": 0.27472031116485596, + "learning_rate": 2.966366666666667e-05, + "loss": 0.0255, + "step": 11013 + }, + { + "epoch": 9.302364864864865, + "grad_norm": 0.37837615609169006, + "learning_rate": 2.9663333333333334e-05, + "loss": 0.0344, + "step": 11014 + }, + { + "epoch": 9.30320945945946, + "grad_norm": 0.37656667828559875, + "learning_rate": 2.9663e-05, + "loss": 0.0338, + "step": 11015 + }, + { + "epoch": 9.304054054054054, + "grad_norm": 0.45607301592826843, + "learning_rate": 2.966266666666667e-05, + "loss": 0.0322, + "step": 11016 + }, + { + "epoch": 9.30489864864865, + "grad_norm": 0.2906550467014313, + "learning_rate": 2.9662333333333332e-05, + "loss": 0.028, + "step": 11017 + }, + { + "epoch": 9.305743243243244, + "grad_norm": 0.3846466839313507, + "learning_rate": 2.9662e-05, + "loss": 0.044, + "step": 11018 + }, + { + "epoch": 9.306587837837839, + "grad_norm": 0.2882506847381592, + "learning_rate": 2.9661666666666664e-05, + "loss": 0.0315, + "step": 11019 + }, + { + "epoch": 9.307432432432432, + "grad_norm": 0.3288223445415497, + "learning_rate": 2.9661333333333333e-05, + "loss": 0.0512, + "step": 11020 + }, + { + "epoch": 9.308277027027026, + "grad_norm": 0.29975324869155884, + "learning_rate": 2.9661000000000003e-05, + "loss": 0.0281, + "step": 11021 + }, + { + "epoch": 9.309121621621621, + "grad_norm": 0.2632562518119812, + "learning_rate": 2.9660666666666665e-05, + "loss": 0.0245, + "step": 11022 + }, + { + "epoch": 9.309966216216216, + "grad_norm": 0.2043425738811493, + "learning_rate": 2.9660333333333335e-05, + "loss": 0.0231, + "step": 11023 + }, + { + "epoch": 9.31081081081081, + "grad_norm": 0.35150402784347534, + "learning_rate": 2.966e-05, + "loss": 0.0411, + "step": 11024 + }, + { + "epoch": 9.311655405405405, + "grad_norm": 0.6026652455329895, + "learning_rate": 2.9659666666666667e-05, + "loss": 0.0459, + "step": 11025 + }, + { + "epoch": 9.3125, + "grad_norm": 0.34211406111717224, + "learning_rate": 2.9659333333333332e-05, + "loss": 0.0302, + "step": 11026 + }, + { + "epoch": 9.313344594594595, + "grad_norm": 0.3988569676876068, + "learning_rate": 2.9659000000000002e-05, + "loss": 0.0337, + "step": 11027 + }, + { + "epoch": 9.31418918918919, + "grad_norm": 0.2960256040096283, + "learning_rate": 2.9658666666666668e-05, + "loss": 0.0378, + "step": 11028 + }, + { + "epoch": 9.315033783783784, + "grad_norm": 0.3145942687988281, + "learning_rate": 2.9658333333333334e-05, + "loss": 0.0331, + "step": 11029 + }, + { + "epoch": 9.315878378378379, + "grad_norm": 0.2309262752532959, + "learning_rate": 2.9658000000000003e-05, + "loss": 0.0179, + "step": 11030 + }, + { + "epoch": 9.316722972972974, + "grad_norm": 0.24445490539073944, + "learning_rate": 2.9657666666666666e-05, + "loss": 0.0183, + "step": 11031 + }, + { + "epoch": 9.317567567567568, + "grad_norm": 0.44063428044319153, + "learning_rate": 2.9657333333333335e-05, + "loss": 0.0309, + "step": 11032 + }, + { + "epoch": 9.318412162162161, + "grad_norm": 0.33461084961891174, + "learning_rate": 2.9657e-05, + "loss": 0.0302, + "step": 11033 + }, + { + "epoch": 9.319256756756756, + "grad_norm": 0.41417577862739563, + "learning_rate": 2.9656666666666667e-05, + "loss": 0.0333, + "step": 11034 + }, + { + "epoch": 9.32010135135135, + "grad_norm": 0.3478659391403198, + "learning_rate": 2.9656333333333333e-05, + "loss": 0.0191, + "step": 11035 + }, + { + "epoch": 9.320945945945946, + "grad_norm": 0.5906790494918823, + "learning_rate": 2.9656000000000002e-05, + "loss": 0.0418, + "step": 11036 + }, + { + "epoch": 9.32179054054054, + "grad_norm": 0.3202957212924957, + "learning_rate": 2.9655666666666668e-05, + "loss": 0.0226, + "step": 11037 + }, + { + "epoch": 9.322635135135135, + "grad_norm": 0.3736867308616638, + "learning_rate": 2.9655333333333334e-05, + "loss": 0.0357, + "step": 11038 + }, + { + "epoch": 9.32347972972973, + "grad_norm": 0.2502497136592865, + "learning_rate": 2.9655000000000003e-05, + "loss": 0.0168, + "step": 11039 + }, + { + "epoch": 9.324324324324325, + "grad_norm": 0.23837479948997498, + "learning_rate": 2.9654666666666666e-05, + "loss": 0.0126, + "step": 11040 + }, + { + "epoch": 9.32516891891892, + "grad_norm": 0.5092124342918396, + "learning_rate": 2.9654333333333335e-05, + "loss": 0.0413, + "step": 11041 + }, + { + "epoch": 9.326013513513514, + "grad_norm": 0.365105539560318, + "learning_rate": 2.9654e-05, + "loss": 0.0233, + "step": 11042 + }, + { + "epoch": 9.326858108108109, + "grad_norm": 0.6763641834259033, + "learning_rate": 2.9653666666666667e-05, + "loss": 0.0549, + "step": 11043 + }, + { + "epoch": 9.327702702702704, + "grad_norm": 0.35501861572265625, + "learning_rate": 2.9653333333333333e-05, + "loss": 0.0243, + "step": 11044 + }, + { + "epoch": 9.328547297297296, + "grad_norm": 0.2850213944911957, + "learning_rate": 2.9653000000000002e-05, + "loss": 0.0253, + "step": 11045 + }, + { + "epoch": 9.329391891891891, + "grad_norm": 0.3831784129142761, + "learning_rate": 2.9652666666666668e-05, + "loss": 0.0267, + "step": 11046 + }, + { + "epoch": 9.330236486486486, + "grad_norm": 0.4933030605316162, + "learning_rate": 2.9652333333333334e-05, + "loss": 0.0277, + "step": 11047 + }, + { + "epoch": 9.33108108108108, + "grad_norm": 0.36041951179504395, + "learning_rate": 2.9652e-05, + "loss": 0.0278, + "step": 11048 + }, + { + "epoch": 9.331925675675675, + "grad_norm": 0.30489444732666016, + "learning_rate": 2.9651666666666666e-05, + "loss": 0.0255, + "step": 11049 + }, + { + "epoch": 9.33277027027027, + "grad_norm": 0.8225983381271362, + "learning_rate": 2.9651333333333335e-05, + "loss": 0.0473, + "step": 11050 + }, + { + "epoch": 9.333614864864865, + "grad_norm": 0.24842455983161926, + "learning_rate": 2.9650999999999998e-05, + "loss": 0.0164, + "step": 11051 + }, + { + "epoch": 9.33445945945946, + "grad_norm": 0.4265466630458832, + "learning_rate": 2.9650666666666667e-05, + "loss": 0.0253, + "step": 11052 + }, + { + "epoch": 9.335304054054054, + "grad_norm": 0.7108823657035828, + "learning_rate": 2.9650333333333336e-05, + "loss": 0.0238, + "step": 11053 + }, + { + "epoch": 9.33614864864865, + "grad_norm": 0.6439202427864075, + "learning_rate": 2.965e-05, + "loss": 0.0338, + "step": 11054 + }, + { + "epoch": 9.336993243243244, + "grad_norm": 1.2778277397155762, + "learning_rate": 2.964966666666667e-05, + "loss": 0.0843, + "step": 11055 + }, + { + "epoch": 9.337837837837839, + "grad_norm": 0.4690075218677521, + "learning_rate": 2.9649333333333334e-05, + "loss": 0.0293, + "step": 11056 + }, + { + "epoch": 9.338682432432432, + "grad_norm": 0.6725248694419861, + "learning_rate": 2.9649e-05, + "loss": 0.2285, + "step": 11057 + }, + { + "epoch": 9.339527027027026, + "grad_norm": 0.4490857422351837, + "learning_rate": 2.9648666666666666e-05, + "loss": 0.1405, + "step": 11058 + }, + { + "epoch": 9.340371621621621, + "grad_norm": 0.36736950278282166, + "learning_rate": 2.9648333333333335e-05, + "loss": 0.0765, + "step": 11059 + }, + { + "epoch": 9.341216216216216, + "grad_norm": 0.4009995460510254, + "learning_rate": 2.9647999999999998e-05, + "loss": 0.0538, + "step": 11060 + }, + { + "epoch": 9.34206081081081, + "grad_norm": 0.24747984111309052, + "learning_rate": 2.9647666666666667e-05, + "loss": 0.0386, + "step": 11061 + }, + { + "epoch": 9.342905405405405, + "grad_norm": 0.30767127871513367, + "learning_rate": 2.9647333333333337e-05, + "loss": 0.0366, + "step": 11062 + }, + { + "epoch": 9.34375, + "grad_norm": 0.2594124674797058, + "learning_rate": 2.9647e-05, + "loss": 0.0224, + "step": 11063 + }, + { + "epoch": 9.344594594594595, + "grad_norm": 0.3432600200176239, + "learning_rate": 2.964666666666667e-05, + "loss": 0.0249, + "step": 11064 + }, + { + "epoch": 9.34543918918919, + "grad_norm": 0.21404781937599182, + "learning_rate": 2.9646333333333334e-05, + "loss": 0.0302, + "step": 11065 + }, + { + "epoch": 9.346283783783784, + "grad_norm": 0.26172900199890137, + "learning_rate": 2.9646e-05, + "loss": 0.0303, + "step": 11066 + }, + { + "epoch": 9.347128378378379, + "grad_norm": 0.3076731264591217, + "learning_rate": 2.9645666666666666e-05, + "loss": 0.0273, + "step": 11067 + }, + { + "epoch": 9.347972972972974, + "grad_norm": 0.3794746398925781, + "learning_rate": 2.9645333333333336e-05, + "loss": 0.0352, + "step": 11068 + }, + { + "epoch": 9.348817567567568, + "grad_norm": 0.25687164068222046, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0304, + "step": 11069 + }, + { + "epoch": 9.349662162162161, + "grad_norm": 0.2527215778827667, + "learning_rate": 2.9644666666666668e-05, + "loss": 0.0293, + "step": 11070 + }, + { + "epoch": 9.350506756756756, + "grad_norm": 0.4036553204059601, + "learning_rate": 2.9644333333333337e-05, + "loss": 0.0395, + "step": 11071 + }, + { + "epoch": 9.35135135135135, + "grad_norm": 0.3577861189842224, + "learning_rate": 2.9644e-05, + "loss": 0.0205, + "step": 11072 + }, + { + "epoch": 9.352195945945946, + "grad_norm": 0.5684479475021362, + "learning_rate": 2.964366666666667e-05, + "loss": 0.0449, + "step": 11073 + }, + { + "epoch": 9.35304054054054, + "grad_norm": 0.321792334318161, + "learning_rate": 2.9643333333333335e-05, + "loss": 0.0375, + "step": 11074 + }, + { + "epoch": 9.353885135135135, + "grad_norm": 0.33270764350891113, + "learning_rate": 2.9643e-05, + "loss": 0.0237, + "step": 11075 + }, + { + "epoch": 9.35472972972973, + "grad_norm": 0.3177965581417084, + "learning_rate": 2.9642666666666667e-05, + "loss": 0.0128, + "step": 11076 + }, + { + "epoch": 9.355574324324325, + "grad_norm": 0.27257654070854187, + "learning_rate": 2.9642333333333336e-05, + "loss": 0.0222, + "step": 11077 + }, + { + "epoch": 9.35641891891892, + "grad_norm": 0.3449542820453644, + "learning_rate": 2.9642000000000002e-05, + "loss": 0.0245, + "step": 11078 + }, + { + "epoch": 9.357263513513514, + "grad_norm": 0.34923475980758667, + "learning_rate": 2.9641666666666668e-05, + "loss": 0.0305, + "step": 11079 + }, + { + "epoch": 9.358108108108109, + "grad_norm": 0.5455548763275146, + "learning_rate": 2.9641333333333334e-05, + "loss": 0.0242, + "step": 11080 + }, + { + "epoch": 9.358952702702704, + "grad_norm": 1.1130579710006714, + "learning_rate": 2.9641e-05, + "loss": 0.0222, + "step": 11081 + }, + { + "epoch": 9.359797297297296, + "grad_norm": 0.5214975476264954, + "learning_rate": 2.964066666666667e-05, + "loss": 0.0374, + "step": 11082 + }, + { + "epoch": 9.360641891891891, + "grad_norm": 0.4830789268016815, + "learning_rate": 2.964033333333333e-05, + "loss": 0.0335, + "step": 11083 + }, + { + "epoch": 9.361486486486486, + "grad_norm": 0.40804243087768555, + "learning_rate": 2.964e-05, + "loss": 0.0298, + "step": 11084 + }, + { + "epoch": 9.36233108108108, + "grad_norm": 0.39163005352020264, + "learning_rate": 2.9639666666666667e-05, + "loss": 0.0273, + "step": 11085 + }, + { + "epoch": 9.363175675675675, + "grad_norm": 0.5452848672866821, + "learning_rate": 2.9639333333333333e-05, + "loss": 0.0293, + "step": 11086 + }, + { + "epoch": 9.36402027027027, + "grad_norm": 0.3675518333911896, + "learning_rate": 2.9639000000000002e-05, + "loss": 0.0235, + "step": 11087 + }, + { + "epoch": 9.364864864864865, + "grad_norm": 0.22459805011749268, + "learning_rate": 2.9638666666666668e-05, + "loss": 0.0157, + "step": 11088 + }, + { + "epoch": 9.36570945945946, + "grad_norm": 0.44825121760368347, + "learning_rate": 2.9638333333333334e-05, + "loss": 0.0394, + "step": 11089 + }, + { + "epoch": 9.366554054054054, + "grad_norm": 0.4214710593223572, + "learning_rate": 2.9638e-05, + "loss": 0.052, + "step": 11090 + }, + { + "epoch": 9.36739864864865, + "grad_norm": 0.34505099058151245, + "learning_rate": 2.963766666666667e-05, + "loss": 0.0238, + "step": 11091 + }, + { + "epoch": 9.368243243243244, + "grad_norm": 0.4222589135169983, + "learning_rate": 2.963733333333333e-05, + "loss": 0.0414, + "step": 11092 + }, + { + "epoch": 9.369087837837839, + "grad_norm": 0.4162959158420563, + "learning_rate": 2.9637e-05, + "loss": 0.0308, + "step": 11093 + }, + { + "epoch": 9.369932432432432, + "grad_norm": 0.3173184096813202, + "learning_rate": 2.9636666666666667e-05, + "loss": 0.0405, + "step": 11094 + }, + { + "epoch": 9.370777027027026, + "grad_norm": 0.30553463101387024, + "learning_rate": 2.9636333333333333e-05, + "loss": 0.0294, + "step": 11095 + }, + { + "epoch": 9.371621621621621, + "grad_norm": 0.28632235527038574, + "learning_rate": 2.9636000000000002e-05, + "loss": 0.0249, + "step": 11096 + }, + { + "epoch": 9.372466216216216, + "grad_norm": 0.6106678247451782, + "learning_rate": 2.9635666666666668e-05, + "loss": 0.0344, + "step": 11097 + }, + { + "epoch": 9.37331081081081, + "grad_norm": 0.4318207800388336, + "learning_rate": 2.9635333333333334e-05, + "loss": 0.0219, + "step": 11098 + }, + { + "epoch": 9.374155405405405, + "grad_norm": 0.47688034176826477, + "learning_rate": 2.9635e-05, + "loss": 0.0444, + "step": 11099 + }, + { + "epoch": 9.375, + "grad_norm": 0.46485787630081177, + "learning_rate": 2.963466666666667e-05, + "loss": 0.0235, + "step": 11100 + }, + { + "epoch": 9.375844594594595, + "grad_norm": 0.33459949493408203, + "learning_rate": 2.9634333333333332e-05, + "loss": 0.0262, + "step": 11101 + }, + { + "epoch": 9.37668918918919, + "grad_norm": 0.40900006890296936, + "learning_rate": 2.9634e-05, + "loss": 0.0456, + "step": 11102 + }, + { + "epoch": 9.377533783783784, + "grad_norm": 0.35467958450317383, + "learning_rate": 2.963366666666667e-05, + "loss": 0.0302, + "step": 11103 + }, + { + "epoch": 9.378378378378379, + "grad_norm": 0.6953525543212891, + "learning_rate": 2.9633333333333333e-05, + "loss": 0.0543, + "step": 11104 + }, + { + "epoch": 9.379222972972974, + "grad_norm": 0.47420525550842285, + "learning_rate": 2.9633000000000002e-05, + "loss": 0.0356, + "step": 11105 + }, + { + "epoch": 9.380067567567568, + "grad_norm": 1.0084341764450073, + "learning_rate": 2.963266666666667e-05, + "loss": 0.0625, + "step": 11106 + }, + { + "epoch": 9.380912162162161, + "grad_norm": 0.6867261528968811, + "learning_rate": 2.9632333333333334e-05, + "loss": 0.2663, + "step": 11107 + }, + { + "epoch": 9.381756756756756, + "grad_norm": 0.5588244199752808, + "learning_rate": 2.9632e-05, + "loss": 0.1743, + "step": 11108 + }, + { + "epoch": 9.38260135135135, + "grad_norm": 0.47695082426071167, + "learning_rate": 2.9631666666666666e-05, + "loss": 0.0689, + "step": 11109 + }, + { + "epoch": 9.383445945945946, + "grad_norm": 0.36543750762939453, + "learning_rate": 2.9631333333333332e-05, + "loss": 0.0522, + "step": 11110 + }, + { + "epoch": 9.38429054054054, + "grad_norm": 0.4917876720428467, + "learning_rate": 2.9631e-05, + "loss": 0.036, + "step": 11111 + }, + { + "epoch": 9.385135135135135, + "grad_norm": 0.4364568889141083, + "learning_rate": 2.9630666666666667e-05, + "loss": 0.046, + "step": 11112 + }, + { + "epoch": 9.38597972972973, + "grad_norm": 0.5043002367019653, + "learning_rate": 2.9630333333333333e-05, + "loss": 0.0414, + "step": 11113 + }, + { + "epoch": 9.386824324324325, + "grad_norm": 0.2789599895477295, + "learning_rate": 2.9630000000000003e-05, + "loss": 0.0338, + "step": 11114 + }, + { + "epoch": 9.38766891891892, + "grad_norm": 0.3641476035118103, + "learning_rate": 2.9629666666666665e-05, + "loss": 0.0294, + "step": 11115 + }, + { + "epoch": 9.388513513513514, + "grad_norm": 0.33702409267425537, + "learning_rate": 2.9629333333333334e-05, + "loss": 0.0228, + "step": 11116 + }, + { + "epoch": 9.389358108108109, + "grad_norm": 0.33868756890296936, + "learning_rate": 2.9629e-05, + "loss": 0.0343, + "step": 11117 + }, + { + "epoch": 9.390202702702704, + "grad_norm": 0.25173768401145935, + "learning_rate": 2.9628666666666666e-05, + "loss": 0.0406, + "step": 11118 + }, + { + "epoch": 9.391047297297296, + "grad_norm": 0.6274950504302979, + "learning_rate": 2.9628333333333332e-05, + "loss": 0.0362, + "step": 11119 + }, + { + "epoch": 9.391891891891891, + "grad_norm": 0.38452401757240295, + "learning_rate": 2.9628e-05, + "loss": 0.0307, + "step": 11120 + }, + { + "epoch": 9.392736486486486, + "grad_norm": 0.2967560291290283, + "learning_rate": 2.9627666666666668e-05, + "loss": 0.0246, + "step": 11121 + }, + { + "epoch": 9.39358108108108, + "grad_norm": 0.3856281638145447, + "learning_rate": 2.9627333333333333e-05, + "loss": 0.022, + "step": 11122 + }, + { + "epoch": 9.394425675675675, + "grad_norm": 0.4468710124492645, + "learning_rate": 2.9627000000000003e-05, + "loss": 0.028, + "step": 11123 + }, + { + "epoch": 9.39527027027027, + "grad_norm": 0.540704607963562, + "learning_rate": 2.9626666666666665e-05, + "loss": 0.0341, + "step": 11124 + }, + { + "epoch": 9.396114864864865, + "grad_norm": 0.23362793028354645, + "learning_rate": 2.9626333333333335e-05, + "loss": 0.0181, + "step": 11125 + }, + { + "epoch": 9.39695945945946, + "grad_norm": 0.2405613362789154, + "learning_rate": 2.9626e-05, + "loss": 0.0323, + "step": 11126 + }, + { + "epoch": 9.397804054054054, + "grad_norm": 0.19434060156345367, + "learning_rate": 2.9625666666666667e-05, + "loss": 0.0233, + "step": 11127 + }, + { + "epoch": 9.39864864864865, + "grad_norm": 0.24441157281398773, + "learning_rate": 2.9625333333333336e-05, + "loss": 0.0264, + "step": 11128 + }, + { + "epoch": 9.399493243243244, + "grad_norm": 0.3641766309738159, + "learning_rate": 2.9625000000000002e-05, + "loss": 0.026, + "step": 11129 + }, + { + "epoch": 9.400337837837839, + "grad_norm": 0.3607533276081085, + "learning_rate": 2.9624666666666668e-05, + "loss": 0.0244, + "step": 11130 + }, + { + "epoch": 9.401182432432432, + "grad_norm": 0.370757132768631, + "learning_rate": 2.9624333333333334e-05, + "loss": 0.0369, + "step": 11131 + }, + { + "epoch": 9.402027027027026, + "grad_norm": 0.32815271615982056, + "learning_rate": 2.9624000000000003e-05, + "loss": 0.0345, + "step": 11132 + }, + { + "epoch": 9.402871621621621, + "grad_norm": 0.3852689564228058, + "learning_rate": 2.9623666666666666e-05, + "loss": 0.0336, + "step": 11133 + }, + { + "epoch": 9.403716216216216, + "grad_norm": 0.23871459066867828, + "learning_rate": 2.9623333333333335e-05, + "loss": 0.0198, + "step": 11134 + }, + { + "epoch": 9.40456081081081, + "grad_norm": 0.29576683044433594, + "learning_rate": 2.9623e-05, + "loss": 0.0155, + "step": 11135 + }, + { + "epoch": 9.405405405405405, + "grad_norm": 0.6007937788963318, + "learning_rate": 2.9622666666666667e-05, + "loss": 0.0303, + "step": 11136 + }, + { + "epoch": 9.40625, + "grad_norm": 0.29970213770866394, + "learning_rate": 2.9622333333333336e-05, + "loss": 0.0238, + "step": 11137 + }, + { + "epoch": 9.407094594594595, + "grad_norm": 0.7456895112991333, + "learning_rate": 2.9622000000000002e-05, + "loss": 0.029, + "step": 11138 + }, + { + "epoch": 9.40793918918919, + "grad_norm": 0.3084583878517151, + "learning_rate": 2.9621666666666668e-05, + "loss": 0.0201, + "step": 11139 + }, + { + "epoch": 9.408783783783784, + "grad_norm": 0.5580081939697266, + "learning_rate": 2.9621333333333334e-05, + "loss": 0.0322, + "step": 11140 + }, + { + "epoch": 9.409628378378379, + "grad_norm": 0.24708634614944458, + "learning_rate": 2.9621e-05, + "loss": 0.0193, + "step": 11141 + }, + { + "epoch": 9.410472972972974, + "grad_norm": 0.30248814821243286, + "learning_rate": 2.9620666666666666e-05, + "loss": 0.0264, + "step": 11142 + }, + { + "epoch": 9.411317567567568, + "grad_norm": 0.3514966368675232, + "learning_rate": 2.9620333333333335e-05, + "loss": 0.0227, + "step": 11143 + }, + { + "epoch": 9.412162162162161, + "grad_norm": 0.5514541268348694, + "learning_rate": 2.9619999999999998e-05, + "loss": 0.0912, + "step": 11144 + }, + { + "epoch": 9.413006756756756, + "grad_norm": 0.4376809597015381, + "learning_rate": 2.9619666666666667e-05, + "loss": 0.0321, + "step": 11145 + }, + { + "epoch": 9.41385135135135, + "grad_norm": 0.2840154767036438, + "learning_rate": 2.9619333333333336e-05, + "loss": 0.0253, + "step": 11146 + }, + { + "epoch": 9.414695945945946, + "grad_norm": 0.401563435792923, + "learning_rate": 2.9619e-05, + "loss": 0.0368, + "step": 11147 + }, + { + "epoch": 9.41554054054054, + "grad_norm": 0.4289030432701111, + "learning_rate": 2.9618666666666668e-05, + "loss": 0.0283, + "step": 11148 + }, + { + "epoch": 9.416385135135135, + "grad_norm": 0.4750906527042389, + "learning_rate": 2.9618333333333334e-05, + "loss": 0.0214, + "step": 11149 + }, + { + "epoch": 9.41722972972973, + "grad_norm": 0.25568369030952454, + "learning_rate": 2.9618e-05, + "loss": 0.0171, + "step": 11150 + }, + { + "epoch": 9.418074324324325, + "grad_norm": 0.4637279808521271, + "learning_rate": 2.9617666666666666e-05, + "loss": 0.0322, + "step": 11151 + }, + { + "epoch": 9.41891891891892, + "grad_norm": 0.516900897026062, + "learning_rate": 2.9617333333333335e-05, + "loss": 0.043, + "step": 11152 + }, + { + "epoch": 9.419763513513514, + "grad_norm": 0.2730914056301117, + "learning_rate": 2.9617e-05, + "loss": 0.0208, + "step": 11153 + }, + { + "epoch": 9.420608108108109, + "grad_norm": 1.3888081312179565, + "learning_rate": 2.9616666666666667e-05, + "loss": 0.0481, + "step": 11154 + }, + { + "epoch": 9.421452702702704, + "grad_norm": 0.6098849773406982, + "learning_rate": 2.9616333333333336e-05, + "loss": 0.0901, + "step": 11155 + }, + { + "epoch": 9.422297297297296, + "grad_norm": 3.0438270568847656, + "learning_rate": 2.9616e-05, + "loss": 0.1312, + "step": 11156 + }, + { + "epoch": 9.423141891891891, + "grad_norm": 1.0262978076934814, + "learning_rate": 2.961566666666667e-05, + "loss": 0.2653, + "step": 11157 + }, + { + "epoch": 9.423986486486486, + "grad_norm": 0.7691695094108582, + "learning_rate": 2.9615333333333334e-05, + "loss": 0.1227, + "step": 11158 + }, + { + "epoch": 9.42483108108108, + "grad_norm": 0.3875310719013214, + "learning_rate": 2.9615e-05, + "loss": 0.0818, + "step": 11159 + }, + { + "epoch": 9.425675675675675, + "grad_norm": 0.4849664866924286, + "learning_rate": 2.9614666666666666e-05, + "loss": 0.0655, + "step": 11160 + }, + { + "epoch": 9.42652027027027, + "grad_norm": 0.48891425132751465, + "learning_rate": 2.9614333333333335e-05, + "loss": 0.038, + "step": 11161 + }, + { + "epoch": 9.427364864864865, + "grad_norm": 0.6273342967033386, + "learning_rate": 2.9614e-05, + "loss": 0.0738, + "step": 11162 + }, + { + "epoch": 9.42820945945946, + "grad_norm": 0.33786818385124207, + "learning_rate": 2.9613666666666667e-05, + "loss": 0.0318, + "step": 11163 + }, + { + "epoch": 9.429054054054054, + "grad_norm": 0.7211613059043884, + "learning_rate": 2.9613333333333337e-05, + "loss": 0.0422, + "step": 11164 + }, + { + "epoch": 9.42989864864865, + "grad_norm": 0.401603102684021, + "learning_rate": 2.9613e-05, + "loss": 0.0432, + "step": 11165 + }, + { + "epoch": 9.430743243243244, + "grad_norm": 0.38382968306541443, + "learning_rate": 2.961266666666667e-05, + "loss": 0.0455, + "step": 11166 + }, + { + "epoch": 9.431587837837839, + "grad_norm": 0.3274305462837219, + "learning_rate": 2.9612333333333334e-05, + "loss": 0.0355, + "step": 11167 + }, + { + "epoch": 9.432432432432432, + "grad_norm": 0.6820510625839233, + "learning_rate": 2.9612e-05, + "loss": 0.0346, + "step": 11168 + }, + { + "epoch": 9.433277027027026, + "grad_norm": 0.304068386554718, + "learning_rate": 2.9611666666666666e-05, + "loss": 0.0261, + "step": 11169 + }, + { + "epoch": 9.434121621621621, + "grad_norm": 0.31258004903793335, + "learning_rate": 2.9611333333333332e-05, + "loss": 0.0451, + "step": 11170 + }, + { + "epoch": 9.434966216216216, + "grad_norm": 0.24877382814884186, + "learning_rate": 2.9611e-05, + "loss": 0.0234, + "step": 11171 + }, + { + "epoch": 9.43581081081081, + "grad_norm": 0.30265888571739197, + "learning_rate": 2.9610666666666668e-05, + "loss": 0.0275, + "step": 11172 + }, + { + "epoch": 9.436655405405405, + "grad_norm": 0.4792634844779968, + "learning_rate": 2.9610333333333333e-05, + "loss": 0.0414, + "step": 11173 + }, + { + "epoch": 9.4375, + "grad_norm": 0.22679780423641205, + "learning_rate": 2.961e-05, + "loss": 0.0227, + "step": 11174 + }, + { + "epoch": 9.438344594594595, + "grad_norm": 0.2533634901046753, + "learning_rate": 2.960966666666667e-05, + "loss": 0.023, + "step": 11175 + }, + { + "epoch": 9.43918918918919, + "grad_norm": 0.6039264798164368, + "learning_rate": 2.960933333333333e-05, + "loss": 0.0342, + "step": 11176 + }, + { + "epoch": 9.440033783783784, + "grad_norm": 0.3693409264087677, + "learning_rate": 2.9609e-05, + "loss": 0.0276, + "step": 11177 + }, + { + "epoch": 9.440878378378379, + "grad_norm": 0.270609587430954, + "learning_rate": 2.960866666666667e-05, + "loss": 0.0185, + "step": 11178 + }, + { + "epoch": 9.441722972972974, + "grad_norm": 0.32764917612075806, + "learning_rate": 2.9608333333333332e-05, + "loss": 0.0315, + "step": 11179 + }, + { + "epoch": 9.442567567567568, + "grad_norm": 0.40087276697158813, + "learning_rate": 2.9608000000000002e-05, + "loss": 0.0216, + "step": 11180 + }, + { + "epoch": 9.443412162162161, + "grad_norm": 0.2954106032848358, + "learning_rate": 2.9607666666666668e-05, + "loss": 0.021, + "step": 11181 + }, + { + "epoch": 9.444256756756756, + "grad_norm": 0.20262278616428375, + "learning_rate": 2.9607333333333334e-05, + "loss": 0.0133, + "step": 11182 + }, + { + "epoch": 9.44510135135135, + "grad_norm": 0.3231744170188904, + "learning_rate": 2.9607e-05, + "loss": 0.029, + "step": 11183 + }, + { + "epoch": 9.445945945945946, + "grad_norm": 0.4111438989639282, + "learning_rate": 2.960666666666667e-05, + "loss": 0.0158, + "step": 11184 + }, + { + "epoch": 9.44679054054054, + "grad_norm": 0.2627154290676117, + "learning_rate": 2.960633333333333e-05, + "loss": 0.0288, + "step": 11185 + }, + { + "epoch": 9.447635135135135, + "grad_norm": 0.31971171498298645, + "learning_rate": 2.9606e-05, + "loss": 0.031, + "step": 11186 + }, + { + "epoch": 9.44847972972973, + "grad_norm": 0.41709527373313904, + "learning_rate": 2.960566666666667e-05, + "loss": 0.0151, + "step": 11187 + }, + { + "epoch": 9.449324324324325, + "grad_norm": 0.4187919497489929, + "learning_rate": 2.9605333333333333e-05, + "loss": 0.0383, + "step": 11188 + }, + { + "epoch": 9.45016891891892, + "grad_norm": 0.454934298992157, + "learning_rate": 2.9605000000000002e-05, + "loss": 0.0322, + "step": 11189 + }, + { + "epoch": 9.451013513513514, + "grad_norm": 0.5432155132293701, + "learning_rate": 2.9604666666666668e-05, + "loss": 0.0399, + "step": 11190 + }, + { + "epoch": 9.451858108108109, + "grad_norm": 0.556195080280304, + "learning_rate": 2.9604333333333334e-05, + "loss": 0.0326, + "step": 11191 + }, + { + "epoch": 9.452702702702704, + "grad_norm": 0.4766179323196411, + "learning_rate": 2.9604e-05, + "loss": 0.0475, + "step": 11192 + }, + { + "epoch": 9.453547297297296, + "grad_norm": 0.3230023980140686, + "learning_rate": 2.960366666666667e-05, + "loss": 0.0199, + "step": 11193 + }, + { + "epoch": 9.454391891891891, + "grad_norm": 0.20374813675880432, + "learning_rate": 2.960333333333333e-05, + "loss": 0.0239, + "step": 11194 + }, + { + "epoch": 9.455236486486486, + "grad_norm": 0.34162595868110657, + "learning_rate": 2.9603e-05, + "loss": 0.0241, + "step": 11195 + }, + { + "epoch": 9.45608108108108, + "grad_norm": 0.4789927303791046, + "learning_rate": 2.960266666666667e-05, + "loss": 0.0263, + "step": 11196 + }, + { + "epoch": 9.456925675675675, + "grad_norm": 0.5198702812194824, + "learning_rate": 2.9602333333333333e-05, + "loss": 0.0296, + "step": 11197 + }, + { + "epoch": 9.45777027027027, + "grad_norm": 0.3754023611545563, + "learning_rate": 2.9602000000000002e-05, + "loss": 0.0217, + "step": 11198 + }, + { + "epoch": 9.458614864864865, + "grad_norm": 0.22397910058498383, + "learning_rate": 2.9601666666666665e-05, + "loss": 0.0234, + "step": 11199 + }, + { + "epoch": 9.45945945945946, + "grad_norm": 0.43403711915016174, + "learning_rate": 2.9601333333333334e-05, + "loss": 0.0265, + "step": 11200 + }, + { + "epoch": 9.460304054054054, + "grad_norm": 0.3790963888168335, + "learning_rate": 2.9601e-05, + "loss": 0.0313, + "step": 11201 + }, + { + "epoch": 9.46114864864865, + "grad_norm": 0.2313002347946167, + "learning_rate": 2.9600666666666666e-05, + "loss": 0.019, + "step": 11202 + }, + { + "epoch": 9.461993243243244, + "grad_norm": 0.7655011415481567, + "learning_rate": 2.9600333333333335e-05, + "loss": 0.0396, + "step": 11203 + }, + { + "epoch": 9.462837837837839, + "grad_norm": 0.49846625328063965, + "learning_rate": 2.96e-05, + "loss": 0.0431, + "step": 11204 + }, + { + "epoch": 9.463682432432432, + "grad_norm": 0.5129923224449158, + "learning_rate": 2.9599666666666667e-05, + "loss": 0.0497, + "step": 11205 + }, + { + "epoch": 9.464527027027026, + "grad_norm": 0.6134381294250488, + "learning_rate": 2.9599333333333333e-05, + "loss": 0.0457, + "step": 11206 + }, + { + "epoch": 9.465371621621621, + "grad_norm": 0.9378283619880676, + "learning_rate": 2.9599000000000002e-05, + "loss": 0.2434, + "step": 11207 + }, + { + "epoch": 9.466216216216216, + "grad_norm": 0.6019265651702881, + "learning_rate": 2.9598666666666665e-05, + "loss": 0.1551, + "step": 11208 + }, + { + "epoch": 9.46706081081081, + "grad_norm": 0.39336588978767395, + "learning_rate": 2.9598333333333334e-05, + "loss": 0.0591, + "step": 11209 + }, + { + "epoch": 9.467905405405405, + "grad_norm": 0.3280636966228485, + "learning_rate": 2.9598e-05, + "loss": 0.0425, + "step": 11210 + }, + { + "epoch": 9.46875, + "grad_norm": 0.39907214045524597, + "learning_rate": 2.9597666666666666e-05, + "loss": 0.062, + "step": 11211 + }, + { + "epoch": 9.469594594594595, + "grad_norm": 0.3348519504070282, + "learning_rate": 2.9597333333333335e-05, + "loss": 0.0498, + "step": 11212 + }, + { + "epoch": 9.47043918918919, + "grad_norm": 0.31673604249954224, + "learning_rate": 2.9597e-05, + "loss": 0.0446, + "step": 11213 + }, + { + "epoch": 9.471283783783784, + "grad_norm": 0.3020336329936981, + "learning_rate": 2.9596666666666667e-05, + "loss": 0.0348, + "step": 11214 + }, + { + "epoch": 9.472128378378379, + "grad_norm": 0.20170505344867706, + "learning_rate": 2.9596333333333333e-05, + "loss": 0.0232, + "step": 11215 + }, + { + "epoch": 9.472972972972974, + "grad_norm": 0.25696828961372375, + "learning_rate": 2.9596000000000003e-05, + "loss": 0.0263, + "step": 11216 + }, + { + "epoch": 9.473817567567568, + "grad_norm": 0.20097090303897858, + "learning_rate": 2.9595666666666665e-05, + "loss": 0.0192, + "step": 11217 + }, + { + "epoch": 9.474662162162161, + "grad_norm": 0.28862109780311584, + "learning_rate": 2.9595333333333334e-05, + "loss": 0.0186, + "step": 11218 + }, + { + "epoch": 9.475506756756756, + "grad_norm": 0.27033767104148865, + "learning_rate": 2.9595e-05, + "loss": 0.0251, + "step": 11219 + }, + { + "epoch": 9.47635135135135, + "grad_norm": 0.1861160546541214, + "learning_rate": 2.9594666666666666e-05, + "loss": 0.0165, + "step": 11220 + }, + { + "epoch": 9.477195945945946, + "grad_norm": 0.23671755194664001, + "learning_rate": 2.9594333333333336e-05, + "loss": 0.03, + "step": 11221 + }, + { + "epoch": 9.47804054054054, + "grad_norm": 0.3896794021129608, + "learning_rate": 2.9594e-05, + "loss": 0.0383, + "step": 11222 + }, + { + "epoch": 9.478885135135135, + "grad_norm": 0.2479865849018097, + "learning_rate": 2.9593666666666668e-05, + "loss": 0.0145, + "step": 11223 + }, + { + "epoch": 9.47972972972973, + "grad_norm": 0.3293443024158478, + "learning_rate": 2.9593333333333333e-05, + "loss": 0.0231, + "step": 11224 + }, + { + "epoch": 9.480574324324325, + "grad_norm": 0.27523329854011536, + "learning_rate": 2.9593000000000003e-05, + "loss": 0.0251, + "step": 11225 + }, + { + "epoch": 9.48141891891892, + "grad_norm": 0.2801060676574707, + "learning_rate": 2.9592666666666665e-05, + "loss": 0.0167, + "step": 11226 + }, + { + "epoch": 9.482263513513514, + "grad_norm": 0.29930585622787476, + "learning_rate": 2.9592333333333335e-05, + "loss": 0.0341, + "step": 11227 + }, + { + "epoch": 9.483108108108109, + "grad_norm": 0.27389082312583923, + "learning_rate": 2.9592000000000004e-05, + "loss": 0.026, + "step": 11228 + }, + { + "epoch": 9.483952702702704, + "grad_norm": 0.4045024812221527, + "learning_rate": 2.9591666666666667e-05, + "loss": 0.0233, + "step": 11229 + }, + { + "epoch": 9.484797297297296, + "grad_norm": 0.44997477531433105, + "learning_rate": 2.9591333333333336e-05, + "loss": 0.0394, + "step": 11230 + }, + { + "epoch": 9.485641891891891, + "grad_norm": 0.40902256965637207, + "learning_rate": 2.9591e-05, + "loss": 0.0183, + "step": 11231 + }, + { + "epoch": 9.486486486486486, + "grad_norm": 0.38438501954078674, + "learning_rate": 2.9590666666666668e-05, + "loss": 0.0262, + "step": 11232 + }, + { + "epoch": 9.48733108108108, + "grad_norm": 0.27349331974983215, + "learning_rate": 2.9590333333333334e-05, + "loss": 0.023, + "step": 11233 + }, + { + "epoch": 9.488175675675675, + "grad_norm": 0.7439749240875244, + "learning_rate": 2.959e-05, + "loss": 0.0266, + "step": 11234 + }, + { + "epoch": 9.48902027027027, + "grad_norm": 0.3812856078147888, + "learning_rate": 2.9589666666666666e-05, + "loss": 0.0309, + "step": 11235 + }, + { + "epoch": 9.489864864864865, + "grad_norm": 0.28219500184059143, + "learning_rate": 2.9589333333333335e-05, + "loss": 0.0318, + "step": 11236 + }, + { + "epoch": 9.49070945945946, + "grad_norm": 0.2242247760295868, + "learning_rate": 2.9589e-05, + "loss": 0.0177, + "step": 11237 + }, + { + "epoch": 9.491554054054054, + "grad_norm": 0.2635655999183655, + "learning_rate": 2.9588666666666667e-05, + "loss": 0.0182, + "step": 11238 + }, + { + "epoch": 9.49239864864865, + "grad_norm": 0.7413722276687622, + "learning_rate": 2.9588333333333336e-05, + "loss": 0.0313, + "step": 11239 + }, + { + "epoch": 9.493243243243244, + "grad_norm": 0.4371053874492645, + "learning_rate": 2.9588e-05, + "loss": 0.0367, + "step": 11240 + }, + { + "epoch": 9.494087837837839, + "grad_norm": 0.30165764689445496, + "learning_rate": 2.9587666666666668e-05, + "loss": 0.0199, + "step": 11241 + }, + { + "epoch": 9.494932432432432, + "grad_norm": 0.31841474771499634, + "learning_rate": 2.9587333333333334e-05, + "loss": 0.0289, + "step": 11242 + }, + { + "epoch": 9.495777027027026, + "grad_norm": 0.2895027697086334, + "learning_rate": 2.9587e-05, + "loss": 0.0242, + "step": 11243 + }, + { + "epoch": 9.496621621621621, + "grad_norm": 0.33921167254447937, + "learning_rate": 2.9586666666666666e-05, + "loss": 0.0284, + "step": 11244 + }, + { + "epoch": 9.497466216216216, + "grad_norm": 0.4645395278930664, + "learning_rate": 2.9586333333333335e-05, + "loss": 0.038, + "step": 11245 + }, + { + "epoch": 9.49831081081081, + "grad_norm": 0.4246661365032196, + "learning_rate": 2.9586e-05, + "loss": 0.0471, + "step": 11246 + }, + { + "epoch": 9.499155405405405, + "grad_norm": 0.5770573616027832, + "learning_rate": 2.9585666666666667e-05, + "loss": 0.0377, + "step": 11247 + }, + { + "epoch": 9.5, + "grad_norm": 0.33884355425834656, + "learning_rate": 2.9585333333333336e-05, + "loss": 0.0322, + "step": 11248 + }, + { + "epoch": 9.500844594594595, + "grad_norm": 0.5296353697776794, + "learning_rate": 2.9585e-05, + "loss": 0.0255, + "step": 11249 + }, + { + "epoch": 9.50168918918919, + "grad_norm": 0.6877767443656921, + "learning_rate": 2.9584666666666668e-05, + "loss": 0.033, + "step": 11250 + }, + { + "epoch": 9.502533783783784, + "grad_norm": 0.6391193270683289, + "learning_rate": 2.9584333333333334e-05, + "loss": 0.0397, + "step": 11251 + }, + { + "epoch": 9.503378378378379, + "grad_norm": 0.3501036465167999, + "learning_rate": 2.9584e-05, + "loss": 0.0258, + "step": 11252 + }, + { + "epoch": 9.504222972972974, + "grad_norm": 0.4731732904911041, + "learning_rate": 2.958366666666667e-05, + "loss": 0.0332, + "step": 11253 + }, + { + "epoch": 9.505067567567568, + "grad_norm": 0.41579464077949524, + "learning_rate": 2.9583333333333335e-05, + "loss": 0.0258, + "step": 11254 + }, + { + "epoch": 9.505912162162161, + "grad_norm": 0.6500418782234192, + "learning_rate": 2.9583e-05, + "loss": 0.0426, + "step": 11255 + }, + { + "epoch": 9.506756756756756, + "grad_norm": 0.5288324356079102, + "learning_rate": 2.9582666666666667e-05, + "loss": 0.0288, + "step": 11256 + }, + { + "epoch": 9.50760135135135, + "grad_norm": 1.1283503770828247, + "learning_rate": 2.9582333333333336e-05, + "loss": 0.3167, + "step": 11257 + }, + { + "epoch": 9.508445945945946, + "grad_norm": 0.46276164054870605, + "learning_rate": 2.9582e-05, + "loss": 0.1077, + "step": 11258 + }, + { + "epoch": 9.50929054054054, + "grad_norm": 0.35491058230400085, + "learning_rate": 2.958166666666667e-05, + "loss": 0.0487, + "step": 11259 + }, + { + "epoch": 9.510135135135135, + "grad_norm": 0.3159835636615753, + "learning_rate": 2.958133333333333e-05, + "loss": 0.0457, + "step": 11260 + }, + { + "epoch": 9.51097972972973, + "grad_norm": 0.33923351764678955, + "learning_rate": 2.9581e-05, + "loss": 0.0412, + "step": 11261 + }, + { + "epoch": 9.511824324324325, + "grad_norm": 0.32122135162353516, + "learning_rate": 2.958066666666667e-05, + "loss": 0.0416, + "step": 11262 + }, + { + "epoch": 9.51266891891892, + "grad_norm": 0.36040830612182617, + "learning_rate": 2.9580333333333332e-05, + "loss": 0.0368, + "step": 11263 + }, + { + "epoch": 9.513513513513514, + "grad_norm": 0.22663386166095734, + "learning_rate": 2.958e-05, + "loss": 0.0236, + "step": 11264 + }, + { + "epoch": 9.514358108108109, + "grad_norm": 0.3028084933757782, + "learning_rate": 2.9579666666666667e-05, + "loss": 0.041, + "step": 11265 + }, + { + "epoch": 9.515202702702704, + "grad_norm": 0.23019060492515564, + "learning_rate": 2.9579333333333333e-05, + "loss": 0.0251, + "step": 11266 + }, + { + "epoch": 9.516047297297296, + "grad_norm": 0.20228755474090576, + "learning_rate": 2.9579e-05, + "loss": 0.0239, + "step": 11267 + }, + { + "epoch": 9.516891891891891, + "grad_norm": 0.20065723359584808, + "learning_rate": 2.957866666666667e-05, + "loss": 0.0197, + "step": 11268 + }, + { + "epoch": 9.517736486486486, + "grad_norm": 0.33131998777389526, + "learning_rate": 2.9578333333333334e-05, + "loss": 0.0346, + "step": 11269 + }, + { + "epoch": 9.51858108108108, + "grad_norm": 0.2811296284198761, + "learning_rate": 2.9578e-05, + "loss": 0.0198, + "step": 11270 + }, + { + "epoch": 9.519425675675675, + "grad_norm": 0.42580458521842957, + "learning_rate": 2.957766666666667e-05, + "loss": 0.0502, + "step": 11271 + }, + { + "epoch": 9.52027027027027, + "grad_norm": 0.2679362893104553, + "learning_rate": 2.9577333333333332e-05, + "loss": 0.0186, + "step": 11272 + }, + { + "epoch": 9.521114864864865, + "grad_norm": 0.3914320170879364, + "learning_rate": 2.9577e-05, + "loss": 0.0221, + "step": 11273 + }, + { + "epoch": 9.52195945945946, + "grad_norm": 0.32749712467193604, + "learning_rate": 2.9576666666666668e-05, + "loss": 0.034, + "step": 11274 + }, + { + "epoch": 9.522804054054054, + "grad_norm": 0.26796093583106995, + "learning_rate": 2.9576333333333333e-05, + "loss": 0.0234, + "step": 11275 + }, + { + "epoch": 9.52364864864865, + "grad_norm": 0.4267657697200775, + "learning_rate": 2.9576e-05, + "loss": 0.0265, + "step": 11276 + }, + { + "epoch": 9.524493243243244, + "grad_norm": 0.5383491516113281, + "learning_rate": 2.957566666666667e-05, + "loss": 0.1229, + "step": 11277 + }, + { + "epoch": 9.525337837837839, + "grad_norm": 0.38602152466773987, + "learning_rate": 2.9575333333333335e-05, + "loss": 0.0347, + "step": 11278 + }, + { + "epoch": 9.526182432432432, + "grad_norm": 0.24718569219112396, + "learning_rate": 2.9575e-05, + "loss": 0.0292, + "step": 11279 + }, + { + "epoch": 9.527027027027026, + "grad_norm": 0.1995132714509964, + "learning_rate": 2.957466666666667e-05, + "loss": 0.0195, + "step": 11280 + }, + { + "epoch": 9.527871621621621, + "grad_norm": 0.2624817490577698, + "learning_rate": 2.9574333333333332e-05, + "loss": 0.0192, + "step": 11281 + }, + { + "epoch": 9.528716216216216, + "grad_norm": 0.41146159172058105, + "learning_rate": 2.9574000000000002e-05, + "loss": 0.047, + "step": 11282 + }, + { + "epoch": 9.52956081081081, + "grad_norm": 0.3564656376838684, + "learning_rate": 2.9573666666666668e-05, + "loss": 0.0353, + "step": 11283 + }, + { + "epoch": 9.530405405405405, + "grad_norm": 0.1829586625099182, + "learning_rate": 2.9573333333333334e-05, + "loss": 0.0198, + "step": 11284 + }, + { + "epoch": 9.53125, + "grad_norm": 0.3971983790397644, + "learning_rate": 2.9573e-05, + "loss": 0.0348, + "step": 11285 + }, + { + "epoch": 9.532094594594595, + "grad_norm": 0.8463064432144165, + "learning_rate": 2.957266666666667e-05, + "loss": 0.0389, + "step": 11286 + }, + { + "epoch": 9.53293918918919, + "grad_norm": 0.43518248200416565, + "learning_rate": 2.9572333333333335e-05, + "loss": 0.0343, + "step": 11287 + }, + { + "epoch": 9.533783783783784, + "grad_norm": 0.2598215341567993, + "learning_rate": 2.9572e-05, + "loss": 0.0284, + "step": 11288 + }, + { + "epoch": 9.534628378378379, + "grad_norm": 0.2791045904159546, + "learning_rate": 2.957166666666667e-05, + "loss": 0.0255, + "step": 11289 + }, + { + "epoch": 9.535472972972974, + "grad_norm": 0.3056836724281311, + "learning_rate": 2.9571333333333333e-05, + "loss": 0.0242, + "step": 11290 + }, + { + "epoch": 9.536317567567568, + "grad_norm": 0.3725433945655823, + "learning_rate": 2.9571000000000002e-05, + "loss": 0.0304, + "step": 11291 + }, + { + "epoch": 9.537162162162161, + "grad_norm": 0.361175537109375, + "learning_rate": 2.9570666666666665e-05, + "loss": 0.0283, + "step": 11292 + }, + { + "epoch": 9.538006756756756, + "grad_norm": 0.3731488883495331, + "learning_rate": 2.9570333333333334e-05, + "loss": 0.0239, + "step": 11293 + }, + { + "epoch": 9.53885135135135, + "grad_norm": 0.2439543455839157, + "learning_rate": 2.957e-05, + "loss": 0.0148, + "step": 11294 + }, + { + "epoch": 9.539695945945946, + "grad_norm": 0.4027840197086334, + "learning_rate": 2.9569666666666666e-05, + "loss": 0.0288, + "step": 11295 + }, + { + "epoch": 9.54054054054054, + "grad_norm": 0.31325462460517883, + "learning_rate": 2.9569333333333335e-05, + "loss": 0.0312, + "step": 11296 + }, + { + "epoch": 9.541385135135135, + "grad_norm": 0.9486468434333801, + "learning_rate": 2.9569e-05, + "loss": 0.0206, + "step": 11297 + }, + { + "epoch": 9.54222972972973, + "grad_norm": 0.677282989025116, + "learning_rate": 2.9568666666666667e-05, + "loss": 0.0467, + "step": 11298 + }, + { + "epoch": 9.543074324324325, + "grad_norm": 0.46225547790527344, + "learning_rate": 2.9568333333333333e-05, + "loss": 0.0335, + "step": 11299 + }, + { + "epoch": 9.54391891891892, + "grad_norm": 1.3648386001586914, + "learning_rate": 2.9568000000000002e-05, + "loss": 0.0408, + "step": 11300 + }, + { + "epoch": 9.544763513513514, + "grad_norm": 0.5371567010879517, + "learning_rate": 2.9567666666666665e-05, + "loss": 0.0362, + "step": 11301 + }, + { + "epoch": 9.545608108108109, + "grad_norm": 0.4141496419906616, + "learning_rate": 2.9567333333333334e-05, + "loss": 0.0253, + "step": 11302 + }, + { + "epoch": 9.546452702702704, + "grad_norm": 0.3131115138530731, + "learning_rate": 2.9567000000000003e-05, + "loss": 0.023, + "step": 11303 + }, + { + "epoch": 9.547297297297296, + "grad_norm": 0.35612013936042786, + "learning_rate": 2.9566666666666666e-05, + "loss": 0.0246, + "step": 11304 + }, + { + "epoch": 9.548141891891891, + "grad_norm": 0.621016800403595, + "learning_rate": 2.9566333333333335e-05, + "loss": 0.0436, + "step": 11305 + }, + { + "epoch": 9.548986486486486, + "grad_norm": 0.6465612053871155, + "learning_rate": 2.9566e-05, + "loss": 0.05, + "step": 11306 + }, + { + "epoch": 9.54983108108108, + "grad_norm": 0.7603992223739624, + "learning_rate": 2.9565666666666667e-05, + "loss": 0.3154, + "step": 11307 + }, + { + "epoch": 9.550675675675675, + "grad_norm": 0.6427009701728821, + "learning_rate": 2.9565333333333333e-05, + "loss": 0.1268, + "step": 11308 + }, + { + "epoch": 9.55152027027027, + "grad_norm": 0.6534337997436523, + "learning_rate": 2.9565000000000002e-05, + "loss": 0.0945, + "step": 11309 + }, + { + "epoch": 9.552364864864865, + "grad_norm": 0.4357595443725586, + "learning_rate": 2.9564666666666665e-05, + "loss": 0.1026, + "step": 11310 + }, + { + "epoch": 9.55320945945946, + "grad_norm": 0.3335844874382019, + "learning_rate": 2.9564333333333334e-05, + "loss": 0.0479, + "step": 11311 + }, + { + "epoch": 9.554054054054054, + "grad_norm": 0.43331748247146606, + "learning_rate": 2.9564000000000004e-05, + "loss": 0.0553, + "step": 11312 + }, + { + "epoch": 9.55489864864865, + "grad_norm": 0.2583531439304352, + "learning_rate": 2.9563666666666666e-05, + "loss": 0.0302, + "step": 11313 + }, + { + "epoch": 9.555743243243244, + "grad_norm": 0.2991441488265991, + "learning_rate": 2.9563333333333335e-05, + "loss": 0.0331, + "step": 11314 + }, + { + "epoch": 9.556587837837839, + "grad_norm": 0.32380491495132446, + "learning_rate": 2.9563e-05, + "loss": 0.0317, + "step": 11315 + }, + { + "epoch": 9.557432432432432, + "grad_norm": 0.6747874617576599, + "learning_rate": 2.9562666666666667e-05, + "loss": 0.0371, + "step": 11316 + }, + { + "epoch": 9.558277027027026, + "grad_norm": 0.33284792304039, + "learning_rate": 2.9562333333333333e-05, + "loss": 0.0345, + "step": 11317 + }, + { + "epoch": 9.559121621621621, + "grad_norm": 0.35188284516334534, + "learning_rate": 2.9562000000000003e-05, + "loss": 0.0387, + "step": 11318 + }, + { + "epoch": 9.559966216216216, + "grad_norm": 0.611024022102356, + "learning_rate": 2.956166666666667e-05, + "loss": 0.036, + "step": 11319 + }, + { + "epoch": 9.56081081081081, + "grad_norm": 0.36075451970100403, + "learning_rate": 2.9561333333333334e-05, + "loss": 0.0321, + "step": 11320 + }, + { + "epoch": 9.561655405405405, + "grad_norm": 0.33073490858078003, + "learning_rate": 2.9561e-05, + "loss": 0.0314, + "step": 11321 + }, + { + "epoch": 9.5625, + "grad_norm": 0.21049170196056366, + "learning_rate": 2.9560666666666666e-05, + "loss": 0.0206, + "step": 11322 + }, + { + "epoch": 9.563344594594595, + "grad_norm": 0.24347630143165588, + "learning_rate": 2.9560333333333336e-05, + "loss": 0.0246, + "step": 11323 + }, + { + "epoch": 9.56418918918919, + "grad_norm": 0.2845936417579651, + "learning_rate": 2.9559999999999998e-05, + "loss": 0.0221, + "step": 11324 + }, + { + "epoch": 9.565033783783784, + "grad_norm": 0.32080337405204773, + "learning_rate": 2.9559666666666668e-05, + "loss": 0.0201, + "step": 11325 + }, + { + "epoch": 9.565878378378379, + "grad_norm": 0.4255790710449219, + "learning_rate": 2.9559333333333333e-05, + "loss": 0.0487, + "step": 11326 + }, + { + "epoch": 9.566722972972974, + "grad_norm": 0.6833961606025696, + "learning_rate": 2.9559e-05, + "loss": 0.0283, + "step": 11327 + }, + { + "epoch": 9.567567567567568, + "grad_norm": 0.3438338339328766, + "learning_rate": 2.955866666666667e-05, + "loss": 0.0116, + "step": 11328 + }, + { + "epoch": 9.568412162162161, + "grad_norm": 0.49211978912353516, + "learning_rate": 2.9558333333333335e-05, + "loss": 0.0215, + "step": 11329 + }, + { + "epoch": 9.569256756756756, + "grad_norm": 0.27255433797836304, + "learning_rate": 2.9558e-05, + "loss": 0.0194, + "step": 11330 + }, + { + "epoch": 9.57010135135135, + "grad_norm": 0.3961300849914551, + "learning_rate": 2.9557666666666667e-05, + "loss": 0.0264, + "step": 11331 + }, + { + "epoch": 9.570945945945946, + "grad_norm": 0.260931134223938, + "learning_rate": 2.9557333333333336e-05, + "loss": 0.0231, + "step": 11332 + }, + { + "epoch": 9.57179054054054, + "grad_norm": 0.507470965385437, + "learning_rate": 2.9557e-05, + "loss": 0.0395, + "step": 11333 + }, + { + "epoch": 9.572635135135135, + "grad_norm": 0.4352559745311737, + "learning_rate": 2.9556666666666668e-05, + "loss": 0.0269, + "step": 11334 + }, + { + "epoch": 9.57347972972973, + "grad_norm": 0.38934123516082764, + "learning_rate": 2.9556333333333334e-05, + "loss": 0.0254, + "step": 11335 + }, + { + "epoch": 9.574324324324325, + "grad_norm": 0.21437323093414307, + "learning_rate": 2.9556e-05, + "loss": 0.0167, + "step": 11336 + }, + { + "epoch": 9.57516891891892, + "grad_norm": 0.7062777876853943, + "learning_rate": 2.955566666666667e-05, + "loss": 0.0265, + "step": 11337 + }, + { + "epoch": 9.576013513513514, + "grad_norm": 0.534184455871582, + "learning_rate": 2.9555333333333335e-05, + "loss": 0.0431, + "step": 11338 + }, + { + "epoch": 9.576858108108109, + "grad_norm": 0.4393843412399292, + "learning_rate": 2.9555e-05, + "loss": 0.0286, + "step": 11339 + }, + { + "epoch": 9.577702702702704, + "grad_norm": 0.44648367166519165, + "learning_rate": 2.9554666666666667e-05, + "loss": 0.0325, + "step": 11340 + }, + { + "epoch": 9.578547297297296, + "grad_norm": 0.24622389674186707, + "learning_rate": 2.9554333333333336e-05, + "loss": 0.0194, + "step": 11341 + }, + { + "epoch": 9.579391891891891, + "grad_norm": 0.25038865208625793, + "learning_rate": 2.9554e-05, + "loss": 0.0198, + "step": 11342 + }, + { + "epoch": 9.580236486486486, + "grad_norm": 0.4007246494293213, + "learning_rate": 2.9553666666666668e-05, + "loss": 0.0351, + "step": 11343 + }, + { + "epoch": 9.58108108108108, + "grad_norm": 0.4156002402305603, + "learning_rate": 2.9553333333333334e-05, + "loss": 0.035, + "step": 11344 + }, + { + "epoch": 9.581925675675675, + "grad_norm": 0.39904943108558655, + "learning_rate": 2.9553e-05, + "loss": 0.0373, + "step": 11345 + }, + { + "epoch": 9.58277027027027, + "grad_norm": 0.5086169242858887, + "learning_rate": 2.955266666666667e-05, + "loss": 0.0425, + "step": 11346 + }, + { + "epoch": 9.583614864864865, + "grad_norm": 0.43593746423721313, + "learning_rate": 2.9552333333333335e-05, + "loss": 0.0301, + "step": 11347 + }, + { + "epoch": 9.58445945945946, + "grad_norm": 0.42667046189308167, + "learning_rate": 2.9552e-05, + "loss": 0.0175, + "step": 11348 + }, + { + "epoch": 9.585304054054054, + "grad_norm": 0.5007052421569824, + "learning_rate": 2.9551666666666667e-05, + "loss": 0.0345, + "step": 11349 + }, + { + "epoch": 9.58614864864865, + "grad_norm": 1.048672080039978, + "learning_rate": 2.9551333333333333e-05, + "loss": 0.035, + "step": 11350 + }, + { + "epoch": 9.586993243243244, + "grad_norm": 0.501357913017273, + "learning_rate": 2.9551e-05, + "loss": 0.0371, + "step": 11351 + }, + { + "epoch": 9.587837837837839, + "grad_norm": 0.6344268321990967, + "learning_rate": 2.9550666666666668e-05, + "loss": 0.0317, + "step": 11352 + }, + { + "epoch": 9.588682432432432, + "grad_norm": 0.42982426285743713, + "learning_rate": 2.9550333333333334e-05, + "loss": 0.0431, + "step": 11353 + }, + { + "epoch": 9.589527027027026, + "grad_norm": 0.6989240050315857, + "learning_rate": 2.955e-05, + "loss": 0.0407, + "step": 11354 + }, + { + "epoch": 9.590371621621621, + "grad_norm": 0.3626875579357147, + "learning_rate": 2.954966666666667e-05, + "loss": 0.0361, + "step": 11355 + }, + { + "epoch": 9.591216216216216, + "grad_norm": 0.6602196097373962, + "learning_rate": 2.9549333333333332e-05, + "loss": 0.0593, + "step": 11356 + }, + { + "epoch": 9.59206081081081, + "grad_norm": 0.7121172547340393, + "learning_rate": 2.9549e-05, + "loss": 0.254, + "step": 11357 + }, + { + "epoch": 9.592905405405405, + "grad_norm": 0.4249216616153717, + "learning_rate": 2.9548666666666667e-05, + "loss": 0.144, + "step": 11358 + }, + { + "epoch": 9.59375, + "grad_norm": 0.3901856541633606, + "learning_rate": 2.9548333333333333e-05, + "loss": 0.0706, + "step": 11359 + }, + { + "epoch": 9.594594594594595, + "grad_norm": 0.4756419360637665, + "learning_rate": 2.9548e-05, + "loss": 0.042, + "step": 11360 + }, + { + "epoch": 9.59543918918919, + "grad_norm": 0.2807675302028656, + "learning_rate": 2.954766666666667e-05, + "loss": 0.0334, + "step": 11361 + }, + { + "epoch": 9.596283783783784, + "grad_norm": 0.9616513252258301, + "learning_rate": 2.9547333333333334e-05, + "loss": 0.0455, + "step": 11362 + }, + { + "epoch": 9.597128378378379, + "grad_norm": 0.38314539194107056, + "learning_rate": 2.9547e-05, + "loss": 0.0396, + "step": 11363 + }, + { + "epoch": 9.597972972972974, + "grad_norm": 0.5575732588768005, + "learning_rate": 2.954666666666667e-05, + "loss": 0.0291, + "step": 11364 + }, + { + "epoch": 9.598817567567568, + "grad_norm": 0.30849942564964294, + "learning_rate": 2.9546333333333332e-05, + "loss": 0.0322, + "step": 11365 + }, + { + "epoch": 9.599662162162161, + "grad_norm": 0.39046308398246765, + "learning_rate": 2.9546e-05, + "loss": 0.0352, + "step": 11366 + }, + { + "epoch": 9.600506756756756, + "grad_norm": 0.3217349052429199, + "learning_rate": 2.9545666666666667e-05, + "loss": 0.0328, + "step": 11367 + }, + { + "epoch": 9.60135135135135, + "grad_norm": 0.341810941696167, + "learning_rate": 2.9545333333333333e-05, + "loss": 0.0326, + "step": 11368 + }, + { + "epoch": 9.602195945945946, + "grad_norm": 0.23820361495018005, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0337, + "step": 11369 + }, + { + "epoch": 9.60304054054054, + "grad_norm": 0.530733585357666, + "learning_rate": 2.954466666666667e-05, + "loss": 0.0298, + "step": 11370 + }, + { + "epoch": 9.603885135135135, + "grad_norm": 0.3042915165424347, + "learning_rate": 2.9544333333333334e-05, + "loss": 0.0277, + "step": 11371 + }, + { + "epoch": 9.60472972972973, + "grad_norm": 0.23607899248600006, + "learning_rate": 2.9544e-05, + "loss": 0.021, + "step": 11372 + }, + { + "epoch": 9.605574324324325, + "grad_norm": 0.42665451765060425, + "learning_rate": 2.954366666666667e-05, + "loss": 0.029, + "step": 11373 + }, + { + "epoch": 9.60641891891892, + "grad_norm": 0.2797597050666809, + "learning_rate": 2.9543333333333332e-05, + "loss": 0.036, + "step": 11374 + }, + { + "epoch": 9.607263513513514, + "grad_norm": 0.24025118350982666, + "learning_rate": 2.9543e-05, + "loss": 0.0322, + "step": 11375 + }, + { + "epoch": 9.608108108108109, + "grad_norm": 0.4292815923690796, + "learning_rate": 2.9542666666666668e-05, + "loss": 0.0228, + "step": 11376 + }, + { + "epoch": 9.608952702702704, + "grad_norm": 0.29215776920318604, + "learning_rate": 2.9542333333333333e-05, + "loss": 0.023, + "step": 11377 + }, + { + "epoch": 9.609797297297296, + "grad_norm": 0.394883930683136, + "learning_rate": 2.9542000000000003e-05, + "loss": 0.0278, + "step": 11378 + }, + { + "epoch": 9.610641891891891, + "grad_norm": 0.3515413999557495, + "learning_rate": 2.954166666666667e-05, + "loss": 0.0246, + "step": 11379 + }, + { + "epoch": 9.611486486486486, + "grad_norm": 0.34574154019355774, + "learning_rate": 2.9541333333333335e-05, + "loss": 0.0226, + "step": 11380 + }, + { + "epoch": 9.61233108108108, + "grad_norm": 0.3851817846298218, + "learning_rate": 2.9541e-05, + "loss": 0.0226, + "step": 11381 + }, + { + "epoch": 9.613175675675675, + "grad_norm": 0.4048345685005188, + "learning_rate": 2.9540666666666667e-05, + "loss": 0.0319, + "step": 11382 + }, + { + "epoch": 9.61402027027027, + "grad_norm": 0.31633469462394714, + "learning_rate": 2.9540333333333332e-05, + "loss": 0.0209, + "step": 11383 + }, + { + "epoch": 9.614864864864865, + "grad_norm": 0.385163813829422, + "learning_rate": 2.9540000000000002e-05, + "loss": 0.0278, + "step": 11384 + }, + { + "epoch": 9.61570945945946, + "grad_norm": 0.2716991901397705, + "learning_rate": 2.9539666666666664e-05, + "loss": 0.0188, + "step": 11385 + }, + { + "epoch": 9.616554054054054, + "grad_norm": 0.36807185411453247, + "learning_rate": 2.9539333333333334e-05, + "loss": 0.0419, + "step": 11386 + }, + { + "epoch": 9.61739864864865, + "grad_norm": 0.1585482954978943, + "learning_rate": 2.9539000000000003e-05, + "loss": 0.012, + "step": 11387 + }, + { + "epoch": 9.618243243243244, + "grad_norm": 0.47241833806037903, + "learning_rate": 2.9538666666666666e-05, + "loss": 0.0357, + "step": 11388 + }, + { + "epoch": 9.619087837837839, + "grad_norm": 0.5102910399436951, + "learning_rate": 2.9538333333333335e-05, + "loss": 0.0328, + "step": 11389 + }, + { + "epoch": 9.619932432432432, + "grad_norm": 0.2987659275531769, + "learning_rate": 2.9538e-05, + "loss": 0.0247, + "step": 11390 + }, + { + "epoch": 9.620777027027026, + "grad_norm": 0.383061021566391, + "learning_rate": 2.9537666666666667e-05, + "loss": 0.0261, + "step": 11391 + }, + { + "epoch": 9.621621621621621, + "grad_norm": 0.349581778049469, + "learning_rate": 2.9537333333333333e-05, + "loss": 0.0188, + "step": 11392 + }, + { + "epoch": 9.622466216216216, + "grad_norm": 0.4604499042034149, + "learning_rate": 2.9537000000000002e-05, + "loss": 0.0911, + "step": 11393 + }, + { + "epoch": 9.62331081081081, + "grad_norm": 0.312861829996109, + "learning_rate": 2.9536666666666668e-05, + "loss": 0.0226, + "step": 11394 + }, + { + "epoch": 9.624155405405405, + "grad_norm": 0.6139096617698669, + "learning_rate": 2.9536333333333334e-05, + "loss": 0.0405, + "step": 11395 + }, + { + "epoch": 9.625, + "grad_norm": 0.3740774989128113, + "learning_rate": 2.9536000000000003e-05, + "loss": 0.0345, + "step": 11396 + }, + { + "epoch": 9.625844594594595, + "grad_norm": 0.3632109761238098, + "learning_rate": 2.9535666666666666e-05, + "loss": 0.04, + "step": 11397 + }, + { + "epoch": 9.62668918918919, + "grad_norm": 0.5118582844734192, + "learning_rate": 2.9535333333333335e-05, + "loss": 0.0246, + "step": 11398 + }, + { + "epoch": 9.627533783783784, + "grad_norm": 0.267941951751709, + "learning_rate": 2.9535e-05, + "loss": 0.0231, + "step": 11399 + }, + { + "epoch": 9.628378378378379, + "grad_norm": 0.26294106245040894, + "learning_rate": 2.9534666666666667e-05, + "loss": 0.0208, + "step": 11400 + }, + { + "epoch": 9.629222972972974, + "grad_norm": 0.26573532819747925, + "learning_rate": 2.9534333333333333e-05, + "loss": 0.0186, + "step": 11401 + }, + { + "epoch": 9.630067567567568, + "grad_norm": 0.5027614235877991, + "learning_rate": 2.9534000000000002e-05, + "loss": 0.0571, + "step": 11402 + }, + { + "epoch": 9.630912162162161, + "grad_norm": 0.44497305154800415, + "learning_rate": 2.9533666666666668e-05, + "loss": 0.035, + "step": 11403 + }, + { + "epoch": 9.631756756756756, + "grad_norm": 0.787971556186676, + "learning_rate": 2.9533333333333334e-05, + "loss": 0.0387, + "step": 11404 + }, + { + "epoch": 9.63260135135135, + "grad_norm": 0.559902548789978, + "learning_rate": 2.9533000000000003e-05, + "loss": 0.0481, + "step": 11405 + }, + { + "epoch": 9.633445945945946, + "grad_norm": 0.96567702293396, + "learning_rate": 2.9532666666666666e-05, + "loss": 0.0554, + "step": 11406 + }, + { + "epoch": 9.63429054054054, + "grad_norm": 1.3060004711151123, + "learning_rate": 2.9532333333333335e-05, + "loss": 0.3244, + "step": 11407 + }, + { + "epoch": 9.635135135135135, + "grad_norm": 0.9757136106491089, + "learning_rate": 2.9532e-05, + "loss": 0.1779, + "step": 11408 + }, + { + "epoch": 9.63597972972973, + "grad_norm": 0.736987829208374, + "learning_rate": 2.9531666666666667e-05, + "loss": 0.0937, + "step": 11409 + }, + { + "epoch": 9.636824324324325, + "grad_norm": 0.3432115614414215, + "learning_rate": 2.9531333333333333e-05, + "loss": 0.0583, + "step": 11410 + }, + { + "epoch": 9.63766891891892, + "grad_norm": 0.3353593945503235, + "learning_rate": 2.9531e-05, + "loss": 0.0396, + "step": 11411 + }, + { + "epoch": 9.638513513513514, + "grad_norm": 0.3473604917526245, + "learning_rate": 2.953066666666667e-05, + "loss": 0.0349, + "step": 11412 + }, + { + "epoch": 9.639358108108109, + "grad_norm": 0.2893134355545044, + "learning_rate": 2.9530333333333334e-05, + "loss": 0.0309, + "step": 11413 + }, + { + "epoch": 9.640202702702704, + "grad_norm": 0.24450011551380157, + "learning_rate": 2.953e-05, + "loss": 0.0264, + "step": 11414 + }, + { + "epoch": 9.641047297297296, + "grad_norm": 0.44710516929626465, + "learning_rate": 2.9529666666666666e-05, + "loss": 0.031, + "step": 11415 + }, + { + "epoch": 9.641891891891891, + "grad_norm": 0.4034963548183441, + "learning_rate": 2.9529333333333335e-05, + "loss": 0.0344, + "step": 11416 + }, + { + "epoch": 9.642736486486486, + "grad_norm": 0.4455912113189697, + "learning_rate": 2.9528999999999998e-05, + "loss": 0.0539, + "step": 11417 + }, + { + "epoch": 9.64358108108108, + "grad_norm": 0.48862963914871216, + "learning_rate": 2.9528666666666667e-05, + "loss": 0.0342, + "step": 11418 + }, + { + "epoch": 9.644425675675675, + "grad_norm": 0.3747498095035553, + "learning_rate": 2.9528333333333337e-05, + "loss": 0.0238, + "step": 11419 + }, + { + "epoch": 9.64527027027027, + "grad_norm": 0.29607683420181274, + "learning_rate": 2.9528e-05, + "loss": 0.0267, + "step": 11420 + }, + { + "epoch": 9.646114864864865, + "grad_norm": 0.5276772379875183, + "learning_rate": 2.952766666666667e-05, + "loss": 0.049, + "step": 11421 + }, + { + "epoch": 9.64695945945946, + "grad_norm": 0.3591160774230957, + "learning_rate": 2.9527333333333334e-05, + "loss": 0.0263, + "step": 11422 + }, + { + "epoch": 9.647804054054054, + "grad_norm": 0.3219457268714905, + "learning_rate": 2.9527e-05, + "loss": 0.0332, + "step": 11423 + }, + { + "epoch": 9.64864864864865, + "grad_norm": 0.33085522055625916, + "learning_rate": 2.9526666666666666e-05, + "loss": 0.0327, + "step": 11424 + }, + { + "epoch": 9.649493243243244, + "grad_norm": 0.41102489829063416, + "learning_rate": 2.9526333333333336e-05, + "loss": 0.034, + "step": 11425 + }, + { + "epoch": 9.650337837837839, + "grad_norm": 0.4157364070415497, + "learning_rate": 2.9525999999999998e-05, + "loss": 0.0372, + "step": 11426 + }, + { + "epoch": 9.651182432432432, + "grad_norm": 0.36931633949279785, + "learning_rate": 2.9525666666666668e-05, + "loss": 0.024, + "step": 11427 + }, + { + "epoch": 9.652027027027026, + "grad_norm": 0.36010292172431946, + "learning_rate": 2.9525333333333337e-05, + "loss": 0.029, + "step": 11428 + }, + { + "epoch": 9.652871621621621, + "grad_norm": 0.46881362795829773, + "learning_rate": 2.9525e-05, + "loss": 0.0257, + "step": 11429 + }, + { + "epoch": 9.653716216216216, + "grad_norm": 0.23775506019592285, + "learning_rate": 2.952466666666667e-05, + "loss": 0.0254, + "step": 11430 + }, + { + "epoch": 9.65456081081081, + "grad_norm": 0.30250558257102966, + "learning_rate": 2.9524333333333335e-05, + "loss": 0.0209, + "step": 11431 + }, + { + "epoch": 9.655405405405405, + "grad_norm": 0.8709969520568848, + "learning_rate": 2.9524e-05, + "loss": 0.0609, + "step": 11432 + }, + { + "epoch": 9.65625, + "grad_norm": 0.2072557955980301, + "learning_rate": 2.9523666666666667e-05, + "loss": 0.0139, + "step": 11433 + }, + { + "epoch": 9.657094594594595, + "grad_norm": 0.307376503944397, + "learning_rate": 2.9523333333333336e-05, + "loss": 0.0237, + "step": 11434 + }, + { + "epoch": 9.65793918918919, + "grad_norm": 0.5518100261688232, + "learning_rate": 2.9523e-05, + "loss": 0.033, + "step": 11435 + }, + { + "epoch": 9.658783783783784, + "grad_norm": 0.6758993268013, + "learning_rate": 2.9522666666666668e-05, + "loss": 0.0241, + "step": 11436 + }, + { + "epoch": 9.659628378378379, + "grad_norm": 0.3310321569442749, + "learning_rate": 2.9522333333333337e-05, + "loss": 0.0292, + "step": 11437 + }, + { + "epoch": 9.660472972972974, + "grad_norm": 0.18595094978809357, + "learning_rate": 2.9522e-05, + "loss": 0.0208, + "step": 11438 + }, + { + "epoch": 9.661317567567568, + "grad_norm": 0.38461834192276, + "learning_rate": 2.952166666666667e-05, + "loss": 0.025, + "step": 11439 + }, + { + "epoch": 9.662162162162161, + "grad_norm": 0.7109621167182922, + "learning_rate": 2.9521333333333335e-05, + "loss": 0.0324, + "step": 11440 + }, + { + "epoch": 9.663006756756756, + "grad_norm": 0.284025102853775, + "learning_rate": 2.9521e-05, + "loss": 0.0225, + "step": 11441 + }, + { + "epoch": 9.66385135135135, + "grad_norm": 0.26468536257743835, + "learning_rate": 2.9520666666666667e-05, + "loss": 0.023, + "step": 11442 + }, + { + "epoch": 9.664695945945946, + "grad_norm": 0.8257609605789185, + "learning_rate": 2.9520333333333333e-05, + "loss": 0.0326, + "step": 11443 + }, + { + "epoch": 9.66554054054054, + "grad_norm": 0.38816604018211365, + "learning_rate": 2.9520000000000002e-05, + "loss": 0.0244, + "step": 11444 + }, + { + "epoch": 9.666385135135135, + "grad_norm": 0.44880160689353943, + "learning_rate": 2.9519666666666668e-05, + "loss": 0.0352, + "step": 11445 + }, + { + "epoch": 9.66722972972973, + "grad_norm": 0.5055591464042664, + "learning_rate": 2.9519333333333334e-05, + "loss": 0.0495, + "step": 11446 + }, + { + "epoch": 9.668074324324325, + "grad_norm": 0.41243913769721985, + "learning_rate": 2.9519e-05, + "loss": 0.0328, + "step": 11447 + }, + { + "epoch": 9.66891891891892, + "grad_norm": 0.33121177554130554, + "learning_rate": 2.951866666666667e-05, + "loss": 0.0279, + "step": 11448 + }, + { + "epoch": 9.669763513513514, + "grad_norm": 0.5126593112945557, + "learning_rate": 2.951833333333333e-05, + "loss": 0.0279, + "step": 11449 + }, + { + "epoch": 9.670608108108109, + "grad_norm": 0.48508578538894653, + "learning_rate": 2.9518e-05, + "loss": 0.0376, + "step": 11450 + }, + { + "epoch": 9.671452702702704, + "grad_norm": 0.7999942302703857, + "learning_rate": 2.9517666666666667e-05, + "loss": 0.0302, + "step": 11451 + }, + { + "epoch": 9.672297297297296, + "grad_norm": 0.4692412316799164, + "learning_rate": 2.9517333333333333e-05, + "loss": 0.0374, + "step": 11452 + }, + { + "epoch": 9.673141891891891, + "grad_norm": 1.2477800846099854, + "learning_rate": 2.9517000000000002e-05, + "loss": 0.0608, + "step": 11453 + }, + { + "epoch": 9.673986486486486, + "grad_norm": 0.3062155842781067, + "learning_rate": 2.9516666666666668e-05, + "loss": 0.0124, + "step": 11454 + }, + { + "epoch": 9.67483108108108, + "grad_norm": 0.6415037512779236, + "learning_rate": 2.9516333333333334e-05, + "loss": 0.0483, + "step": 11455 + }, + { + "epoch": 9.675675675675675, + "grad_norm": 1.6434268951416016, + "learning_rate": 2.9516e-05, + "loss": 0.0722, + "step": 11456 + }, + { + "epoch": 9.67652027027027, + "grad_norm": 3.769953966140747, + "learning_rate": 2.951566666666667e-05, + "loss": 0.354, + "step": 11457 + }, + { + "epoch": 9.677364864864865, + "grad_norm": 0.6609859466552734, + "learning_rate": 2.9515333333333332e-05, + "loss": 0.1842, + "step": 11458 + }, + { + "epoch": 9.67820945945946, + "grad_norm": 0.6849238276481628, + "learning_rate": 2.9515e-05, + "loss": 0.0763, + "step": 11459 + }, + { + "epoch": 9.679054054054054, + "grad_norm": 0.43508628010749817, + "learning_rate": 2.9514666666666667e-05, + "loss": 0.0462, + "step": 11460 + }, + { + "epoch": 9.67989864864865, + "grad_norm": 0.29056304693222046, + "learning_rate": 2.9514333333333333e-05, + "loss": 0.0317, + "step": 11461 + }, + { + "epoch": 9.680743243243244, + "grad_norm": 0.4525752067565918, + "learning_rate": 2.9514000000000002e-05, + "loss": 0.0499, + "step": 11462 + }, + { + "epoch": 9.681587837837839, + "grad_norm": 0.4072585701942444, + "learning_rate": 2.9513666666666668e-05, + "loss": 0.0348, + "step": 11463 + }, + { + "epoch": 9.682432432432432, + "grad_norm": 0.3569250702857971, + "learning_rate": 2.9513333333333334e-05, + "loss": 0.0383, + "step": 11464 + }, + { + "epoch": 9.683277027027026, + "grad_norm": 0.4011498689651489, + "learning_rate": 2.9513e-05, + "loss": 0.023, + "step": 11465 + }, + { + "epoch": 9.684121621621621, + "grad_norm": 0.3174823224544525, + "learning_rate": 2.951266666666667e-05, + "loss": 0.0261, + "step": 11466 + }, + { + "epoch": 9.684966216216216, + "grad_norm": 0.21519342064857483, + "learning_rate": 2.9512333333333332e-05, + "loss": 0.0203, + "step": 11467 + }, + { + "epoch": 9.68581081081081, + "grad_norm": 0.2867100238800049, + "learning_rate": 2.9512e-05, + "loss": 0.0279, + "step": 11468 + }, + { + "epoch": 9.686655405405405, + "grad_norm": 0.34842005372047424, + "learning_rate": 2.951166666666667e-05, + "loss": 0.0349, + "step": 11469 + }, + { + "epoch": 9.6875, + "grad_norm": 0.28309935331344604, + "learning_rate": 2.9511333333333333e-05, + "loss": 0.0244, + "step": 11470 + }, + { + "epoch": 9.688344594594595, + "grad_norm": 0.3747639060020447, + "learning_rate": 2.9511000000000003e-05, + "loss": 0.0308, + "step": 11471 + }, + { + "epoch": 9.68918918918919, + "grad_norm": 0.31990617513656616, + "learning_rate": 2.9510666666666665e-05, + "loss": 0.0288, + "step": 11472 + }, + { + "epoch": 9.690033783783784, + "grad_norm": 0.4888480007648468, + "learning_rate": 2.9510333333333334e-05, + "loss": 0.0638, + "step": 11473 + }, + { + "epoch": 9.690878378378379, + "grad_norm": 0.3779074549674988, + "learning_rate": 2.951e-05, + "loss": 0.0352, + "step": 11474 + }, + { + "epoch": 9.691722972972974, + "grad_norm": 0.39540356397628784, + "learning_rate": 2.9509666666666666e-05, + "loss": 0.0315, + "step": 11475 + }, + { + "epoch": 9.692567567567568, + "grad_norm": 0.45179638266563416, + "learning_rate": 2.9509333333333332e-05, + "loss": 0.0383, + "step": 11476 + }, + { + "epoch": 9.693412162162161, + "grad_norm": 0.4401050806045532, + "learning_rate": 2.9509e-05, + "loss": 0.0213, + "step": 11477 + }, + { + "epoch": 9.694256756756756, + "grad_norm": 0.2957468032836914, + "learning_rate": 2.9508666666666668e-05, + "loss": 0.0277, + "step": 11478 + }, + { + "epoch": 9.69510135135135, + "grad_norm": 0.3171297609806061, + "learning_rate": 2.9508333333333333e-05, + "loss": 0.0252, + "step": 11479 + }, + { + "epoch": 9.695945945945946, + "grad_norm": 0.2573707103729248, + "learning_rate": 2.9508000000000003e-05, + "loss": 0.0254, + "step": 11480 + }, + { + "epoch": 9.69679054054054, + "grad_norm": 0.43789684772491455, + "learning_rate": 2.9507666666666665e-05, + "loss": 0.0326, + "step": 11481 + }, + { + "epoch": 9.697635135135135, + "grad_norm": 0.32628974318504333, + "learning_rate": 2.9507333333333335e-05, + "loss": 0.0262, + "step": 11482 + }, + { + "epoch": 9.69847972972973, + "grad_norm": 0.8105659484863281, + "learning_rate": 2.9507e-05, + "loss": 0.0506, + "step": 11483 + }, + { + "epoch": 9.699324324324325, + "grad_norm": 0.22184620797634125, + "learning_rate": 2.9506666666666667e-05, + "loss": 0.0142, + "step": 11484 + }, + { + "epoch": 9.70016891891892, + "grad_norm": 0.4186660051345825, + "learning_rate": 2.9506333333333332e-05, + "loss": 0.019, + "step": 11485 + }, + { + "epoch": 9.701013513513514, + "grad_norm": 1.3705110549926758, + "learning_rate": 2.9506000000000002e-05, + "loss": 0.0523, + "step": 11486 + }, + { + "epoch": 9.701858108108109, + "grad_norm": 0.3118041753768921, + "learning_rate": 2.9505666666666668e-05, + "loss": 0.0296, + "step": 11487 + }, + { + "epoch": 9.702702702702704, + "grad_norm": 0.2923065721988678, + "learning_rate": 2.9505333333333334e-05, + "loss": 0.0341, + "step": 11488 + }, + { + "epoch": 9.703547297297296, + "grad_norm": 0.3332690894603729, + "learning_rate": 2.9505000000000003e-05, + "loss": 0.0243, + "step": 11489 + }, + { + "epoch": 9.704391891891891, + "grad_norm": 0.2891739308834076, + "learning_rate": 2.9504666666666666e-05, + "loss": 0.0356, + "step": 11490 + }, + { + "epoch": 9.705236486486486, + "grad_norm": 0.340045690536499, + "learning_rate": 2.9504333333333335e-05, + "loss": 0.0229, + "step": 11491 + }, + { + "epoch": 9.70608108108108, + "grad_norm": 0.6871470212936401, + "learning_rate": 2.9504e-05, + "loss": 0.0507, + "step": 11492 + }, + { + "epoch": 9.706925675675675, + "grad_norm": 0.46037691831588745, + "learning_rate": 2.9503666666666667e-05, + "loss": 0.0571, + "step": 11493 + }, + { + "epoch": 9.70777027027027, + "grad_norm": 0.40681615471839905, + "learning_rate": 2.9503333333333336e-05, + "loss": 0.0341, + "step": 11494 + }, + { + "epoch": 9.708614864864865, + "grad_norm": 0.40819698572158813, + "learning_rate": 2.9503000000000002e-05, + "loss": 0.0346, + "step": 11495 + }, + { + "epoch": 9.70945945945946, + "grad_norm": 0.29732921719551086, + "learning_rate": 2.9502666666666668e-05, + "loss": 0.0236, + "step": 11496 + }, + { + "epoch": 9.710304054054054, + "grad_norm": 0.5468436479568481, + "learning_rate": 2.9502333333333334e-05, + "loss": 0.0332, + "step": 11497 + }, + { + "epoch": 9.71114864864865, + "grad_norm": 0.43107396364212036, + "learning_rate": 2.9502000000000003e-05, + "loss": 0.0221, + "step": 11498 + }, + { + "epoch": 9.711993243243244, + "grad_norm": 0.31395313143730164, + "learning_rate": 2.9501666666666666e-05, + "loss": 0.0288, + "step": 11499 + }, + { + "epoch": 9.712837837837839, + "grad_norm": 0.6530475616455078, + "learning_rate": 2.9501333333333335e-05, + "loss": 0.0326, + "step": 11500 + }, + { + "epoch": 9.713682432432432, + "grad_norm": 0.49552273750305176, + "learning_rate": 2.9500999999999998e-05, + "loss": 0.0189, + "step": 11501 + }, + { + "epoch": 9.714527027027026, + "grad_norm": 0.39611247181892395, + "learning_rate": 2.9500666666666667e-05, + "loss": 0.0327, + "step": 11502 + }, + { + "epoch": 9.715371621621621, + "grad_norm": 0.5159072875976562, + "learning_rate": 2.9500333333333336e-05, + "loss": 0.0441, + "step": 11503 + }, + { + "epoch": 9.716216216216216, + "grad_norm": 0.6900811195373535, + "learning_rate": 2.95e-05, + "loss": 0.0506, + "step": 11504 + }, + { + "epoch": 9.71706081081081, + "grad_norm": 0.6861036419868469, + "learning_rate": 2.9499666666666668e-05, + "loss": 0.0467, + "step": 11505 + }, + { + "epoch": 9.717905405405405, + "grad_norm": 2.1838951110839844, + "learning_rate": 2.9499333333333334e-05, + "loss": 0.0855, + "step": 11506 + }, + { + "epoch": 9.71875, + "grad_norm": 0.7937610745429993, + "learning_rate": 2.9499e-05, + "loss": 0.2403, + "step": 11507 + }, + { + "epoch": 9.719594594594595, + "grad_norm": 0.5127743482589722, + "learning_rate": 2.9498666666666666e-05, + "loss": 0.1452, + "step": 11508 + }, + { + "epoch": 9.72043918918919, + "grad_norm": 0.43610283732414246, + "learning_rate": 2.9498333333333335e-05, + "loss": 0.0746, + "step": 11509 + }, + { + "epoch": 9.721283783783784, + "grad_norm": 0.613092303276062, + "learning_rate": 2.9497999999999998e-05, + "loss": 0.0563, + "step": 11510 + }, + { + "epoch": 9.722128378378379, + "grad_norm": 0.43245989084243774, + "learning_rate": 2.9497666666666667e-05, + "loss": 0.0345, + "step": 11511 + }, + { + "epoch": 9.722972972972974, + "grad_norm": 0.45245635509490967, + "learning_rate": 2.9497333333333336e-05, + "loss": 0.0319, + "step": 11512 + }, + { + "epoch": 9.723817567567568, + "grad_norm": 0.3154793679714203, + "learning_rate": 2.9497e-05, + "loss": 0.0228, + "step": 11513 + }, + { + "epoch": 9.724662162162161, + "grad_norm": 0.3564782440662384, + "learning_rate": 2.9496666666666668e-05, + "loss": 0.0357, + "step": 11514 + }, + { + "epoch": 9.725506756756756, + "grad_norm": 0.7160162925720215, + "learning_rate": 2.9496333333333334e-05, + "loss": 0.0383, + "step": 11515 + }, + { + "epoch": 9.72635135135135, + "grad_norm": 0.4627059996128082, + "learning_rate": 2.9496e-05, + "loss": 0.055, + "step": 11516 + }, + { + "epoch": 9.727195945945946, + "grad_norm": 0.3711988925933838, + "learning_rate": 2.9495666666666666e-05, + "loss": 0.0338, + "step": 11517 + }, + { + "epoch": 9.72804054054054, + "grad_norm": 0.42199715971946716, + "learning_rate": 2.9495333333333335e-05, + "loss": 0.0264, + "step": 11518 + }, + { + "epoch": 9.728885135135135, + "grad_norm": 0.2485896795988083, + "learning_rate": 2.9495e-05, + "loss": 0.0272, + "step": 11519 + }, + { + "epoch": 9.72972972972973, + "grad_norm": 0.5064272880554199, + "learning_rate": 2.9494666666666667e-05, + "loss": 0.0339, + "step": 11520 + }, + { + "epoch": 9.730574324324325, + "grad_norm": 0.3249787986278534, + "learning_rate": 2.9494333333333337e-05, + "loss": 0.0287, + "step": 11521 + }, + { + "epoch": 9.73141891891892, + "grad_norm": 0.23609650135040283, + "learning_rate": 2.9494e-05, + "loss": 0.0194, + "step": 11522 + }, + { + "epoch": 9.732263513513514, + "grad_norm": 0.40190500020980835, + "learning_rate": 2.949366666666667e-05, + "loss": 0.0365, + "step": 11523 + }, + { + "epoch": 9.733108108108109, + "grad_norm": 0.5424790978431702, + "learning_rate": 2.9493333333333334e-05, + "loss": 0.0231, + "step": 11524 + }, + { + "epoch": 9.733952702702704, + "grad_norm": 0.2626858949661255, + "learning_rate": 2.9493e-05, + "loss": 0.0194, + "step": 11525 + }, + { + "epoch": 9.734797297297296, + "grad_norm": 0.3007703125476837, + "learning_rate": 2.9492666666666666e-05, + "loss": 0.0453, + "step": 11526 + }, + { + "epoch": 9.735641891891891, + "grad_norm": 0.32798805832862854, + "learning_rate": 2.9492333333333336e-05, + "loss": 0.0286, + "step": 11527 + }, + { + "epoch": 9.736486486486486, + "grad_norm": 0.3720741271972656, + "learning_rate": 2.9492e-05, + "loss": 0.0341, + "step": 11528 + }, + { + "epoch": 9.73733108108108, + "grad_norm": 0.39734676480293274, + "learning_rate": 2.9491666666666667e-05, + "loss": 0.0347, + "step": 11529 + }, + { + "epoch": 9.738175675675675, + "grad_norm": 0.36536991596221924, + "learning_rate": 2.9491333333333337e-05, + "loss": 0.0353, + "step": 11530 + }, + { + "epoch": 9.73902027027027, + "grad_norm": 0.4988326132297516, + "learning_rate": 2.9491e-05, + "loss": 0.0485, + "step": 11531 + }, + { + "epoch": 9.739864864864865, + "grad_norm": 0.41347596049308777, + "learning_rate": 2.949066666666667e-05, + "loss": 0.0316, + "step": 11532 + }, + { + "epoch": 9.74070945945946, + "grad_norm": 0.4053260385990143, + "learning_rate": 2.949033333333333e-05, + "loss": 0.0255, + "step": 11533 + }, + { + "epoch": 9.741554054054054, + "grad_norm": 0.3491417169570923, + "learning_rate": 2.949e-05, + "loss": 0.0235, + "step": 11534 + }, + { + "epoch": 9.74239864864865, + "grad_norm": 0.37287598848342896, + "learning_rate": 2.9489666666666666e-05, + "loss": 0.0185, + "step": 11535 + }, + { + "epoch": 9.743243243243244, + "grad_norm": 0.5061323046684265, + "learning_rate": 2.9489333333333332e-05, + "loss": 0.0548, + "step": 11536 + }, + { + "epoch": 9.744087837837839, + "grad_norm": 0.3142039477825165, + "learning_rate": 2.9489000000000002e-05, + "loss": 0.0256, + "step": 11537 + }, + { + "epoch": 9.744932432432432, + "grad_norm": 0.34030330181121826, + "learning_rate": 2.9488666666666668e-05, + "loss": 0.0315, + "step": 11538 + }, + { + "epoch": 9.745777027027026, + "grad_norm": 0.5685747265815735, + "learning_rate": 2.9488333333333334e-05, + "loss": 0.0345, + "step": 11539 + }, + { + "epoch": 9.746621621621621, + "grad_norm": 0.4604559540748596, + "learning_rate": 2.9488e-05, + "loss": 0.04, + "step": 11540 + }, + { + "epoch": 9.747466216216216, + "grad_norm": 0.37917643785476685, + "learning_rate": 2.948766666666667e-05, + "loss": 0.0128, + "step": 11541 + }, + { + "epoch": 9.74831081081081, + "grad_norm": 0.29924476146698, + "learning_rate": 2.948733333333333e-05, + "loss": 0.0328, + "step": 11542 + }, + { + "epoch": 9.749155405405405, + "grad_norm": 0.973938524723053, + "learning_rate": 2.9487e-05, + "loss": 0.0329, + "step": 11543 + }, + { + "epoch": 9.75, + "grad_norm": 0.3047575354576111, + "learning_rate": 2.948666666666667e-05, + "loss": 0.026, + "step": 11544 + }, + { + "epoch": 9.750844594594595, + "grad_norm": 0.4979707896709442, + "learning_rate": 2.9486333333333333e-05, + "loss": 0.0339, + "step": 11545 + }, + { + "epoch": 9.75168918918919, + "grad_norm": 0.2997751235961914, + "learning_rate": 2.9486000000000002e-05, + "loss": 0.029, + "step": 11546 + }, + { + "epoch": 9.752533783783784, + "grad_norm": 0.6338545083999634, + "learning_rate": 2.9485666666666668e-05, + "loss": 0.0487, + "step": 11547 + }, + { + "epoch": 9.753378378378379, + "grad_norm": 0.3480209410190582, + "learning_rate": 2.9485333333333334e-05, + "loss": 0.0198, + "step": 11548 + }, + { + "epoch": 9.754222972972974, + "grad_norm": 0.42820343375205994, + "learning_rate": 2.9485e-05, + "loss": 0.0289, + "step": 11549 + }, + { + "epoch": 9.755067567567568, + "grad_norm": 0.6857536435127258, + "learning_rate": 2.948466666666667e-05, + "loss": 0.0419, + "step": 11550 + }, + { + "epoch": 9.755912162162161, + "grad_norm": 0.5036264061927795, + "learning_rate": 2.948433333333333e-05, + "loss": 0.0795, + "step": 11551 + }, + { + "epoch": 9.756756756756756, + "grad_norm": 0.4717893600463867, + "learning_rate": 2.9484e-05, + "loss": 0.0378, + "step": 11552 + }, + { + "epoch": 9.75760135135135, + "grad_norm": 0.6484730243682861, + "learning_rate": 2.948366666666667e-05, + "loss": 0.0548, + "step": 11553 + }, + { + "epoch": 9.758445945945946, + "grad_norm": 1.3461549282073975, + "learning_rate": 2.9483333333333333e-05, + "loss": 0.0606, + "step": 11554 + }, + { + "epoch": 9.75929054054054, + "grad_norm": 0.3906731903553009, + "learning_rate": 2.9483000000000002e-05, + "loss": 0.0522, + "step": 11555 + }, + { + "epoch": 9.760135135135135, + "grad_norm": 1.7401694059371948, + "learning_rate": 2.9482666666666668e-05, + "loss": 0.0597, + "step": 11556 + }, + { + "epoch": 9.76097972972973, + "grad_norm": 0.816433846950531, + "learning_rate": 2.9482333333333334e-05, + "loss": 0.2762, + "step": 11557 + }, + { + "epoch": 9.761824324324325, + "grad_norm": 0.550595223903656, + "learning_rate": 2.9482e-05, + "loss": 0.1367, + "step": 11558 + }, + { + "epoch": 9.76266891891892, + "grad_norm": 0.3858853578567505, + "learning_rate": 2.948166666666667e-05, + "loss": 0.0704, + "step": 11559 + }, + { + "epoch": 9.763513513513514, + "grad_norm": 0.5336297154426575, + "learning_rate": 2.9481333333333332e-05, + "loss": 0.0652, + "step": 11560 + }, + { + "epoch": 9.764358108108109, + "grad_norm": 0.29236096143722534, + "learning_rate": 2.9481e-05, + "loss": 0.0317, + "step": 11561 + }, + { + "epoch": 9.765202702702704, + "grad_norm": 0.41388845443725586, + "learning_rate": 2.9480666666666667e-05, + "loss": 0.0418, + "step": 11562 + }, + { + "epoch": 9.766047297297296, + "grad_norm": 0.460034042596817, + "learning_rate": 2.9480333333333333e-05, + "loss": 0.0309, + "step": 11563 + }, + { + "epoch": 9.766891891891891, + "grad_norm": 0.5000514984130859, + "learning_rate": 2.9480000000000002e-05, + "loss": 0.0353, + "step": 11564 + }, + { + "epoch": 9.767736486486486, + "grad_norm": 0.4920153021812439, + "learning_rate": 2.9479666666666665e-05, + "loss": 0.0521, + "step": 11565 + }, + { + "epoch": 9.76858108108108, + "grad_norm": 0.4267587661743164, + "learning_rate": 2.9479333333333334e-05, + "loss": 0.0502, + "step": 11566 + }, + { + "epoch": 9.769425675675675, + "grad_norm": 0.38450661301612854, + "learning_rate": 2.9479e-05, + "loss": 0.0582, + "step": 11567 + }, + { + "epoch": 9.77027027027027, + "grad_norm": 0.5109238028526306, + "learning_rate": 2.9478666666666666e-05, + "loss": 0.0335, + "step": 11568 + }, + { + "epoch": 9.771114864864865, + "grad_norm": 0.30025118589401245, + "learning_rate": 2.9478333333333335e-05, + "loss": 0.0265, + "step": 11569 + }, + { + "epoch": 9.77195945945946, + "grad_norm": 0.4837532937526703, + "learning_rate": 2.9478e-05, + "loss": 0.0448, + "step": 11570 + }, + { + "epoch": 9.772804054054054, + "grad_norm": 0.3684430420398712, + "learning_rate": 2.9477666666666667e-05, + "loss": 0.0359, + "step": 11571 + }, + { + "epoch": 9.77364864864865, + "grad_norm": 0.6108958125114441, + "learning_rate": 2.9477333333333333e-05, + "loss": 0.0456, + "step": 11572 + }, + { + "epoch": 9.774493243243244, + "grad_norm": 0.3734204173088074, + "learning_rate": 2.9477000000000003e-05, + "loss": 0.035, + "step": 11573 + }, + { + "epoch": 9.775337837837839, + "grad_norm": 0.45653587579727173, + "learning_rate": 2.9476666666666665e-05, + "loss": 0.0352, + "step": 11574 + }, + { + "epoch": 9.776182432432432, + "grad_norm": 0.3801581561565399, + "learning_rate": 2.9476333333333334e-05, + "loss": 0.0303, + "step": 11575 + }, + { + "epoch": 9.777027027027026, + "grad_norm": 0.3115236759185791, + "learning_rate": 2.9476e-05, + "loss": 0.0229, + "step": 11576 + }, + { + "epoch": 9.777871621621621, + "grad_norm": 0.3072820007801056, + "learning_rate": 2.9475666666666666e-05, + "loss": 0.0219, + "step": 11577 + }, + { + "epoch": 9.778716216216216, + "grad_norm": 0.27425187826156616, + "learning_rate": 2.9475333333333336e-05, + "loss": 0.0189, + "step": 11578 + }, + { + "epoch": 9.77956081081081, + "grad_norm": 0.43212759494781494, + "learning_rate": 2.9475e-05, + "loss": 0.0324, + "step": 11579 + }, + { + "epoch": 9.780405405405405, + "grad_norm": 0.344857394695282, + "learning_rate": 2.9474666666666667e-05, + "loss": 0.0372, + "step": 11580 + }, + { + "epoch": 9.78125, + "grad_norm": 0.2836344540119171, + "learning_rate": 2.9474333333333333e-05, + "loss": 0.0151, + "step": 11581 + }, + { + "epoch": 9.782094594594595, + "grad_norm": 0.40009811520576477, + "learning_rate": 2.9474000000000003e-05, + "loss": 0.0485, + "step": 11582 + }, + { + "epoch": 9.78293918918919, + "grad_norm": 0.5030900239944458, + "learning_rate": 2.9473666666666665e-05, + "loss": 0.0154, + "step": 11583 + }, + { + "epoch": 9.783783783783784, + "grad_norm": 0.40995514392852783, + "learning_rate": 2.9473333333333335e-05, + "loss": 0.0285, + "step": 11584 + }, + { + "epoch": 9.784628378378379, + "grad_norm": 0.45917394757270813, + "learning_rate": 2.9473e-05, + "loss": 0.0363, + "step": 11585 + }, + { + "epoch": 9.785472972972974, + "grad_norm": 0.5122460722923279, + "learning_rate": 2.9472666666666666e-05, + "loss": 0.0376, + "step": 11586 + }, + { + "epoch": 9.786317567567568, + "grad_norm": 0.581379234790802, + "learning_rate": 2.9472333333333336e-05, + "loss": 0.0329, + "step": 11587 + }, + { + "epoch": 9.787162162162161, + "grad_norm": 0.29735156893730164, + "learning_rate": 2.9472000000000002e-05, + "loss": 0.0398, + "step": 11588 + }, + { + "epoch": 9.788006756756756, + "grad_norm": 0.27382129430770874, + "learning_rate": 2.9471666666666668e-05, + "loss": 0.0304, + "step": 11589 + }, + { + "epoch": 9.78885135135135, + "grad_norm": 0.22817151248455048, + "learning_rate": 2.9471333333333334e-05, + "loss": 0.0223, + "step": 11590 + }, + { + "epoch": 9.789695945945946, + "grad_norm": 0.3414807617664337, + "learning_rate": 2.9471000000000003e-05, + "loss": 0.031, + "step": 11591 + }, + { + "epoch": 9.79054054054054, + "grad_norm": 0.33237096667289734, + "learning_rate": 2.9470666666666665e-05, + "loss": 0.0244, + "step": 11592 + }, + { + "epoch": 9.791385135135135, + "grad_norm": 0.426766961812973, + "learning_rate": 2.9470333333333335e-05, + "loss": 0.0251, + "step": 11593 + }, + { + "epoch": 9.79222972972973, + "grad_norm": 0.5166851878166199, + "learning_rate": 2.947e-05, + "loss": 0.043, + "step": 11594 + }, + { + "epoch": 9.793074324324325, + "grad_norm": 0.3372207581996918, + "learning_rate": 2.9469666666666667e-05, + "loss": 0.0314, + "step": 11595 + }, + { + "epoch": 9.79391891891892, + "grad_norm": 0.3481924533843994, + "learning_rate": 2.9469333333333336e-05, + "loss": 0.0401, + "step": 11596 + }, + { + "epoch": 9.794763513513514, + "grad_norm": 0.5960468649864197, + "learning_rate": 2.9469e-05, + "loss": 0.0428, + "step": 11597 + }, + { + "epoch": 9.795608108108109, + "grad_norm": 0.3024349510669708, + "learning_rate": 2.9468666666666668e-05, + "loss": 0.0189, + "step": 11598 + }, + { + "epoch": 9.796452702702704, + "grad_norm": 0.36046478152275085, + "learning_rate": 2.9468333333333334e-05, + "loss": 0.0238, + "step": 11599 + }, + { + "epoch": 9.797297297297296, + "grad_norm": 0.3725576400756836, + "learning_rate": 2.9468e-05, + "loss": 0.0164, + "step": 11600 + }, + { + "epoch": 9.798141891891891, + "grad_norm": 0.45045459270477295, + "learning_rate": 2.9467666666666666e-05, + "loss": 0.025, + "step": 11601 + }, + { + "epoch": 9.798986486486486, + "grad_norm": 0.5007177591323853, + "learning_rate": 2.9467333333333335e-05, + "loss": 0.0328, + "step": 11602 + }, + { + "epoch": 9.79983108108108, + "grad_norm": 0.4552939534187317, + "learning_rate": 2.9467e-05, + "loss": 0.045, + "step": 11603 + }, + { + "epoch": 9.800675675675675, + "grad_norm": 0.39114105701446533, + "learning_rate": 2.9466666666666667e-05, + "loss": 0.0251, + "step": 11604 + }, + { + "epoch": 9.80152027027027, + "grad_norm": 0.4210653603076935, + "learning_rate": 2.9466333333333336e-05, + "loss": 0.0373, + "step": 11605 + }, + { + "epoch": 9.802364864864865, + "grad_norm": 0.8820524215698242, + "learning_rate": 2.9466e-05, + "loss": 0.0474, + "step": 11606 + }, + { + "epoch": 9.80320945945946, + "grad_norm": 0.7457113265991211, + "learning_rate": 2.9465666666666668e-05, + "loss": 0.2938, + "step": 11607 + }, + { + "epoch": 9.804054054054054, + "grad_norm": 0.5946821570396423, + "learning_rate": 2.9465333333333334e-05, + "loss": 0.1428, + "step": 11608 + }, + { + "epoch": 9.80489864864865, + "grad_norm": 0.7460871934890747, + "learning_rate": 2.9465e-05, + "loss": 0.1136, + "step": 11609 + }, + { + "epoch": 9.805743243243244, + "grad_norm": 0.34222736954689026, + "learning_rate": 2.9464666666666666e-05, + "loss": 0.0606, + "step": 11610 + }, + { + "epoch": 9.806587837837839, + "grad_norm": 0.31980910897254944, + "learning_rate": 2.9464333333333335e-05, + "loss": 0.0434, + "step": 11611 + }, + { + "epoch": 9.807432432432432, + "grad_norm": 0.6057620048522949, + "learning_rate": 2.9464e-05, + "loss": 0.0403, + "step": 11612 + }, + { + "epoch": 9.808277027027026, + "grad_norm": 0.34464699029922485, + "learning_rate": 2.9463666666666667e-05, + "loss": 0.0187, + "step": 11613 + }, + { + "epoch": 9.809121621621621, + "grad_norm": 0.27097222208976746, + "learning_rate": 2.9463333333333336e-05, + "loss": 0.0331, + "step": 11614 + }, + { + "epoch": 9.809966216216216, + "grad_norm": 0.19653314352035522, + "learning_rate": 2.9463e-05, + "loss": 0.0153, + "step": 11615 + }, + { + "epoch": 9.81081081081081, + "grad_norm": 0.3244788348674774, + "learning_rate": 2.9462666666666668e-05, + "loss": 0.0347, + "step": 11616 + }, + { + "epoch": 9.811655405405405, + "grad_norm": 0.47863641381263733, + "learning_rate": 2.9462333333333334e-05, + "loss": 0.0409, + "step": 11617 + }, + { + "epoch": 9.8125, + "grad_norm": 0.2476755529642105, + "learning_rate": 2.9462e-05, + "loss": 0.0266, + "step": 11618 + }, + { + "epoch": 9.813344594594595, + "grad_norm": 0.29827287793159485, + "learning_rate": 2.946166666666667e-05, + "loss": 0.0388, + "step": 11619 + }, + { + "epoch": 9.81418918918919, + "grad_norm": 0.35479024052619934, + "learning_rate": 2.9461333333333335e-05, + "loss": 0.026, + "step": 11620 + }, + { + "epoch": 9.815033783783784, + "grad_norm": 0.330532968044281, + "learning_rate": 2.9461e-05, + "loss": 0.0272, + "step": 11621 + }, + { + "epoch": 9.815878378378379, + "grad_norm": 0.332044780254364, + "learning_rate": 2.9460666666666667e-05, + "loss": 0.0306, + "step": 11622 + }, + { + "epoch": 9.816722972972974, + "grad_norm": 0.30589213967323303, + "learning_rate": 2.9460333333333333e-05, + "loss": 0.0323, + "step": 11623 + }, + { + "epoch": 9.817567567567568, + "grad_norm": 0.34282752871513367, + "learning_rate": 2.946e-05, + "loss": 0.027, + "step": 11624 + }, + { + "epoch": 9.818412162162161, + "grad_norm": 0.18972864747047424, + "learning_rate": 2.945966666666667e-05, + "loss": 0.0133, + "step": 11625 + }, + { + "epoch": 9.819256756756756, + "grad_norm": 0.2317439466714859, + "learning_rate": 2.945933333333333e-05, + "loss": 0.016, + "step": 11626 + }, + { + "epoch": 9.82010135135135, + "grad_norm": 0.2715020477771759, + "learning_rate": 2.9459e-05, + "loss": 0.0302, + "step": 11627 + }, + { + "epoch": 9.820945945945946, + "grad_norm": 0.345406711101532, + "learning_rate": 2.945866666666667e-05, + "loss": 0.0338, + "step": 11628 + }, + { + "epoch": 9.82179054054054, + "grad_norm": 0.41448649764060974, + "learning_rate": 2.9458333333333332e-05, + "loss": 0.0304, + "step": 11629 + }, + { + "epoch": 9.822635135135135, + "grad_norm": 0.2881683111190796, + "learning_rate": 2.9458e-05, + "loss": 0.0339, + "step": 11630 + }, + { + "epoch": 9.82347972972973, + "grad_norm": 0.2375568002462387, + "learning_rate": 2.9457666666666667e-05, + "loss": 0.0292, + "step": 11631 + }, + { + "epoch": 9.824324324324325, + "grad_norm": 0.5967161059379578, + "learning_rate": 2.9457333333333333e-05, + "loss": 0.024, + "step": 11632 + }, + { + "epoch": 9.82516891891892, + "grad_norm": 0.2860706150531769, + "learning_rate": 2.9457e-05, + "loss": 0.0281, + "step": 11633 + }, + { + "epoch": 9.826013513513514, + "grad_norm": 0.3341441750526428, + "learning_rate": 2.945666666666667e-05, + "loss": 0.0184, + "step": 11634 + }, + { + "epoch": 9.826858108108109, + "grad_norm": 0.42909589409828186, + "learning_rate": 2.945633333333333e-05, + "loss": 0.0306, + "step": 11635 + }, + { + "epoch": 9.827702702702704, + "grad_norm": 0.2985861599445343, + "learning_rate": 2.9456e-05, + "loss": 0.0372, + "step": 11636 + }, + { + "epoch": 9.828547297297296, + "grad_norm": 0.30486130714416504, + "learning_rate": 2.945566666666667e-05, + "loss": 0.0403, + "step": 11637 + }, + { + "epoch": 9.829391891891891, + "grad_norm": 0.24939383566379547, + "learning_rate": 2.9455333333333332e-05, + "loss": 0.0168, + "step": 11638 + }, + { + "epoch": 9.830236486486486, + "grad_norm": 0.42539694905281067, + "learning_rate": 2.9455000000000002e-05, + "loss": 0.0422, + "step": 11639 + }, + { + "epoch": 9.83108108108108, + "grad_norm": 0.5015351176261902, + "learning_rate": 2.9454666666666668e-05, + "loss": 0.031, + "step": 11640 + }, + { + "epoch": 9.831925675675675, + "grad_norm": 0.34932583570480347, + "learning_rate": 2.9454333333333334e-05, + "loss": 0.0368, + "step": 11641 + }, + { + "epoch": 9.83277027027027, + "grad_norm": 0.3832683861255646, + "learning_rate": 2.9454e-05, + "loss": 0.0316, + "step": 11642 + }, + { + "epoch": 9.833614864864865, + "grad_norm": 0.4169631600379944, + "learning_rate": 2.945366666666667e-05, + "loss": 0.0265, + "step": 11643 + }, + { + "epoch": 9.83445945945946, + "grad_norm": 0.5007480382919312, + "learning_rate": 2.9453333333333335e-05, + "loss": 0.0333, + "step": 11644 + }, + { + "epoch": 9.835304054054054, + "grad_norm": 0.44211119413375854, + "learning_rate": 2.9453e-05, + "loss": 0.0195, + "step": 11645 + }, + { + "epoch": 9.83614864864865, + "grad_norm": 0.3528074622154236, + "learning_rate": 2.945266666666667e-05, + "loss": 0.0272, + "step": 11646 + }, + { + "epoch": 9.836993243243244, + "grad_norm": 0.6358624696731567, + "learning_rate": 2.9452333333333333e-05, + "loss": 0.0365, + "step": 11647 + }, + { + "epoch": 9.837837837837839, + "grad_norm": 0.22900459170341492, + "learning_rate": 2.9452000000000002e-05, + "loss": 0.0236, + "step": 11648 + }, + { + "epoch": 9.838682432432432, + "grad_norm": 0.2601896822452545, + "learning_rate": 2.9451666666666668e-05, + "loss": 0.0251, + "step": 11649 + }, + { + "epoch": 9.839527027027026, + "grad_norm": 0.3694901466369629, + "learning_rate": 2.9451333333333334e-05, + "loss": 0.0319, + "step": 11650 + }, + { + "epoch": 9.840371621621621, + "grad_norm": 0.4978238344192505, + "learning_rate": 2.9451e-05, + "loss": 0.0325, + "step": 11651 + }, + { + "epoch": 9.841216216216216, + "grad_norm": 0.5228124260902405, + "learning_rate": 2.945066666666667e-05, + "loss": 0.0376, + "step": 11652 + }, + { + "epoch": 9.84206081081081, + "grad_norm": 0.3409355580806732, + "learning_rate": 2.9450333333333335e-05, + "loss": 0.0363, + "step": 11653 + }, + { + "epoch": 9.842905405405405, + "grad_norm": 0.5205913186073303, + "learning_rate": 2.945e-05, + "loss": 0.0434, + "step": 11654 + }, + { + "epoch": 9.84375, + "grad_norm": 0.4332854151725769, + "learning_rate": 2.9449666666666667e-05, + "loss": 0.0383, + "step": 11655 + }, + { + "epoch": 9.844594594594595, + "grad_norm": 1.4719290733337402, + "learning_rate": 2.9449333333333333e-05, + "loss": 0.0499, + "step": 11656 + }, + { + "epoch": 9.84543918918919, + "grad_norm": 0.7727394104003906, + "learning_rate": 2.9449000000000002e-05, + "loss": 0.3111, + "step": 11657 + }, + { + "epoch": 9.846283783783784, + "grad_norm": 0.794484555721283, + "learning_rate": 2.9448666666666665e-05, + "loss": 0.1984, + "step": 11658 + }, + { + "epoch": 9.847128378378379, + "grad_norm": 0.5306296944618225, + "learning_rate": 2.9448333333333334e-05, + "loss": 0.1437, + "step": 11659 + }, + { + "epoch": 9.847972972972974, + "grad_norm": 0.3885365426540375, + "learning_rate": 2.9448e-05, + "loss": 0.066, + "step": 11660 + }, + { + "epoch": 9.848817567567568, + "grad_norm": 0.4399680495262146, + "learning_rate": 2.9447666666666666e-05, + "loss": 0.0492, + "step": 11661 + }, + { + "epoch": 9.849662162162161, + "grad_norm": 0.6149874925613403, + "learning_rate": 2.9447333333333335e-05, + "loss": 0.0424, + "step": 11662 + }, + { + "epoch": 9.850506756756756, + "grad_norm": 0.4546169936656952, + "learning_rate": 2.9447e-05, + "loss": 0.0373, + "step": 11663 + }, + { + "epoch": 9.85135135135135, + "grad_norm": 0.43821480870246887, + "learning_rate": 2.9446666666666667e-05, + "loss": 0.0569, + "step": 11664 + }, + { + "epoch": 9.852195945945946, + "grad_norm": 0.3460332453250885, + "learning_rate": 2.9446333333333333e-05, + "loss": 0.0353, + "step": 11665 + }, + { + "epoch": 9.85304054054054, + "grad_norm": 0.49706119298934937, + "learning_rate": 2.9446000000000002e-05, + "loss": 0.0314, + "step": 11666 + }, + { + "epoch": 9.853885135135135, + "grad_norm": 0.42345309257507324, + "learning_rate": 2.9445666666666665e-05, + "loss": 0.0473, + "step": 11667 + }, + { + "epoch": 9.85472972972973, + "grad_norm": 0.3382772207260132, + "learning_rate": 2.9445333333333334e-05, + "loss": 0.0432, + "step": 11668 + }, + { + "epoch": 9.855574324324325, + "grad_norm": 0.2876442074775696, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0238, + "step": 11669 + }, + { + "epoch": 9.85641891891892, + "grad_norm": 0.2508351802825928, + "learning_rate": 2.9444666666666666e-05, + "loss": 0.0228, + "step": 11670 + }, + { + "epoch": 9.857263513513514, + "grad_norm": 0.3397003412246704, + "learning_rate": 2.9444333333333335e-05, + "loss": 0.0312, + "step": 11671 + }, + { + "epoch": 9.858108108108109, + "grad_norm": 0.304871529340744, + "learning_rate": 2.9444e-05, + "loss": 0.0439, + "step": 11672 + }, + { + "epoch": 9.858952702702704, + "grad_norm": 0.30062341690063477, + "learning_rate": 2.9443666666666667e-05, + "loss": 0.0288, + "step": 11673 + }, + { + "epoch": 9.859797297297296, + "grad_norm": 0.18864195048809052, + "learning_rate": 2.9443333333333333e-05, + "loss": 0.023, + "step": 11674 + }, + { + "epoch": 9.860641891891891, + "grad_norm": 0.2527582049369812, + "learning_rate": 2.9443000000000003e-05, + "loss": 0.0293, + "step": 11675 + }, + { + "epoch": 9.861486486486486, + "grad_norm": 0.3321996033191681, + "learning_rate": 2.9442666666666665e-05, + "loss": 0.0259, + "step": 11676 + }, + { + "epoch": 9.86233108108108, + "grad_norm": 0.5244556665420532, + "learning_rate": 2.9442333333333334e-05, + "loss": 0.0416, + "step": 11677 + }, + { + "epoch": 9.863175675675675, + "grad_norm": 0.17376354336738586, + "learning_rate": 2.9442000000000004e-05, + "loss": 0.0148, + "step": 11678 + }, + { + "epoch": 9.86402027027027, + "grad_norm": 0.29542237520217896, + "learning_rate": 2.9441666666666666e-05, + "loss": 0.0327, + "step": 11679 + }, + { + "epoch": 9.864864864864865, + "grad_norm": 0.7350746393203735, + "learning_rate": 2.9441333333333336e-05, + "loss": 0.0333, + "step": 11680 + }, + { + "epoch": 9.86570945945946, + "grad_norm": 0.247551828622818, + "learning_rate": 2.9441e-05, + "loss": 0.02, + "step": 11681 + }, + { + "epoch": 9.866554054054054, + "grad_norm": 0.3189364969730377, + "learning_rate": 2.9440666666666667e-05, + "loss": 0.0308, + "step": 11682 + }, + { + "epoch": 9.86739864864865, + "grad_norm": 0.44633468985557556, + "learning_rate": 2.9440333333333333e-05, + "loss": 0.0424, + "step": 11683 + }, + { + "epoch": 9.868243243243244, + "grad_norm": 0.2726702094078064, + "learning_rate": 2.944e-05, + "loss": 0.0226, + "step": 11684 + }, + { + "epoch": 9.869087837837839, + "grad_norm": 0.2997015714645386, + "learning_rate": 2.9439666666666665e-05, + "loss": 0.018, + "step": 11685 + }, + { + "epoch": 9.869932432432432, + "grad_norm": 0.33302661776542664, + "learning_rate": 2.9439333333333335e-05, + "loss": 0.0373, + "step": 11686 + }, + { + "epoch": 9.870777027027026, + "grad_norm": 0.3673003911972046, + "learning_rate": 2.9439e-05, + "loss": 0.0356, + "step": 11687 + }, + { + "epoch": 9.871621621621621, + "grad_norm": 0.36031845211982727, + "learning_rate": 2.9438666666666666e-05, + "loss": 0.0316, + "step": 11688 + }, + { + "epoch": 9.872466216216216, + "grad_norm": 0.5690703988075256, + "learning_rate": 2.9438333333333336e-05, + "loss": 0.0301, + "step": 11689 + }, + { + "epoch": 9.87331081081081, + "grad_norm": 0.3150649666786194, + "learning_rate": 2.9438e-05, + "loss": 0.0265, + "step": 11690 + }, + { + "epoch": 9.874155405405405, + "grad_norm": 0.4565725028514862, + "learning_rate": 2.9437666666666668e-05, + "loss": 0.0339, + "step": 11691 + }, + { + "epoch": 9.875, + "grad_norm": 0.5222131609916687, + "learning_rate": 2.9437333333333334e-05, + "loss": 0.0294, + "step": 11692 + }, + { + "epoch": 9.875844594594595, + "grad_norm": 0.32187026739120483, + "learning_rate": 2.9437e-05, + "loss": 0.0253, + "step": 11693 + }, + { + "epoch": 9.87668918918919, + "grad_norm": 0.3127923905849457, + "learning_rate": 2.943666666666667e-05, + "loss": 0.0263, + "step": 11694 + }, + { + "epoch": 9.877533783783784, + "grad_norm": 0.3759300112724304, + "learning_rate": 2.9436333333333335e-05, + "loss": 0.0287, + "step": 11695 + }, + { + "epoch": 9.878378378378379, + "grad_norm": 0.7834782600402832, + "learning_rate": 2.9436e-05, + "loss": 0.0451, + "step": 11696 + }, + { + "epoch": 9.879222972972974, + "grad_norm": 0.6577222943305969, + "learning_rate": 2.9435666666666667e-05, + "loss": 0.036, + "step": 11697 + }, + { + "epoch": 9.880067567567568, + "grad_norm": 0.523807942867279, + "learning_rate": 2.9435333333333336e-05, + "loss": 0.0281, + "step": 11698 + }, + { + "epoch": 9.880912162162161, + "grad_norm": 0.3653657138347626, + "learning_rate": 2.9435e-05, + "loss": 0.0285, + "step": 11699 + }, + { + "epoch": 9.881756756756756, + "grad_norm": 0.2777361571788788, + "learning_rate": 2.9434666666666668e-05, + "loss": 0.0313, + "step": 11700 + }, + { + "epoch": 9.88260135135135, + "grad_norm": 0.7698178887367249, + "learning_rate": 2.9434333333333334e-05, + "loss": 0.0376, + "step": 11701 + }, + { + "epoch": 9.883445945945946, + "grad_norm": 0.7226001620292664, + "learning_rate": 2.9434e-05, + "loss": 0.0471, + "step": 11702 + }, + { + "epoch": 9.88429054054054, + "grad_norm": 0.33473941683769226, + "learning_rate": 2.943366666666667e-05, + "loss": 0.0268, + "step": 11703 + }, + { + "epoch": 9.885135135135135, + "grad_norm": 0.6177241802215576, + "learning_rate": 2.9433333333333335e-05, + "loss": 0.0564, + "step": 11704 + }, + { + "epoch": 9.88597972972973, + "grad_norm": 0.6432079076766968, + "learning_rate": 2.9433e-05, + "loss": 0.05, + "step": 11705 + }, + { + "epoch": 9.886824324324325, + "grad_norm": 0.9464672207832336, + "learning_rate": 2.9432666666666667e-05, + "loss": 0.0391, + "step": 11706 + }, + { + "epoch": 9.88766891891892, + "grad_norm": 0.7225475907325745, + "learning_rate": 2.9432333333333336e-05, + "loss": 0.2087, + "step": 11707 + }, + { + "epoch": 9.888513513513514, + "grad_norm": 0.7199450135231018, + "learning_rate": 2.9432e-05, + "loss": 0.2221, + "step": 11708 + }, + { + "epoch": 9.889358108108109, + "grad_norm": 0.5561912059783936, + "learning_rate": 2.9431666666666668e-05, + "loss": 0.0794, + "step": 11709 + }, + { + "epoch": 9.890202702702704, + "grad_norm": 0.3432846665382385, + "learning_rate": 2.9431333333333334e-05, + "loss": 0.0384, + "step": 11710 + }, + { + "epoch": 9.891047297297296, + "grad_norm": 0.3818921148777008, + "learning_rate": 2.9431e-05, + "loss": 0.0566, + "step": 11711 + }, + { + "epoch": 9.891891891891891, + "grad_norm": 0.3042685389518738, + "learning_rate": 2.943066666666667e-05, + "loss": 0.0359, + "step": 11712 + }, + { + "epoch": 9.892736486486486, + "grad_norm": 0.28423017263412476, + "learning_rate": 2.9430333333333332e-05, + "loss": 0.0318, + "step": 11713 + }, + { + "epoch": 9.89358108108108, + "grad_norm": 0.30477485060691833, + "learning_rate": 2.943e-05, + "loss": 0.029, + "step": 11714 + }, + { + "epoch": 9.894425675675675, + "grad_norm": 0.6365103721618652, + "learning_rate": 2.9429666666666667e-05, + "loss": 0.0365, + "step": 11715 + }, + { + "epoch": 9.89527027027027, + "grad_norm": 0.26579082012176514, + "learning_rate": 2.9429333333333333e-05, + "loss": 0.0276, + "step": 11716 + }, + { + "epoch": 9.896114864864865, + "grad_norm": 0.2075004279613495, + "learning_rate": 2.9429e-05, + "loss": 0.0274, + "step": 11717 + }, + { + "epoch": 9.89695945945946, + "grad_norm": 0.42918112874031067, + "learning_rate": 2.9428666666666668e-05, + "loss": 0.0353, + "step": 11718 + }, + { + "epoch": 9.897804054054054, + "grad_norm": 0.21017642319202423, + "learning_rate": 2.9428333333333334e-05, + "loss": 0.0208, + "step": 11719 + }, + { + "epoch": 9.89864864864865, + "grad_norm": 0.32073718309402466, + "learning_rate": 2.9428e-05, + "loss": 0.0455, + "step": 11720 + }, + { + "epoch": 9.899493243243244, + "grad_norm": 0.27709296345710754, + "learning_rate": 2.942766666666667e-05, + "loss": 0.0246, + "step": 11721 + }, + { + "epoch": 9.900337837837839, + "grad_norm": 0.3211188316345215, + "learning_rate": 2.9427333333333332e-05, + "loss": 0.03, + "step": 11722 + }, + { + "epoch": 9.901182432432432, + "grad_norm": 0.40468889474868774, + "learning_rate": 2.9427e-05, + "loss": 0.0309, + "step": 11723 + }, + { + "epoch": 9.902027027027026, + "grad_norm": 0.2132161259651184, + "learning_rate": 2.9426666666666667e-05, + "loss": 0.017, + "step": 11724 + }, + { + "epoch": 9.902871621621621, + "grad_norm": 0.2701455354690552, + "learning_rate": 2.9426333333333333e-05, + "loss": 0.0257, + "step": 11725 + }, + { + "epoch": 9.903716216216216, + "grad_norm": 0.49376538395881653, + "learning_rate": 2.9426e-05, + "loss": 0.0356, + "step": 11726 + }, + { + "epoch": 9.90456081081081, + "grad_norm": 0.31789153814315796, + "learning_rate": 2.942566666666667e-05, + "loss": 0.0214, + "step": 11727 + }, + { + "epoch": 9.905405405405405, + "grad_norm": 0.4995053708553314, + "learning_rate": 2.9425333333333334e-05, + "loss": 0.0389, + "step": 11728 + }, + { + "epoch": 9.90625, + "grad_norm": 0.4722929000854492, + "learning_rate": 2.9425e-05, + "loss": 0.0372, + "step": 11729 + }, + { + "epoch": 9.907094594594595, + "grad_norm": 0.49319419264793396, + "learning_rate": 2.942466666666667e-05, + "loss": 0.046, + "step": 11730 + }, + { + "epoch": 9.90793918918919, + "grad_norm": 0.2797389626502991, + "learning_rate": 2.9424333333333332e-05, + "loss": 0.0277, + "step": 11731 + }, + { + "epoch": 9.908783783783784, + "grad_norm": 0.6054284572601318, + "learning_rate": 2.9424e-05, + "loss": 0.0329, + "step": 11732 + }, + { + "epoch": 9.909628378378379, + "grad_norm": 0.36010390520095825, + "learning_rate": 2.9423666666666667e-05, + "loss": 0.0331, + "step": 11733 + }, + { + "epoch": 9.910472972972974, + "grad_norm": 0.15791743993759155, + "learning_rate": 2.9423333333333333e-05, + "loss": 0.015, + "step": 11734 + }, + { + "epoch": 9.911317567567568, + "grad_norm": 0.385572224855423, + "learning_rate": 2.9423e-05, + "loss": 0.0178, + "step": 11735 + }, + { + "epoch": 9.912162162162161, + "grad_norm": 0.4404248297214508, + "learning_rate": 2.942266666666667e-05, + "loss": 0.0369, + "step": 11736 + }, + { + "epoch": 9.913006756756756, + "grad_norm": 0.5306886434555054, + "learning_rate": 2.9422333333333335e-05, + "loss": 0.0383, + "step": 11737 + }, + { + "epoch": 9.91385135135135, + "grad_norm": 0.4040183126926422, + "learning_rate": 2.9422e-05, + "loss": 0.0281, + "step": 11738 + }, + { + "epoch": 9.914695945945946, + "grad_norm": 0.5994480848312378, + "learning_rate": 2.942166666666667e-05, + "loss": 0.0378, + "step": 11739 + }, + { + "epoch": 9.91554054054054, + "grad_norm": 0.6326051950454712, + "learning_rate": 2.9421333333333332e-05, + "loss": 0.0385, + "step": 11740 + }, + { + "epoch": 9.916385135135135, + "grad_norm": 0.19747164845466614, + "learning_rate": 2.9421000000000002e-05, + "loss": 0.0228, + "step": 11741 + }, + { + "epoch": 9.91722972972973, + "grad_norm": 0.2919124960899353, + "learning_rate": 2.9420666666666668e-05, + "loss": 0.0325, + "step": 11742 + }, + { + "epoch": 9.918074324324325, + "grad_norm": 1.2331953048706055, + "learning_rate": 2.9420333333333334e-05, + "loss": 0.0548, + "step": 11743 + }, + { + "epoch": 9.91891891891892, + "grad_norm": 0.2878147065639496, + "learning_rate": 2.9420000000000003e-05, + "loss": 0.0251, + "step": 11744 + }, + { + "epoch": 9.919763513513514, + "grad_norm": 0.3637610375881195, + "learning_rate": 2.9419666666666665e-05, + "loss": 0.026, + "step": 11745 + }, + { + "epoch": 9.920608108108109, + "grad_norm": 0.3464789092540741, + "learning_rate": 2.9419333333333335e-05, + "loss": 0.032, + "step": 11746 + }, + { + "epoch": 9.921452702702704, + "grad_norm": 0.27382710576057434, + "learning_rate": 2.9419e-05, + "loss": 0.0288, + "step": 11747 + }, + { + "epoch": 9.922297297297296, + "grad_norm": 0.31655412912368774, + "learning_rate": 2.9418666666666667e-05, + "loss": 0.0215, + "step": 11748 + }, + { + "epoch": 9.923141891891891, + "grad_norm": 0.4988168478012085, + "learning_rate": 2.9418333333333333e-05, + "loss": 0.0378, + "step": 11749 + }, + { + "epoch": 9.923986486486486, + "grad_norm": 0.36525845527648926, + "learning_rate": 2.9418000000000002e-05, + "loss": 0.0334, + "step": 11750 + }, + { + "epoch": 9.92483108108108, + "grad_norm": 0.38505876064300537, + "learning_rate": 2.9417666666666664e-05, + "loss": 0.0282, + "step": 11751 + }, + { + "epoch": 9.925675675675675, + "grad_norm": 0.543749213218689, + "learning_rate": 2.9417333333333334e-05, + "loss": 0.0292, + "step": 11752 + }, + { + "epoch": 9.92652027027027, + "grad_norm": 0.4579499661922455, + "learning_rate": 2.9417000000000003e-05, + "loss": 0.0292, + "step": 11753 + }, + { + "epoch": 9.927364864864865, + "grad_norm": 0.9478399753570557, + "learning_rate": 2.9416666666666666e-05, + "loss": 0.0541, + "step": 11754 + }, + { + "epoch": 9.92820945945946, + "grad_norm": 0.4599696099758148, + "learning_rate": 2.9416333333333335e-05, + "loss": 0.0308, + "step": 11755 + }, + { + "epoch": 9.929054054054054, + "grad_norm": 0.6282275915145874, + "learning_rate": 2.9416e-05, + "loss": 0.0387, + "step": 11756 + }, + { + "epoch": 9.92989864864865, + "grad_norm": 0.8580427169799805, + "learning_rate": 2.9415666666666667e-05, + "loss": 0.2787, + "step": 11757 + }, + { + "epoch": 9.930743243243244, + "grad_norm": 0.643423318862915, + "learning_rate": 2.9415333333333333e-05, + "loss": 0.1631, + "step": 11758 + }, + { + "epoch": 9.931587837837839, + "grad_norm": 0.5287588238716125, + "learning_rate": 2.9415000000000002e-05, + "loss": 0.0778, + "step": 11759 + }, + { + "epoch": 9.932432432432432, + "grad_norm": 0.4152839481830597, + "learning_rate": 2.9414666666666668e-05, + "loss": 0.0572, + "step": 11760 + }, + { + "epoch": 9.933277027027026, + "grad_norm": 0.42786353826522827, + "learning_rate": 2.9414333333333334e-05, + "loss": 0.05, + "step": 11761 + }, + { + "epoch": 9.934121621621621, + "grad_norm": 0.406929075717926, + "learning_rate": 2.9414000000000003e-05, + "loss": 0.0395, + "step": 11762 + }, + { + "epoch": 9.934966216216216, + "grad_norm": 0.47469377517700195, + "learning_rate": 2.9413666666666666e-05, + "loss": 0.0381, + "step": 11763 + }, + { + "epoch": 9.93581081081081, + "grad_norm": 0.28942033648490906, + "learning_rate": 2.9413333333333335e-05, + "loss": 0.0335, + "step": 11764 + }, + { + "epoch": 9.936655405405405, + "grad_norm": 0.2880221903324127, + "learning_rate": 2.9413e-05, + "loss": 0.0327, + "step": 11765 + }, + { + "epoch": 9.9375, + "grad_norm": 0.24721795320510864, + "learning_rate": 2.9412666666666667e-05, + "loss": 0.025, + "step": 11766 + }, + { + "epoch": 9.938344594594595, + "grad_norm": 0.32463619112968445, + "learning_rate": 2.9412333333333333e-05, + "loss": 0.0322, + "step": 11767 + }, + { + "epoch": 9.93918918918919, + "grad_norm": 0.24652191996574402, + "learning_rate": 2.9412000000000002e-05, + "loss": 0.0219, + "step": 11768 + }, + { + "epoch": 9.940033783783784, + "grad_norm": 0.5580655336380005, + "learning_rate": 2.9411666666666668e-05, + "loss": 0.0342, + "step": 11769 + }, + { + "epoch": 9.940878378378379, + "grad_norm": 0.27187392115592957, + "learning_rate": 2.9411333333333334e-05, + "loss": 0.0229, + "step": 11770 + }, + { + "epoch": 9.941722972972974, + "grad_norm": 0.29285964369773865, + "learning_rate": 2.9411000000000004e-05, + "loss": 0.0363, + "step": 11771 + }, + { + "epoch": 9.942567567567568, + "grad_norm": 0.3352312743663788, + "learning_rate": 2.9410666666666666e-05, + "loss": 0.0451, + "step": 11772 + }, + { + "epoch": 9.943412162162161, + "grad_norm": 0.4905226528644562, + "learning_rate": 2.9410333333333335e-05, + "loss": 0.0225, + "step": 11773 + }, + { + "epoch": 9.944256756756756, + "grad_norm": 0.38668572902679443, + "learning_rate": 2.9409999999999998e-05, + "loss": 0.0263, + "step": 11774 + }, + { + "epoch": 9.94510135135135, + "grad_norm": 0.4170123040676117, + "learning_rate": 2.9409666666666667e-05, + "loss": 0.03, + "step": 11775 + }, + { + "epoch": 9.945945945945946, + "grad_norm": 0.31699636578559875, + "learning_rate": 2.9409333333333333e-05, + "loss": 0.0266, + "step": 11776 + }, + { + "epoch": 9.94679054054054, + "grad_norm": 0.3641918897628784, + "learning_rate": 2.9409e-05, + "loss": 0.0209, + "step": 11777 + }, + { + "epoch": 9.947635135135135, + "grad_norm": 0.3110513687133789, + "learning_rate": 2.940866666666667e-05, + "loss": 0.0346, + "step": 11778 + }, + { + "epoch": 9.94847972972973, + "grad_norm": 0.3155041038990021, + "learning_rate": 2.9408333333333334e-05, + "loss": 0.0288, + "step": 11779 + }, + { + "epoch": 9.949324324324325, + "grad_norm": 0.3467773497104645, + "learning_rate": 2.9408e-05, + "loss": 0.0305, + "step": 11780 + }, + { + "epoch": 9.95016891891892, + "grad_norm": 0.4772398769855499, + "learning_rate": 2.9407666666666666e-05, + "loss": 0.0265, + "step": 11781 + }, + { + "epoch": 9.951013513513514, + "grad_norm": 0.24341896176338196, + "learning_rate": 2.9407333333333336e-05, + "loss": 0.0286, + "step": 11782 + }, + { + "epoch": 9.951858108108109, + "grad_norm": 0.3575499355792999, + "learning_rate": 2.9406999999999998e-05, + "loss": 0.0309, + "step": 11783 + }, + { + "epoch": 9.952702702702704, + "grad_norm": 0.34669721126556396, + "learning_rate": 2.9406666666666667e-05, + "loss": 0.0233, + "step": 11784 + }, + { + "epoch": 9.953547297297296, + "grad_norm": 0.24252764880657196, + "learning_rate": 2.9406333333333333e-05, + "loss": 0.0188, + "step": 11785 + }, + { + "epoch": 9.954391891891891, + "grad_norm": 0.3555050790309906, + "learning_rate": 2.9406e-05, + "loss": 0.0331, + "step": 11786 + }, + { + "epoch": 9.955236486486486, + "grad_norm": 0.4937962591648102, + "learning_rate": 2.940566666666667e-05, + "loss": 0.0247, + "step": 11787 + }, + { + "epoch": 9.95608108108108, + "grad_norm": 0.5707792043685913, + "learning_rate": 2.9405333333333335e-05, + "loss": 0.0355, + "step": 11788 + }, + { + "epoch": 9.956925675675675, + "grad_norm": 0.2542950212955475, + "learning_rate": 2.9405e-05, + "loss": 0.021, + "step": 11789 + }, + { + "epoch": 9.95777027027027, + "grad_norm": 0.3032192289829254, + "learning_rate": 2.9404666666666666e-05, + "loss": 0.03, + "step": 11790 + }, + { + "epoch": 9.958614864864865, + "grad_norm": 0.48721617460250854, + "learning_rate": 2.9404333333333336e-05, + "loss": 0.0354, + "step": 11791 + }, + { + "epoch": 9.95945945945946, + "grad_norm": 0.21637624502182007, + "learning_rate": 2.9404e-05, + "loss": 0.0227, + "step": 11792 + }, + { + "epoch": 9.960304054054054, + "grad_norm": 0.48675912618637085, + "learning_rate": 2.9403666666666668e-05, + "loss": 0.0268, + "step": 11793 + }, + { + "epoch": 9.96114864864865, + "grad_norm": 0.4575657844543457, + "learning_rate": 2.9403333333333337e-05, + "loss": 0.0273, + "step": 11794 + }, + { + "epoch": 9.961993243243244, + "grad_norm": 0.3216952681541443, + "learning_rate": 2.9403e-05, + "loss": 0.0273, + "step": 11795 + }, + { + "epoch": 9.962837837837839, + "grad_norm": 0.34302523732185364, + "learning_rate": 2.940266666666667e-05, + "loss": 0.0322, + "step": 11796 + }, + { + "epoch": 9.963682432432432, + "grad_norm": 0.5384007096290588, + "learning_rate": 2.9402333333333335e-05, + "loss": 0.0448, + "step": 11797 + }, + { + "epoch": 9.964527027027026, + "grad_norm": 0.23746849596500397, + "learning_rate": 2.9402e-05, + "loss": 0.0215, + "step": 11798 + }, + { + "epoch": 9.965371621621621, + "grad_norm": 0.3877687454223633, + "learning_rate": 2.9401666666666667e-05, + "loss": 0.0244, + "step": 11799 + }, + { + "epoch": 9.966216216216216, + "grad_norm": 0.2959543764591217, + "learning_rate": 2.9401333333333336e-05, + "loss": 0.0331, + "step": 11800 + }, + { + "epoch": 9.96706081081081, + "grad_norm": 0.5795984268188477, + "learning_rate": 2.9401e-05, + "loss": 0.0809, + "step": 11801 + }, + { + "epoch": 9.967905405405405, + "grad_norm": 0.43380722403526306, + "learning_rate": 2.9400666666666668e-05, + "loss": 0.0299, + "step": 11802 + }, + { + "epoch": 9.96875, + "grad_norm": 0.2142593115568161, + "learning_rate": 2.9400333333333337e-05, + "loss": 0.0179, + "step": 11803 + }, + { + "epoch": 9.969594594594595, + "grad_norm": 0.4604625701904297, + "learning_rate": 2.94e-05, + "loss": 0.0532, + "step": 11804 + }, + { + "epoch": 9.97043918918919, + "grad_norm": 0.42353951930999756, + "learning_rate": 2.939966666666667e-05, + "loss": 0.032, + "step": 11805 + }, + { + "epoch": 9.971283783783784, + "grad_norm": 1.561930775642395, + "learning_rate": 2.939933333333333e-05, + "loss": 0.0554, + "step": 11806 + }, + { + "epoch": 9.972128378378379, + "grad_norm": 0.6768043637275696, + "learning_rate": 2.9399e-05, + "loss": 0.2102, + "step": 11807 + }, + { + "epoch": 9.972972972972974, + "grad_norm": 0.3771132230758667, + "learning_rate": 2.9398666666666667e-05, + "loss": 0.0983, + "step": 11808 + }, + { + "epoch": 9.973817567567568, + "grad_norm": 0.3459983170032501, + "learning_rate": 2.9398333333333333e-05, + "loss": 0.0502, + "step": 11809 + }, + { + "epoch": 9.974662162162161, + "grad_norm": 0.24158114194869995, + "learning_rate": 2.9398000000000002e-05, + "loss": 0.041, + "step": 11810 + }, + { + "epoch": 9.975506756756756, + "grad_norm": 0.37489810585975647, + "learning_rate": 2.9397666666666668e-05, + "loss": 0.0348, + "step": 11811 + }, + { + "epoch": 9.97635135135135, + "grad_norm": 0.4057409465312958, + "learning_rate": 2.9397333333333334e-05, + "loss": 0.0332, + "step": 11812 + }, + { + "epoch": 9.977195945945946, + "grad_norm": 0.3321476876735687, + "learning_rate": 2.9397e-05, + "loss": 0.0411, + "step": 11813 + }, + { + "epoch": 9.97804054054054, + "grad_norm": 0.31103530526161194, + "learning_rate": 2.939666666666667e-05, + "loss": 0.0396, + "step": 11814 + }, + { + "epoch": 9.978885135135135, + "grad_norm": 0.4445159137248993, + "learning_rate": 2.9396333333333332e-05, + "loss": 0.0688, + "step": 11815 + }, + { + "epoch": 9.97972972972973, + "grad_norm": 0.32054516673088074, + "learning_rate": 2.9396e-05, + "loss": 0.0379, + "step": 11816 + }, + { + "epoch": 9.980574324324325, + "grad_norm": 0.3814144730567932, + "learning_rate": 2.9395666666666667e-05, + "loss": 0.0424, + "step": 11817 + }, + { + "epoch": 9.98141891891892, + "grad_norm": 0.26839184761047363, + "learning_rate": 2.9395333333333333e-05, + "loss": 0.0409, + "step": 11818 + }, + { + "epoch": 9.982263513513514, + "grad_norm": 0.3951376974582672, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0383, + "step": 11819 + }, + { + "epoch": 9.983108108108109, + "grad_norm": 0.45736539363861084, + "learning_rate": 2.9394666666666668e-05, + "loss": 0.021, + "step": 11820 + }, + { + "epoch": 9.983952702702704, + "grad_norm": 0.7607112526893616, + "learning_rate": 2.9394333333333334e-05, + "loss": 0.0242, + "step": 11821 + }, + { + "epoch": 9.984797297297296, + "grad_norm": 0.42081329226493835, + "learning_rate": 2.9394e-05, + "loss": 0.0342, + "step": 11822 + }, + { + "epoch": 9.985641891891891, + "grad_norm": 0.5048511624336243, + "learning_rate": 2.939366666666667e-05, + "loss": 0.0275, + "step": 11823 + }, + { + "epoch": 9.986486486486486, + "grad_norm": 0.6649622917175293, + "learning_rate": 2.9393333333333332e-05, + "loss": 0.0257, + "step": 11824 + }, + { + "epoch": 9.98733108108108, + "grad_norm": 0.3721824288368225, + "learning_rate": 2.9393e-05, + "loss": 0.0393, + "step": 11825 + }, + { + "epoch": 9.988175675675675, + "grad_norm": 0.4308610260486603, + "learning_rate": 2.9392666666666667e-05, + "loss": 0.03, + "step": 11826 + }, + { + "epoch": 9.98902027027027, + "grad_norm": 0.4900721311569214, + "learning_rate": 2.9392333333333333e-05, + "loss": 0.0315, + "step": 11827 + }, + { + "epoch": 9.989864864864865, + "grad_norm": 0.6424446105957031, + "learning_rate": 2.9392000000000003e-05, + "loss": 0.0313, + "step": 11828 + }, + { + "epoch": 9.99070945945946, + "grad_norm": 0.22641873359680176, + "learning_rate": 2.939166666666667e-05, + "loss": 0.0295, + "step": 11829 + }, + { + "epoch": 9.991554054054054, + "grad_norm": 0.4048631489276886, + "learning_rate": 2.9391333333333334e-05, + "loss": 0.0486, + "step": 11830 + }, + { + "epoch": 9.99239864864865, + "grad_norm": 0.24754992127418518, + "learning_rate": 2.9391e-05, + "loss": 0.0207, + "step": 11831 + }, + { + "epoch": 9.993243243243244, + "grad_norm": 0.3140896260738373, + "learning_rate": 2.939066666666667e-05, + "loss": 0.0222, + "step": 11832 + }, + { + "epoch": 9.994087837837839, + "grad_norm": 0.23966814577579498, + "learning_rate": 2.9390333333333332e-05, + "loss": 0.0291, + "step": 11833 + }, + { + "epoch": 9.994932432432432, + "grad_norm": 0.4005562663078308, + "learning_rate": 2.939e-05, + "loss": 0.0421, + "step": 11834 + }, + { + "epoch": 9.995777027027026, + "grad_norm": 0.41102200746536255, + "learning_rate": 2.9389666666666667e-05, + "loss": 0.0294, + "step": 11835 + }, + { + "epoch": 9.996621621621621, + "grad_norm": 0.8831378221511841, + "learning_rate": 2.9389333333333333e-05, + "loss": 0.0611, + "step": 11836 + }, + { + "epoch": 9.997466216216216, + "grad_norm": 0.49550968408584595, + "learning_rate": 2.9389000000000003e-05, + "loss": 0.0514, + "step": 11837 + }, + { + "epoch": 9.99831081081081, + "grad_norm": 0.43015411496162415, + "learning_rate": 2.9388666666666665e-05, + "loss": 0.0288, + "step": 11838 + }, + { + "epoch": 9.999155405405405, + "grad_norm": 0.7826573252677917, + "learning_rate": 2.9388333333333335e-05, + "loss": 0.0488, + "step": 11839 + }, + { + "epoch": 10.0, + "grad_norm": 1.473915457725525, + "learning_rate": 2.9388e-05, + "loss": 0.1036, + "step": 11840 + }, + { + "epoch": 10.000844594594595, + "grad_norm": 0.8362059593200684, + "learning_rate": 2.9387666666666666e-05, + "loss": 0.2738, + "step": 11841 + }, + { + "epoch": 10.00168918918919, + "grad_norm": 0.47969695925712585, + "learning_rate": 2.9387333333333332e-05, + "loss": 0.1414, + "step": 11842 + }, + { + "epoch": 10.002533783783784, + "grad_norm": 0.4127149283885956, + "learning_rate": 2.9387000000000002e-05, + "loss": 0.1053, + "step": 11843 + }, + { + "epoch": 10.003378378378379, + "grad_norm": 0.4352947175502777, + "learning_rate": 2.9386666666666668e-05, + "loss": 0.0501, + "step": 11844 + }, + { + "epoch": 10.004222972972974, + "grad_norm": 0.3463134467601776, + "learning_rate": 2.9386333333333334e-05, + "loss": 0.0277, + "step": 11845 + }, + { + "epoch": 10.005067567567568, + "grad_norm": 0.6470978260040283, + "learning_rate": 2.9386000000000003e-05, + "loss": 0.0403, + "step": 11846 + }, + { + "epoch": 10.005912162162161, + "grad_norm": 0.508667528629303, + "learning_rate": 2.9385666666666665e-05, + "loss": 0.0369, + "step": 11847 + }, + { + "epoch": 10.006756756756756, + "grad_norm": 0.22615651786327362, + "learning_rate": 2.9385333333333335e-05, + "loss": 0.0188, + "step": 11848 + }, + { + "epoch": 10.00760135135135, + "grad_norm": 0.665334165096283, + "learning_rate": 2.9385e-05, + "loss": 0.0447, + "step": 11849 + }, + { + "epoch": 10.008445945945946, + "grad_norm": 0.5281597971916199, + "learning_rate": 2.9384666666666667e-05, + "loss": 0.0292, + "step": 11850 + }, + { + "epoch": 10.00929054054054, + "grad_norm": 0.41893619298934937, + "learning_rate": 2.9384333333333333e-05, + "loss": 0.0354, + "step": 11851 + }, + { + "epoch": 10.010135135135135, + "grad_norm": 0.25346776843070984, + "learning_rate": 2.9384000000000002e-05, + "loss": 0.0379, + "step": 11852 + }, + { + "epoch": 10.01097972972973, + "grad_norm": 0.45310837030410767, + "learning_rate": 2.9383666666666668e-05, + "loss": 0.0336, + "step": 11853 + }, + { + "epoch": 10.011824324324325, + "grad_norm": 0.3042244017124176, + "learning_rate": 2.9383333333333334e-05, + "loss": 0.024, + "step": 11854 + }, + { + "epoch": 10.01266891891892, + "grad_norm": 0.4451376497745514, + "learning_rate": 2.9383000000000003e-05, + "loss": 0.0206, + "step": 11855 + }, + { + "epoch": 10.013513513513514, + "grad_norm": 0.31650903820991516, + "learning_rate": 2.9382666666666666e-05, + "loss": 0.0234, + "step": 11856 + }, + { + "epoch": 10.014358108108109, + "grad_norm": 0.40015092492103577, + "learning_rate": 2.9382333333333335e-05, + "loss": 0.0236, + "step": 11857 + }, + { + "epoch": 10.015202702702704, + "grad_norm": 0.5003324747085571, + "learning_rate": 2.9382e-05, + "loss": 0.0294, + "step": 11858 + }, + { + "epoch": 10.016047297297296, + "grad_norm": 0.3041222393512726, + "learning_rate": 2.9381666666666667e-05, + "loss": 0.0215, + "step": 11859 + }, + { + "epoch": 10.016891891891891, + "grad_norm": 0.22514912486076355, + "learning_rate": 2.9381333333333336e-05, + "loss": 0.0197, + "step": 11860 + }, + { + "epoch": 10.017736486486486, + "grad_norm": 0.22279801964759827, + "learning_rate": 2.9381000000000002e-05, + "loss": 0.0185, + "step": 11861 + }, + { + "epoch": 10.01858108108108, + "grad_norm": 0.36346760392189026, + "learning_rate": 2.9380666666666668e-05, + "loss": 0.0355, + "step": 11862 + }, + { + "epoch": 10.019425675675675, + "grad_norm": 0.41963544487953186, + "learning_rate": 2.9380333333333334e-05, + "loss": 0.0243, + "step": 11863 + }, + { + "epoch": 10.02027027027027, + "grad_norm": 0.47192323207855225, + "learning_rate": 2.938e-05, + "loss": 0.0206, + "step": 11864 + }, + { + "epoch": 10.021114864864865, + "grad_norm": 0.3494402766227722, + "learning_rate": 2.9379666666666666e-05, + "loss": 0.0214, + "step": 11865 + }, + { + "epoch": 10.02195945945946, + "grad_norm": 0.3588979244232178, + "learning_rate": 2.9379333333333335e-05, + "loss": 0.0293, + "step": 11866 + }, + { + "epoch": 10.022804054054054, + "grad_norm": 0.2656156122684479, + "learning_rate": 2.9378999999999998e-05, + "loss": 0.0224, + "step": 11867 + }, + { + "epoch": 10.02364864864865, + "grad_norm": 0.4317682385444641, + "learning_rate": 2.9378666666666667e-05, + "loss": 0.0162, + "step": 11868 + }, + { + "epoch": 10.024493243243244, + "grad_norm": 0.36744314432144165, + "learning_rate": 2.9378333333333336e-05, + "loss": 0.0249, + "step": 11869 + }, + { + "epoch": 10.025337837837839, + "grad_norm": 0.36004501581192017, + "learning_rate": 2.9378e-05, + "loss": 0.0284, + "step": 11870 + }, + { + "epoch": 10.026182432432432, + "grad_norm": 0.20185042917728424, + "learning_rate": 2.9377666666666668e-05, + "loss": 0.0137, + "step": 11871 + }, + { + "epoch": 10.027027027027026, + "grad_norm": 0.40887686610221863, + "learning_rate": 2.9377333333333334e-05, + "loss": 0.0575, + "step": 11872 + }, + { + "epoch": 10.027871621621621, + "grad_norm": 0.3084328770637512, + "learning_rate": 2.9377e-05, + "loss": 0.0144, + "step": 11873 + }, + { + "epoch": 10.028716216216216, + "grad_norm": 0.4962264597415924, + "learning_rate": 2.9376666666666666e-05, + "loss": 0.0294, + "step": 11874 + }, + { + "epoch": 10.02956081081081, + "grad_norm": 0.9315268397331238, + "learning_rate": 2.9376333333333335e-05, + "loss": 0.0316, + "step": 11875 + }, + { + "epoch": 10.030405405405405, + "grad_norm": 0.22535189986228943, + "learning_rate": 2.9375999999999998e-05, + "loss": 0.0156, + "step": 11876 + }, + { + "epoch": 10.03125, + "grad_norm": 0.5492666363716125, + "learning_rate": 2.9375666666666667e-05, + "loss": 0.0345, + "step": 11877 + }, + { + "epoch": 10.032094594594595, + "grad_norm": 0.42344972491264343, + "learning_rate": 2.9375333333333337e-05, + "loss": 0.0499, + "step": 11878 + }, + { + "epoch": 10.03293918918919, + "grad_norm": 0.27067750692367554, + "learning_rate": 2.9375e-05, + "loss": 0.0159, + "step": 11879 + }, + { + "epoch": 10.033783783783784, + "grad_norm": 0.2559014558792114, + "learning_rate": 2.937466666666667e-05, + "loss": 0.0308, + "step": 11880 + }, + { + "epoch": 10.034628378378379, + "grad_norm": 0.5326773524284363, + "learning_rate": 2.9374333333333334e-05, + "loss": 0.0351, + "step": 11881 + }, + { + "epoch": 10.035472972972974, + "grad_norm": 0.3792746961116791, + "learning_rate": 2.9374e-05, + "loss": 0.0351, + "step": 11882 + }, + { + "epoch": 10.036317567567568, + "grad_norm": 0.6365184783935547, + "learning_rate": 2.9373666666666666e-05, + "loss": 0.0301, + "step": 11883 + }, + { + "epoch": 10.037162162162161, + "grad_norm": 0.38584455847740173, + "learning_rate": 2.9373333333333336e-05, + "loss": 0.0249, + "step": 11884 + }, + { + "epoch": 10.038006756756756, + "grad_norm": 0.2867867946624756, + "learning_rate": 2.9373e-05, + "loss": 0.0214, + "step": 11885 + }, + { + "epoch": 10.03885135135135, + "grad_norm": 0.5126237869262695, + "learning_rate": 2.9372666666666667e-05, + "loss": 0.0279, + "step": 11886 + }, + { + "epoch": 10.039695945945946, + "grad_norm": 0.5958481431007385, + "learning_rate": 2.9372333333333337e-05, + "loss": 0.0247, + "step": 11887 + }, + { + "epoch": 10.04054054054054, + "grad_norm": 0.7697275280952454, + "learning_rate": 2.9372e-05, + "loss": 0.0363, + "step": 11888 + }, + { + "epoch": 10.041385135135135, + "grad_norm": 0.36424520611763, + "learning_rate": 2.937166666666667e-05, + "loss": 0.0302, + "step": 11889 + }, + { + "epoch": 10.04222972972973, + "grad_norm": 1.0245145559310913, + "learning_rate": 2.9371333333333335e-05, + "loss": 0.0297, + "step": 11890 + }, + { + "epoch": 10.043074324324325, + "grad_norm": 0.7835915088653564, + "learning_rate": 2.9371e-05, + "loss": 0.261, + "step": 11891 + }, + { + "epoch": 10.04391891891892, + "grad_norm": 0.5979470014572144, + "learning_rate": 2.9370666666666666e-05, + "loss": 0.1195, + "step": 11892 + }, + { + "epoch": 10.044763513513514, + "grad_norm": 0.30003246665000916, + "learning_rate": 2.9370333333333336e-05, + "loss": 0.0336, + "step": 11893 + }, + { + "epoch": 10.045608108108109, + "grad_norm": 0.2045544981956482, + "learning_rate": 2.9370000000000002e-05, + "loss": 0.033, + "step": 11894 + }, + { + "epoch": 10.046452702702704, + "grad_norm": 0.2859458029270172, + "learning_rate": 2.9369666666666668e-05, + "loss": 0.0301, + "step": 11895 + }, + { + "epoch": 10.047297297297296, + "grad_norm": 0.27175113558769226, + "learning_rate": 2.9369333333333334e-05, + "loss": 0.0364, + "step": 11896 + }, + { + "epoch": 10.048141891891891, + "grad_norm": 0.370034396648407, + "learning_rate": 2.9369e-05, + "loss": 0.0319, + "step": 11897 + }, + { + "epoch": 10.048986486486486, + "grad_norm": 0.3927510678768158, + "learning_rate": 2.936866666666667e-05, + "loss": 0.0363, + "step": 11898 + }, + { + "epoch": 10.04983108108108, + "grad_norm": 0.5420047640800476, + "learning_rate": 2.936833333333333e-05, + "loss": 0.0414, + "step": 11899 + }, + { + "epoch": 10.050675675675675, + "grad_norm": 0.3329138159751892, + "learning_rate": 2.9368e-05, + "loss": 0.0277, + "step": 11900 + }, + { + "epoch": 10.05152027027027, + "grad_norm": 0.23481543362140656, + "learning_rate": 2.9367666666666667e-05, + "loss": 0.0213, + "step": 11901 + }, + { + "epoch": 10.052364864864865, + "grad_norm": 0.6257069110870361, + "learning_rate": 2.9367333333333333e-05, + "loss": 0.0332, + "step": 11902 + }, + { + "epoch": 10.05320945945946, + "grad_norm": 0.3377346098423004, + "learning_rate": 2.9367000000000002e-05, + "loss": 0.03, + "step": 11903 + }, + { + "epoch": 10.054054054054054, + "grad_norm": 0.305429071187973, + "learning_rate": 2.9366666666666668e-05, + "loss": 0.0423, + "step": 11904 + }, + { + "epoch": 10.05489864864865, + "grad_norm": 0.24986466765403748, + "learning_rate": 2.9366333333333334e-05, + "loss": 0.0244, + "step": 11905 + }, + { + "epoch": 10.055743243243244, + "grad_norm": 0.39359378814697266, + "learning_rate": 2.9366e-05, + "loss": 0.0216, + "step": 11906 + }, + { + "epoch": 10.056587837837839, + "grad_norm": 0.20508712530136108, + "learning_rate": 2.936566666666667e-05, + "loss": 0.0201, + "step": 11907 + }, + { + "epoch": 10.057432432432432, + "grad_norm": 0.432590514421463, + "learning_rate": 2.936533333333333e-05, + "loss": 0.0254, + "step": 11908 + }, + { + "epoch": 10.058277027027026, + "grad_norm": 0.36931994557380676, + "learning_rate": 2.9365e-05, + "loss": 0.0172, + "step": 11909 + }, + { + "epoch": 10.059121621621621, + "grad_norm": 0.35390231013298035, + "learning_rate": 2.936466666666667e-05, + "loss": 0.021, + "step": 11910 + }, + { + "epoch": 10.059966216216216, + "grad_norm": 0.4521443545818329, + "learning_rate": 2.9364333333333333e-05, + "loss": 0.0275, + "step": 11911 + }, + { + "epoch": 10.06081081081081, + "grad_norm": 0.48420682549476624, + "learning_rate": 2.9364000000000002e-05, + "loss": 0.0257, + "step": 11912 + }, + { + "epoch": 10.061655405405405, + "grad_norm": 0.39983728528022766, + "learning_rate": 2.9363666666666668e-05, + "loss": 0.0269, + "step": 11913 + }, + { + "epoch": 10.0625, + "grad_norm": 0.3435817360877991, + "learning_rate": 2.9363333333333334e-05, + "loss": 0.0272, + "step": 11914 + }, + { + "epoch": 10.063344594594595, + "grad_norm": 0.2701273560523987, + "learning_rate": 2.9363e-05, + "loss": 0.0169, + "step": 11915 + }, + { + "epoch": 10.06418918918919, + "grad_norm": 0.8513690233230591, + "learning_rate": 2.936266666666667e-05, + "loss": 0.031, + "step": 11916 + }, + { + "epoch": 10.065033783783784, + "grad_norm": 0.3091111183166504, + "learning_rate": 2.9362333333333332e-05, + "loss": 0.0183, + "step": 11917 + }, + { + "epoch": 10.065878378378379, + "grad_norm": 0.296501100063324, + "learning_rate": 2.9362e-05, + "loss": 0.0174, + "step": 11918 + }, + { + "epoch": 10.066722972972974, + "grad_norm": 0.29378411173820496, + "learning_rate": 2.936166666666667e-05, + "loss": 0.0136, + "step": 11919 + }, + { + "epoch": 10.067567567567568, + "grad_norm": 0.4561079144477844, + "learning_rate": 2.9361333333333333e-05, + "loss": 0.0317, + "step": 11920 + }, + { + "epoch": 10.068412162162161, + "grad_norm": 0.28683096170425415, + "learning_rate": 2.9361000000000002e-05, + "loss": 0.0166, + "step": 11921 + }, + { + "epoch": 10.069256756756756, + "grad_norm": 0.33273541927337646, + "learning_rate": 2.9360666666666668e-05, + "loss": 0.0258, + "step": 11922 + }, + { + "epoch": 10.07010135135135, + "grad_norm": 0.3021653890609741, + "learning_rate": 2.9360333333333334e-05, + "loss": 0.0207, + "step": 11923 + }, + { + "epoch": 10.070945945945946, + "grad_norm": 0.4175640642642975, + "learning_rate": 2.936e-05, + "loss": 0.0235, + "step": 11924 + }, + { + "epoch": 10.07179054054054, + "grad_norm": 0.39327478408813477, + "learning_rate": 2.9359666666666666e-05, + "loss": 0.0332, + "step": 11925 + }, + { + "epoch": 10.072635135135135, + "grad_norm": 0.39367812871932983, + "learning_rate": 2.9359333333333332e-05, + "loss": 0.0318, + "step": 11926 + }, + { + "epoch": 10.07347972972973, + "grad_norm": 0.40831416845321655, + "learning_rate": 2.9359e-05, + "loss": 0.027, + "step": 11927 + }, + { + "epoch": 10.074324324324325, + "grad_norm": 0.46488437056541443, + "learning_rate": 2.9358666666666667e-05, + "loss": 0.028, + "step": 11928 + }, + { + "epoch": 10.07516891891892, + "grad_norm": 0.5061922073364258, + "learning_rate": 2.9358333333333333e-05, + "loss": 0.0329, + "step": 11929 + }, + { + "epoch": 10.076013513513514, + "grad_norm": 0.3716853857040405, + "learning_rate": 2.9358000000000003e-05, + "loss": 0.0302, + "step": 11930 + }, + { + "epoch": 10.076858108108109, + "grad_norm": 0.4729728400707245, + "learning_rate": 2.9357666666666665e-05, + "loss": 0.0341, + "step": 11931 + }, + { + "epoch": 10.077702702702704, + "grad_norm": 0.43465757369995117, + "learning_rate": 2.9357333333333334e-05, + "loss": 0.0273, + "step": 11932 + }, + { + "epoch": 10.078547297297296, + "grad_norm": 0.781123161315918, + "learning_rate": 2.9357e-05, + "loss": 0.027, + "step": 11933 + }, + { + "epoch": 10.079391891891891, + "grad_norm": 0.3783429265022278, + "learning_rate": 2.9356666666666666e-05, + "loss": 0.0174, + "step": 11934 + }, + { + "epoch": 10.080236486486486, + "grad_norm": 0.27428844571113586, + "learning_rate": 2.9356333333333336e-05, + "loss": 0.024, + "step": 11935 + }, + { + "epoch": 10.08108108108108, + "grad_norm": 0.6146356463432312, + "learning_rate": 2.9356e-05, + "loss": 0.0206, + "step": 11936 + }, + { + "epoch": 10.081925675675675, + "grad_norm": 0.758490800857544, + "learning_rate": 2.9355666666666667e-05, + "loss": 0.0308, + "step": 11937 + }, + { + "epoch": 10.08277027027027, + "grad_norm": 0.3809660077095032, + "learning_rate": 2.9355333333333333e-05, + "loss": 0.0305, + "step": 11938 + }, + { + "epoch": 10.083614864864865, + "grad_norm": 0.8332499861717224, + "learning_rate": 2.9355000000000003e-05, + "loss": 0.0608, + "step": 11939 + }, + { + "epoch": 10.08445945945946, + "grad_norm": 0.7054195404052734, + "learning_rate": 2.9354666666666665e-05, + "loss": 0.0528, + "step": 11940 + }, + { + "epoch": 10.085304054054054, + "grad_norm": 0.9247394800186157, + "learning_rate": 2.9354333333333335e-05, + "loss": 0.2688, + "step": 11941 + }, + { + "epoch": 10.08614864864865, + "grad_norm": 0.5790601372718811, + "learning_rate": 2.9354e-05, + "loss": 0.1497, + "step": 11942 + }, + { + "epoch": 10.086993243243244, + "grad_norm": 0.35896751284599304, + "learning_rate": 2.9353666666666666e-05, + "loss": 0.0612, + "step": 11943 + }, + { + "epoch": 10.087837837837839, + "grad_norm": 0.3853261172771454, + "learning_rate": 2.9353333333333336e-05, + "loss": 0.0446, + "step": 11944 + }, + { + "epoch": 10.088682432432432, + "grad_norm": 0.31121084094047546, + "learning_rate": 2.9353000000000002e-05, + "loss": 0.043, + "step": 11945 + }, + { + "epoch": 10.089527027027026, + "grad_norm": 0.3186139464378357, + "learning_rate": 2.9352666666666668e-05, + "loss": 0.0414, + "step": 11946 + }, + { + "epoch": 10.090371621621621, + "grad_norm": 0.8235874176025391, + "learning_rate": 2.9352333333333334e-05, + "loss": 0.0534, + "step": 11947 + }, + { + "epoch": 10.091216216216216, + "grad_norm": 0.5583106875419617, + "learning_rate": 2.9352000000000003e-05, + "loss": 0.0225, + "step": 11948 + }, + { + "epoch": 10.09206081081081, + "grad_norm": 0.2655808925628662, + "learning_rate": 2.9351666666666665e-05, + "loss": 0.0253, + "step": 11949 + }, + { + "epoch": 10.092905405405405, + "grad_norm": 0.32925429940223694, + "learning_rate": 2.9351333333333335e-05, + "loss": 0.027, + "step": 11950 + }, + { + "epoch": 10.09375, + "grad_norm": 0.2109832465648651, + "learning_rate": 2.9351e-05, + "loss": 0.0159, + "step": 11951 + }, + { + "epoch": 10.094594594594595, + "grad_norm": 0.34093132615089417, + "learning_rate": 2.9350666666666667e-05, + "loss": 0.0247, + "step": 11952 + }, + { + "epoch": 10.09543918918919, + "grad_norm": 0.32372382283210754, + "learning_rate": 2.9350333333333336e-05, + "loss": 0.0246, + "step": 11953 + }, + { + "epoch": 10.096283783783784, + "grad_norm": 0.2596915364265442, + "learning_rate": 2.9350000000000002e-05, + "loss": 0.0192, + "step": 11954 + }, + { + "epoch": 10.097128378378379, + "grad_norm": 0.31535986065864563, + "learning_rate": 2.9349666666666668e-05, + "loss": 0.028, + "step": 11955 + }, + { + "epoch": 10.097972972972974, + "grad_norm": 0.2912137806415558, + "learning_rate": 2.9349333333333334e-05, + "loss": 0.032, + "step": 11956 + }, + { + "epoch": 10.098817567567568, + "grad_norm": 0.4096211791038513, + "learning_rate": 2.9349e-05, + "loss": 0.052, + "step": 11957 + }, + { + "epoch": 10.099662162162161, + "grad_norm": 0.4209945499897003, + "learning_rate": 2.9348666666666666e-05, + "loss": 0.0184, + "step": 11958 + }, + { + "epoch": 10.100506756756756, + "grad_norm": 0.37796512246131897, + "learning_rate": 2.9348333333333335e-05, + "loss": 0.0389, + "step": 11959 + }, + { + "epoch": 10.10135135135135, + "grad_norm": 0.5019893646240234, + "learning_rate": 2.9348e-05, + "loss": 0.0311, + "step": 11960 + }, + { + "epoch": 10.102195945945946, + "grad_norm": 0.4594051241874695, + "learning_rate": 2.9347666666666667e-05, + "loss": 0.0202, + "step": 11961 + }, + { + "epoch": 10.10304054054054, + "grad_norm": 0.3238101899623871, + "learning_rate": 2.9347333333333336e-05, + "loss": 0.0227, + "step": 11962 + }, + { + "epoch": 10.103885135135135, + "grad_norm": 0.3750319480895996, + "learning_rate": 2.9347e-05, + "loss": 0.035, + "step": 11963 + }, + { + "epoch": 10.10472972972973, + "grad_norm": 0.2728874981403351, + "learning_rate": 2.9346666666666668e-05, + "loss": 0.0178, + "step": 11964 + }, + { + "epoch": 10.105574324324325, + "grad_norm": 0.25437963008880615, + "learning_rate": 2.9346333333333334e-05, + "loss": 0.0188, + "step": 11965 + }, + { + "epoch": 10.10641891891892, + "grad_norm": 0.5920263528823853, + "learning_rate": 2.9346e-05, + "loss": 0.0215, + "step": 11966 + }, + { + "epoch": 10.107263513513514, + "grad_norm": 0.2648269236087799, + "learning_rate": 2.9345666666666666e-05, + "loss": 0.0247, + "step": 11967 + }, + { + "epoch": 10.108108108108109, + "grad_norm": 0.35127517580986023, + "learning_rate": 2.9345333333333335e-05, + "loss": 0.0211, + "step": 11968 + }, + { + "epoch": 10.108952702702704, + "grad_norm": 0.42201152443885803, + "learning_rate": 2.9345e-05, + "loss": 0.0175, + "step": 11969 + }, + { + "epoch": 10.109797297297296, + "grad_norm": 0.35019341111183167, + "learning_rate": 2.9344666666666667e-05, + "loss": 0.0356, + "step": 11970 + }, + { + "epoch": 10.110641891891891, + "grad_norm": 0.40118661522865295, + "learning_rate": 2.9344333333333336e-05, + "loss": 0.0193, + "step": 11971 + }, + { + "epoch": 10.111486486486486, + "grad_norm": 0.27749982476234436, + "learning_rate": 2.9344e-05, + "loss": 0.0247, + "step": 11972 + }, + { + "epoch": 10.11233108108108, + "grad_norm": 0.3485490381717682, + "learning_rate": 2.9343666666666668e-05, + "loss": 0.0304, + "step": 11973 + }, + { + "epoch": 10.113175675675675, + "grad_norm": 0.6329979300498962, + "learning_rate": 2.9343333333333334e-05, + "loss": 0.0357, + "step": 11974 + }, + { + "epoch": 10.11402027027027, + "grad_norm": 0.405967652797699, + "learning_rate": 2.9343e-05, + "loss": 0.0194, + "step": 11975 + }, + { + "epoch": 10.114864864864865, + "grad_norm": 0.5029829144477844, + "learning_rate": 2.9342666666666666e-05, + "loss": 0.0162, + "step": 11976 + }, + { + "epoch": 10.11570945945946, + "grad_norm": 0.31915774941444397, + "learning_rate": 2.9342333333333335e-05, + "loss": 0.0224, + "step": 11977 + }, + { + "epoch": 10.116554054054054, + "grad_norm": 0.31109002232551575, + "learning_rate": 2.9342e-05, + "loss": 0.0295, + "step": 11978 + }, + { + "epoch": 10.11739864864865, + "grad_norm": 0.4180866777896881, + "learning_rate": 2.9341666666666667e-05, + "loss": 0.0241, + "step": 11979 + }, + { + "epoch": 10.118243243243244, + "grad_norm": 0.33766287565231323, + "learning_rate": 2.9341333333333337e-05, + "loss": 0.0224, + "step": 11980 + }, + { + "epoch": 10.119087837837839, + "grad_norm": 0.30255642533302307, + "learning_rate": 2.9341e-05, + "loss": 0.0264, + "step": 11981 + }, + { + "epoch": 10.119932432432432, + "grad_norm": 0.3350462317466736, + "learning_rate": 2.934066666666667e-05, + "loss": 0.0192, + "step": 11982 + }, + { + "epoch": 10.120777027027026, + "grad_norm": 0.2709514796733856, + "learning_rate": 2.9340333333333334e-05, + "loss": 0.0184, + "step": 11983 + }, + { + "epoch": 10.121621621621621, + "grad_norm": 0.3526303172111511, + "learning_rate": 2.934e-05, + "loss": 0.0268, + "step": 11984 + }, + { + "epoch": 10.122466216216216, + "grad_norm": 0.4748207628726959, + "learning_rate": 2.933966666666667e-05, + "loss": 0.024, + "step": 11985 + }, + { + "epoch": 10.12331081081081, + "grad_norm": 0.31808874011039734, + "learning_rate": 2.9339333333333332e-05, + "loss": 0.0371, + "step": 11986 + }, + { + "epoch": 10.124155405405405, + "grad_norm": 0.7657219171524048, + "learning_rate": 2.9339e-05, + "loss": 0.033, + "step": 11987 + }, + { + "epoch": 10.125, + "grad_norm": 0.4327652156352997, + "learning_rate": 2.9338666666666667e-05, + "loss": 0.025, + "step": 11988 + }, + { + "epoch": 10.125844594594595, + "grad_norm": 0.41516247391700745, + "learning_rate": 2.9338333333333333e-05, + "loss": 0.034, + "step": 11989 + }, + { + "epoch": 10.12668918918919, + "grad_norm": 0.5950279831886292, + "learning_rate": 2.9338e-05, + "loss": 0.0352, + "step": 11990 + }, + { + "epoch": 10.127533783783784, + "grad_norm": 1.7853492498397827, + "learning_rate": 2.933766666666667e-05, + "loss": 0.3079, + "step": 11991 + }, + { + "epoch": 10.128378378378379, + "grad_norm": 1.0373075008392334, + "learning_rate": 2.933733333333333e-05, + "loss": 0.144, + "step": 11992 + }, + { + "epoch": 10.129222972972974, + "grad_norm": 0.5094127058982849, + "learning_rate": 2.9337e-05, + "loss": 0.0937, + "step": 11993 + }, + { + "epoch": 10.130067567567568, + "grad_norm": 0.3961866497993469, + "learning_rate": 2.933666666666667e-05, + "loss": 0.0467, + "step": 11994 + }, + { + "epoch": 10.130912162162161, + "grad_norm": 0.3201410174369812, + "learning_rate": 2.9336333333333332e-05, + "loss": 0.0443, + "step": 11995 + }, + { + "epoch": 10.131756756756756, + "grad_norm": 0.34755340218544006, + "learning_rate": 2.9336000000000002e-05, + "loss": 0.0491, + "step": 11996 + }, + { + "epoch": 10.13260135135135, + "grad_norm": 0.46177056431770325, + "learning_rate": 2.9335666666666668e-05, + "loss": 0.0295, + "step": 11997 + }, + { + "epoch": 10.133445945945946, + "grad_norm": 0.3537154197692871, + "learning_rate": 2.9335333333333334e-05, + "loss": 0.0367, + "step": 11998 + }, + { + "epoch": 10.13429054054054, + "grad_norm": 0.24014350771903992, + "learning_rate": 2.9335e-05, + "loss": 0.0293, + "step": 11999 + }, + { + "epoch": 10.135135135135135, + "grad_norm": 0.25935879349708557, + "learning_rate": 2.933466666666667e-05, + "loss": 0.026, + "step": 12000 + }, + { + "epoch": 10.135135135135135, + "eval_cer": 0.10489117339972694, + "eval_loss": 0.27887648344039917, + "eval_runtime": 8.5442, + "eval_samples_per_second": 11.821, + "eval_steps_per_second": 0.468, + "eval_wer": 0.2958553791887125, + "step": 12000 + }, + { + "epoch": 10.13597972972973, + "grad_norm": 0.2489563226699829, + "learning_rate": 2.933433333333333e-05, + "loss": 0.0214, + "step": 12001 + }, + { + "epoch": 10.136824324324325, + "grad_norm": 0.17294032871723175, + "learning_rate": 2.9334e-05, + "loss": 0.0126, + "step": 12002 + }, + { + "epoch": 10.13766891891892, + "grad_norm": 0.36961230635643005, + "learning_rate": 2.933366666666667e-05, + "loss": 0.0517, + "step": 12003 + }, + { + "epoch": 10.138513513513514, + "grad_norm": 0.37647899985313416, + "learning_rate": 2.9333333333333333e-05, + "loss": 0.0284, + "step": 12004 + }, + { + "epoch": 10.139358108108109, + "grad_norm": 0.22465664148330688, + "learning_rate": 2.9333000000000002e-05, + "loss": 0.0248, + "step": 12005 + }, + { + "epoch": 10.140202702702704, + "grad_norm": 0.2222723513841629, + "learning_rate": 2.9332666666666668e-05, + "loss": 0.0269, + "step": 12006 + }, + { + "epoch": 10.141047297297296, + "grad_norm": 0.2755671441555023, + "learning_rate": 2.9332333333333334e-05, + "loss": 0.0163, + "step": 12007 + }, + { + "epoch": 10.141891891891891, + "grad_norm": 0.30090031027793884, + "learning_rate": 2.9332e-05, + "loss": 0.02, + "step": 12008 + }, + { + "epoch": 10.142736486486486, + "grad_norm": 0.25441697239875793, + "learning_rate": 2.933166666666667e-05, + "loss": 0.0194, + "step": 12009 + }, + { + "epoch": 10.14358108108108, + "grad_norm": 0.2017701119184494, + "learning_rate": 2.9331333333333335e-05, + "loss": 0.0199, + "step": 12010 + }, + { + "epoch": 10.144425675675675, + "grad_norm": 0.2898290753364563, + "learning_rate": 2.9331e-05, + "loss": 0.0401, + "step": 12011 + }, + { + "epoch": 10.14527027027027, + "grad_norm": 0.30026596784591675, + "learning_rate": 2.933066666666667e-05, + "loss": 0.0309, + "step": 12012 + }, + { + "epoch": 10.146114864864865, + "grad_norm": 0.6261517405509949, + "learning_rate": 2.9330333333333333e-05, + "loss": 0.0244, + "step": 12013 + }, + { + "epoch": 10.14695945945946, + "grad_norm": 0.33825474977493286, + "learning_rate": 2.9330000000000002e-05, + "loss": 0.0326, + "step": 12014 + }, + { + "epoch": 10.147804054054054, + "grad_norm": 0.20082873106002808, + "learning_rate": 2.9329666666666668e-05, + "loss": 0.0128, + "step": 12015 + }, + { + "epoch": 10.14864864864865, + "grad_norm": 0.36086323857307434, + "learning_rate": 2.9329333333333334e-05, + "loss": 0.034, + "step": 12016 + }, + { + "epoch": 10.149493243243244, + "grad_norm": 0.2500097155570984, + "learning_rate": 2.9329e-05, + "loss": 0.0176, + "step": 12017 + }, + { + "epoch": 10.150337837837839, + "grad_norm": 0.4529547691345215, + "learning_rate": 2.9328666666666666e-05, + "loss": 0.0485, + "step": 12018 + }, + { + "epoch": 10.151182432432432, + "grad_norm": 0.5602414608001709, + "learning_rate": 2.9328333333333335e-05, + "loss": 0.0182, + "step": 12019 + }, + { + "epoch": 10.152027027027026, + "grad_norm": 0.7491174340248108, + "learning_rate": 2.9328e-05, + "loss": 0.036, + "step": 12020 + }, + { + "epoch": 10.152871621621621, + "grad_norm": 0.2968389391899109, + "learning_rate": 2.9327666666666667e-05, + "loss": 0.0247, + "step": 12021 + }, + { + "epoch": 10.153716216216216, + "grad_norm": 0.4221189022064209, + "learning_rate": 2.9327333333333333e-05, + "loss": 0.0204, + "step": 12022 + }, + { + "epoch": 10.15456081081081, + "grad_norm": 0.31234580278396606, + "learning_rate": 2.9327000000000002e-05, + "loss": 0.0255, + "step": 12023 + }, + { + "epoch": 10.155405405405405, + "grad_norm": 0.9289035201072693, + "learning_rate": 2.9326666666666665e-05, + "loss": 0.0493, + "step": 12024 + }, + { + "epoch": 10.15625, + "grad_norm": 0.36695319414138794, + "learning_rate": 2.9326333333333334e-05, + "loss": 0.0284, + "step": 12025 + }, + { + "epoch": 10.157094594594595, + "grad_norm": 0.5482935309410095, + "learning_rate": 2.9326e-05, + "loss": 0.0439, + "step": 12026 + }, + { + "epoch": 10.15793918918919, + "grad_norm": 0.275471031665802, + "learning_rate": 2.9325666666666666e-05, + "loss": 0.0329, + "step": 12027 + }, + { + "epoch": 10.158783783783784, + "grad_norm": 0.20425571501255035, + "learning_rate": 2.9325333333333335e-05, + "loss": 0.02, + "step": 12028 + }, + { + "epoch": 10.159628378378379, + "grad_norm": 0.6625102758407593, + "learning_rate": 2.9325e-05, + "loss": 0.0319, + "step": 12029 + }, + { + "epoch": 10.160472972972974, + "grad_norm": 0.3483542203903198, + "learning_rate": 2.9324666666666667e-05, + "loss": 0.0462, + "step": 12030 + }, + { + "epoch": 10.161317567567568, + "grad_norm": 0.5576221346855164, + "learning_rate": 2.9324333333333333e-05, + "loss": 0.0462, + "step": 12031 + }, + { + "epoch": 10.162162162162161, + "grad_norm": 0.3420105576515198, + "learning_rate": 2.9324000000000002e-05, + "loss": 0.0231, + "step": 12032 + }, + { + "epoch": 10.163006756756756, + "grad_norm": 0.24994820356369019, + "learning_rate": 2.9323666666666665e-05, + "loss": 0.0205, + "step": 12033 + }, + { + "epoch": 10.16385135135135, + "grad_norm": 0.2645058035850525, + "learning_rate": 2.9323333333333334e-05, + "loss": 0.0225, + "step": 12034 + }, + { + "epoch": 10.164695945945946, + "grad_norm": 0.5850998759269714, + "learning_rate": 2.9323000000000004e-05, + "loss": 0.0676, + "step": 12035 + }, + { + "epoch": 10.16554054054054, + "grad_norm": 0.26026976108551025, + "learning_rate": 2.9322666666666666e-05, + "loss": 0.0295, + "step": 12036 + }, + { + "epoch": 10.166385135135135, + "grad_norm": 0.3550553023815155, + "learning_rate": 2.9322333333333336e-05, + "loss": 0.0254, + "step": 12037 + }, + { + "epoch": 10.16722972972973, + "grad_norm": 0.2914535105228424, + "learning_rate": 2.9322e-05, + "loss": 0.0287, + "step": 12038 + }, + { + "epoch": 10.168074324324325, + "grad_norm": 0.8738174438476562, + "learning_rate": 2.9321666666666667e-05, + "loss": 0.0363, + "step": 12039 + }, + { + "epoch": 10.16891891891892, + "grad_norm": 0.964074432849884, + "learning_rate": 2.9321333333333333e-05, + "loss": 0.0469, + "step": 12040 + }, + { + "epoch": 10.169763513513514, + "grad_norm": 0.8601141571998596, + "learning_rate": 2.9321000000000003e-05, + "loss": 0.2458, + "step": 12041 + }, + { + "epoch": 10.170608108108109, + "grad_norm": 0.8480502963066101, + "learning_rate": 2.9320666666666665e-05, + "loss": 0.1383, + "step": 12042 + }, + { + "epoch": 10.171452702702704, + "grad_norm": 0.33249443769454956, + "learning_rate": 2.9320333333333335e-05, + "loss": 0.0742, + "step": 12043 + }, + { + "epoch": 10.172297297297296, + "grad_norm": 0.5167922377586365, + "learning_rate": 2.9320000000000004e-05, + "loss": 0.0624, + "step": 12044 + }, + { + "epoch": 10.173141891891891, + "grad_norm": 0.3912670910358429, + "learning_rate": 2.9319666666666666e-05, + "loss": 0.0452, + "step": 12045 + }, + { + "epoch": 10.173986486486486, + "grad_norm": 0.35363662242889404, + "learning_rate": 2.9319333333333336e-05, + "loss": 0.0378, + "step": 12046 + }, + { + "epoch": 10.17483108108108, + "grad_norm": 0.32132354378700256, + "learning_rate": 2.9318999999999998e-05, + "loss": 0.0246, + "step": 12047 + }, + { + "epoch": 10.175675675675675, + "grad_norm": 0.44734492897987366, + "learning_rate": 2.9318666666666668e-05, + "loss": 0.0396, + "step": 12048 + }, + { + "epoch": 10.17652027027027, + "grad_norm": 0.4068070948123932, + "learning_rate": 2.9318333333333334e-05, + "loss": 0.0333, + "step": 12049 + }, + { + "epoch": 10.177364864864865, + "grad_norm": 0.2838825285434723, + "learning_rate": 2.9318e-05, + "loss": 0.0348, + "step": 12050 + }, + { + "epoch": 10.17820945945946, + "grad_norm": 0.4107498824596405, + "learning_rate": 2.9317666666666665e-05, + "loss": 0.0349, + "step": 12051 + }, + { + "epoch": 10.179054054054054, + "grad_norm": 0.29581668972969055, + "learning_rate": 2.9317333333333335e-05, + "loss": 0.0216, + "step": 12052 + }, + { + "epoch": 10.17989864864865, + "grad_norm": 0.32919400930404663, + "learning_rate": 2.9317e-05, + "loss": 0.0284, + "step": 12053 + }, + { + "epoch": 10.180743243243244, + "grad_norm": 0.30296608805656433, + "learning_rate": 2.9316666666666667e-05, + "loss": 0.0375, + "step": 12054 + }, + { + "epoch": 10.181587837837839, + "grad_norm": 0.5391029119491577, + "learning_rate": 2.9316333333333336e-05, + "loss": 0.0235, + "step": 12055 + }, + { + "epoch": 10.182432432432432, + "grad_norm": 0.29064929485321045, + "learning_rate": 2.9316e-05, + "loss": 0.028, + "step": 12056 + }, + { + "epoch": 10.183277027027026, + "grad_norm": 0.2185813933610916, + "learning_rate": 2.9315666666666668e-05, + "loss": 0.0227, + "step": 12057 + }, + { + "epoch": 10.184121621621621, + "grad_norm": 0.2922849655151367, + "learning_rate": 2.9315333333333334e-05, + "loss": 0.0281, + "step": 12058 + }, + { + "epoch": 10.184966216216216, + "grad_norm": 0.32499051094055176, + "learning_rate": 2.9315e-05, + "loss": 0.0247, + "step": 12059 + }, + { + "epoch": 10.18581081081081, + "grad_norm": 0.16652175784111023, + "learning_rate": 2.931466666666667e-05, + "loss": 0.0138, + "step": 12060 + }, + { + "epoch": 10.186655405405405, + "grad_norm": 0.2396145612001419, + "learning_rate": 2.9314333333333335e-05, + "loss": 0.0218, + "step": 12061 + }, + { + "epoch": 10.1875, + "grad_norm": 0.3327319025993347, + "learning_rate": 2.9314e-05, + "loss": 0.0298, + "step": 12062 + }, + { + "epoch": 10.188344594594595, + "grad_norm": 0.2736503481864929, + "learning_rate": 2.9313666666666667e-05, + "loss": 0.0273, + "step": 12063 + }, + { + "epoch": 10.18918918918919, + "grad_norm": 0.1634538322687149, + "learning_rate": 2.9313333333333336e-05, + "loss": 0.0129, + "step": 12064 + }, + { + "epoch": 10.190033783783784, + "grad_norm": 0.43018728494644165, + "learning_rate": 2.9313e-05, + "loss": 0.0223, + "step": 12065 + }, + { + "epoch": 10.190878378378379, + "grad_norm": 0.30009108781814575, + "learning_rate": 2.9312666666666668e-05, + "loss": 0.0134, + "step": 12066 + }, + { + "epoch": 10.191722972972974, + "grad_norm": 0.4703797996044159, + "learning_rate": 2.9312333333333334e-05, + "loss": 0.0405, + "step": 12067 + }, + { + "epoch": 10.192567567567568, + "grad_norm": 0.4337528347969055, + "learning_rate": 2.9312e-05, + "loss": 0.0358, + "step": 12068 + }, + { + "epoch": 10.193412162162161, + "grad_norm": 0.22930294275283813, + "learning_rate": 2.931166666666667e-05, + "loss": 0.0193, + "step": 12069 + }, + { + "epoch": 10.194256756756756, + "grad_norm": 0.23761974275112152, + "learning_rate": 2.9311333333333335e-05, + "loss": 0.0217, + "step": 12070 + }, + { + "epoch": 10.19510135135135, + "grad_norm": 0.6103181838989258, + "learning_rate": 2.9311e-05, + "loss": 0.0271, + "step": 12071 + }, + { + "epoch": 10.195945945945946, + "grad_norm": 0.3547978699207306, + "learning_rate": 2.9310666666666667e-05, + "loss": 0.0246, + "step": 12072 + }, + { + "epoch": 10.19679054054054, + "grad_norm": 0.36566442251205444, + "learning_rate": 2.9310333333333336e-05, + "loss": 0.023, + "step": 12073 + }, + { + "epoch": 10.197635135135135, + "grad_norm": 0.665097713470459, + "learning_rate": 2.931e-05, + "loss": 0.0359, + "step": 12074 + }, + { + "epoch": 10.19847972972973, + "grad_norm": 0.21396979689598083, + "learning_rate": 2.9309666666666668e-05, + "loss": 0.0179, + "step": 12075 + }, + { + "epoch": 10.199324324324325, + "grad_norm": 0.7284930348396301, + "learning_rate": 2.930933333333333e-05, + "loss": 0.0277, + "step": 12076 + }, + { + "epoch": 10.20016891891892, + "grad_norm": 0.24794575572013855, + "learning_rate": 2.9309e-05, + "loss": 0.0573, + "step": 12077 + }, + { + "epoch": 10.201013513513514, + "grad_norm": 0.25861427187919617, + "learning_rate": 2.930866666666667e-05, + "loss": 0.0231, + "step": 12078 + }, + { + "epoch": 10.201858108108109, + "grad_norm": 0.35630640387535095, + "learning_rate": 2.9308333333333332e-05, + "loss": 0.0236, + "step": 12079 + }, + { + "epoch": 10.202702702702704, + "grad_norm": 0.23185394704341888, + "learning_rate": 2.9308e-05, + "loss": 0.0175, + "step": 12080 + }, + { + "epoch": 10.203547297297296, + "grad_norm": 0.5255040526390076, + "learning_rate": 2.9307666666666667e-05, + "loss": 0.0397, + "step": 12081 + }, + { + "epoch": 10.204391891891891, + "grad_norm": 0.3418085277080536, + "learning_rate": 2.9307333333333333e-05, + "loss": 0.0385, + "step": 12082 + }, + { + "epoch": 10.205236486486486, + "grad_norm": 0.22197717428207397, + "learning_rate": 2.9307e-05, + "loss": 0.0166, + "step": 12083 + }, + { + "epoch": 10.20608108108108, + "grad_norm": 0.5002927780151367, + "learning_rate": 2.930666666666667e-05, + "loss": 0.0367, + "step": 12084 + }, + { + "epoch": 10.206925675675675, + "grad_norm": 0.2572418451309204, + "learning_rate": 2.9306333333333334e-05, + "loss": 0.0212, + "step": 12085 + }, + { + "epoch": 10.20777027027027, + "grad_norm": 0.3771834373474121, + "learning_rate": 2.9306e-05, + "loss": 0.0302, + "step": 12086 + }, + { + "epoch": 10.208614864864865, + "grad_norm": 0.25723129510879517, + "learning_rate": 2.930566666666667e-05, + "loss": 0.0213, + "step": 12087 + }, + { + "epoch": 10.20945945945946, + "grad_norm": 0.3535037934780121, + "learning_rate": 2.9305333333333332e-05, + "loss": 0.0267, + "step": 12088 + }, + { + "epoch": 10.210304054054054, + "grad_norm": 0.4486699402332306, + "learning_rate": 2.9305e-05, + "loss": 0.0332, + "step": 12089 + }, + { + "epoch": 10.21114864864865, + "grad_norm": 0.8905884623527527, + "learning_rate": 2.9304666666666667e-05, + "loss": 0.0556, + "step": 12090 + }, + { + "epoch": 10.211993243243244, + "grad_norm": 1.1556349992752075, + "learning_rate": 2.9304333333333333e-05, + "loss": 0.3241, + "step": 12091 + }, + { + "epoch": 10.212837837837839, + "grad_norm": 0.8432889580726624, + "learning_rate": 2.9304e-05, + "loss": 0.242, + "step": 12092 + }, + { + "epoch": 10.213682432432432, + "grad_norm": 0.4666464328765869, + "learning_rate": 2.930366666666667e-05, + "loss": 0.0767, + "step": 12093 + }, + { + "epoch": 10.214527027027026, + "grad_norm": 0.44178080558776855, + "learning_rate": 2.9303333333333335e-05, + "loss": 0.0644, + "step": 12094 + }, + { + "epoch": 10.215371621621621, + "grad_norm": 0.418793648481369, + "learning_rate": 2.9303e-05, + "loss": 0.0536, + "step": 12095 + }, + { + "epoch": 10.216216216216216, + "grad_norm": 0.3137243092060089, + "learning_rate": 2.930266666666667e-05, + "loss": 0.0355, + "step": 12096 + }, + { + "epoch": 10.21706081081081, + "grad_norm": 0.26123714447021484, + "learning_rate": 2.9302333333333332e-05, + "loss": 0.0318, + "step": 12097 + }, + { + "epoch": 10.217905405405405, + "grad_norm": 0.274448424577713, + "learning_rate": 2.9302e-05, + "loss": 0.0301, + "step": 12098 + }, + { + "epoch": 10.21875, + "grad_norm": 0.41370677947998047, + "learning_rate": 2.9301666666666668e-05, + "loss": 0.0424, + "step": 12099 + }, + { + "epoch": 10.219594594594595, + "grad_norm": 0.3336068391799927, + "learning_rate": 2.9301333333333334e-05, + "loss": 0.0232, + "step": 12100 + }, + { + "epoch": 10.22043918918919, + "grad_norm": 0.3425310254096985, + "learning_rate": 2.9301e-05, + "loss": 0.0299, + "step": 12101 + }, + { + "epoch": 10.221283783783784, + "grad_norm": 0.30067870020866394, + "learning_rate": 2.930066666666667e-05, + "loss": 0.0321, + "step": 12102 + }, + { + "epoch": 10.222128378378379, + "grad_norm": 0.342621773481369, + "learning_rate": 2.9300333333333335e-05, + "loss": 0.0246, + "step": 12103 + }, + { + "epoch": 10.222972972972974, + "grad_norm": 0.358667254447937, + "learning_rate": 2.93e-05, + "loss": 0.02, + "step": 12104 + }, + { + "epoch": 10.223817567567568, + "grad_norm": 0.27546927332878113, + "learning_rate": 2.929966666666667e-05, + "loss": 0.0203, + "step": 12105 + }, + { + "epoch": 10.224662162162161, + "grad_norm": 0.346628874540329, + "learning_rate": 2.9299333333333333e-05, + "loss": 0.0247, + "step": 12106 + }, + { + "epoch": 10.225506756756756, + "grad_norm": 0.24324730038642883, + "learning_rate": 2.9299000000000002e-05, + "loss": 0.0252, + "step": 12107 + }, + { + "epoch": 10.22635135135135, + "grad_norm": 0.27995580434799194, + "learning_rate": 2.9298666666666664e-05, + "loss": 0.0256, + "step": 12108 + }, + { + "epoch": 10.227195945945946, + "grad_norm": 0.3076736629009247, + "learning_rate": 2.9298333333333334e-05, + "loss": 0.013, + "step": 12109 + }, + { + "epoch": 10.22804054054054, + "grad_norm": 0.2524840533733368, + "learning_rate": 2.9298000000000003e-05, + "loss": 0.0172, + "step": 12110 + }, + { + "epoch": 10.228885135135135, + "grad_norm": 0.40792474150657654, + "learning_rate": 2.9297666666666666e-05, + "loss": 0.0286, + "step": 12111 + }, + { + "epoch": 10.22972972972973, + "grad_norm": 0.2740190029144287, + "learning_rate": 2.9297333333333335e-05, + "loss": 0.0189, + "step": 12112 + }, + { + "epoch": 10.230574324324325, + "grad_norm": 0.22641442716121674, + "learning_rate": 2.9297e-05, + "loss": 0.0166, + "step": 12113 + }, + { + "epoch": 10.23141891891892, + "grad_norm": 0.32183268666267395, + "learning_rate": 2.9296666666666667e-05, + "loss": 0.0256, + "step": 12114 + }, + { + "epoch": 10.232263513513514, + "grad_norm": 0.8929100632667542, + "learning_rate": 2.9296333333333333e-05, + "loss": 0.0256, + "step": 12115 + }, + { + "epoch": 10.233108108108109, + "grad_norm": 0.40967363119125366, + "learning_rate": 2.9296000000000002e-05, + "loss": 0.0571, + "step": 12116 + }, + { + "epoch": 10.233952702702704, + "grad_norm": 0.4226244390010834, + "learning_rate": 2.9295666666666665e-05, + "loss": 0.0342, + "step": 12117 + }, + { + "epoch": 10.234797297297296, + "grad_norm": 0.6273697018623352, + "learning_rate": 2.9295333333333334e-05, + "loss": 0.0308, + "step": 12118 + }, + { + "epoch": 10.235641891891891, + "grad_norm": 0.3896539807319641, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.022, + "step": 12119 + }, + { + "epoch": 10.236486486486486, + "grad_norm": 0.47841739654541016, + "learning_rate": 2.9294666666666666e-05, + "loss": 0.019, + "step": 12120 + }, + { + "epoch": 10.23733108108108, + "grad_norm": 0.4693341851234436, + "learning_rate": 2.9294333333333335e-05, + "loss": 0.036, + "step": 12121 + }, + { + "epoch": 10.238175675675675, + "grad_norm": 0.4532187879085541, + "learning_rate": 2.9294e-05, + "loss": 0.0242, + "step": 12122 + }, + { + "epoch": 10.23902027027027, + "grad_norm": 0.5872882008552551, + "learning_rate": 2.9293666666666667e-05, + "loss": 0.0358, + "step": 12123 + }, + { + "epoch": 10.239864864864865, + "grad_norm": 0.2153579443693161, + "learning_rate": 2.9293333333333333e-05, + "loss": 0.0141, + "step": 12124 + }, + { + "epoch": 10.24070945945946, + "grad_norm": 0.26488032937049866, + "learning_rate": 2.9293000000000002e-05, + "loss": 0.0261, + "step": 12125 + }, + { + "epoch": 10.241554054054054, + "grad_norm": 0.2849927842617035, + "learning_rate": 2.9292666666666665e-05, + "loss": 0.0282, + "step": 12126 + }, + { + "epoch": 10.24239864864865, + "grad_norm": 0.2854515016078949, + "learning_rate": 2.9292333333333334e-05, + "loss": 0.024, + "step": 12127 + }, + { + "epoch": 10.243243243243244, + "grad_norm": 0.8658056855201721, + "learning_rate": 2.9292000000000003e-05, + "loss": 0.0179, + "step": 12128 + }, + { + "epoch": 10.244087837837839, + "grad_norm": 0.3915572762489319, + "learning_rate": 2.9291666666666666e-05, + "loss": 0.0389, + "step": 12129 + }, + { + "epoch": 10.244932432432432, + "grad_norm": 0.392465740442276, + "learning_rate": 2.9291333333333335e-05, + "loss": 0.0273, + "step": 12130 + }, + { + "epoch": 10.245777027027026, + "grad_norm": 0.5610549449920654, + "learning_rate": 2.9291e-05, + "loss": 0.0318, + "step": 12131 + }, + { + "epoch": 10.246621621621621, + "grad_norm": 0.37195831537246704, + "learning_rate": 2.9290666666666667e-05, + "loss": 0.0188, + "step": 12132 + }, + { + "epoch": 10.247466216216216, + "grad_norm": 0.43663936853408813, + "learning_rate": 2.9290333333333333e-05, + "loss": 0.0256, + "step": 12133 + }, + { + "epoch": 10.24831081081081, + "grad_norm": 0.37725719809532166, + "learning_rate": 2.9290000000000002e-05, + "loss": 0.0359, + "step": 12134 + }, + { + "epoch": 10.249155405405405, + "grad_norm": 0.20622976124286652, + "learning_rate": 2.928966666666667e-05, + "loss": 0.0137, + "step": 12135 + }, + { + "epoch": 10.25, + "grad_norm": 0.2891639471054077, + "learning_rate": 2.9289333333333334e-05, + "loss": 0.0358, + "step": 12136 + }, + { + "epoch": 10.250844594594595, + "grad_norm": 0.39632540941238403, + "learning_rate": 2.9289e-05, + "loss": 0.0235, + "step": 12137 + }, + { + "epoch": 10.25168918918919, + "grad_norm": 0.8981592059135437, + "learning_rate": 2.9288666666666666e-05, + "loss": 0.0487, + "step": 12138 + }, + { + "epoch": 10.252533783783784, + "grad_norm": 0.4703606367111206, + "learning_rate": 2.9288333333333336e-05, + "loss": 0.0294, + "step": 12139 + }, + { + "epoch": 10.253378378378379, + "grad_norm": 0.593924343585968, + "learning_rate": 2.9287999999999998e-05, + "loss": 0.0441, + "step": 12140 + }, + { + "epoch": 10.254222972972974, + "grad_norm": 1.0740039348602295, + "learning_rate": 2.9287666666666667e-05, + "loss": 0.2531, + "step": 12141 + }, + { + "epoch": 10.255067567567568, + "grad_norm": 0.5120143890380859, + "learning_rate": 2.9287333333333333e-05, + "loss": 0.1543, + "step": 12142 + }, + { + "epoch": 10.255912162162161, + "grad_norm": 0.3915368616580963, + "learning_rate": 2.9287e-05, + "loss": 0.0784, + "step": 12143 + }, + { + "epoch": 10.256756756756756, + "grad_norm": 0.3453342914581299, + "learning_rate": 2.928666666666667e-05, + "loss": 0.051, + "step": 12144 + }, + { + "epoch": 10.25760135135135, + "grad_norm": 0.2877572476863861, + "learning_rate": 2.9286333333333335e-05, + "loss": 0.0272, + "step": 12145 + }, + { + "epoch": 10.258445945945946, + "grad_norm": 0.26134297251701355, + "learning_rate": 2.9286e-05, + "loss": 0.0286, + "step": 12146 + }, + { + "epoch": 10.25929054054054, + "grad_norm": 0.514521062374115, + "learning_rate": 2.9285666666666666e-05, + "loss": 0.0381, + "step": 12147 + }, + { + "epoch": 10.260135135135135, + "grad_norm": 0.32221996784210205, + "learning_rate": 2.9285333333333336e-05, + "loss": 0.0194, + "step": 12148 + }, + { + "epoch": 10.26097972972973, + "grad_norm": 0.3395960330963135, + "learning_rate": 2.9284999999999998e-05, + "loss": 0.0354, + "step": 12149 + }, + { + "epoch": 10.261824324324325, + "grad_norm": 0.4697655439376831, + "learning_rate": 2.9284666666666668e-05, + "loss": 0.0306, + "step": 12150 + }, + { + "epoch": 10.26266891891892, + "grad_norm": 0.35610976815223694, + "learning_rate": 2.9284333333333334e-05, + "loss": 0.0288, + "step": 12151 + }, + { + "epoch": 10.263513513513514, + "grad_norm": 0.24573679268360138, + "learning_rate": 2.9284e-05, + "loss": 0.0345, + "step": 12152 + }, + { + "epoch": 10.264358108108109, + "grad_norm": 0.3528389036655426, + "learning_rate": 2.928366666666667e-05, + "loss": 0.0423, + "step": 12153 + }, + { + "epoch": 10.265202702702704, + "grad_norm": 0.4050474464893341, + "learning_rate": 2.9283333333333335e-05, + "loss": 0.0347, + "step": 12154 + }, + { + "epoch": 10.266047297297296, + "grad_norm": 0.30225545167922974, + "learning_rate": 2.9283e-05, + "loss": 0.0271, + "step": 12155 + }, + { + "epoch": 10.266891891891891, + "grad_norm": 0.2960115671157837, + "learning_rate": 2.9282666666666667e-05, + "loss": 0.0319, + "step": 12156 + }, + { + "epoch": 10.267736486486486, + "grad_norm": 0.33983975648880005, + "learning_rate": 2.9282333333333336e-05, + "loss": 0.0279, + "step": 12157 + }, + { + "epoch": 10.26858108108108, + "grad_norm": 0.2965575158596039, + "learning_rate": 2.9282e-05, + "loss": 0.0361, + "step": 12158 + }, + { + "epoch": 10.269425675675675, + "grad_norm": 0.3342045843601227, + "learning_rate": 2.9281666666666668e-05, + "loss": 0.037, + "step": 12159 + }, + { + "epoch": 10.27027027027027, + "grad_norm": 0.27293092012405396, + "learning_rate": 2.9281333333333337e-05, + "loss": 0.0238, + "step": 12160 + }, + { + "epoch": 10.271114864864865, + "grad_norm": 0.3181976079940796, + "learning_rate": 2.9281e-05, + "loss": 0.0221, + "step": 12161 + }, + { + "epoch": 10.27195945945946, + "grad_norm": 0.45941615104675293, + "learning_rate": 2.928066666666667e-05, + "loss": 0.0211, + "step": 12162 + }, + { + "epoch": 10.272804054054054, + "grad_norm": 0.2357138842344284, + "learning_rate": 2.9280333333333335e-05, + "loss": 0.0282, + "step": 12163 + }, + { + "epoch": 10.27364864864865, + "grad_norm": 0.2375466525554657, + "learning_rate": 2.928e-05, + "loss": 0.0232, + "step": 12164 + }, + { + "epoch": 10.274493243243244, + "grad_norm": 0.20336785912513733, + "learning_rate": 2.9279666666666667e-05, + "loss": 0.0176, + "step": 12165 + }, + { + "epoch": 10.275337837837839, + "grad_norm": 0.24324505031108856, + "learning_rate": 2.9279333333333336e-05, + "loss": 0.0185, + "step": 12166 + }, + { + "epoch": 10.276182432432432, + "grad_norm": 0.27911725640296936, + "learning_rate": 2.9279e-05, + "loss": 0.0235, + "step": 12167 + }, + { + "epoch": 10.277027027027026, + "grad_norm": 0.2252526581287384, + "learning_rate": 2.9278666666666668e-05, + "loss": 0.0266, + "step": 12168 + }, + { + "epoch": 10.277871621621621, + "grad_norm": 0.2428465038537979, + "learning_rate": 2.9278333333333334e-05, + "loss": 0.0215, + "step": 12169 + }, + { + "epoch": 10.278716216216216, + "grad_norm": 0.27195414900779724, + "learning_rate": 2.9278e-05, + "loss": 0.0202, + "step": 12170 + }, + { + "epoch": 10.27956081081081, + "grad_norm": 0.8204793334007263, + "learning_rate": 2.927766666666667e-05, + "loss": 0.028, + "step": 12171 + }, + { + "epoch": 10.280405405405405, + "grad_norm": 0.28673285245895386, + "learning_rate": 2.9277333333333332e-05, + "loss": 0.0215, + "step": 12172 + }, + { + "epoch": 10.28125, + "grad_norm": 0.2657383680343628, + "learning_rate": 2.9277e-05, + "loss": 0.0225, + "step": 12173 + }, + { + "epoch": 10.282094594594595, + "grad_norm": 0.45380836725234985, + "learning_rate": 2.9276666666666667e-05, + "loss": 0.0256, + "step": 12174 + }, + { + "epoch": 10.28293918918919, + "grad_norm": 0.6528440713882446, + "learning_rate": 2.9276333333333333e-05, + "loss": 0.0369, + "step": 12175 + }, + { + "epoch": 10.283783783783784, + "grad_norm": 0.2913818359375, + "learning_rate": 2.9276e-05, + "loss": 0.0134, + "step": 12176 + }, + { + "epoch": 10.284628378378379, + "grad_norm": 0.5983127355575562, + "learning_rate": 2.9275666666666668e-05, + "loss": 0.0298, + "step": 12177 + }, + { + "epoch": 10.285472972972974, + "grad_norm": 0.3421265482902527, + "learning_rate": 2.9275333333333334e-05, + "loss": 0.0209, + "step": 12178 + }, + { + "epoch": 10.286317567567568, + "grad_norm": 0.4183584749698639, + "learning_rate": 2.9275e-05, + "loss": 0.0379, + "step": 12179 + }, + { + "epoch": 10.287162162162161, + "grad_norm": 0.37724679708480835, + "learning_rate": 2.927466666666667e-05, + "loss": 0.0259, + "step": 12180 + }, + { + "epoch": 10.288006756756756, + "grad_norm": 0.36715391278266907, + "learning_rate": 2.9274333333333332e-05, + "loss": 0.0233, + "step": 12181 + }, + { + "epoch": 10.28885135135135, + "grad_norm": 0.39792975783348083, + "learning_rate": 2.9274e-05, + "loss": 0.0385, + "step": 12182 + }, + { + "epoch": 10.289695945945946, + "grad_norm": 0.3947053551673889, + "learning_rate": 2.9273666666666667e-05, + "loss": 0.0226, + "step": 12183 + }, + { + "epoch": 10.29054054054054, + "grad_norm": 0.3892548382282257, + "learning_rate": 2.9273333333333333e-05, + "loss": 0.0204, + "step": 12184 + }, + { + "epoch": 10.291385135135135, + "grad_norm": 0.5592049956321716, + "learning_rate": 2.9273000000000002e-05, + "loss": 0.0551, + "step": 12185 + }, + { + "epoch": 10.29222972972973, + "grad_norm": 0.49351873993873596, + "learning_rate": 2.927266666666667e-05, + "loss": 0.0434, + "step": 12186 + }, + { + "epoch": 10.293074324324325, + "grad_norm": 0.3510099947452545, + "learning_rate": 2.9272333333333334e-05, + "loss": 0.0372, + "step": 12187 + }, + { + "epoch": 10.29391891891892, + "grad_norm": 0.7953169941902161, + "learning_rate": 2.9272e-05, + "loss": 0.0446, + "step": 12188 + }, + { + "epoch": 10.294763513513514, + "grad_norm": 0.48534467816352844, + "learning_rate": 2.927166666666667e-05, + "loss": 0.0377, + "step": 12189 + }, + { + "epoch": 10.295608108108109, + "grad_norm": 0.9511626958847046, + "learning_rate": 2.9271333333333332e-05, + "loss": 0.0531, + "step": 12190 + }, + { + "epoch": 10.296452702702704, + "grad_norm": 1.426512598991394, + "learning_rate": 2.9271e-05, + "loss": 0.3055, + "step": 12191 + }, + { + "epoch": 10.297297297297296, + "grad_norm": 0.6866361498832703, + "learning_rate": 2.9270666666666667e-05, + "loss": 0.191, + "step": 12192 + }, + { + "epoch": 10.298141891891891, + "grad_norm": 0.5692123770713806, + "learning_rate": 2.9270333333333333e-05, + "loss": 0.0822, + "step": 12193 + }, + { + "epoch": 10.298986486486486, + "grad_norm": 0.4274158179759979, + "learning_rate": 2.9270000000000003e-05, + "loss": 0.0396, + "step": 12194 + }, + { + "epoch": 10.29983108108108, + "grad_norm": 0.35044413805007935, + "learning_rate": 2.926966666666667e-05, + "loss": 0.0518, + "step": 12195 + }, + { + "epoch": 10.300675675675675, + "grad_norm": 0.32566800713539124, + "learning_rate": 2.9269333333333335e-05, + "loss": 0.0345, + "step": 12196 + }, + { + "epoch": 10.30152027027027, + "grad_norm": 0.36165401339530945, + "learning_rate": 2.9269e-05, + "loss": 0.0359, + "step": 12197 + }, + { + "epoch": 10.302364864864865, + "grad_norm": 0.3234684467315674, + "learning_rate": 2.9268666666666666e-05, + "loss": 0.0271, + "step": 12198 + }, + { + "epoch": 10.30320945945946, + "grad_norm": 0.3065427541732788, + "learning_rate": 2.9268333333333332e-05, + "loss": 0.0376, + "step": 12199 + }, + { + "epoch": 10.304054054054054, + "grad_norm": 0.25475654006004333, + "learning_rate": 2.9268e-05, + "loss": 0.0195, + "step": 12200 + }, + { + "epoch": 10.30489864864865, + "grad_norm": 0.24762913584709167, + "learning_rate": 2.9267666666666664e-05, + "loss": 0.0246, + "step": 12201 + }, + { + "epoch": 10.305743243243244, + "grad_norm": 0.3001789450645447, + "learning_rate": 2.9267333333333334e-05, + "loss": 0.0297, + "step": 12202 + }, + { + "epoch": 10.306587837837839, + "grad_norm": 0.5273482203483582, + "learning_rate": 2.9267000000000003e-05, + "loss": 0.0389, + "step": 12203 + }, + { + "epoch": 10.307432432432432, + "grad_norm": 0.2860933244228363, + "learning_rate": 2.9266666666666665e-05, + "loss": 0.0221, + "step": 12204 + }, + { + "epoch": 10.308277027027026, + "grad_norm": 0.2857808470726013, + "learning_rate": 2.9266333333333335e-05, + "loss": 0.0293, + "step": 12205 + }, + { + "epoch": 10.309121621621621, + "grad_norm": 0.24440619349479675, + "learning_rate": 2.9266e-05, + "loss": 0.0197, + "step": 12206 + }, + { + "epoch": 10.309966216216216, + "grad_norm": 0.2751956284046173, + "learning_rate": 2.9265666666666667e-05, + "loss": 0.0244, + "step": 12207 + }, + { + "epoch": 10.31081081081081, + "grad_norm": 0.24487519264221191, + "learning_rate": 2.9265333333333333e-05, + "loss": 0.0214, + "step": 12208 + }, + { + "epoch": 10.311655405405405, + "grad_norm": 0.2615402936935425, + "learning_rate": 2.9265000000000002e-05, + "loss": 0.0149, + "step": 12209 + }, + { + "epoch": 10.3125, + "grad_norm": 0.4036824703216553, + "learning_rate": 2.9264666666666668e-05, + "loss": 0.0197, + "step": 12210 + }, + { + "epoch": 10.313344594594595, + "grad_norm": 0.3630652129650116, + "learning_rate": 2.9264333333333334e-05, + "loss": 0.0277, + "step": 12211 + }, + { + "epoch": 10.31418918918919, + "grad_norm": 0.46564793586730957, + "learning_rate": 2.9264000000000003e-05, + "loss": 0.0399, + "step": 12212 + }, + { + "epoch": 10.315033783783784, + "grad_norm": 0.33285170793533325, + "learning_rate": 2.9263666666666666e-05, + "loss": 0.0188, + "step": 12213 + }, + { + "epoch": 10.315878378378379, + "grad_norm": 0.6711264252662659, + "learning_rate": 2.9263333333333335e-05, + "loss": 0.0272, + "step": 12214 + }, + { + "epoch": 10.316722972972974, + "grad_norm": 0.34707313776016235, + "learning_rate": 2.9263e-05, + "loss": 0.0283, + "step": 12215 + }, + { + "epoch": 10.317567567567568, + "grad_norm": 0.4912678003311157, + "learning_rate": 2.9262666666666667e-05, + "loss": 0.0243, + "step": 12216 + }, + { + "epoch": 10.318412162162161, + "grad_norm": 0.22226819396018982, + "learning_rate": 2.9262333333333333e-05, + "loss": 0.0195, + "step": 12217 + }, + { + "epoch": 10.319256756756756, + "grad_norm": 0.20132088661193848, + "learning_rate": 2.9262000000000002e-05, + "loss": 0.0116, + "step": 12218 + }, + { + "epoch": 10.32010135135135, + "grad_norm": 0.41732704639434814, + "learning_rate": 2.9261666666666668e-05, + "loss": 0.0292, + "step": 12219 + }, + { + "epoch": 10.320945945945946, + "grad_norm": 0.7472606301307678, + "learning_rate": 2.9261333333333334e-05, + "loss": 0.0292, + "step": 12220 + }, + { + "epoch": 10.32179054054054, + "grad_norm": 0.6285501718521118, + "learning_rate": 2.9261000000000003e-05, + "loss": 0.035, + "step": 12221 + }, + { + "epoch": 10.322635135135135, + "grad_norm": 0.1886579692363739, + "learning_rate": 2.9260666666666666e-05, + "loss": 0.0169, + "step": 12222 + }, + { + "epoch": 10.32347972972973, + "grad_norm": 0.3097766637802124, + "learning_rate": 2.9260333333333335e-05, + "loss": 0.0187, + "step": 12223 + }, + { + "epoch": 10.324324324324325, + "grad_norm": 0.47955992817878723, + "learning_rate": 2.926e-05, + "loss": 0.0257, + "step": 12224 + }, + { + "epoch": 10.32516891891892, + "grad_norm": 0.512060284614563, + "learning_rate": 2.9259666666666667e-05, + "loss": 0.0198, + "step": 12225 + }, + { + "epoch": 10.326013513513514, + "grad_norm": 0.33768194913864136, + "learning_rate": 2.9259333333333333e-05, + "loss": 0.0238, + "step": 12226 + }, + { + "epoch": 10.326858108108109, + "grad_norm": 0.36717310547828674, + "learning_rate": 2.9259e-05, + "loss": 0.0254, + "step": 12227 + }, + { + "epoch": 10.327702702702704, + "grad_norm": 0.22091421484947205, + "learning_rate": 2.9258666666666668e-05, + "loss": 0.0186, + "step": 12228 + }, + { + "epoch": 10.328547297297296, + "grad_norm": 0.5557757019996643, + "learning_rate": 2.9258333333333334e-05, + "loss": 0.0219, + "step": 12229 + }, + { + "epoch": 10.329391891891891, + "grad_norm": 0.4166153371334076, + "learning_rate": 2.9258e-05, + "loss": 0.042, + "step": 12230 + }, + { + "epoch": 10.330236486486486, + "grad_norm": 0.7503197193145752, + "learning_rate": 2.9257666666666666e-05, + "loss": 0.0339, + "step": 12231 + }, + { + "epoch": 10.33108108108108, + "grad_norm": 0.31568431854248047, + "learning_rate": 2.9257333333333335e-05, + "loss": 0.0168, + "step": 12232 + }, + { + "epoch": 10.331925675675675, + "grad_norm": 0.9386552572250366, + "learning_rate": 2.9256999999999998e-05, + "loss": 0.0352, + "step": 12233 + }, + { + "epoch": 10.33277027027027, + "grad_norm": 0.3154095411300659, + "learning_rate": 2.9256666666666667e-05, + "loss": 0.0252, + "step": 12234 + }, + { + "epoch": 10.333614864864865, + "grad_norm": 1.1921470165252686, + "learning_rate": 2.9256333333333337e-05, + "loss": 0.0289, + "step": 12235 + }, + { + "epoch": 10.33445945945946, + "grad_norm": 0.5586099028587341, + "learning_rate": 2.9256e-05, + "loss": 0.0312, + "step": 12236 + }, + { + "epoch": 10.335304054054054, + "grad_norm": 0.3295399844646454, + "learning_rate": 2.925566666666667e-05, + "loss": 0.0336, + "step": 12237 + }, + { + "epoch": 10.33614864864865, + "grad_norm": 0.8192151784896851, + "learning_rate": 2.9255333333333334e-05, + "loss": 0.0589, + "step": 12238 + }, + { + "epoch": 10.336993243243244, + "grad_norm": 0.7531377077102661, + "learning_rate": 2.9255e-05, + "loss": 0.0955, + "step": 12239 + }, + { + "epoch": 10.337837837837839, + "grad_norm": 0.39076536893844604, + "learning_rate": 2.9254666666666666e-05, + "loss": 0.0343, + "step": 12240 + }, + { + "epoch": 10.338682432432432, + "grad_norm": 0.7579339146614075, + "learning_rate": 2.9254333333333336e-05, + "loss": 0.2627, + "step": 12241 + }, + { + "epoch": 10.339527027027026, + "grad_norm": 0.6227624416351318, + "learning_rate": 2.9253999999999998e-05, + "loss": 0.1842, + "step": 12242 + }, + { + "epoch": 10.340371621621621, + "grad_norm": 0.45487087965011597, + "learning_rate": 2.9253666666666667e-05, + "loss": 0.0863, + "step": 12243 + }, + { + "epoch": 10.341216216216216, + "grad_norm": 0.5448442697525024, + "learning_rate": 2.9253333333333337e-05, + "loss": 0.103, + "step": 12244 + }, + { + "epoch": 10.34206081081081, + "grad_norm": 0.28892621397972107, + "learning_rate": 2.9253e-05, + "loss": 0.0353, + "step": 12245 + }, + { + "epoch": 10.342905405405405, + "grad_norm": 0.27871283888816833, + "learning_rate": 2.925266666666667e-05, + "loss": 0.0275, + "step": 12246 + }, + { + "epoch": 10.34375, + "grad_norm": 0.436007022857666, + "learning_rate": 2.9252333333333335e-05, + "loss": 0.039, + "step": 12247 + }, + { + "epoch": 10.344594594594595, + "grad_norm": 0.24390052258968353, + "learning_rate": 2.9252e-05, + "loss": 0.0222, + "step": 12248 + }, + { + "epoch": 10.34543918918919, + "grad_norm": 0.49638059735298157, + "learning_rate": 2.9251666666666666e-05, + "loss": 0.0322, + "step": 12249 + }, + { + "epoch": 10.346283783783784, + "grad_norm": 0.33726921677589417, + "learning_rate": 2.9251333333333336e-05, + "loss": 0.0256, + "step": 12250 + }, + { + "epoch": 10.347128378378379, + "grad_norm": 0.31618571281433105, + "learning_rate": 2.9250999999999998e-05, + "loss": 0.0274, + "step": 12251 + }, + { + "epoch": 10.347972972972974, + "grad_norm": 0.6803858280181885, + "learning_rate": 2.9250666666666668e-05, + "loss": 0.0314, + "step": 12252 + }, + { + "epoch": 10.348817567567568, + "grad_norm": 0.27590474486351013, + "learning_rate": 2.9250333333333337e-05, + "loss": 0.0231, + "step": 12253 + }, + { + "epoch": 10.349662162162161, + "grad_norm": 0.3305881917476654, + "learning_rate": 2.925e-05, + "loss": 0.0426, + "step": 12254 + }, + { + "epoch": 10.350506756756756, + "grad_norm": 0.33427560329437256, + "learning_rate": 2.924966666666667e-05, + "loss": 0.0326, + "step": 12255 + }, + { + "epoch": 10.35135135135135, + "grad_norm": 0.32823607325553894, + "learning_rate": 2.9249333333333335e-05, + "loss": 0.0416, + "step": 12256 + }, + { + "epoch": 10.352195945945946, + "grad_norm": 0.2338120937347412, + "learning_rate": 2.9249e-05, + "loss": 0.0253, + "step": 12257 + }, + { + "epoch": 10.35304054054054, + "grad_norm": 0.6339359879493713, + "learning_rate": 2.9248666666666667e-05, + "loss": 0.0257, + "step": 12258 + }, + { + "epoch": 10.353885135135135, + "grad_norm": 0.39218229055404663, + "learning_rate": 2.9248333333333333e-05, + "loss": 0.0295, + "step": 12259 + }, + { + "epoch": 10.35472972972973, + "grad_norm": 0.31045493483543396, + "learning_rate": 2.9248000000000002e-05, + "loss": 0.0329, + "step": 12260 + }, + { + "epoch": 10.355574324324325, + "grad_norm": 0.33373650908470154, + "learning_rate": 2.9247666666666668e-05, + "loss": 0.0201, + "step": 12261 + }, + { + "epoch": 10.35641891891892, + "grad_norm": 0.5097512006759644, + "learning_rate": 2.9247333333333334e-05, + "loss": 0.0253, + "step": 12262 + }, + { + "epoch": 10.357263513513514, + "grad_norm": 0.2865590751171112, + "learning_rate": 2.9247e-05, + "loss": 0.0226, + "step": 12263 + }, + { + "epoch": 10.358108108108109, + "grad_norm": 1.2197729349136353, + "learning_rate": 2.924666666666667e-05, + "loss": 0.1137, + "step": 12264 + }, + { + "epoch": 10.358952702702704, + "grad_norm": 0.4519363343715668, + "learning_rate": 2.924633333333333e-05, + "loss": 0.031, + "step": 12265 + }, + { + "epoch": 10.359797297297296, + "grad_norm": 0.24410288035869598, + "learning_rate": 2.9246e-05, + "loss": 0.0175, + "step": 12266 + }, + { + "epoch": 10.360641891891891, + "grad_norm": 0.3417820930480957, + "learning_rate": 2.9245666666666667e-05, + "loss": 0.0228, + "step": 12267 + }, + { + "epoch": 10.361486486486486, + "grad_norm": 0.4446077346801758, + "learning_rate": 2.9245333333333333e-05, + "loss": 0.0353, + "step": 12268 + }, + { + "epoch": 10.36233108108108, + "grad_norm": 0.28968510031700134, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0172, + "step": 12269 + }, + { + "epoch": 10.363175675675675, + "grad_norm": 0.42882826924324036, + "learning_rate": 2.9244666666666668e-05, + "loss": 0.0212, + "step": 12270 + }, + { + "epoch": 10.36402027027027, + "grad_norm": 0.5405979156494141, + "learning_rate": 2.9244333333333334e-05, + "loss": 0.0217, + "step": 12271 + }, + { + "epoch": 10.364864864864865, + "grad_norm": 0.4799449145793915, + "learning_rate": 2.9244e-05, + "loss": 0.0425, + "step": 12272 + }, + { + "epoch": 10.36570945945946, + "grad_norm": 0.3663110136985779, + "learning_rate": 2.924366666666667e-05, + "loss": 0.0253, + "step": 12273 + }, + { + "epoch": 10.366554054054054, + "grad_norm": 0.5181928873062134, + "learning_rate": 2.9243333333333332e-05, + "loss": 0.0257, + "step": 12274 + }, + { + "epoch": 10.36739864864865, + "grad_norm": 0.3898049592971802, + "learning_rate": 2.9243e-05, + "loss": 0.0253, + "step": 12275 + }, + { + "epoch": 10.368243243243244, + "grad_norm": 0.7334952354431152, + "learning_rate": 2.9242666666666667e-05, + "loss": 0.0525, + "step": 12276 + }, + { + "epoch": 10.369087837837839, + "grad_norm": 0.41446173191070557, + "learning_rate": 2.9242333333333333e-05, + "loss": 0.0273, + "step": 12277 + }, + { + "epoch": 10.369932432432432, + "grad_norm": 0.4810672104358673, + "learning_rate": 2.9242000000000002e-05, + "loss": 0.0414, + "step": 12278 + }, + { + "epoch": 10.370777027027026, + "grad_norm": 0.3387177884578705, + "learning_rate": 2.9241666666666668e-05, + "loss": 0.0346, + "step": 12279 + }, + { + "epoch": 10.371621621621621, + "grad_norm": 0.3893691599369049, + "learning_rate": 2.9241333333333334e-05, + "loss": 0.0236, + "step": 12280 + }, + { + "epoch": 10.372466216216216, + "grad_norm": 0.5675857663154602, + "learning_rate": 2.9241e-05, + "loss": 0.0371, + "step": 12281 + }, + { + "epoch": 10.37331081081081, + "grad_norm": 0.3897525668144226, + "learning_rate": 2.924066666666667e-05, + "loss": 0.027, + "step": 12282 + }, + { + "epoch": 10.374155405405405, + "grad_norm": 0.4603312909603119, + "learning_rate": 2.9240333333333332e-05, + "loss": 0.0357, + "step": 12283 + }, + { + "epoch": 10.375, + "grad_norm": 0.6267913579940796, + "learning_rate": 2.924e-05, + "loss": 0.0468, + "step": 12284 + }, + { + "epoch": 10.375844594594595, + "grad_norm": 0.38703224062919617, + "learning_rate": 2.923966666666667e-05, + "loss": 0.0279, + "step": 12285 + }, + { + "epoch": 10.37668918918919, + "grad_norm": 0.3235868215560913, + "learning_rate": 2.9239333333333333e-05, + "loss": 0.0274, + "step": 12286 + }, + { + "epoch": 10.377533783783784, + "grad_norm": 0.4269748032093048, + "learning_rate": 2.9239000000000002e-05, + "loss": 0.0315, + "step": 12287 + }, + { + "epoch": 10.378378378378379, + "grad_norm": 0.4156336486339569, + "learning_rate": 2.9238666666666665e-05, + "loss": 0.0289, + "step": 12288 + }, + { + "epoch": 10.379222972972974, + "grad_norm": 0.5619422793388367, + "learning_rate": 2.9238333333333334e-05, + "loss": 0.0421, + "step": 12289 + }, + { + "epoch": 10.380067567567568, + "grad_norm": 0.7795421481132507, + "learning_rate": 2.9238e-05, + "loss": 0.0514, + "step": 12290 + }, + { + "epoch": 10.380912162162161, + "grad_norm": 0.8368946313858032, + "learning_rate": 2.9237666666666666e-05, + "loss": 0.2294, + "step": 12291 + }, + { + "epoch": 10.381756756756756, + "grad_norm": 0.762021541595459, + "learning_rate": 2.9237333333333332e-05, + "loss": 0.1713, + "step": 12292 + }, + { + "epoch": 10.38260135135135, + "grad_norm": 0.3836497962474823, + "learning_rate": 2.9237e-05, + "loss": 0.0655, + "step": 12293 + }, + { + "epoch": 10.383445945945946, + "grad_norm": 0.33937427401542664, + "learning_rate": 2.9236666666666667e-05, + "loss": 0.0324, + "step": 12294 + }, + { + "epoch": 10.38429054054054, + "grad_norm": 0.4055727422237396, + "learning_rate": 2.9236333333333333e-05, + "loss": 0.0306, + "step": 12295 + }, + { + "epoch": 10.385135135135135, + "grad_norm": 0.46806126832962036, + "learning_rate": 2.9236000000000003e-05, + "loss": 0.0354, + "step": 12296 + }, + { + "epoch": 10.38597972972973, + "grad_norm": 0.32432544231414795, + "learning_rate": 2.9235666666666665e-05, + "loss": 0.0449, + "step": 12297 + }, + { + "epoch": 10.386824324324325, + "grad_norm": 0.5084580779075623, + "learning_rate": 2.9235333333333335e-05, + "loss": 0.0329, + "step": 12298 + }, + { + "epoch": 10.38766891891892, + "grad_norm": 0.21827717125415802, + "learning_rate": 2.9235e-05, + "loss": 0.0121, + "step": 12299 + }, + { + "epoch": 10.388513513513514, + "grad_norm": 0.33113113045692444, + "learning_rate": 2.9234666666666666e-05, + "loss": 0.0291, + "step": 12300 + }, + { + "epoch": 10.389358108108109, + "grad_norm": 0.32961520552635193, + "learning_rate": 2.9234333333333332e-05, + "loss": 0.0237, + "step": 12301 + }, + { + "epoch": 10.390202702702704, + "grad_norm": 0.4170404374599457, + "learning_rate": 2.9234e-05, + "loss": 0.0279, + "step": 12302 + }, + { + "epoch": 10.391047297297296, + "grad_norm": 0.42288270592689514, + "learning_rate": 2.9233666666666668e-05, + "loss": 0.0353, + "step": 12303 + }, + { + "epoch": 10.391891891891891, + "grad_norm": 0.3289206624031067, + "learning_rate": 2.9233333333333334e-05, + "loss": 0.0242, + "step": 12304 + }, + { + "epoch": 10.392736486486486, + "grad_norm": 0.3095571994781494, + "learning_rate": 2.9233000000000003e-05, + "loss": 0.0242, + "step": 12305 + }, + { + "epoch": 10.39358108108108, + "grad_norm": 0.293293833732605, + "learning_rate": 2.9232666666666665e-05, + "loss": 0.0205, + "step": 12306 + }, + { + "epoch": 10.394425675675675, + "grad_norm": 0.580608069896698, + "learning_rate": 2.9232333333333335e-05, + "loss": 0.0348, + "step": 12307 + }, + { + "epoch": 10.39527027027027, + "grad_norm": 0.36254602670669556, + "learning_rate": 2.9232e-05, + "loss": 0.0356, + "step": 12308 + }, + { + "epoch": 10.396114864864865, + "grad_norm": 0.3115309774875641, + "learning_rate": 2.9231666666666667e-05, + "loss": 0.0284, + "step": 12309 + }, + { + "epoch": 10.39695945945946, + "grad_norm": 0.5476272106170654, + "learning_rate": 2.9231333333333336e-05, + "loss": 0.0271, + "step": 12310 + }, + { + "epoch": 10.397804054054054, + "grad_norm": 0.38888174295425415, + "learning_rate": 2.9231000000000002e-05, + "loss": 0.026, + "step": 12311 + }, + { + "epoch": 10.39864864864865, + "grad_norm": 0.5566068887710571, + "learning_rate": 2.9230666666666668e-05, + "loss": 0.0192, + "step": 12312 + }, + { + "epoch": 10.399493243243244, + "grad_norm": 0.24522650241851807, + "learning_rate": 2.9230333333333334e-05, + "loss": 0.0221, + "step": 12313 + }, + { + "epoch": 10.400337837837839, + "grad_norm": 0.23858368396759033, + "learning_rate": 2.9230000000000003e-05, + "loss": 0.015, + "step": 12314 + }, + { + "epoch": 10.401182432432432, + "grad_norm": 0.2598164677619934, + "learning_rate": 2.9229666666666666e-05, + "loss": 0.0176, + "step": 12315 + }, + { + "epoch": 10.402027027027026, + "grad_norm": 0.35105961561203003, + "learning_rate": 2.9229333333333335e-05, + "loss": 0.024, + "step": 12316 + }, + { + "epoch": 10.402871621621621, + "grad_norm": 0.6137009263038635, + "learning_rate": 2.9229e-05, + "loss": 0.0328, + "step": 12317 + }, + { + "epoch": 10.403716216216216, + "grad_norm": 0.3406740128993988, + "learning_rate": 2.9228666666666667e-05, + "loss": 0.0227, + "step": 12318 + }, + { + "epoch": 10.40456081081081, + "grad_norm": 1.0714919567108154, + "learning_rate": 2.9228333333333336e-05, + "loss": 0.0124, + "step": 12319 + }, + { + "epoch": 10.405405405405405, + "grad_norm": 0.7433798313140869, + "learning_rate": 2.9228e-05, + "loss": 0.029, + "step": 12320 + }, + { + "epoch": 10.40625, + "grad_norm": 0.4106453061103821, + "learning_rate": 2.9227666666666668e-05, + "loss": 0.0213, + "step": 12321 + }, + { + "epoch": 10.407094594594595, + "grad_norm": 0.2716028094291687, + "learning_rate": 2.9227333333333334e-05, + "loss": 0.0166, + "step": 12322 + }, + { + "epoch": 10.40793918918919, + "grad_norm": 0.4038941562175751, + "learning_rate": 2.9227e-05, + "loss": 0.0298, + "step": 12323 + }, + { + "epoch": 10.408783783783784, + "grad_norm": 0.3064543604850769, + "learning_rate": 2.9226666666666666e-05, + "loss": 0.0278, + "step": 12324 + }, + { + "epoch": 10.409628378378379, + "grad_norm": 0.5131444334983826, + "learning_rate": 2.9226333333333335e-05, + "loss": 0.0365, + "step": 12325 + }, + { + "epoch": 10.410472972972974, + "grad_norm": 0.473193883895874, + "learning_rate": 2.9226e-05, + "loss": 0.0208, + "step": 12326 + }, + { + "epoch": 10.411317567567568, + "grad_norm": 0.5105886459350586, + "learning_rate": 2.9225666666666667e-05, + "loss": 0.0322, + "step": 12327 + }, + { + "epoch": 10.412162162162161, + "grad_norm": 1.0444574356079102, + "learning_rate": 2.9225333333333336e-05, + "loss": 0.0462, + "step": 12328 + }, + { + "epoch": 10.413006756756756, + "grad_norm": 0.39119961857795715, + "learning_rate": 2.9225e-05, + "loss": 0.0264, + "step": 12329 + }, + { + "epoch": 10.41385135135135, + "grad_norm": 0.42556488513946533, + "learning_rate": 2.9224666666666668e-05, + "loss": 0.0225, + "step": 12330 + }, + { + "epoch": 10.414695945945946, + "grad_norm": 0.28647366166114807, + "learning_rate": 2.9224333333333334e-05, + "loss": 0.0273, + "step": 12331 + }, + { + "epoch": 10.41554054054054, + "grad_norm": 0.5060071349143982, + "learning_rate": 2.9224e-05, + "loss": 0.0295, + "step": 12332 + }, + { + "epoch": 10.416385135135135, + "grad_norm": 1.2556090354919434, + "learning_rate": 2.9223666666666666e-05, + "loss": 0.0446, + "step": 12333 + }, + { + "epoch": 10.41722972972973, + "grad_norm": 0.33550554513931274, + "learning_rate": 2.9223333333333335e-05, + "loss": 0.0344, + "step": 12334 + }, + { + "epoch": 10.418074324324325, + "grad_norm": 0.9622490406036377, + "learning_rate": 2.9223e-05, + "loss": 0.0435, + "step": 12335 + }, + { + "epoch": 10.41891891891892, + "grad_norm": 0.33309727907180786, + "learning_rate": 2.9222666666666667e-05, + "loss": 0.0406, + "step": 12336 + }, + { + "epoch": 10.419763513513514, + "grad_norm": 0.28604209423065186, + "learning_rate": 2.9222333333333337e-05, + "loss": 0.0156, + "step": 12337 + }, + { + "epoch": 10.420608108108109, + "grad_norm": 0.8154877424240112, + "learning_rate": 2.9222e-05, + "loss": 0.054, + "step": 12338 + }, + { + "epoch": 10.421452702702704, + "grad_norm": 0.614705502986908, + "learning_rate": 2.922166666666667e-05, + "loss": 0.0424, + "step": 12339 + }, + { + "epoch": 10.422297297297296, + "grad_norm": 0.8875071406364441, + "learning_rate": 2.9221333333333334e-05, + "loss": 0.0674, + "step": 12340 + }, + { + "epoch": 10.423141891891891, + "grad_norm": 0.6347576379776001, + "learning_rate": 2.9221e-05, + "loss": 0.2167, + "step": 12341 + }, + { + "epoch": 10.423986486486486, + "grad_norm": 0.5941317081451416, + "learning_rate": 2.9220666666666666e-05, + "loss": 0.1265, + "step": 12342 + }, + { + "epoch": 10.42483108108108, + "grad_norm": 0.5065725445747375, + "learning_rate": 2.9220333333333336e-05, + "loss": 0.0842, + "step": 12343 + }, + { + "epoch": 10.425675675675675, + "grad_norm": 0.4466271996498108, + "learning_rate": 2.922e-05, + "loss": 0.047, + "step": 12344 + }, + { + "epoch": 10.42652027027027, + "grad_norm": 0.3099254369735718, + "learning_rate": 2.9219666666666667e-05, + "loss": 0.0266, + "step": 12345 + }, + { + "epoch": 10.427364864864865, + "grad_norm": 0.3165457546710968, + "learning_rate": 2.9219333333333337e-05, + "loss": 0.0385, + "step": 12346 + }, + { + "epoch": 10.42820945945946, + "grad_norm": 0.46108317375183105, + "learning_rate": 2.9219e-05, + "loss": 0.0295, + "step": 12347 + }, + { + "epoch": 10.429054054054054, + "grad_norm": 0.1973700374364853, + "learning_rate": 2.921866666666667e-05, + "loss": 0.0253, + "step": 12348 + }, + { + "epoch": 10.42989864864865, + "grad_norm": 0.29045993089675903, + "learning_rate": 2.921833333333333e-05, + "loss": 0.0335, + "step": 12349 + }, + { + "epoch": 10.430743243243244, + "grad_norm": 0.4001398980617523, + "learning_rate": 2.9218e-05, + "loss": 0.0315, + "step": 12350 + }, + { + "epoch": 10.431587837837839, + "grad_norm": 0.39834287762641907, + "learning_rate": 2.9217666666666666e-05, + "loss": 0.0366, + "step": 12351 + }, + { + "epoch": 10.432432432432432, + "grad_norm": 0.2932445704936981, + "learning_rate": 2.9217333333333332e-05, + "loss": 0.023, + "step": 12352 + }, + { + "epoch": 10.433277027027026, + "grad_norm": 0.41611453890800476, + "learning_rate": 2.9217e-05, + "loss": 0.0183, + "step": 12353 + }, + { + "epoch": 10.434121621621621, + "grad_norm": 0.4119648337364197, + "learning_rate": 2.9216666666666668e-05, + "loss": 0.0457, + "step": 12354 + }, + { + "epoch": 10.434966216216216, + "grad_norm": 0.34216102957725525, + "learning_rate": 2.9216333333333334e-05, + "loss": 0.0226, + "step": 12355 + }, + { + "epoch": 10.43581081081081, + "grad_norm": 0.4690122604370117, + "learning_rate": 2.9216e-05, + "loss": 0.0232, + "step": 12356 + }, + { + "epoch": 10.436655405405405, + "grad_norm": 0.26480525732040405, + "learning_rate": 2.921566666666667e-05, + "loss": 0.0189, + "step": 12357 + }, + { + "epoch": 10.4375, + "grad_norm": 0.28444531559944153, + "learning_rate": 2.921533333333333e-05, + "loss": 0.0216, + "step": 12358 + }, + { + "epoch": 10.438344594594595, + "grad_norm": 0.2739647626876831, + "learning_rate": 2.9215e-05, + "loss": 0.0173, + "step": 12359 + }, + { + "epoch": 10.43918918918919, + "grad_norm": 0.5039196610450745, + "learning_rate": 2.921466666666667e-05, + "loss": 0.0299, + "step": 12360 + }, + { + "epoch": 10.440033783783784, + "grad_norm": 0.5720238089561462, + "learning_rate": 2.9214333333333333e-05, + "loss": 0.025, + "step": 12361 + }, + { + "epoch": 10.440878378378379, + "grad_norm": 0.24256594479084015, + "learning_rate": 2.9214000000000002e-05, + "loss": 0.0168, + "step": 12362 + }, + { + "epoch": 10.441722972972974, + "grad_norm": 0.5239944458007812, + "learning_rate": 2.9213666666666668e-05, + "loss": 0.0279, + "step": 12363 + }, + { + "epoch": 10.442567567567568, + "grad_norm": 1.0003111362457275, + "learning_rate": 2.9213333333333334e-05, + "loss": 0.0197, + "step": 12364 + }, + { + "epoch": 10.443412162162161, + "grad_norm": 0.2698504626750946, + "learning_rate": 2.9213e-05, + "loss": 0.0245, + "step": 12365 + }, + { + "epoch": 10.444256756756756, + "grad_norm": 0.571602463722229, + "learning_rate": 2.921266666666667e-05, + "loss": 0.0309, + "step": 12366 + }, + { + "epoch": 10.44510135135135, + "grad_norm": 0.4501730501651764, + "learning_rate": 2.921233333333333e-05, + "loss": 0.0225, + "step": 12367 + }, + { + "epoch": 10.445945945945946, + "grad_norm": 0.314355731010437, + "learning_rate": 2.9212e-05, + "loss": 0.026, + "step": 12368 + }, + { + "epoch": 10.44679054054054, + "grad_norm": 0.31399333477020264, + "learning_rate": 2.921166666666667e-05, + "loss": 0.0237, + "step": 12369 + }, + { + "epoch": 10.447635135135135, + "grad_norm": 0.19966533780097961, + "learning_rate": 2.9211333333333333e-05, + "loss": 0.0148, + "step": 12370 + }, + { + "epoch": 10.44847972972973, + "grad_norm": 0.45640116930007935, + "learning_rate": 2.9211000000000002e-05, + "loss": 0.028, + "step": 12371 + }, + { + "epoch": 10.449324324324325, + "grad_norm": 0.45231395959854126, + "learning_rate": 2.9210666666666668e-05, + "loss": 0.0266, + "step": 12372 + }, + { + "epoch": 10.45016891891892, + "grad_norm": 0.3697686791419983, + "learning_rate": 2.9210333333333334e-05, + "loss": 0.0403, + "step": 12373 + }, + { + "epoch": 10.451013513513514, + "grad_norm": 0.5688126087188721, + "learning_rate": 2.921e-05, + "loss": 0.031, + "step": 12374 + }, + { + "epoch": 10.451858108108109, + "grad_norm": 0.6012576222419739, + "learning_rate": 2.920966666666667e-05, + "loss": 0.0276, + "step": 12375 + }, + { + "epoch": 10.452702702702704, + "grad_norm": 0.8666161298751831, + "learning_rate": 2.9209333333333335e-05, + "loss": 0.0171, + "step": 12376 + }, + { + "epoch": 10.453547297297296, + "grad_norm": 0.36579689383506775, + "learning_rate": 2.9209e-05, + "loss": 0.0218, + "step": 12377 + }, + { + "epoch": 10.454391891891891, + "grad_norm": 0.5102797746658325, + "learning_rate": 2.9208666666666667e-05, + "loss": 0.027, + "step": 12378 + }, + { + "epoch": 10.455236486486486, + "grad_norm": 0.2612493634223938, + "learning_rate": 2.9208333333333333e-05, + "loss": 0.012, + "step": 12379 + }, + { + "epoch": 10.45608108108108, + "grad_norm": 0.3519825339317322, + "learning_rate": 2.9208000000000002e-05, + "loss": 0.0275, + "step": 12380 + }, + { + "epoch": 10.456925675675675, + "grad_norm": 0.34935593605041504, + "learning_rate": 2.9207666666666665e-05, + "loss": 0.0227, + "step": 12381 + }, + { + "epoch": 10.45777027027027, + "grad_norm": 0.5663378238677979, + "learning_rate": 2.9207333333333334e-05, + "loss": 0.0241, + "step": 12382 + }, + { + "epoch": 10.458614864864865, + "grad_norm": 0.40630871057510376, + "learning_rate": 2.9207e-05, + "loss": 0.0287, + "step": 12383 + }, + { + "epoch": 10.45945945945946, + "grad_norm": 0.4497697353363037, + "learning_rate": 2.9206666666666666e-05, + "loss": 0.0458, + "step": 12384 + }, + { + "epoch": 10.460304054054054, + "grad_norm": 0.38914740085601807, + "learning_rate": 2.9206333333333335e-05, + "loss": 0.031, + "step": 12385 + }, + { + "epoch": 10.46114864864865, + "grad_norm": 0.4216226041316986, + "learning_rate": 2.9206e-05, + "loss": 0.0253, + "step": 12386 + }, + { + "epoch": 10.461993243243244, + "grad_norm": 0.547930121421814, + "learning_rate": 2.9205666666666667e-05, + "loss": 0.0362, + "step": 12387 + }, + { + "epoch": 10.462837837837839, + "grad_norm": 0.6220450401306152, + "learning_rate": 2.9205333333333333e-05, + "loss": 0.0205, + "step": 12388 + }, + { + "epoch": 10.463682432432432, + "grad_norm": 0.6675661206245422, + "learning_rate": 2.9205000000000002e-05, + "loss": 0.0402, + "step": 12389 + }, + { + "epoch": 10.464527027027026, + "grad_norm": 0.9751723408699036, + "learning_rate": 2.9204666666666665e-05, + "loss": 0.1025, + "step": 12390 + }, + { + "epoch": 10.465371621621621, + "grad_norm": 0.6192664504051208, + "learning_rate": 2.9204333333333334e-05, + "loss": 0.2325, + "step": 12391 + }, + { + "epoch": 10.466216216216216, + "grad_norm": 0.5706810355186462, + "learning_rate": 2.9204e-05, + "loss": 0.1475, + "step": 12392 + }, + { + "epoch": 10.46706081081081, + "grad_norm": 0.40981560945510864, + "learning_rate": 2.9203666666666666e-05, + "loss": 0.0661, + "step": 12393 + }, + { + "epoch": 10.467905405405405, + "grad_norm": 0.4587197005748749, + "learning_rate": 2.9203333333333336e-05, + "loss": 0.065, + "step": 12394 + }, + { + "epoch": 10.46875, + "grad_norm": 0.3302710950374603, + "learning_rate": 2.9203e-05, + "loss": 0.0323, + "step": 12395 + }, + { + "epoch": 10.469594594594595, + "grad_norm": 0.35095420479774475, + "learning_rate": 2.9202666666666667e-05, + "loss": 0.0292, + "step": 12396 + }, + { + "epoch": 10.47043918918919, + "grad_norm": 0.37354838848114014, + "learning_rate": 2.9202333333333333e-05, + "loss": 0.0344, + "step": 12397 + }, + { + "epoch": 10.471283783783784, + "grad_norm": 0.6092918515205383, + "learning_rate": 2.9202000000000003e-05, + "loss": 0.0319, + "step": 12398 + }, + { + "epoch": 10.472128378378379, + "grad_norm": 0.3917221426963806, + "learning_rate": 2.9201666666666665e-05, + "loss": 0.0298, + "step": 12399 + }, + { + "epoch": 10.472972972972974, + "grad_norm": 0.35576754808425903, + "learning_rate": 2.9201333333333335e-05, + "loss": 0.0357, + "step": 12400 + }, + { + "epoch": 10.473817567567568, + "grad_norm": 0.26636603474617004, + "learning_rate": 2.9201e-05, + "loss": 0.0236, + "step": 12401 + }, + { + "epoch": 10.474662162162161, + "grad_norm": 0.35455265641212463, + "learning_rate": 2.9200666666666666e-05, + "loss": 0.0409, + "step": 12402 + }, + { + "epoch": 10.475506756756756, + "grad_norm": 0.3121139705181122, + "learning_rate": 2.9200333333333336e-05, + "loss": 0.0333, + "step": 12403 + }, + { + "epoch": 10.47635135135135, + "grad_norm": 0.2441842257976532, + "learning_rate": 2.92e-05, + "loss": 0.0309, + "step": 12404 + }, + { + "epoch": 10.477195945945946, + "grad_norm": 0.43959909677505493, + "learning_rate": 2.9199666666666668e-05, + "loss": 0.0331, + "step": 12405 + }, + { + "epoch": 10.47804054054054, + "grad_norm": 0.6053375601768494, + "learning_rate": 2.9199333333333334e-05, + "loss": 0.03, + "step": 12406 + }, + { + "epoch": 10.478885135135135, + "grad_norm": 0.6757490634918213, + "learning_rate": 2.9199000000000003e-05, + "loss": 0.0236, + "step": 12407 + }, + { + "epoch": 10.47972972972973, + "grad_norm": 0.34523406624794006, + "learning_rate": 2.9198666666666665e-05, + "loss": 0.0388, + "step": 12408 + }, + { + "epoch": 10.480574324324325, + "grad_norm": 0.49294576048851013, + "learning_rate": 2.9198333333333335e-05, + "loss": 0.0282, + "step": 12409 + }, + { + "epoch": 10.48141891891892, + "grad_norm": 0.18472221493721008, + "learning_rate": 2.9198e-05, + "loss": 0.0148, + "step": 12410 + }, + { + "epoch": 10.482263513513514, + "grad_norm": 0.34946393966674805, + "learning_rate": 2.9197666666666667e-05, + "loss": 0.0355, + "step": 12411 + }, + { + "epoch": 10.483108108108109, + "grad_norm": 0.42689797282218933, + "learning_rate": 2.9197333333333336e-05, + "loss": 0.0265, + "step": 12412 + }, + { + "epoch": 10.483952702702704, + "grad_norm": 0.6001972556114197, + "learning_rate": 2.9197e-05, + "loss": 0.0243, + "step": 12413 + }, + { + "epoch": 10.484797297297296, + "grad_norm": 0.6725383400917053, + "learning_rate": 2.9196666666666668e-05, + "loss": 0.0355, + "step": 12414 + }, + { + "epoch": 10.485641891891891, + "grad_norm": 0.3140304982662201, + "learning_rate": 2.9196333333333334e-05, + "loss": 0.027, + "step": 12415 + }, + { + "epoch": 10.486486486486486, + "grad_norm": 0.4509124159812927, + "learning_rate": 2.9196e-05, + "loss": 0.0262, + "step": 12416 + }, + { + "epoch": 10.48733108108108, + "grad_norm": 0.2286706119775772, + "learning_rate": 2.9195666666666666e-05, + "loss": 0.0163, + "step": 12417 + }, + { + "epoch": 10.488175675675675, + "grad_norm": 0.650479257106781, + "learning_rate": 2.9195333333333335e-05, + "loss": 0.0252, + "step": 12418 + }, + { + "epoch": 10.48902027027027, + "grad_norm": 0.33236202597618103, + "learning_rate": 2.9195e-05, + "loss": 0.0294, + "step": 12419 + }, + { + "epoch": 10.489864864864865, + "grad_norm": 0.3375228941440582, + "learning_rate": 2.9194666666666667e-05, + "loss": 0.0266, + "step": 12420 + }, + { + "epoch": 10.49070945945946, + "grad_norm": 0.3180883228778839, + "learning_rate": 2.9194333333333336e-05, + "loss": 0.0283, + "step": 12421 + }, + { + "epoch": 10.491554054054054, + "grad_norm": 0.5238485932350159, + "learning_rate": 2.9194e-05, + "loss": 0.0353, + "step": 12422 + }, + { + "epoch": 10.49239864864865, + "grad_norm": 0.36607810854911804, + "learning_rate": 2.9193666666666668e-05, + "loss": 0.038, + "step": 12423 + }, + { + "epoch": 10.493243243243244, + "grad_norm": 0.30456745624542236, + "learning_rate": 2.9193333333333334e-05, + "loss": 0.0209, + "step": 12424 + }, + { + "epoch": 10.494087837837839, + "grad_norm": 0.5028946399688721, + "learning_rate": 2.9193e-05, + "loss": 0.022, + "step": 12425 + }, + { + "epoch": 10.494932432432432, + "grad_norm": 0.7062841057777405, + "learning_rate": 2.919266666666667e-05, + "loss": 0.0233, + "step": 12426 + }, + { + "epoch": 10.495777027027026, + "grad_norm": 0.48140859603881836, + "learning_rate": 2.9192333333333335e-05, + "loss": 0.023, + "step": 12427 + }, + { + "epoch": 10.496621621621621, + "grad_norm": 0.3524928390979767, + "learning_rate": 2.9192e-05, + "loss": 0.0199, + "step": 12428 + }, + { + "epoch": 10.497466216216216, + "grad_norm": 0.5206747651100159, + "learning_rate": 2.9191666666666667e-05, + "loss": 0.0232, + "step": 12429 + }, + { + "epoch": 10.49831081081081, + "grad_norm": 0.4041655659675598, + "learning_rate": 2.9191333333333336e-05, + "loss": 0.0405, + "step": 12430 + }, + { + "epoch": 10.499155405405405, + "grad_norm": 0.27803120017051697, + "learning_rate": 2.9191e-05, + "loss": 0.023, + "step": 12431 + }, + { + "epoch": 10.5, + "grad_norm": 0.5196408629417419, + "learning_rate": 2.9190666666666668e-05, + "loss": 0.0257, + "step": 12432 + }, + { + "epoch": 10.500844594594595, + "grad_norm": 0.3235853612422943, + "learning_rate": 2.9190333333333334e-05, + "loss": 0.0345, + "step": 12433 + }, + { + "epoch": 10.50168918918919, + "grad_norm": 0.2584573030471802, + "learning_rate": 2.919e-05, + "loss": 0.0246, + "step": 12434 + }, + { + "epoch": 10.502533783783784, + "grad_norm": 0.35490453243255615, + "learning_rate": 2.918966666666667e-05, + "loss": 0.0211, + "step": 12435 + }, + { + "epoch": 10.503378378378379, + "grad_norm": 0.5540328621864319, + "learning_rate": 2.9189333333333335e-05, + "loss": 0.0405, + "step": 12436 + }, + { + "epoch": 10.504222972972974, + "grad_norm": 0.5631927847862244, + "learning_rate": 2.9189e-05, + "loss": 0.0287, + "step": 12437 + }, + { + "epoch": 10.505067567567568, + "grad_norm": 0.3887176215648651, + "learning_rate": 2.9188666666666667e-05, + "loss": 0.0226, + "step": 12438 + }, + { + "epoch": 10.505912162162161, + "grad_norm": 0.5449908375740051, + "learning_rate": 2.9188333333333333e-05, + "loss": 0.0474, + "step": 12439 + }, + { + "epoch": 10.506756756756756, + "grad_norm": 1.0505112409591675, + "learning_rate": 2.9188e-05, + "loss": 0.066, + "step": 12440 + }, + { + "epoch": 10.50760135135135, + "grad_norm": 0.8857782483100891, + "learning_rate": 2.918766666666667e-05, + "loss": 0.2515, + "step": 12441 + }, + { + "epoch": 10.508445945945946, + "grad_norm": 0.5107197761535645, + "learning_rate": 2.918733333333333e-05, + "loss": 0.1575, + "step": 12442 + }, + { + "epoch": 10.50929054054054, + "grad_norm": 0.4337444007396698, + "learning_rate": 2.9187e-05, + "loss": 0.0746, + "step": 12443 + }, + { + "epoch": 10.510135135135135, + "grad_norm": 0.3394027650356293, + "learning_rate": 2.918666666666667e-05, + "loss": 0.0427, + "step": 12444 + }, + { + "epoch": 10.51097972972973, + "grad_norm": 0.39434272050857544, + "learning_rate": 2.9186333333333332e-05, + "loss": 0.0404, + "step": 12445 + }, + { + "epoch": 10.511824324324325, + "grad_norm": 0.3200964033603668, + "learning_rate": 2.9186e-05, + "loss": 0.0367, + "step": 12446 + }, + { + "epoch": 10.51266891891892, + "grad_norm": 0.2708476185798645, + "learning_rate": 2.9185666666666667e-05, + "loss": 0.0288, + "step": 12447 + }, + { + "epoch": 10.513513513513514, + "grad_norm": 0.3715165853500366, + "learning_rate": 2.9185333333333333e-05, + "loss": 0.0342, + "step": 12448 + }, + { + "epoch": 10.514358108108109, + "grad_norm": 0.3358181118965149, + "learning_rate": 2.9185e-05, + "loss": 0.0329, + "step": 12449 + }, + { + "epoch": 10.515202702702704, + "grad_norm": 0.27996551990509033, + "learning_rate": 2.918466666666667e-05, + "loss": 0.0257, + "step": 12450 + }, + { + "epoch": 10.516047297297296, + "grad_norm": 0.37230026721954346, + "learning_rate": 2.9184333333333335e-05, + "loss": 0.0571, + "step": 12451 + }, + { + "epoch": 10.516891891891891, + "grad_norm": 0.3636334538459778, + "learning_rate": 2.9184e-05, + "loss": 0.0226, + "step": 12452 + }, + { + "epoch": 10.517736486486486, + "grad_norm": 0.18670327961444855, + "learning_rate": 2.918366666666667e-05, + "loss": 0.0207, + "step": 12453 + }, + { + "epoch": 10.51858108108108, + "grad_norm": 0.3424517512321472, + "learning_rate": 2.9183333333333332e-05, + "loss": 0.0247, + "step": 12454 + }, + { + "epoch": 10.519425675675675, + "grad_norm": 0.3963881731033325, + "learning_rate": 2.9183e-05, + "loss": 0.0621, + "step": 12455 + }, + { + "epoch": 10.52027027027027, + "grad_norm": 0.6052885055541992, + "learning_rate": 2.9182666666666668e-05, + "loss": 0.0183, + "step": 12456 + }, + { + "epoch": 10.521114864864865, + "grad_norm": 0.3519333600997925, + "learning_rate": 2.9182333333333334e-05, + "loss": 0.0201, + "step": 12457 + }, + { + "epoch": 10.52195945945946, + "grad_norm": 0.3974546492099762, + "learning_rate": 2.9182e-05, + "loss": 0.0233, + "step": 12458 + }, + { + "epoch": 10.522804054054054, + "grad_norm": 0.26514798402786255, + "learning_rate": 2.918166666666667e-05, + "loss": 0.0208, + "step": 12459 + }, + { + "epoch": 10.52364864864865, + "grad_norm": 0.3957194685935974, + "learning_rate": 2.9181333333333335e-05, + "loss": 0.0469, + "step": 12460 + }, + { + "epoch": 10.524493243243244, + "grad_norm": 0.35952460765838623, + "learning_rate": 2.9181e-05, + "loss": 0.0261, + "step": 12461 + }, + { + "epoch": 10.525337837837839, + "grad_norm": 0.34886711835861206, + "learning_rate": 2.918066666666667e-05, + "loss": 0.0344, + "step": 12462 + }, + { + "epoch": 10.526182432432432, + "grad_norm": 0.3172694444656372, + "learning_rate": 2.9180333333333333e-05, + "loss": 0.0223, + "step": 12463 + }, + { + "epoch": 10.527027027027026, + "grad_norm": 0.4673287868499756, + "learning_rate": 2.9180000000000002e-05, + "loss": 0.035, + "step": 12464 + }, + { + "epoch": 10.527871621621621, + "grad_norm": 0.22729679942131042, + "learning_rate": 2.9179666666666668e-05, + "loss": 0.0331, + "step": 12465 + }, + { + "epoch": 10.528716216216216, + "grad_norm": 0.584180474281311, + "learning_rate": 2.9179333333333334e-05, + "loss": 0.0188, + "step": 12466 + }, + { + "epoch": 10.52956081081081, + "grad_norm": 0.43235692381858826, + "learning_rate": 2.9179e-05, + "loss": 0.0389, + "step": 12467 + }, + { + "epoch": 10.530405405405405, + "grad_norm": 0.2636573314666748, + "learning_rate": 2.917866666666667e-05, + "loss": 0.0153, + "step": 12468 + }, + { + "epoch": 10.53125, + "grad_norm": 0.22104942798614502, + "learning_rate": 2.9178333333333335e-05, + "loss": 0.0219, + "step": 12469 + }, + { + "epoch": 10.532094594594595, + "grad_norm": 0.3466954529285431, + "learning_rate": 2.9178e-05, + "loss": 0.0309, + "step": 12470 + }, + { + "epoch": 10.53293918918919, + "grad_norm": 0.4725640118122101, + "learning_rate": 2.9177666666666667e-05, + "loss": 0.0404, + "step": 12471 + }, + { + "epoch": 10.533783783783784, + "grad_norm": 0.41078823804855347, + "learning_rate": 2.9177333333333333e-05, + "loss": 0.0248, + "step": 12472 + }, + { + "epoch": 10.534628378378379, + "grad_norm": 0.29325681924819946, + "learning_rate": 2.9177000000000002e-05, + "loss": 0.0153, + "step": 12473 + }, + { + "epoch": 10.535472972972974, + "grad_norm": 0.541157066822052, + "learning_rate": 2.9176666666666665e-05, + "loss": 0.0296, + "step": 12474 + }, + { + "epoch": 10.536317567567568, + "grad_norm": 0.47935423254966736, + "learning_rate": 2.9176333333333334e-05, + "loss": 0.0253, + "step": 12475 + }, + { + "epoch": 10.537162162162161, + "grad_norm": 0.6662607192993164, + "learning_rate": 2.9176000000000003e-05, + "loss": 0.0451, + "step": 12476 + }, + { + "epoch": 10.538006756756756, + "grad_norm": 0.5698843598365784, + "learning_rate": 2.9175666666666666e-05, + "loss": 0.0304, + "step": 12477 + }, + { + "epoch": 10.53885135135135, + "grad_norm": 0.35773298144340515, + "learning_rate": 2.9175333333333335e-05, + "loss": 0.0301, + "step": 12478 + }, + { + "epoch": 10.539695945945946, + "grad_norm": 0.26782891154289246, + "learning_rate": 2.9175e-05, + "loss": 0.0253, + "step": 12479 + }, + { + "epoch": 10.54054054054054, + "grad_norm": 0.9941923022270203, + "learning_rate": 2.9174666666666667e-05, + "loss": 0.0202, + "step": 12480 + }, + { + "epoch": 10.541385135135135, + "grad_norm": 0.4838036894798279, + "learning_rate": 2.9174333333333333e-05, + "loss": 0.0348, + "step": 12481 + }, + { + "epoch": 10.54222972972973, + "grad_norm": 0.45727819204330444, + "learning_rate": 2.9174000000000002e-05, + "loss": 0.0229, + "step": 12482 + }, + { + "epoch": 10.543074324324325, + "grad_norm": 0.5283264517784119, + "learning_rate": 2.9173666666666665e-05, + "loss": 0.0326, + "step": 12483 + }, + { + "epoch": 10.54391891891892, + "grad_norm": 0.44751545786857605, + "learning_rate": 2.9173333333333334e-05, + "loss": 0.0279, + "step": 12484 + }, + { + "epoch": 10.544763513513514, + "grad_norm": 0.5441763997077942, + "learning_rate": 2.9173000000000003e-05, + "loss": 0.0307, + "step": 12485 + }, + { + "epoch": 10.545608108108109, + "grad_norm": 0.6512041091918945, + "learning_rate": 2.9172666666666666e-05, + "loss": 0.0516, + "step": 12486 + }, + { + "epoch": 10.546452702702704, + "grad_norm": 0.6228252649307251, + "learning_rate": 2.9172333333333335e-05, + "loss": 0.0319, + "step": 12487 + }, + { + "epoch": 10.547297297297296, + "grad_norm": 0.43884962797164917, + "learning_rate": 2.9172e-05, + "loss": 0.0319, + "step": 12488 + }, + { + "epoch": 10.548141891891891, + "grad_norm": 0.3756852149963379, + "learning_rate": 2.9171666666666667e-05, + "loss": 0.0236, + "step": 12489 + }, + { + "epoch": 10.548986486486486, + "grad_norm": 0.9141204953193665, + "learning_rate": 2.9171333333333333e-05, + "loss": 0.0413, + "step": 12490 + }, + { + "epoch": 10.54983108108108, + "grad_norm": 0.8481151461601257, + "learning_rate": 2.9171000000000002e-05, + "loss": 0.292, + "step": 12491 + }, + { + "epoch": 10.550675675675675, + "grad_norm": 0.505742609500885, + "learning_rate": 2.9170666666666665e-05, + "loss": 0.1735, + "step": 12492 + }, + { + "epoch": 10.55152027027027, + "grad_norm": 0.5539576411247253, + "learning_rate": 2.9170333333333334e-05, + "loss": 0.1001, + "step": 12493 + }, + { + "epoch": 10.552364864864865, + "grad_norm": 0.2226545661687851, + "learning_rate": 2.9170000000000004e-05, + "loss": 0.0448, + "step": 12494 + }, + { + "epoch": 10.55320945945946, + "grad_norm": 0.2899037301540375, + "learning_rate": 2.9169666666666666e-05, + "loss": 0.0442, + "step": 12495 + }, + { + "epoch": 10.554054054054054, + "grad_norm": 0.24327483773231506, + "learning_rate": 2.9169333333333335e-05, + "loss": 0.0342, + "step": 12496 + }, + { + "epoch": 10.55489864864865, + "grad_norm": 0.39524975419044495, + "learning_rate": 2.9169e-05, + "loss": 0.028, + "step": 12497 + }, + { + "epoch": 10.555743243243244, + "grad_norm": 0.30760321021080017, + "learning_rate": 2.9168666666666667e-05, + "loss": 0.0218, + "step": 12498 + }, + { + "epoch": 10.556587837837839, + "grad_norm": 1.3366340398788452, + "learning_rate": 2.9168333333333333e-05, + "loss": 0.0333, + "step": 12499 + }, + { + "epoch": 10.557432432432432, + "grad_norm": 0.48808711767196655, + "learning_rate": 2.9168e-05, + "loss": 0.0305, + "step": 12500 + }, + { + "epoch": 10.558277027027026, + "grad_norm": 0.3041754364967346, + "learning_rate": 2.916766666666667e-05, + "loss": 0.0266, + "step": 12501 + }, + { + "epoch": 10.559121621621621, + "grad_norm": 0.2826143801212311, + "learning_rate": 2.9167333333333334e-05, + "loss": 0.0304, + "step": 12502 + }, + { + "epoch": 10.559966216216216, + "grad_norm": 0.3483152389526367, + "learning_rate": 2.9167e-05, + "loss": 0.0323, + "step": 12503 + }, + { + "epoch": 10.56081081081081, + "grad_norm": 0.26811328530311584, + "learning_rate": 2.9166666666666666e-05, + "loss": 0.0254, + "step": 12504 + }, + { + "epoch": 10.561655405405405, + "grad_norm": 0.2358369529247284, + "learning_rate": 2.9166333333333336e-05, + "loss": 0.0214, + "step": 12505 + }, + { + "epoch": 10.5625, + "grad_norm": 0.2830599844455719, + "learning_rate": 2.9165999999999998e-05, + "loss": 0.05, + "step": 12506 + }, + { + "epoch": 10.563344594594595, + "grad_norm": 0.2379494309425354, + "learning_rate": 2.9165666666666668e-05, + "loss": 0.0238, + "step": 12507 + }, + { + "epoch": 10.56418918918919, + "grad_norm": 0.3904242515563965, + "learning_rate": 2.9165333333333334e-05, + "loss": 0.0294, + "step": 12508 + }, + { + "epoch": 10.565033783783784, + "grad_norm": 0.2926868796348572, + "learning_rate": 2.9165e-05, + "loss": 0.035, + "step": 12509 + }, + { + "epoch": 10.565878378378379, + "grad_norm": 0.30068710446357727, + "learning_rate": 2.916466666666667e-05, + "loss": 0.0277, + "step": 12510 + }, + { + "epoch": 10.566722972972974, + "grad_norm": 0.3413609564304352, + "learning_rate": 2.9164333333333335e-05, + "loss": 0.0241, + "step": 12511 + }, + { + "epoch": 10.567567567567568, + "grad_norm": 0.2893957793712616, + "learning_rate": 2.9164e-05, + "loss": 0.0239, + "step": 12512 + }, + { + "epoch": 10.568412162162161, + "grad_norm": 0.5643206238746643, + "learning_rate": 2.9163666666666667e-05, + "loss": 0.0258, + "step": 12513 + }, + { + "epoch": 10.569256756756756, + "grad_norm": 0.262525737285614, + "learning_rate": 2.9163333333333336e-05, + "loss": 0.0187, + "step": 12514 + }, + { + "epoch": 10.57010135135135, + "grad_norm": 0.25825944542884827, + "learning_rate": 2.9163e-05, + "loss": 0.0218, + "step": 12515 + }, + { + "epoch": 10.570945945945946, + "grad_norm": 0.2420961707830429, + "learning_rate": 2.9162666666666668e-05, + "loss": 0.0161, + "step": 12516 + }, + { + "epoch": 10.57179054054054, + "grad_norm": 0.2885952889919281, + "learning_rate": 2.9162333333333334e-05, + "loss": 0.0219, + "step": 12517 + }, + { + "epoch": 10.572635135135135, + "grad_norm": 0.3522258996963501, + "learning_rate": 2.9162e-05, + "loss": 0.0277, + "step": 12518 + }, + { + "epoch": 10.57347972972973, + "grad_norm": 0.46111440658569336, + "learning_rate": 2.916166666666667e-05, + "loss": 0.0325, + "step": 12519 + }, + { + "epoch": 10.574324324324325, + "grad_norm": 0.42702269554138184, + "learning_rate": 2.9161333333333335e-05, + "loss": 0.0341, + "step": 12520 + }, + { + "epoch": 10.57516891891892, + "grad_norm": 0.6464988589286804, + "learning_rate": 2.9161e-05, + "loss": 0.0414, + "step": 12521 + }, + { + "epoch": 10.576013513513514, + "grad_norm": 0.29178643226623535, + "learning_rate": 2.9160666666666667e-05, + "loss": 0.0291, + "step": 12522 + }, + { + "epoch": 10.576858108108109, + "grad_norm": 0.30884110927581787, + "learning_rate": 2.9160333333333336e-05, + "loss": 0.0298, + "step": 12523 + }, + { + "epoch": 10.577702702702704, + "grad_norm": 0.42287570238113403, + "learning_rate": 2.916e-05, + "loss": 0.0386, + "step": 12524 + }, + { + "epoch": 10.578547297297296, + "grad_norm": 0.7902531623840332, + "learning_rate": 2.9159666666666668e-05, + "loss": 0.0226, + "step": 12525 + }, + { + "epoch": 10.579391891891891, + "grad_norm": 0.3440299332141876, + "learning_rate": 2.9159333333333337e-05, + "loss": 0.018, + "step": 12526 + }, + { + "epoch": 10.580236486486486, + "grad_norm": 0.28478556871414185, + "learning_rate": 2.9159e-05, + "loss": 0.0246, + "step": 12527 + }, + { + "epoch": 10.58108108108108, + "grad_norm": 0.38284558057785034, + "learning_rate": 2.915866666666667e-05, + "loss": 0.0304, + "step": 12528 + }, + { + "epoch": 10.581925675675675, + "grad_norm": 0.6338142156600952, + "learning_rate": 2.9158333333333335e-05, + "loss": 0.0393, + "step": 12529 + }, + { + "epoch": 10.58277027027027, + "grad_norm": 0.439596563577652, + "learning_rate": 2.9158e-05, + "loss": 0.0261, + "step": 12530 + }, + { + "epoch": 10.583614864864865, + "grad_norm": 0.3038300573825836, + "learning_rate": 2.9157666666666667e-05, + "loss": 0.0315, + "step": 12531 + }, + { + "epoch": 10.58445945945946, + "grad_norm": 0.23306195437908173, + "learning_rate": 2.9157333333333333e-05, + "loss": 0.0137, + "step": 12532 + }, + { + "epoch": 10.585304054054054, + "grad_norm": 0.6435199975967407, + "learning_rate": 2.9157e-05, + "loss": 0.0297, + "step": 12533 + }, + { + "epoch": 10.58614864864865, + "grad_norm": 0.3098912239074707, + "learning_rate": 2.9156666666666668e-05, + "loss": 0.0301, + "step": 12534 + }, + { + "epoch": 10.586993243243244, + "grad_norm": 0.565555989742279, + "learning_rate": 2.9156333333333334e-05, + "loss": 0.0384, + "step": 12535 + }, + { + "epoch": 10.587837837837839, + "grad_norm": 0.5286461710929871, + "learning_rate": 2.9156e-05, + "loss": 0.0439, + "step": 12536 + }, + { + "epoch": 10.588682432432432, + "grad_norm": 0.3040394186973572, + "learning_rate": 2.915566666666667e-05, + "loss": 0.0175, + "step": 12537 + }, + { + "epoch": 10.589527027027026, + "grad_norm": 0.5591830611228943, + "learning_rate": 2.9155333333333332e-05, + "loss": 0.0258, + "step": 12538 + }, + { + "epoch": 10.590371621621621, + "grad_norm": 0.5415881872177124, + "learning_rate": 2.9155e-05, + "loss": 0.0334, + "step": 12539 + }, + { + "epoch": 10.591216216216216, + "grad_norm": 0.8984895348548889, + "learning_rate": 2.9154666666666667e-05, + "loss": 0.0584, + "step": 12540 + }, + { + "epoch": 10.59206081081081, + "grad_norm": 0.6091243028640747, + "learning_rate": 2.9154333333333333e-05, + "loss": 0.2545, + "step": 12541 + }, + { + "epoch": 10.592905405405405, + "grad_norm": 0.6307113766670227, + "learning_rate": 2.9154e-05, + "loss": 0.144, + "step": 12542 + }, + { + "epoch": 10.59375, + "grad_norm": 0.30879107117652893, + "learning_rate": 2.915366666666667e-05, + "loss": 0.0597, + "step": 12543 + }, + { + "epoch": 10.594594594594595, + "grad_norm": 0.5105152726173401, + "learning_rate": 2.9153333333333334e-05, + "loss": 0.0398, + "step": 12544 + }, + { + "epoch": 10.59543918918919, + "grad_norm": 0.5527247190475464, + "learning_rate": 2.9153e-05, + "loss": 0.07, + "step": 12545 + }, + { + "epoch": 10.596283783783784, + "grad_norm": 0.2259741872549057, + "learning_rate": 2.915266666666667e-05, + "loss": 0.0349, + "step": 12546 + }, + { + "epoch": 10.597128378378379, + "grad_norm": 0.2679794728755951, + "learning_rate": 2.9152333333333332e-05, + "loss": 0.0371, + "step": 12547 + }, + { + "epoch": 10.597972972972974, + "grad_norm": 0.6378389000892639, + "learning_rate": 2.9152e-05, + "loss": 0.0333, + "step": 12548 + }, + { + "epoch": 10.598817567567568, + "grad_norm": 0.22061532735824585, + "learning_rate": 2.9151666666666667e-05, + "loss": 0.0202, + "step": 12549 + }, + { + "epoch": 10.599662162162161, + "grad_norm": 0.26438677310943604, + "learning_rate": 2.9151333333333333e-05, + "loss": 0.0319, + "step": 12550 + }, + { + "epoch": 10.600506756756756, + "grad_norm": 0.24983233213424683, + "learning_rate": 2.9151000000000003e-05, + "loss": 0.0242, + "step": 12551 + }, + { + "epoch": 10.60135135135135, + "grad_norm": 0.24751132726669312, + "learning_rate": 2.915066666666667e-05, + "loss": 0.027, + "step": 12552 + }, + { + "epoch": 10.602195945945946, + "grad_norm": 0.3840431272983551, + "learning_rate": 2.9150333333333334e-05, + "loss": 0.0335, + "step": 12553 + }, + { + "epoch": 10.60304054054054, + "grad_norm": 0.38071101903915405, + "learning_rate": 2.915e-05, + "loss": 0.0323, + "step": 12554 + }, + { + "epoch": 10.603885135135135, + "grad_norm": 0.48300468921661377, + "learning_rate": 2.914966666666667e-05, + "loss": 0.0412, + "step": 12555 + }, + { + "epoch": 10.60472972972973, + "grad_norm": 0.22869260609149933, + "learning_rate": 2.9149333333333332e-05, + "loss": 0.0154, + "step": 12556 + }, + { + "epoch": 10.605574324324325, + "grad_norm": 0.4303664565086365, + "learning_rate": 2.9149e-05, + "loss": 0.0344, + "step": 12557 + }, + { + "epoch": 10.60641891891892, + "grad_norm": 0.1761706918478012, + "learning_rate": 2.9148666666666668e-05, + "loss": 0.0158, + "step": 12558 + }, + { + "epoch": 10.607263513513514, + "grad_norm": 0.24030227959156036, + "learning_rate": 2.9148333333333333e-05, + "loss": 0.0261, + "step": 12559 + }, + { + "epoch": 10.608108108108109, + "grad_norm": 0.21278473734855652, + "learning_rate": 2.9148000000000003e-05, + "loss": 0.0259, + "step": 12560 + }, + { + "epoch": 10.608952702702704, + "grad_norm": 0.29136186838150024, + "learning_rate": 2.9147666666666665e-05, + "loss": 0.0219, + "step": 12561 + }, + { + "epoch": 10.609797297297296, + "grad_norm": 0.40854698419570923, + "learning_rate": 2.9147333333333335e-05, + "loss": 0.0399, + "step": 12562 + }, + { + "epoch": 10.610641891891891, + "grad_norm": 0.2724388539791107, + "learning_rate": 2.9147e-05, + "loss": 0.0278, + "step": 12563 + }, + { + "epoch": 10.611486486486486, + "grad_norm": 0.38178086280822754, + "learning_rate": 2.9146666666666667e-05, + "loss": 0.0364, + "step": 12564 + }, + { + "epoch": 10.61233108108108, + "grad_norm": 0.5322302579879761, + "learning_rate": 2.9146333333333332e-05, + "loss": 0.0271, + "step": 12565 + }, + { + "epoch": 10.613175675675675, + "grad_norm": 0.2570520341396332, + "learning_rate": 2.9146000000000002e-05, + "loss": 0.0185, + "step": 12566 + }, + { + "epoch": 10.61402027027027, + "grad_norm": 0.23592104017734528, + "learning_rate": 2.9145666666666664e-05, + "loss": 0.014, + "step": 12567 + }, + { + "epoch": 10.614864864864865, + "grad_norm": 0.4172877073287964, + "learning_rate": 2.9145333333333334e-05, + "loss": 0.0232, + "step": 12568 + }, + { + "epoch": 10.61570945945946, + "grad_norm": 0.4188760817050934, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0327, + "step": 12569 + }, + { + "epoch": 10.616554054054054, + "grad_norm": 0.20228546857833862, + "learning_rate": 2.9144666666666666e-05, + "loss": 0.0278, + "step": 12570 + }, + { + "epoch": 10.61739864864865, + "grad_norm": 0.5428280830383301, + "learning_rate": 2.9144333333333335e-05, + "loss": 0.0408, + "step": 12571 + }, + { + "epoch": 10.618243243243244, + "grad_norm": 0.17419563233852386, + "learning_rate": 2.9144e-05, + "loss": 0.014, + "step": 12572 + }, + { + "epoch": 10.619087837837839, + "grad_norm": 0.46452605724334717, + "learning_rate": 2.9143666666666667e-05, + "loss": 0.0324, + "step": 12573 + }, + { + "epoch": 10.619932432432432, + "grad_norm": 0.5079410672187805, + "learning_rate": 2.9143333333333333e-05, + "loss": 0.0266, + "step": 12574 + }, + { + "epoch": 10.620777027027026, + "grad_norm": 0.3822384774684906, + "learning_rate": 2.9143000000000002e-05, + "loss": 0.039, + "step": 12575 + }, + { + "epoch": 10.621621621621621, + "grad_norm": 0.2885979115962982, + "learning_rate": 2.9142666666666668e-05, + "loss": 0.0209, + "step": 12576 + }, + { + "epoch": 10.622466216216216, + "grad_norm": 0.4038766026496887, + "learning_rate": 2.9142333333333334e-05, + "loss": 0.0319, + "step": 12577 + }, + { + "epoch": 10.62331081081081, + "grad_norm": 0.30678948760032654, + "learning_rate": 2.9142000000000003e-05, + "loss": 0.0297, + "step": 12578 + }, + { + "epoch": 10.624155405405405, + "grad_norm": 0.2588624656200409, + "learning_rate": 2.9141666666666666e-05, + "loss": 0.0176, + "step": 12579 + }, + { + "epoch": 10.625, + "grad_norm": 0.4605541229248047, + "learning_rate": 2.9141333333333335e-05, + "loss": 0.0266, + "step": 12580 + }, + { + "epoch": 10.625844594594595, + "grad_norm": 0.35416966676712036, + "learning_rate": 2.9141e-05, + "loss": 0.0261, + "step": 12581 + }, + { + "epoch": 10.62668918918919, + "grad_norm": 0.4608687460422516, + "learning_rate": 2.9140666666666667e-05, + "loss": 0.0244, + "step": 12582 + }, + { + "epoch": 10.627533783783784, + "grad_norm": 0.2870256304740906, + "learning_rate": 2.9140333333333333e-05, + "loss": 0.0308, + "step": 12583 + }, + { + "epoch": 10.628378378378379, + "grad_norm": 0.26740583777427673, + "learning_rate": 2.9140000000000002e-05, + "loss": 0.015, + "step": 12584 + }, + { + "epoch": 10.629222972972974, + "grad_norm": 0.528544545173645, + "learning_rate": 2.9139666666666668e-05, + "loss": 0.0604, + "step": 12585 + }, + { + "epoch": 10.630067567567568, + "grad_norm": 0.7736045718193054, + "learning_rate": 2.9139333333333334e-05, + "loss": 0.0397, + "step": 12586 + }, + { + "epoch": 10.630912162162161, + "grad_norm": 0.38708603382110596, + "learning_rate": 2.9139000000000003e-05, + "loss": 0.031, + "step": 12587 + }, + { + "epoch": 10.631756756756756, + "grad_norm": 1.492484450340271, + "learning_rate": 2.9138666666666666e-05, + "loss": 0.0513, + "step": 12588 + }, + { + "epoch": 10.63260135135135, + "grad_norm": 2.685563087463379, + "learning_rate": 2.9138333333333335e-05, + "loss": 0.1577, + "step": 12589 + }, + { + "epoch": 10.633445945945946, + "grad_norm": 2.4047389030456543, + "learning_rate": 2.9137999999999998e-05, + "loss": 0.0769, + "step": 12590 + }, + { + "epoch": 10.63429054054054, + "grad_norm": 0.9556952118873596, + "learning_rate": 2.9137666666666667e-05, + "loss": 0.2118, + "step": 12591 + }, + { + "epoch": 10.635135135135135, + "grad_norm": 0.5177783370018005, + "learning_rate": 2.9137333333333333e-05, + "loss": 0.0919, + "step": 12592 + }, + { + "epoch": 10.63597972972973, + "grad_norm": 0.5057559609413147, + "learning_rate": 2.9137e-05, + "loss": 0.0654, + "step": 12593 + }, + { + "epoch": 10.636824324324325, + "grad_norm": 0.33580002188682556, + "learning_rate": 2.913666666666667e-05, + "loss": 0.0318, + "step": 12594 + }, + { + "epoch": 10.63766891891892, + "grad_norm": 0.27052944898605347, + "learning_rate": 2.9136333333333334e-05, + "loss": 0.0317, + "step": 12595 + }, + { + "epoch": 10.638513513513514, + "grad_norm": 0.37242162227630615, + "learning_rate": 2.9136e-05, + "loss": 0.0327, + "step": 12596 + }, + { + "epoch": 10.639358108108109, + "grad_norm": 0.3319549858570099, + "learning_rate": 2.9135666666666666e-05, + "loss": 0.028, + "step": 12597 + }, + { + "epoch": 10.640202702702704, + "grad_norm": 0.38946983218193054, + "learning_rate": 2.9135333333333335e-05, + "loss": 0.0476, + "step": 12598 + }, + { + "epoch": 10.641047297297296, + "grad_norm": 0.4233160614967346, + "learning_rate": 2.9134999999999998e-05, + "loss": 0.0254, + "step": 12599 + }, + { + "epoch": 10.641891891891891, + "grad_norm": 0.3484310209751129, + "learning_rate": 2.9134666666666667e-05, + "loss": 0.0279, + "step": 12600 + }, + { + "epoch": 10.642736486486486, + "grad_norm": 0.2807140648365021, + "learning_rate": 2.9134333333333337e-05, + "loss": 0.032, + "step": 12601 + }, + { + "epoch": 10.64358108108108, + "grad_norm": 0.7104345560073853, + "learning_rate": 2.9134e-05, + "loss": 0.0402, + "step": 12602 + }, + { + "epoch": 10.644425675675675, + "grad_norm": 0.731503427028656, + "learning_rate": 2.913366666666667e-05, + "loss": 0.0386, + "step": 12603 + }, + { + "epoch": 10.64527027027027, + "grad_norm": 0.5448463559150696, + "learning_rate": 2.9133333333333334e-05, + "loss": 0.0281, + "step": 12604 + }, + { + "epoch": 10.646114864864865, + "grad_norm": 0.6058053374290466, + "learning_rate": 2.9133e-05, + "loss": 0.024, + "step": 12605 + }, + { + "epoch": 10.64695945945946, + "grad_norm": 0.38442039489746094, + "learning_rate": 2.9132666666666666e-05, + "loss": 0.0356, + "step": 12606 + }, + { + "epoch": 10.647804054054054, + "grad_norm": 0.31877580285072327, + "learning_rate": 2.9132333333333336e-05, + "loss": 0.0173, + "step": 12607 + }, + { + "epoch": 10.64864864864865, + "grad_norm": 0.5637972354888916, + "learning_rate": 2.9131999999999998e-05, + "loss": 0.033, + "step": 12608 + }, + { + "epoch": 10.649493243243244, + "grad_norm": 0.22701528668403625, + "learning_rate": 2.9131666666666668e-05, + "loss": 0.0189, + "step": 12609 + }, + { + "epoch": 10.650337837837839, + "grad_norm": 0.5608363747596741, + "learning_rate": 2.9131333333333337e-05, + "loss": 0.0345, + "step": 12610 + }, + { + "epoch": 10.651182432432432, + "grad_norm": 0.35592055320739746, + "learning_rate": 2.9131e-05, + "loss": 0.0269, + "step": 12611 + }, + { + "epoch": 10.652027027027026, + "grad_norm": 0.38524240255355835, + "learning_rate": 2.913066666666667e-05, + "loss": 0.0689, + "step": 12612 + }, + { + "epoch": 10.652871621621621, + "grad_norm": 0.25575700402259827, + "learning_rate": 2.9130333333333335e-05, + "loss": 0.0179, + "step": 12613 + }, + { + "epoch": 10.653716216216216, + "grad_norm": 0.3232392966747284, + "learning_rate": 2.913e-05, + "loss": 0.0274, + "step": 12614 + }, + { + "epoch": 10.65456081081081, + "grad_norm": 0.18418028950691223, + "learning_rate": 2.9129666666666667e-05, + "loss": 0.015, + "step": 12615 + }, + { + "epoch": 10.655405405405405, + "grad_norm": 0.2682207226753235, + "learning_rate": 2.9129333333333336e-05, + "loss": 0.0284, + "step": 12616 + }, + { + "epoch": 10.65625, + "grad_norm": 0.3660833537578583, + "learning_rate": 2.9129e-05, + "loss": 0.038, + "step": 12617 + }, + { + "epoch": 10.657094594594595, + "grad_norm": 0.44784343242645264, + "learning_rate": 2.9128666666666668e-05, + "loss": 0.0227, + "step": 12618 + }, + { + "epoch": 10.65793918918919, + "grad_norm": 0.46886780858039856, + "learning_rate": 2.9128333333333337e-05, + "loss": 0.0302, + "step": 12619 + }, + { + "epoch": 10.658783783783784, + "grad_norm": 0.5818700790405273, + "learning_rate": 2.9128e-05, + "loss": 0.0277, + "step": 12620 + }, + { + "epoch": 10.659628378378379, + "grad_norm": 0.3335019052028656, + "learning_rate": 2.912766666666667e-05, + "loss": 0.026, + "step": 12621 + }, + { + "epoch": 10.660472972972974, + "grad_norm": 0.4575708508491516, + "learning_rate": 2.912733333333333e-05, + "loss": 0.0345, + "step": 12622 + }, + { + "epoch": 10.661317567567568, + "grad_norm": 0.3521083891391754, + "learning_rate": 2.9127e-05, + "loss": 0.016, + "step": 12623 + }, + { + "epoch": 10.662162162162161, + "grad_norm": 0.3182884454727173, + "learning_rate": 2.9126666666666667e-05, + "loss": 0.0225, + "step": 12624 + }, + { + "epoch": 10.663006756756756, + "grad_norm": 0.30594056844711304, + "learning_rate": 2.9126333333333333e-05, + "loss": 0.0175, + "step": 12625 + }, + { + "epoch": 10.66385135135135, + "grad_norm": 0.8694934844970703, + "learning_rate": 2.9126000000000002e-05, + "loss": 0.0584, + "step": 12626 + }, + { + "epoch": 10.664695945945946, + "grad_norm": 0.6267538666725159, + "learning_rate": 2.9125666666666668e-05, + "loss": 0.0313, + "step": 12627 + }, + { + "epoch": 10.66554054054054, + "grad_norm": 0.39756080508232117, + "learning_rate": 2.9125333333333334e-05, + "loss": 0.0167, + "step": 12628 + }, + { + "epoch": 10.666385135135135, + "grad_norm": 0.6028510928153992, + "learning_rate": 2.9125e-05, + "loss": 0.0269, + "step": 12629 + }, + { + "epoch": 10.66722972972973, + "grad_norm": 0.5216915607452393, + "learning_rate": 2.912466666666667e-05, + "loss": 0.0204, + "step": 12630 + }, + { + "epoch": 10.668074324324325, + "grad_norm": 0.5379849076271057, + "learning_rate": 2.912433333333333e-05, + "loss": 0.0254, + "step": 12631 + }, + { + "epoch": 10.66891891891892, + "grad_norm": 0.2755392789840698, + "learning_rate": 2.9124e-05, + "loss": 0.0179, + "step": 12632 + }, + { + "epoch": 10.669763513513514, + "grad_norm": 0.24093064665794373, + "learning_rate": 2.9123666666666667e-05, + "loss": 0.0213, + "step": 12633 + }, + { + "epoch": 10.670608108108109, + "grad_norm": 0.39272746443748474, + "learning_rate": 2.9123333333333333e-05, + "loss": 0.0244, + "step": 12634 + }, + { + "epoch": 10.671452702702704, + "grad_norm": 1.032461166381836, + "learning_rate": 2.9123000000000002e-05, + "loss": 0.0207, + "step": 12635 + }, + { + "epoch": 10.672297297297296, + "grad_norm": 0.29211854934692383, + "learning_rate": 2.9122666666666668e-05, + "loss": 0.0229, + "step": 12636 + }, + { + "epoch": 10.673141891891891, + "grad_norm": 0.4506415128707886, + "learning_rate": 2.9122333333333334e-05, + "loss": 0.0162, + "step": 12637 + }, + { + "epoch": 10.673986486486486, + "grad_norm": 0.7017631530761719, + "learning_rate": 2.9122e-05, + "loss": 0.0294, + "step": 12638 + }, + { + "epoch": 10.67483108108108, + "grad_norm": 0.5766967535018921, + "learning_rate": 2.912166666666667e-05, + "loss": 0.0542, + "step": 12639 + }, + { + "epoch": 10.675675675675675, + "grad_norm": 0.7411516308784485, + "learning_rate": 2.9121333333333332e-05, + "loss": 0.0494, + "step": 12640 + }, + { + "epoch": 10.67652027027027, + "grad_norm": 0.7781756520271301, + "learning_rate": 2.9121e-05, + "loss": 0.2548, + "step": 12641 + }, + { + "epoch": 10.677364864864865, + "grad_norm": 0.7205067873001099, + "learning_rate": 2.9120666666666667e-05, + "loss": 0.1519, + "step": 12642 + }, + { + "epoch": 10.67820945945946, + "grad_norm": 0.3858988881111145, + "learning_rate": 2.9120333333333333e-05, + "loss": 0.0834, + "step": 12643 + }, + { + "epoch": 10.679054054054054, + "grad_norm": 0.41803425550460815, + "learning_rate": 2.9120000000000002e-05, + "loss": 0.0593, + "step": 12644 + }, + { + "epoch": 10.67989864864865, + "grad_norm": 0.332795113325119, + "learning_rate": 2.911966666666667e-05, + "loss": 0.0441, + "step": 12645 + }, + { + "epoch": 10.680743243243244, + "grad_norm": 0.39461198449134827, + "learning_rate": 2.9119333333333334e-05, + "loss": 0.0416, + "step": 12646 + }, + { + "epoch": 10.681587837837839, + "grad_norm": 0.40641653537750244, + "learning_rate": 2.9119e-05, + "loss": 0.0274, + "step": 12647 + }, + { + "epoch": 10.682432432432432, + "grad_norm": 0.3105248808860779, + "learning_rate": 2.911866666666667e-05, + "loss": 0.0338, + "step": 12648 + }, + { + "epoch": 10.683277027027026, + "grad_norm": 0.23175810277462006, + "learning_rate": 2.9118333333333332e-05, + "loss": 0.0216, + "step": 12649 + }, + { + "epoch": 10.684121621621621, + "grad_norm": 0.3485006093978882, + "learning_rate": 2.9118e-05, + "loss": 0.0302, + "step": 12650 + }, + { + "epoch": 10.684966216216216, + "grad_norm": 0.2625131607055664, + "learning_rate": 2.9117666666666667e-05, + "loss": 0.0296, + "step": 12651 + }, + { + "epoch": 10.68581081081081, + "grad_norm": 0.42669662833213806, + "learning_rate": 2.9117333333333333e-05, + "loss": 0.0369, + "step": 12652 + }, + { + "epoch": 10.686655405405405, + "grad_norm": 0.3243633806705475, + "learning_rate": 2.9117000000000003e-05, + "loss": 0.0326, + "step": 12653 + }, + { + "epoch": 10.6875, + "grad_norm": 0.34049662947654724, + "learning_rate": 2.9116666666666665e-05, + "loss": 0.026, + "step": 12654 + }, + { + "epoch": 10.688344594594595, + "grad_norm": 0.41378331184387207, + "learning_rate": 2.9116333333333334e-05, + "loss": 0.0229, + "step": 12655 + }, + { + "epoch": 10.68918918918919, + "grad_norm": 0.21160167455673218, + "learning_rate": 2.9116e-05, + "loss": 0.0155, + "step": 12656 + }, + { + "epoch": 10.690033783783784, + "grad_norm": 0.6802726984024048, + "learning_rate": 2.9115666666666666e-05, + "loss": 0.0442, + "step": 12657 + }, + { + "epoch": 10.690878378378379, + "grad_norm": 0.3739222586154938, + "learning_rate": 2.9115333333333332e-05, + "loss": 0.0242, + "step": 12658 + }, + { + "epoch": 10.691722972972974, + "grad_norm": 0.2502087652683258, + "learning_rate": 2.9115e-05, + "loss": 0.0132, + "step": 12659 + }, + { + "epoch": 10.692567567567568, + "grad_norm": 0.25839826464653015, + "learning_rate": 2.9114666666666668e-05, + "loss": 0.0245, + "step": 12660 + }, + { + "epoch": 10.693412162162161, + "grad_norm": 0.4908178150653839, + "learning_rate": 2.9114333333333333e-05, + "loss": 0.0425, + "step": 12661 + }, + { + "epoch": 10.694256756756756, + "grad_norm": 0.25615596771240234, + "learning_rate": 2.9114000000000003e-05, + "loss": 0.0177, + "step": 12662 + }, + { + "epoch": 10.69510135135135, + "grad_norm": 0.3410649299621582, + "learning_rate": 2.9113666666666665e-05, + "loss": 0.0238, + "step": 12663 + }, + { + "epoch": 10.695945945945946, + "grad_norm": 0.5553067922592163, + "learning_rate": 2.9113333333333335e-05, + "loss": 0.0259, + "step": 12664 + }, + { + "epoch": 10.69679054054054, + "grad_norm": 0.351265549659729, + "learning_rate": 2.9113e-05, + "loss": 0.0245, + "step": 12665 + }, + { + "epoch": 10.697635135135135, + "grad_norm": 0.4940446615219116, + "learning_rate": 2.9112666666666667e-05, + "loss": 0.0279, + "step": 12666 + }, + { + "epoch": 10.69847972972973, + "grad_norm": 0.2313889116048813, + "learning_rate": 2.9112333333333332e-05, + "loss": 0.0193, + "step": 12667 + }, + { + "epoch": 10.699324324324325, + "grad_norm": 0.43843162059783936, + "learning_rate": 2.9112000000000002e-05, + "loss": 0.0214, + "step": 12668 + }, + { + "epoch": 10.70016891891892, + "grad_norm": 0.549274206161499, + "learning_rate": 2.9111666666666668e-05, + "loss": 0.0224, + "step": 12669 + }, + { + "epoch": 10.701013513513514, + "grad_norm": 0.3711833655834198, + "learning_rate": 2.9111333333333334e-05, + "loss": 0.0239, + "step": 12670 + }, + { + "epoch": 10.701858108108109, + "grad_norm": 0.28308600187301636, + "learning_rate": 2.9111000000000003e-05, + "loss": 0.0169, + "step": 12671 + }, + { + "epoch": 10.702702702702704, + "grad_norm": 0.6591829061508179, + "learning_rate": 2.9110666666666666e-05, + "loss": 0.0317, + "step": 12672 + }, + { + "epoch": 10.703547297297296, + "grad_norm": 1.0476057529449463, + "learning_rate": 2.9110333333333335e-05, + "loss": 0.0251, + "step": 12673 + }, + { + "epoch": 10.704391891891891, + "grad_norm": 0.3629162609577179, + "learning_rate": 2.911e-05, + "loss": 0.0731, + "step": 12674 + }, + { + "epoch": 10.705236486486486, + "grad_norm": 0.4693553149700165, + "learning_rate": 2.9109666666666667e-05, + "loss": 0.0311, + "step": 12675 + }, + { + "epoch": 10.70608108108108, + "grad_norm": 0.34941574931144714, + "learning_rate": 2.9109333333333336e-05, + "loss": 0.0175, + "step": 12676 + }, + { + "epoch": 10.706925675675675, + "grad_norm": 0.7560368776321411, + "learning_rate": 2.9109000000000002e-05, + "loss": 0.0263, + "step": 12677 + }, + { + "epoch": 10.70777027027027, + "grad_norm": 0.33371591567993164, + "learning_rate": 2.9108666666666668e-05, + "loss": 0.0208, + "step": 12678 + }, + { + "epoch": 10.708614864864865, + "grad_norm": 0.2985753118991852, + "learning_rate": 2.9108333333333334e-05, + "loss": 0.0292, + "step": 12679 + }, + { + "epoch": 10.70945945945946, + "grad_norm": 0.4241466522216797, + "learning_rate": 2.9108000000000003e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 10.710304054054054, + "grad_norm": 0.4055080711841583, + "learning_rate": 2.9107666666666666e-05, + "loss": 0.0193, + "step": 12681 + }, + { + "epoch": 10.71114864864865, + "grad_norm": 0.8203215599060059, + "learning_rate": 2.9107333333333335e-05, + "loss": 0.0319, + "step": 12682 + }, + { + "epoch": 10.711993243243244, + "grad_norm": 0.42398300766944885, + "learning_rate": 2.9106999999999998e-05, + "loss": 0.033, + "step": 12683 + }, + { + "epoch": 10.712837837837839, + "grad_norm": 0.5320850014686584, + "learning_rate": 2.9106666666666667e-05, + "loss": 0.0295, + "step": 12684 + }, + { + "epoch": 10.713682432432432, + "grad_norm": 0.38367393612861633, + "learning_rate": 2.9106333333333336e-05, + "loss": 0.0256, + "step": 12685 + }, + { + "epoch": 10.714527027027026, + "grad_norm": 0.3368259370326996, + "learning_rate": 2.9106e-05, + "loss": 0.0354, + "step": 12686 + }, + { + "epoch": 10.715371621621621, + "grad_norm": 0.41906246542930603, + "learning_rate": 2.9105666666666668e-05, + "loss": 0.0249, + "step": 12687 + }, + { + "epoch": 10.716216216216216, + "grad_norm": 0.48309648036956787, + "learning_rate": 2.9105333333333334e-05, + "loss": 0.0374, + "step": 12688 + }, + { + "epoch": 10.71706081081081, + "grad_norm": 0.4643124043941498, + "learning_rate": 2.9105e-05, + "loss": 0.0297, + "step": 12689 + }, + { + "epoch": 10.717905405405405, + "grad_norm": 1.0205894708633423, + "learning_rate": 2.9104666666666666e-05, + "loss": 0.0442, + "step": 12690 + }, + { + "epoch": 10.71875, + "grad_norm": 0.6471207737922668, + "learning_rate": 2.9104333333333335e-05, + "loss": 0.2175, + "step": 12691 + }, + { + "epoch": 10.719594594594595, + "grad_norm": 0.6229130029678345, + "learning_rate": 2.9103999999999998e-05, + "loss": 0.1321, + "step": 12692 + }, + { + "epoch": 10.72043918918919, + "grad_norm": 0.44534602761268616, + "learning_rate": 2.9103666666666667e-05, + "loss": 0.0896, + "step": 12693 + }, + { + "epoch": 10.721283783783784, + "grad_norm": 0.4692506790161133, + "learning_rate": 2.9103333333333336e-05, + "loss": 0.0556, + "step": 12694 + }, + { + "epoch": 10.722128378378379, + "grad_norm": 0.5128887891769409, + "learning_rate": 2.9103e-05, + "loss": 0.0329, + "step": 12695 + }, + { + "epoch": 10.722972972972974, + "grad_norm": 0.49916818737983704, + "learning_rate": 2.910266666666667e-05, + "loss": 0.0511, + "step": 12696 + }, + { + "epoch": 10.723817567567568, + "grad_norm": 0.35552430152893066, + "learning_rate": 2.9102333333333334e-05, + "loss": 0.0399, + "step": 12697 + }, + { + "epoch": 10.724662162162161, + "grad_norm": 0.3356742858886719, + "learning_rate": 2.9102e-05, + "loss": 0.0201, + "step": 12698 + }, + { + "epoch": 10.725506756756756, + "grad_norm": 0.44608619809150696, + "learning_rate": 2.9101666666666666e-05, + "loss": 0.0286, + "step": 12699 + }, + { + "epoch": 10.72635135135135, + "grad_norm": 0.6344843506813049, + "learning_rate": 2.9101333333333335e-05, + "loss": 0.0362, + "step": 12700 + }, + { + "epoch": 10.727195945945946, + "grad_norm": 0.3190571069717407, + "learning_rate": 2.9101e-05, + "loss": 0.046, + "step": 12701 + }, + { + "epoch": 10.72804054054054, + "grad_norm": 0.4241401255130768, + "learning_rate": 2.9100666666666667e-05, + "loss": 0.0394, + "step": 12702 + }, + { + "epoch": 10.728885135135135, + "grad_norm": 0.2642906904220581, + "learning_rate": 2.9100333333333337e-05, + "loss": 0.0353, + "step": 12703 + }, + { + "epoch": 10.72972972972973, + "grad_norm": 0.6101384162902832, + "learning_rate": 2.91e-05, + "loss": 0.025, + "step": 12704 + }, + { + "epoch": 10.730574324324325, + "grad_norm": 0.22098033130168915, + "learning_rate": 2.909966666666667e-05, + "loss": 0.029, + "step": 12705 + }, + { + "epoch": 10.73141891891892, + "grad_norm": 0.3737427294254303, + "learning_rate": 2.9099333333333334e-05, + "loss": 0.0294, + "step": 12706 + }, + { + "epoch": 10.732263513513514, + "grad_norm": 0.36516067385673523, + "learning_rate": 2.9099e-05, + "loss": 0.0307, + "step": 12707 + }, + { + "epoch": 10.733108108108109, + "grad_norm": 0.25167277455329895, + "learning_rate": 2.9098666666666666e-05, + "loss": 0.0228, + "step": 12708 + }, + { + "epoch": 10.733952702702704, + "grad_norm": 0.3571414053440094, + "learning_rate": 2.9098333333333336e-05, + "loss": 0.0299, + "step": 12709 + }, + { + "epoch": 10.734797297297296, + "grad_norm": 0.27992600202560425, + "learning_rate": 2.9098e-05, + "loss": 0.0268, + "step": 12710 + }, + { + "epoch": 10.735641891891891, + "grad_norm": 0.26986420154571533, + "learning_rate": 2.9097666666666668e-05, + "loss": 0.0158, + "step": 12711 + }, + { + "epoch": 10.736486486486486, + "grad_norm": 0.2498490810394287, + "learning_rate": 2.9097333333333333e-05, + "loss": 0.0268, + "step": 12712 + }, + { + "epoch": 10.73733108108108, + "grad_norm": 0.23340757191181183, + "learning_rate": 2.9097e-05, + "loss": 0.0193, + "step": 12713 + }, + { + "epoch": 10.738175675675675, + "grad_norm": 0.334849089384079, + "learning_rate": 2.909666666666667e-05, + "loss": 0.0294, + "step": 12714 + }, + { + "epoch": 10.73902027027027, + "grad_norm": 0.36311304569244385, + "learning_rate": 2.909633333333333e-05, + "loss": 0.0243, + "step": 12715 + }, + { + "epoch": 10.739864864864865, + "grad_norm": 0.37138110399246216, + "learning_rate": 2.9096e-05, + "loss": 0.0295, + "step": 12716 + }, + { + "epoch": 10.74070945945946, + "grad_norm": 0.45649346709251404, + "learning_rate": 2.9095666666666667e-05, + "loss": 0.0275, + "step": 12717 + }, + { + "epoch": 10.741554054054054, + "grad_norm": 0.7037672400474548, + "learning_rate": 2.9095333333333332e-05, + "loss": 0.0311, + "step": 12718 + }, + { + "epoch": 10.74239864864865, + "grad_norm": 0.5108250379562378, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0346, + "step": 12719 + }, + { + "epoch": 10.743243243243244, + "grad_norm": 0.5456815958023071, + "learning_rate": 2.9094666666666668e-05, + "loss": 0.0318, + "step": 12720 + }, + { + "epoch": 10.744087837837839, + "grad_norm": 0.4567168951034546, + "learning_rate": 2.9094333333333334e-05, + "loss": 0.0315, + "step": 12721 + }, + { + "epoch": 10.744932432432432, + "grad_norm": 0.26951053738594055, + "learning_rate": 2.9094e-05, + "loss": 0.0182, + "step": 12722 + }, + { + "epoch": 10.745777027027026, + "grad_norm": 0.25412899255752563, + "learning_rate": 2.909366666666667e-05, + "loss": 0.0316, + "step": 12723 + }, + { + "epoch": 10.746621621621621, + "grad_norm": 0.4152126610279083, + "learning_rate": 2.909333333333333e-05, + "loss": 0.0307, + "step": 12724 + }, + { + "epoch": 10.747466216216216, + "grad_norm": 0.35105690360069275, + "learning_rate": 2.9093e-05, + "loss": 0.0246, + "step": 12725 + }, + { + "epoch": 10.74831081081081, + "grad_norm": 0.3616734743118286, + "learning_rate": 2.909266666666667e-05, + "loss": 0.0426, + "step": 12726 + }, + { + "epoch": 10.749155405405405, + "grad_norm": 0.5016709566116333, + "learning_rate": 2.9092333333333333e-05, + "loss": 0.0532, + "step": 12727 + }, + { + "epoch": 10.75, + "grad_norm": 0.7325279712677002, + "learning_rate": 2.9092000000000002e-05, + "loss": 0.0292, + "step": 12728 + }, + { + "epoch": 10.750844594594595, + "grad_norm": 0.38472962379455566, + "learning_rate": 2.9091666666666668e-05, + "loss": 0.0185, + "step": 12729 + }, + { + "epoch": 10.75168918918919, + "grad_norm": 0.330827921628952, + "learning_rate": 2.9091333333333334e-05, + "loss": 0.0185, + "step": 12730 + }, + { + "epoch": 10.752533783783784, + "grad_norm": 0.5310145020484924, + "learning_rate": 2.9091e-05, + "loss": 0.0395, + "step": 12731 + }, + { + "epoch": 10.753378378378379, + "grad_norm": 0.22948673367500305, + "learning_rate": 2.909066666666667e-05, + "loss": 0.0205, + "step": 12732 + }, + { + "epoch": 10.754222972972974, + "grad_norm": 0.9176730513572693, + "learning_rate": 2.909033333333333e-05, + "loss": 0.0669, + "step": 12733 + }, + { + "epoch": 10.755067567567568, + "grad_norm": 0.48519253730773926, + "learning_rate": 2.909e-05, + "loss": 0.0315, + "step": 12734 + }, + { + "epoch": 10.755912162162161, + "grad_norm": 0.32316920161247253, + "learning_rate": 2.908966666666667e-05, + "loss": 0.0344, + "step": 12735 + }, + { + "epoch": 10.756756756756756, + "grad_norm": 0.2748912572860718, + "learning_rate": 2.9089333333333333e-05, + "loss": 0.0189, + "step": 12736 + }, + { + "epoch": 10.75760135135135, + "grad_norm": 0.9001589417457581, + "learning_rate": 2.9089000000000002e-05, + "loss": 0.0434, + "step": 12737 + }, + { + "epoch": 10.758445945945946, + "grad_norm": 0.6350423097610474, + "learning_rate": 2.9088666666666668e-05, + "loss": 0.0321, + "step": 12738 + }, + { + "epoch": 10.75929054054054, + "grad_norm": 0.3067616820335388, + "learning_rate": 2.9088333333333334e-05, + "loss": 0.0276, + "step": 12739 + }, + { + "epoch": 10.760135135135135, + "grad_norm": 0.8286774158477783, + "learning_rate": 2.9088e-05, + "loss": 0.0544, + "step": 12740 + }, + { + "epoch": 10.76097972972973, + "grad_norm": 0.9170118570327759, + "learning_rate": 2.9087666666666666e-05, + "loss": 0.2772, + "step": 12741 + }, + { + "epoch": 10.761824324324325, + "grad_norm": 0.4715818762779236, + "learning_rate": 2.9087333333333332e-05, + "loss": 0.1315, + "step": 12742 + }, + { + "epoch": 10.76266891891892, + "grad_norm": 0.47588229179382324, + "learning_rate": 2.9087e-05, + "loss": 0.0738, + "step": 12743 + }, + { + "epoch": 10.763513513513514, + "grad_norm": 0.44817906618118286, + "learning_rate": 2.9086666666666667e-05, + "loss": 0.061, + "step": 12744 + }, + { + "epoch": 10.764358108108109, + "grad_norm": 0.2964242994785309, + "learning_rate": 2.9086333333333333e-05, + "loss": 0.0314, + "step": 12745 + }, + { + "epoch": 10.765202702702704, + "grad_norm": 0.3645385503768921, + "learning_rate": 2.9086000000000002e-05, + "loss": 0.0475, + "step": 12746 + }, + { + "epoch": 10.766047297297296, + "grad_norm": 0.2759977877140045, + "learning_rate": 2.9085666666666665e-05, + "loss": 0.0255, + "step": 12747 + }, + { + "epoch": 10.766891891891891, + "grad_norm": 0.33743903040885925, + "learning_rate": 2.9085333333333334e-05, + "loss": 0.0357, + "step": 12748 + }, + { + "epoch": 10.767736486486486, + "grad_norm": 0.3277914524078369, + "learning_rate": 2.9085e-05, + "loss": 0.0225, + "step": 12749 + }, + { + "epoch": 10.76858108108108, + "grad_norm": 0.6816125512123108, + "learning_rate": 2.9084666666666666e-05, + "loss": 0.0248, + "step": 12750 + }, + { + "epoch": 10.769425675675675, + "grad_norm": 0.44166770577430725, + "learning_rate": 2.9084333333333335e-05, + "loss": 0.0375, + "step": 12751 + }, + { + "epoch": 10.77027027027027, + "grad_norm": 0.5416327714920044, + "learning_rate": 2.9084e-05, + "loss": 0.0281, + "step": 12752 + }, + { + "epoch": 10.771114864864865, + "grad_norm": 0.2186601608991623, + "learning_rate": 2.9083666666666667e-05, + "loss": 0.0169, + "step": 12753 + }, + { + "epoch": 10.77195945945946, + "grad_norm": 0.2657105028629303, + "learning_rate": 2.9083333333333333e-05, + "loss": 0.026, + "step": 12754 + }, + { + "epoch": 10.772804054054054, + "grad_norm": 0.3633258044719696, + "learning_rate": 2.9083000000000003e-05, + "loss": 0.0257, + "step": 12755 + }, + { + "epoch": 10.77364864864865, + "grad_norm": 0.20306475460529327, + "learning_rate": 2.9082666666666665e-05, + "loss": 0.0176, + "step": 12756 + }, + { + "epoch": 10.774493243243244, + "grad_norm": 0.3463284373283386, + "learning_rate": 2.9082333333333334e-05, + "loss": 0.0292, + "step": 12757 + }, + { + "epoch": 10.775337837837839, + "grad_norm": 0.3382990062236786, + "learning_rate": 2.9082e-05, + "loss": 0.0264, + "step": 12758 + }, + { + "epoch": 10.776182432432432, + "grad_norm": 0.3053520917892456, + "learning_rate": 2.9081666666666666e-05, + "loss": 0.0255, + "step": 12759 + }, + { + "epoch": 10.777027027027026, + "grad_norm": 0.3042345345020294, + "learning_rate": 2.9081333333333336e-05, + "loss": 0.0258, + "step": 12760 + }, + { + "epoch": 10.777871621621621, + "grad_norm": 0.34135702252388, + "learning_rate": 2.9081e-05, + "loss": 0.031, + "step": 12761 + }, + { + "epoch": 10.778716216216216, + "grad_norm": 0.2684811055660248, + "learning_rate": 2.9080666666666668e-05, + "loss": 0.0195, + "step": 12762 + }, + { + "epoch": 10.77956081081081, + "grad_norm": 0.6675364971160889, + "learning_rate": 2.9080333333333333e-05, + "loss": 0.0387, + "step": 12763 + }, + { + "epoch": 10.780405405405405, + "grad_norm": 0.2818337082862854, + "learning_rate": 2.9080000000000003e-05, + "loss": 0.016, + "step": 12764 + }, + { + "epoch": 10.78125, + "grad_norm": 0.3055327236652374, + "learning_rate": 2.9079666666666665e-05, + "loss": 0.0272, + "step": 12765 + }, + { + "epoch": 10.782094594594595, + "grad_norm": 0.34700947999954224, + "learning_rate": 2.9079333333333335e-05, + "loss": 0.0441, + "step": 12766 + }, + { + "epoch": 10.78293918918919, + "grad_norm": 0.2876169681549072, + "learning_rate": 2.9079e-05, + "loss": 0.0227, + "step": 12767 + }, + { + "epoch": 10.783783783783784, + "grad_norm": 0.2817254066467285, + "learning_rate": 2.9078666666666667e-05, + "loss": 0.0247, + "step": 12768 + }, + { + "epoch": 10.784628378378379, + "grad_norm": 0.34366679191589355, + "learning_rate": 2.9078333333333336e-05, + "loss": 0.0213, + "step": 12769 + }, + { + "epoch": 10.785472972972974, + "grad_norm": 0.5446111559867859, + "learning_rate": 2.9078000000000002e-05, + "loss": 0.0202, + "step": 12770 + }, + { + "epoch": 10.786317567567568, + "grad_norm": 0.24312859773635864, + "learning_rate": 2.9077666666666668e-05, + "loss": 0.0205, + "step": 12771 + }, + { + "epoch": 10.787162162162161, + "grad_norm": 0.7937974333763123, + "learning_rate": 2.9077333333333334e-05, + "loss": 0.0391, + "step": 12772 + }, + { + "epoch": 10.788006756756756, + "grad_norm": 0.3539845943450928, + "learning_rate": 2.9077e-05, + "loss": 0.0248, + "step": 12773 + }, + { + "epoch": 10.78885135135135, + "grad_norm": 0.9564263820648193, + "learning_rate": 2.9076666666666666e-05, + "loss": 0.0527, + "step": 12774 + }, + { + "epoch": 10.789695945945946, + "grad_norm": 0.318453311920166, + "learning_rate": 2.9076333333333335e-05, + "loss": 0.0297, + "step": 12775 + }, + { + "epoch": 10.79054054054054, + "grad_norm": 0.19866648316383362, + "learning_rate": 2.9076e-05, + "loss": 0.0123, + "step": 12776 + }, + { + "epoch": 10.791385135135135, + "grad_norm": 0.4483756422996521, + "learning_rate": 2.9075666666666667e-05, + "loss": 0.0293, + "step": 12777 + }, + { + "epoch": 10.79222972972973, + "grad_norm": 0.3368069529533386, + "learning_rate": 2.9075333333333336e-05, + "loss": 0.0231, + "step": 12778 + }, + { + "epoch": 10.793074324324325, + "grad_norm": 0.6684255599975586, + "learning_rate": 2.9075e-05, + "loss": 0.0216, + "step": 12779 + }, + { + "epoch": 10.79391891891892, + "grad_norm": 0.3699893057346344, + "learning_rate": 2.9074666666666668e-05, + "loss": 0.0284, + "step": 12780 + }, + { + "epoch": 10.794763513513514, + "grad_norm": 0.40618735551834106, + "learning_rate": 2.9074333333333334e-05, + "loss": 0.0251, + "step": 12781 + }, + { + "epoch": 10.795608108108109, + "grad_norm": 0.254670649766922, + "learning_rate": 2.9074e-05, + "loss": 0.0173, + "step": 12782 + }, + { + "epoch": 10.796452702702704, + "grad_norm": 0.4657425284385681, + "learning_rate": 2.9073666666666666e-05, + "loss": 0.0316, + "step": 12783 + }, + { + "epoch": 10.797297297297296, + "grad_norm": 0.4800550043582916, + "learning_rate": 2.9073333333333335e-05, + "loss": 0.0699, + "step": 12784 + }, + { + "epoch": 10.798141891891891, + "grad_norm": 0.18296536803245544, + "learning_rate": 2.9073e-05, + "loss": 0.0158, + "step": 12785 + }, + { + "epoch": 10.798986486486486, + "grad_norm": 0.34267890453338623, + "learning_rate": 2.9072666666666667e-05, + "loss": 0.0296, + "step": 12786 + }, + { + "epoch": 10.79983108108108, + "grad_norm": 0.6419066190719604, + "learning_rate": 2.9072333333333336e-05, + "loss": 0.0595, + "step": 12787 + }, + { + "epoch": 10.800675675675675, + "grad_norm": 0.3799419105052948, + "learning_rate": 2.9072e-05, + "loss": 0.0323, + "step": 12788 + }, + { + "epoch": 10.80152027027027, + "grad_norm": 0.399372935295105, + "learning_rate": 2.9071666666666668e-05, + "loss": 0.0462, + "step": 12789 + }, + { + "epoch": 10.802364864864865, + "grad_norm": 1.005149006843567, + "learning_rate": 2.9071333333333334e-05, + "loss": 0.0351, + "step": 12790 + }, + { + "epoch": 10.80320945945946, + "grad_norm": 0.9447376132011414, + "learning_rate": 2.9071e-05, + "loss": 0.2457, + "step": 12791 + }, + { + "epoch": 10.804054054054054, + "grad_norm": 0.49498632550239563, + "learning_rate": 2.9070666666666666e-05, + "loss": 0.1488, + "step": 12792 + }, + { + "epoch": 10.80489864864865, + "grad_norm": 0.4464951157569885, + "learning_rate": 2.9070333333333335e-05, + "loss": 0.0551, + "step": 12793 + }, + { + "epoch": 10.805743243243244, + "grad_norm": 0.3237608075141907, + "learning_rate": 2.907e-05, + "loss": 0.0273, + "step": 12794 + }, + { + "epoch": 10.806587837837839, + "grad_norm": 0.27456140518188477, + "learning_rate": 2.9069666666666667e-05, + "loss": 0.0325, + "step": 12795 + }, + { + "epoch": 10.807432432432432, + "grad_norm": 0.3059849143028259, + "learning_rate": 2.9069333333333336e-05, + "loss": 0.0233, + "step": 12796 + }, + { + "epoch": 10.808277027027026, + "grad_norm": 0.29946181178092957, + "learning_rate": 2.9069e-05, + "loss": 0.027, + "step": 12797 + }, + { + "epoch": 10.809121621621621, + "grad_norm": 0.42673081159591675, + "learning_rate": 2.906866666666667e-05, + "loss": 0.025, + "step": 12798 + }, + { + "epoch": 10.809966216216216, + "grad_norm": 0.2545531392097473, + "learning_rate": 2.9068333333333334e-05, + "loss": 0.0229, + "step": 12799 + }, + { + "epoch": 10.81081081081081, + "grad_norm": 0.27241063117980957, + "learning_rate": 2.9068e-05, + "loss": 0.021, + "step": 12800 + }, + { + "epoch": 10.811655405405405, + "grad_norm": 0.23197568953037262, + "learning_rate": 2.906766666666667e-05, + "loss": 0.0223, + "step": 12801 + }, + { + "epoch": 10.8125, + "grad_norm": 0.2883169949054718, + "learning_rate": 2.9067333333333332e-05, + "loss": 0.029, + "step": 12802 + }, + { + "epoch": 10.813344594594595, + "grad_norm": 0.21445608139038086, + "learning_rate": 2.9067e-05, + "loss": 0.025, + "step": 12803 + }, + { + "epoch": 10.81418918918919, + "grad_norm": 0.30817344784736633, + "learning_rate": 2.9066666666666667e-05, + "loss": 0.0344, + "step": 12804 + }, + { + "epoch": 10.815033783783784, + "grad_norm": 0.2916944622993469, + "learning_rate": 2.9066333333333333e-05, + "loss": 0.0232, + "step": 12805 + }, + { + "epoch": 10.815878378378379, + "grad_norm": 0.43822813034057617, + "learning_rate": 2.9066e-05, + "loss": 0.0332, + "step": 12806 + }, + { + "epoch": 10.816722972972974, + "grad_norm": 0.486015260219574, + "learning_rate": 2.906566666666667e-05, + "loss": 0.0327, + "step": 12807 + }, + { + "epoch": 10.817567567567568, + "grad_norm": 0.3106563687324524, + "learning_rate": 2.906533333333333e-05, + "loss": 0.02, + "step": 12808 + }, + { + "epoch": 10.818412162162161, + "grad_norm": 0.19089540839195251, + "learning_rate": 2.9065e-05, + "loss": 0.016, + "step": 12809 + }, + { + "epoch": 10.819256756756756, + "grad_norm": 0.3288341164588928, + "learning_rate": 2.906466666666667e-05, + "loss": 0.0287, + "step": 12810 + }, + { + "epoch": 10.82010135135135, + "grad_norm": 0.20858746767044067, + "learning_rate": 2.9064333333333332e-05, + "loss": 0.0188, + "step": 12811 + }, + { + "epoch": 10.820945945945946, + "grad_norm": 0.25373920798301697, + "learning_rate": 2.9064e-05, + "loss": 0.0226, + "step": 12812 + }, + { + "epoch": 10.82179054054054, + "grad_norm": 0.3567250669002533, + "learning_rate": 2.9063666666666668e-05, + "loss": 0.0241, + "step": 12813 + }, + { + "epoch": 10.822635135135135, + "grad_norm": 0.38649782538414, + "learning_rate": 2.9063333333333333e-05, + "loss": 0.0199, + "step": 12814 + }, + { + "epoch": 10.82347972972973, + "grad_norm": 0.2840172052383423, + "learning_rate": 2.9063e-05, + "loss": 0.0253, + "step": 12815 + }, + { + "epoch": 10.824324324324325, + "grad_norm": 0.23922987282276154, + "learning_rate": 2.906266666666667e-05, + "loss": 0.0327, + "step": 12816 + }, + { + "epoch": 10.82516891891892, + "grad_norm": 0.15379182994365692, + "learning_rate": 2.9062333333333335e-05, + "loss": 0.014, + "step": 12817 + }, + { + "epoch": 10.826013513513514, + "grad_norm": 0.48334449529647827, + "learning_rate": 2.9062e-05, + "loss": 0.022, + "step": 12818 + }, + { + "epoch": 10.826858108108109, + "grad_norm": 0.39927029609680176, + "learning_rate": 2.906166666666667e-05, + "loss": 0.0201, + "step": 12819 + }, + { + "epoch": 10.827702702702704, + "grad_norm": 1.5238701105117798, + "learning_rate": 2.9061333333333332e-05, + "loss": 0.0304, + "step": 12820 + }, + { + "epoch": 10.828547297297296, + "grad_norm": 0.1944616734981537, + "learning_rate": 2.9061000000000002e-05, + "loss": 0.0198, + "step": 12821 + }, + { + "epoch": 10.829391891891891, + "grad_norm": 0.27634260058403015, + "learning_rate": 2.9060666666666668e-05, + "loss": 0.0216, + "step": 12822 + }, + { + "epoch": 10.830236486486486, + "grad_norm": 0.33690115809440613, + "learning_rate": 2.9060333333333334e-05, + "loss": 0.0189, + "step": 12823 + }, + { + "epoch": 10.83108108108108, + "grad_norm": 0.23619608581066132, + "learning_rate": 2.906e-05, + "loss": 0.0156, + "step": 12824 + }, + { + "epoch": 10.831925675675675, + "grad_norm": 0.3021117150783539, + "learning_rate": 2.905966666666667e-05, + "loss": 0.0274, + "step": 12825 + }, + { + "epoch": 10.83277027027027, + "grad_norm": 0.3140360414981842, + "learning_rate": 2.9059333333333335e-05, + "loss": 0.0231, + "step": 12826 + }, + { + "epoch": 10.833614864864865, + "grad_norm": 0.7040709257125854, + "learning_rate": 2.9059e-05, + "loss": 0.0376, + "step": 12827 + }, + { + "epoch": 10.83445945945946, + "grad_norm": 0.35596543550491333, + "learning_rate": 2.905866666666667e-05, + "loss": 0.0268, + "step": 12828 + }, + { + "epoch": 10.835304054054054, + "grad_norm": 0.26941582560539246, + "learning_rate": 2.9058333333333333e-05, + "loss": 0.037, + "step": 12829 + }, + { + "epoch": 10.83614864864865, + "grad_norm": 0.25664255023002625, + "learning_rate": 2.9058000000000002e-05, + "loss": 0.0206, + "step": 12830 + }, + { + "epoch": 10.836993243243244, + "grad_norm": 0.49544402956962585, + "learning_rate": 2.9057666666666668e-05, + "loss": 0.0221, + "step": 12831 + }, + { + "epoch": 10.837837837837839, + "grad_norm": 0.5047066807746887, + "learning_rate": 2.9057333333333334e-05, + "loss": 0.0213, + "step": 12832 + }, + { + "epoch": 10.838682432432432, + "grad_norm": 0.4460195004940033, + "learning_rate": 2.9057e-05, + "loss": 0.0312, + "step": 12833 + }, + { + "epoch": 10.839527027027026, + "grad_norm": 0.7132643461227417, + "learning_rate": 2.9056666666666666e-05, + "loss": 0.0309, + "step": 12834 + }, + { + "epoch": 10.840371621621621, + "grad_norm": 0.32407423853874207, + "learning_rate": 2.9056333333333335e-05, + "loss": 0.0202, + "step": 12835 + }, + { + "epoch": 10.841216216216216, + "grad_norm": 0.3288262188434601, + "learning_rate": 2.9056e-05, + "loss": 0.0255, + "step": 12836 + }, + { + "epoch": 10.84206081081081, + "grad_norm": 0.44193515181541443, + "learning_rate": 2.9055666666666667e-05, + "loss": 0.0322, + "step": 12837 + }, + { + "epoch": 10.842905405405405, + "grad_norm": 0.31606870889663696, + "learning_rate": 2.9055333333333333e-05, + "loss": 0.0276, + "step": 12838 + }, + { + "epoch": 10.84375, + "grad_norm": 0.5308084487915039, + "learning_rate": 2.9055000000000002e-05, + "loss": 0.0292, + "step": 12839 + }, + { + "epoch": 10.844594594594595, + "grad_norm": 0.9485588669776917, + "learning_rate": 2.9054666666666665e-05, + "loss": 0.0667, + "step": 12840 + }, + { + "epoch": 10.84543918918919, + "grad_norm": 0.6544814705848694, + "learning_rate": 2.9054333333333334e-05, + "loss": 0.2127, + "step": 12841 + }, + { + "epoch": 10.846283783783784, + "grad_norm": 0.5698652863502502, + "learning_rate": 2.9054e-05, + "loss": 0.1326, + "step": 12842 + }, + { + "epoch": 10.847128378378379, + "grad_norm": 0.41489654779434204, + "learning_rate": 2.9053666666666666e-05, + "loss": 0.0751, + "step": 12843 + }, + { + "epoch": 10.847972972972974, + "grad_norm": 0.2875814437866211, + "learning_rate": 2.9053333333333335e-05, + "loss": 0.0431, + "step": 12844 + }, + { + "epoch": 10.848817567567568, + "grad_norm": 0.3531750738620758, + "learning_rate": 2.9053e-05, + "loss": 0.0285, + "step": 12845 + }, + { + "epoch": 10.849662162162161, + "grad_norm": 0.4056904911994934, + "learning_rate": 2.9052666666666667e-05, + "loss": 0.035, + "step": 12846 + }, + { + "epoch": 10.850506756756756, + "grad_norm": 0.24361485242843628, + "learning_rate": 2.9052333333333333e-05, + "loss": 0.0269, + "step": 12847 + }, + { + "epoch": 10.85135135135135, + "grad_norm": 0.36733609437942505, + "learning_rate": 2.9052000000000002e-05, + "loss": 0.036, + "step": 12848 + }, + { + "epoch": 10.852195945945946, + "grad_norm": 0.5397789478302002, + "learning_rate": 2.9051666666666665e-05, + "loss": 0.0369, + "step": 12849 + }, + { + "epoch": 10.85304054054054, + "grad_norm": 0.19255217909812927, + "learning_rate": 2.9051333333333334e-05, + "loss": 0.0267, + "step": 12850 + }, + { + "epoch": 10.853885135135135, + "grad_norm": 0.2046056091785431, + "learning_rate": 2.9051000000000004e-05, + "loss": 0.0232, + "step": 12851 + }, + { + "epoch": 10.85472972972973, + "grad_norm": 0.259150892496109, + "learning_rate": 2.9050666666666666e-05, + "loss": 0.0224, + "step": 12852 + }, + { + "epoch": 10.855574324324325, + "grad_norm": 0.2701674699783325, + "learning_rate": 2.9050333333333335e-05, + "loss": 0.027, + "step": 12853 + }, + { + "epoch": 10.85641891891892, + "grad_norm": 0.4368049204349518, + "learning_rate": 2.905e-05, + "loss": 0.0479, + "step": 12854 + }, + { + "epoch": 10.857263513513514, + "grad_norm": 0.47728216648101807, + "learning_rate": 2.9049666666666667e-05, + "loss": 0.0281, + "step": 12855 + }, + { + "epoch": 10.858108108108109, + "grad_norm": 0.25011146068573, + "learning_rate": 2.9049333333333333e-05, + "loss": 0.0231, + "step": 12856 + }, + { + "epoch": 10.858952702702704, + "grad_norm": 0.3266584575176239, + "learning_rate": 2.9049000000000003e-05, + "loss": 0.0384, + "step": 12857 + }, + { + "epoch": 10.859797297297296, + "grad_norm": 0.3072369694709778, + "learning_rate": 2.9048666666666665e-05, + "loss": 0.0334, + "step": 12858 + }, + { + "epoch": 10.860641891891891, + "grad_norm": 1.4529600143432617, + "learning_rate": 2.9048333333333334e-05, + "loss": 0.0622, + "step": 12859 + }, + { + "epoch": 10.861486486486486, + "grad_norm": 0.22464071214199066, + "learning_rate": 2.9048000000000004e-05, + "loss": 0.0183, + "step": 12860 + }, + { + "epoch": 10.86233108108108, + "grad_norm": 0.29495781660079956, + "learning_rate": 2.9047666666666666e-05, + "loss": 0.0255, + "step": 12861 + }, + { + "epoch": 10.863175675675675, + "grad_norm": 0.2705918550491333, + "learning_rate": 2.9047333333333336e-05, + "loss": 0.0239, + "step": 12862 + }, + { + "epoch": 10.86402027027027, + "grad_norm": 0.9711025953292847, + "learning_rate": 2.9046999999999998e-05, + "loss": 0.0292, + "step": 12863 + }, + { + "epoch": 10.864864864864865, + "grad_norm": 0.44311749935150146, + "learning_rate": 2.9046666666666668e-05, + "loss": 0.0354, + "step": 12864 + }, + { + "epoch": 10.86570945945946, + "grad_norm": 0.24457992613315582, + "learning_rate": 2.9046333333333333e-05, + "loss": 0.0239, + "step": 12865 + }, + { + "epoch": 10.866554054054054, + "grad_norm": 0.7754105925559998, + "learning_rate": 2.9046e-05, + "loss": 0.0384, + "step": 12866 + }, + { + "epoch": 10.86739864864865, + "grad_norm": 0.34004318714141846, + "learning_rate": 2.904566666666667e-05, + "loss": 0.0204, + "step": 12867 + }, + { + "epoch": 10.868243243243244, + "grad_norm": 0.7331631779670715, + "learning_rate": 2.9045333333333335e-05, + "loss": 0.0224, + "step": 12868 + }, + { + "epoch": 10.869087837837839, + "grad_norm": 0.3728622794151306, + "learning_rate": 2.9045e-05, + "loss": 0.0372, + "step": 12869 + }, + { + "epoch": 10.869932432432432, + "grad_norm": 0.2517010569572449, + "learning_rate": 2.9044666666666667e-05, + "loss": 0.0202, + "step": 12870 + }, + { + "epoch": 10.870777027027026, + "grad_norm": 0.5058990120887756, + "learning_rate": 2.9044333333333336e-05, + "loss": 0.0375, + "step": 12871 + }, + { + "epoch": 10.871621621621621, + "grad_norm": 0.23983320593833923, + "learning_rate": 2.9044e-05, + "loss": 0.0147, + "step": 12872 + }, + { + "epoch": 10.872466216216216, + "grad_norm": 0.31110623478889465, + "learning_rate": 2.9043666666666668e-05, + "loss": 0.0259, + "step": 12873 + }, + { + "epoch": 10.87331081081081, + "grad_norm": 0.6114042401313782, + "learning_rate": 2.9043333333333334e-05, + "loss": 0.0269, + "step": 12874 + }, + { + "epoch": 10.874155405405405, + "grad_norm": 0.46779143810272217, + "learning_rate": 2.9043e-05, + "loss": 0.0298, + "step": 12875 + }, + { + "epoch": 10.875, + "grad_norm": 0.42560094594955444, + "learning_rate": 2.904266666666667e-05, + "loss": 0.0361, + "step": 12876 + }, + { + "epoch": 10.875844594594595, + "grad_norm": 0.44796663522720337, + "learning_rate": 2.9042333333333335e-05, + "loss": 0.0774, + "step": 12877 + }, + { + "epoch": 10.87668918918919, + "grad_norm": 0.4114105999469757, + "learning_rate": 2.9042e-05, + "loss": 0.0328, + "step": 12878 + }, + { + "epoch": 10.877533783783784, + "grad_norm": 0.3434268534183502, + "learning_rate": 2.9041666666666667e-05, + "loss": 0.0175, + "step": 12879 + }, + { + "epoch": 10.878378378378379, + "grad_norm": 0.6839203834533691, + "learning_rate": 2.9041333333333336e-05, + "loss": 0.0406, + "step": 12880 + }, + { + "epoch": 10.879222972972974, + "grad_norm": 0.29519084095954895, + "learning_rate": 2.9041e-05, + "loss": 0.0189, + "step": 12881 + }, + { + "epoch": 10.880067567567568, + "grad_norm": 0.31569406390190125, + "learning_rate": 2.9040666666666668e-05, + "loss": 0.0254, + "step": 12882 + }, + { + "epoch": 10.880912162162161, + "grad_norm": 0.7577753663063049, + "learning_rate": 2.9040333333333334e-05, + "loss": 0.0323, + "step": 12883 + }, + { + "epoch": 10.881756756756756, + "grad_norm": 0.6038687229156494, + "learning_rate": 2.904e-05, + "loss": 0.0351, + "step": 12884 + }, + { + "epoch": 10.88260135135135, + "grad_norm": 0.5950882434844971, + "learning_rate": 2.903966666666667e-05, + "loss": 0.0339, + "step": 12885 + }, + { + "epoch": 10.883445945945946, + "grad_norm": 0.3309486508369446, + "learning_rate": 2.9039333333333335e-05, + "loss": 0.0231, + "step": 12886 + }, + { + "epoch": 10.88429054054054, + "grad_norm": 0.5911192893981934, + "learning_rate": 2.9039e-05, + "loss": 0.0427, + "step": 12887 + }, + { + "epoch": 10.885135135135135, + "grad_norm": 0.6123714447021484, + "learning_rate": 2.9038666666666667e-05, + "loss": 0.0301, + "step": 12888 + }, + { + "epoch": 10.88597972972973, + "grad_norm": 0.4603399932384491, + "learning_rate": 2.9038333333333336e-05, + "loss": 0.0407, + "step": 12889 + }, + { + "epoch": 10.886824324324325, + "grad_norm": 0.7202482223510742, + "learning_rate": 2.9038e-05, + "loss": 0.0843, + "step": 12890 + }, + { + "epoch": 10.88766891891892, + "grad_norm": 0.7273654937744141, + "learning_rate": 2.9037666666666668e-05, + "loss": 0.2438, + "step": 12891 + }, + { + "epoch": 10.888513513513514, + "grad_norm": 0.4420076906681061, + "learning_rate": 2.9037333333333334e-05, + "loss": 0.1484, + "step": 12892 + }, + { + "epoch": 10.889358108108109, + "grad_norm": 1.4891853332519531, + "learning_rate": 2.9037e-05, + "loss": 0.0923, + "step": 12893 + }, + { + "epoch": 10.890202702702704, + "grad_norm": 0.3757818937301636, + "learning_rate": 2.903666666666667e-05, + "loss": 0.0387, + "step": 12894 + }, + { + "epoch": 10.891047297297296, + "grad_norm": 0.47574037313461304, + "learning_rate": 2.9036333333333332e-05, + "loss": 0.0577, + "step": 12895 + }, + { + "epoch": 10.891891891891891, + "grad_norm": 0.44803670048713684, + "learning_rate": 2.9036e-05, + "loss": 0.0328, + "step": 12896 + }, + { + "epoch": 10.892736486486486, + "grad_norm": 0.666714608669281, + "learning_rate": 2.9035666666666667e-05, + "loss": 0.0474, + "step": 12897 + }, + { + "epoch": 10.89358108108108, + "grad_norm": 0.3223460018634796, + "learning_rate": 2.9035333333333333e-05, + "loss": 0.0261, + "step": 12898 + }, + { + "epoch": 10.894425675675675, + "grad_norm": 0.227498397231102, + "learning_rate": 2.9035e-05, + "loss": 0.0195, + "step": 12899 + }, + { + "epoch": 10.89527027027027, + "grad_norm": 0.5029867887496948, + "learning_rate": 2.903466666666667e-05, + "loss": 0.0398, + "step": 12900 + }, + { + "epoch": 10.896114864864865, + "grad_norm": 0.22226768732070923, + "learning_rate": 2.9034333333333334e-05, + "loss": 0.0241, + "step": 12901 + }, + { + "epoch": 10.89695945945946, + "grad_norm": 0.309396892786026, + "learning_rate": 2.9034e-05, + "loss": 0.0386, + "step": 12902 + }, + { + "epoch": 10.897804054054054, + "grad_norm": 0.22736915946006775, + "learning_rate": 2.903366666666667e-05, + "loss": 0.0244, + "step": 12903 + }, + { + "epoch": 10.89864864864865, + "grad_norm": 0.45120954513549805, + "learning_rate": 2.9033333333333332e-05, + "loss": 0.032, + "step": 12904 + }, + { + "epoch": 10.899493243243244, + "grad_norm": 0.6230713129043579, + "learning_rate": 2.9033e-05, + "loss": 0.026, + "step": 12905 + }, + { + "epoch": 10.900337837837839, + "grad_norm": 0.291129469871521, + "learning_rate": 2.9032666666666667e-05, + "loss": 0.0278, + "step": 12906 + }, + { + "epoch": 10.901182432432432, + "grad_norm": 0.2799825966358185, + "learning_rate": 2.9032333333333333e-05, + "loss": 0.0145, + "step": 12907 + }, + { + "epoch": 10.902027027027026, + "grad_norm": 0.2517739236354828, + "learning_rate": 2.9032e-05, + "loss": 0.0257, + "step": 12908 + }, + { + "epoch": 10.902871621621621, + "grad_norm": 0.2646135091781616, + "learning_rate": 2.903166666666667e-05, + "loss": 0.0283, + "step": 12909 + }, + { + "epoch": 10.903716216216216, + "grad_norm": 0.2615620791912079, + "learning_rate": 2.9031333333333334e-05, + "loss": 0.0189, + "step": 12910 + }, + { + "epoch": 10.90456081081081, + "grad_norm": 0.47005942463874817, + "learning_rate": 2.9031e-05, + "loss": 0.1126, + "step": 12911 + }, + { + "epoch": 10.905405405405405, + "grad_norm": 0.24402165412902832, + "learning_rate": 2.903066666666667e-05, + "loss": 0.0242, + "step": 12912 + }, + { + "epoch": 10.90625, + "grad_norm": 0.292060524225235, + "learning_rate": 2.9030333333333332e-05, + "loss": 0.0324, + "step": 12913 + }, + { + "epoch": 10.907094594594595, + "grad_norm": 0.30748727917671204, + "learning_rate": 2.903e-05, + "loss": 0.0208, + "step": 12914 + }, + { + "epoch": 10.90793918918919, + "grad_norm": 0.28650084137916565, + "learning_rate": 2.9029666666666668e-05, + "loss": 0.0239, + "step": 12915 + }, + { + "epoch": 10.908783783783784, + "grad_norm": 0.2907694876194, + "learning_rate": 2.9029333333333333e-05, + "loss": 0.0232, + "step": 12916 + }, + { + "epoch": 10.909628378378379, + "grad_norm": 0.34407052397727966, + "learning_rate": 2.9029000000000003e-05, + "loss": 0.0258, + "step": 12917 + }, + { + "epoch": 10.910472972972974, + "grad_norm": 0.29245293140411377, + "learning_rate": 2.902866666666667e-05, + "loss": 0.0301, + "step": 12918 + }, + { + "epoch": 10.911317567567568, + "grad_norm": 0.2615477740764618, + "learning_rate": 2.9028333333333335e-05, + "loss": 0.02, + "step": 12919 + }, + { + "epoch": 10.912162162162161, + "grad_norm": 0.39302346110343933, + "learning_rate": 2.9028e-05, + "loss": 0.0277, + "step": 12920 + }, + { + "epoch": 10.913006756756756, + "grad_norm": 0.2716119587421417, + "learning_rate": 2.902766666666667e-05, + "loss": 0.024, + "step": 12921 + }, + { + "epoch": 10.91385135135135, + "grad_norm": 0.4313150644302368, + "learning_rate": 2.9027333333333332e-05, + "loss": 0.0279, + "step": 12922 + }, + { + "epoch": 10.914695945945946, + "grad_norm": 0.30989402532577515, + "learning_rate": 2.9027000000000002e-05, + "loss": 0.0284, + "step": 12923 + }, + { + "epoch": 10.91554054054054, + "grad_norm": 0.34687575697898865, + "learning_rate": 2.9026666666666664e-05, + "loss": 0.0293, + "step": 12924 + }, + { + "epoch": 10.916385135135135, + "grad_norm": 0.3585902750492096, + "learning_rate": 2.9026333333333334e-05, + "loss": 0.0149, + "step": 12925 + }, + { + "epoch": 10.91722972972973, + "grad_norm": 0.31095439195632935, + "learning_rate": 2.9026000000000003e-05, + "loss": 0.0254, + "step": 12926 + }, + { + "epoch": 10.918074324324325, + "grad_norm": 0.3344234526157379, + "learning_rate": 2.9025666666666666e-05, + "loss": 0.0207, + "step": 12927 + }, + { + "epoch": 10.91891891891892, + "grad_norm": 0.433745801448822, + "learning_rate": 2.9025333333333335e-05, + "loss": 0.068, + "step": 12928 + }, + { + "epoch": 10.919763513513514, + "grad_norm": 0.23253023624420166, + "learning_rate": 2.9025e-05, + "loss": 0.0239, + "step": 12929 + }, + { + "epoch": 10.920608108108109, + "grad_norm": 0.3090877830982208, + "learning_rate": 2.9024666666666667e-05, + "loss": 0.0193, + "step": 12930 + }, + { + "epoch": 10.921452702702704, + "grad_norm": 0.7797053456306458, + "learning_rate": 2.9024333333333333e-05, + "loss": 0.0355, + "step": 12931 + }, + { + "epoch": 10.922297297297296, + "grad_norm": 0.7398427724838257, + "learning_rate": 2.9024000000000002e-05, + "loss": 0.0284, + "step": 12932 + }, + { + "epoch": 10.923141891891891, + "grad_norm": 0.25028541684150696, + "learning_rate": 2.9023666666666665e-05, + "loss": 0.0197, + "step": 12933 + }, + { + "epoch": 10.923986486486486, + "grad_norm": 0.5617284774780273, + "learning_rate": 2.9023333333333334e-05, + "loss": 0.0202, + "step": 12934 + }, + { + "epoch": 10.92483108108108, + "grad_norm": 0.4660979211330414, + "learning_rate": 2.9023000000000003e-05, + "loss": 0.022, + "step": 12935 + }, + { + "epoch": 10.925675675675675, + "grad_norm": 0.5298300385475159, + "learning_rate": 2.9022666666666666e-05, + "loss": 0.0286, + "step": 12936 + }, + { + "epoch": 10.92652027027027, + "grad_norm": 0.34113195538520813, + "learning_rate": 2.9022333333333335e-05, + "loss": 0.023, + "step": 12937 + }, + { + "epoch": 10.927364864864865, + "grad_norm": 0.7230084538459778, + "learning_rate": 2.9022e-05, + "loss": 0.0268, + "step": 12938 + }, + { + "epoch": 10.92820945945946, + "grad_norm": 0.68904048204422, + "learning_rate": 2.9021666666666667e-05, + "loss": 0.0353, + "step": 12939 + }, + { + "epoch": 10.929054054054054, + "grad_norm": 2.8735570907592773, + "learning_rate": 2.9021333333333333e-05, + "loss": 0.0679, + "step": 12940 + }, + { + "epoch": 10.92989864864865, + "grad_norm": 0.5993403792381287, + "learning_rate": 2.9021000000000002e-05, + "loss": 0.2311, + "step": 12941 + }, + { + "epoch": 10.930743243243244, + "grad_norm": 0.8288267254829407, + "learning_rate": 2.9020666666666668e-05, + "loss": 0.171, + "step": 12942 + }, + { + "epoch": 10.931587837837839, + "grad_norm": 0.45730385184288025, + "learning_rate": 2.9020333333333334e-05, + "loss": 0.0857, + "step": 12943 + }, + { + "epoch": 10.932432432432432, + "grad_norm": 0.5529173016548157, + "learning_rate": 2.9020000000000003e-05, + "loss": 0.0825, + "step": 12944 + }, + { + "epoch": 10.933277027027026, + "grad_norm": 0.474933922290802, + "learning_rate": 2.9019666666666666e-05, + "loss": 0.0371, + "step": 12945 + }, + { + "epoch": 10.934121621621621, + "grad_norm": 0.23538659512996674, + "learning_rate": 2.9019333333333335e-05, + "loss": 0.0231, + "step": 12946 + }, + { + "epoch": 10.934966216216216, + "grad_norm": 0.47416260838508606, + "learning_rate": 2.9019e-05, + "loss": 0.037, + "step": 12947 + }, + { + "epoch": 10.93581081081081, + "grad_norm": 0.39667263627052307, + "learning_rate": 2.9018666666666667e-05, + "loss": 0.026, + "step": 12948 + }, + { + "epoch": 10.936655405405405, + "grad_norm": 0.3090662360191345, + "learning_rate": 2.9018333333333333e-05, + "loss": 0.0311, + "step": 12949 + }, + { + "epoch": 10.9375, + "grad_norm": 0.5291633009910583, + "learning_rate": 2.9018000000000002e-05, + "loss": 0.031, + "step": 12950 + }, + { + "epoch": 10.938344594594595, + "grad_norm": 0.3731946051120758, + "learning_rate": 2.9017666666666668e-05, + "loss": 0.0258, + "step": 12951 + }, + { + "epoch": 10.93918918918919, + "grad_norm": 0.3258332312107086, + "learning_rate": 2.9017333333333334e-05, + "loss": 0.0188, + "step": 12952 + }, + { + "epoch": 10.940033783783784, + "grad_norm": 0.44655099511146545, + "learning_rate": 2.9017e-05, + "loss": 0.024, + "step": 12953 + }, + { + "epoch": 10.940878378378379, + "grad_norm": 0.20809654891490936, + "learning_rate": 2.9016666666666666e-05, + "loss": 0.0196, + "step": 12954 + }, + { + "epoch": 10.941722972972974, + "grad_norm": 0.33591458201408386, + "learning_rate": 2.9016333333333335e-05, + "loss": 0.0336, + "step": 12955 + }, + { + "epoch": 10.942567567567568, + "grad_norm": 0.22067435085773468, + "learning_rate": 2.9015999999999998e-05, + "loss": 0.0179, + "step": 12956 + }, + { + "epoch": 10.943412162162161, + "grad_norm": 0.2255731076002121, + "learning_rate": 2.9015666666666667e-05, + "loss": 0.0215, + "step": 12957 + }, + { + "epoch": 10.944256756756756, + "grad_norm": 0.23849934339523315, + "learning_rate": 2.9015333333333333e-05, + "loss": 0.0254, + "step": 12958 + }, + { + "epoch": 10.94510135135135, + "grad_norm": 0.3089723289012909, + "learning_rate": 2.9015e-05, + "loss": 0.0432, + "step": 12959 + }, + { + "epoch": 10.945945945945946, + "grad_norm": 0.5168803930282593, + "learning_rate": 2.901466666666667e-05, + "loss": 0.0332, + "step": 12960 + }, + { + "epoch": 10.94679054054054, + "grad_norm": 0.3839878737926483, + "learning_rate": 2.9014333333333334e-05, + "loss": 0.0374, + "step": 12961 + }, + { + "epoch": 10.947635135135135, + "grad_norm": 0.5386906862258911, + "learning_rate": 2.9014e-05, + "loss": 0.032, + "step": 12962 + }, + { + "epoch": 10.94847972972973, + "grad_norm": 0.4027230441570282, + "learning_rate": 2.9013666666666666e-05, + "loss": 0.0251, + "step": 12963 + }, + { + "epoch": 10.949324324324325, + "grad_norm": 0.2271745353937149, + "learning_rate": 2.9013333333333336e-05, + "loss": 0.0174, + "step": 12964 + }, + { + "epoch": 10.95016891891892, + "grad_norm": 0.32971563935279846, + "learning_rate": 2.9012999999999998e-05, + "loss": 0.0341, + "step": 12965 + }, + { + "epoch": 10.951013513513514, + "grad_norm": 0.2861926555633545, + "learning_rate": 2.9012666666666668e-05, + "loss": 0.0256, + "step": 12966 + }, + { + "epoch": 10.951858108108109, + "grad_norm": 0.30563709139823914, + "learning_rate": 2.9012333333333337e-05, + "loss": 0.0182, + "step": 12967 + }, + { + "epoch": 10.952702702702704, + "grad_norm": 0.4720655679702759, + "learning_rate": 2.9012e-05, + "loss": 0.0245, + "step": 12968 + }, + { + "epoch": 10.953547297297296, + "grad_norm": 0.23523834347724915, + "learning_rate": 2.901166666666667e-05, + "loss": 0.0209, + "step": 12969 + }, + { + "epoch": 10.954391891891891, + "grad_norm": 0.32697659730911255, + "learning_rate": 2.9011333333333335e-05, + "loss": 0.0316, + "step": 12970 + }, + { + "epoch": 10.955236486486486, + "grad_norm": 0.3401971459388733, + "learning_rate": 2.9011e-05, + "loss": 0.0249, + "step": 12971 + }, + { + "epoch": 10.95608108108108, + "grad_norm": 0.3692456781864166, + "learning_rate": 2.9010666666666667e-05, + "loss": 0.0271, + "step": 12972 + }, + { + "epoch": 10.956925675675675, + "grad_norm": 0.44006168842315674, + "learning_rate": 2.9010333333333336e-05, + "loss": 0.0217, + "step": 12973 + }, + { + "epoch": 10.95777027027027, + "grad_norm": 0.5450272560119629, + "learning_rate": 2.901e-05, + "loss": 0.033, + "step": 12974 + }, + { + "epoch": 10.958614864864865, + "grad_norm": 0.3640427589416504, + "learning_rate": 2.9009666666666668e-05, + "loss": 0.0186, + "step": 12975 + }, + { + "epoch": 10.95945945945946, + "grad_norm": 0.4115029275417328, + "learning_rate": 2.9009333333333337e-05, + "loss": 0.0404, + "step": 12976 + }, + { + "epoch": 10.960304054054054, + "grad_norm": 0.3066667318344116, + "learning_rate": 2.9009e-05, + "loss": 0.0358, + "step": 12977 + }, + { + "epoch": 10.96114864864865, + "grad_norm": 0.47105729579925537, + "learning_rate": 2.900866666666667e-05, + "loss": 0.024, + "step": 12978 + }, + { + "epoch": 10.961993243243244, + "grad_norm": 0.30846017599105835, + "learning_rate": 2.9008333333333335e-05, + "loss": 0.0209, + "step": 12979 + }, + { + "epoch": 10.962837837837839, + "grad_norm": 0.4006567597389221, + "learning_rate": 2.9008e-05, + "loss": 0.0305, + "step": 12980 + }, + { + "epoch": 10.963682432432432, + "grad_norm": 0.358094185590744, + "learning_rate": 2.9007666666666667e-05, + "loss": 0.023, + "step": 12981 + }, + { + "epoch": 10.964527027027026, + "grad_norm": 0.282778799533844, + "learning_rate": 2.9007333333333336e-05, + "loss": 0.0209, + "step": 12982 + }, + { + "epoch": 10.965371621621621, + "grad_norm": 0.4195542335510254, + "learning_rate": 2.9007e-05, + "loss": 0.0194, + "step": 12983 + }, + { + "epoch": 10.966216216216216, + "grad_norm": 0.2738286554813385, + "learning_rate": 2.9006666666666668e-05, + "loss": 0.0188, + "step": 12984 + }, + { + "epoch": 10.96706081081081, + "grad_norm": 0.3047526478767395, + "learning_rate": 2.9006333333333334e-05, + "loss": 0.0262, + "step": 12985 + }, + { + "epoch": 10.967905405405405, + "grad_norm": 0.6347941756248474, + "learning_rate": 2.9006e-05, + "loss": 0.0408, + "step": 12986 + }, + { + "epoch": 10.96875, + "grad_norm": 0.538526713848114, + "learning_rate": 2.900566666666667e-05, + "loss": 0.0388, + "step": 12987 + }, + { + "epoch": 10.969594594594595, + "grad_norm": 0.4803716838359833, + "learning_rate": 2.900533333333333e-05, + "loss": 0.0283, + "step": 12988 + }, + { + "epoch": 10.97043918918919, + "grad_norm": 0.43449562788009644, + "learning_rate": 2.9005e-05, + "loss": 0.025, + "step": 12989 + }, + { + "epoch": 10.971283783783784, + "grad_norm": 0.49529191851615906, + "learning_rate": 2.9004666666666667e-05, + "loss": 0.0219, + "step": 12990 + }, + { + "epoch": 10.972128378378379, + "grad_norm": 0.6800267100334167, + "learning_rate": 2.9004333333333333e-05, + "loss": 0.1945, + "step": 12991 + }, + { + "epoch": 10.972972972972974, + "grad_norm": 0.5419362187385559, + "learning_rate": 2.9004000000000002e-05, + "loss": 0.0832, + "step": 12992 + }, + { + "epoch": 10.973817567567568, + "grad_norm": 0.2953452169895172, + "learning_rate": 2.9003666666666668e-05, + "loss": 0.0268, + "step": 12993 + }, + { + "epoch": 10.974662162162161, + "grad_norm": 0.383439302444458, + "learning_rate": 2.9003333333333334e-05, + "loss": 0.0318, + "step": 12994 + }, + { + "epoch": 10.975506756756756, + "grad_norm": 0.5055800676345825, + "learning_rate": 2.9003e-05, + "loss": 0.0443, + "step": 12995 + }, + { + "epoch": 10.97635135135135, + "grad_norm": 0.41903233528137207, + "learning_rate": 2.900266666666667e-05, + "loss": 0.0372, + "step": 12996 + }, + { + "epoch": 10.977195945945946, + "grad_norm": 0.25429970026016235, + "learning_rate": 2.9002333333333332e-05, + "loss": 0.0246, + "step": 12997 + }, + { + "epoch": 10.97804054054054, + "grad_norm": 0.2645382285118103, + "learning_rate": 2.9002e-05, + "loss": 0.0257, + "step": 12998 + }, + { + "epoch": 10.978885135135135, + "grad_norm": 0.23204098641872406, + "learning_rate": 2.9001666666666667e-05, + "loss": 0.0191, + "step": 12999 + }, + { + "epoch": 10.97972972972973, + "grad_norm": 0.3564077317714691, + "learning_rate": 2.9001333333333333e-05, + "loss": 0.0274, + "step": 13000 + }, + { + "epoch": 10.97972972972973, + "eval_cer": 0.09220142960404787, + "eval_loss": 0.25424641370773315, + "eval_runtime": 8.4928, + "eval_samples_per_second": 11.892, + "eval_steps_per_second": 0.471, + "eval_wer": 0.27821869488536155, + "step": 13000 + }, + { + "epoch": 10.980574324324325, + "grad_norm": 0.469705194234848, + "learning_rate": 2.9001000000000002e-05, + "loss": 0.0299, + "step": 13001 + }, + { + "epoch": 10.98141891891892, + "grad_norm": 0.2777026295661926, + "learning_rate": 2.9000666666666668e-05, + "loss": 0.0336, + "step": 13002 + }, + { + "epoch": 10.982263513513514, + "grad_norm": 0.3871004581451416, + "learning_rate": 2.9000333333333334e-05, + "loss": 0.0189, + "step": 13003 + }, + { + "epoch": 10.983108108108109, + "grad_norm": 0.4260302484035492, + "learning_rate": 2.9e-05, + "loss": 0.0267, + "step": 13004 + }, + { + "epoch": 10.983952702702704, + "grad_norm": 0.30915549397468567, + "learning_rate": 2.899966666666667e-05, + "loss": 0.033, + "step": 13005 + }, + { + "epoch": 10.984797297297296, + "grad_norm": 0.27798277139663696, + "learning_rate": 2.8999333333333332e-05, + "loss": 0.0226, + "step": 13006 + }, + { + "epoch": 10.985641891891891, + "grad_norm": 0.478039413690567, + "learning_rate": 2.8999e-05, + "loss": 0.0282, + "step": 13007 + }, + { + "epoch": 10.986486486486486, + "grad_norm": 0.2692088782787323, + "learning_rate": 2.8998666666666667e-05, + "loss": 0.0293, + "step": 13008 + }, + { + "epoch": 10.98733108108108, + "grad_norm": 0.27698543667793274, + "learning_rate": 2.8998333333333333e-05, + "loss": 0.0226, + "step": 13009 + }, + { + "epoch": 10.988175675675675, + "grad_norm": 0.5040289759635925, + "learning_rate": 2.8998000000000003e-05, + "loss": 0.0431, + "step": 13010 + }, + { + "epoch": 10.98902027027027, + "grad_norm": 0.3771674633026123, + "learning_rate": 2.899766666666667e-05, + "loss": 0.0318, + "step": 13011 + }, + { + "epoch": 10.989864864864865, + "grad_norm": 0.4419075548648834, + "learning_rate": 2.8997333333333334e-05, + "loss": 0.0232, + "step": 13012 + }, + { + "epoch": 10.99070945945946, + "grad_norm": 0.3358815908432007, + "learning_rate": 2.8997e-05, + "loss": 0.0371, + "step": 13013 + }, + { + "epoch": 10.991554054054054, + "grad_norm": 0.2987058162689209, + "learning_rate": 2.8996666666666666e-05, + "loss": 0.028, + "step": 13014 + }, + { + "epoch": 10.99239864864865, + "grad_norm": 0.7163323163986206, + "learning_rate": 2.8996333333333332e-05, + "loss": 0.0299, + "step": 13015 + }, + { + "epoch": 10.993243243243244, + "grad_norm": 0.3455723822116852, + "learning_rate": 2.8996e-05, + "loss": 0.0278, + "step": 13016 + }, + { + "epoch": 10.994087837837839, + "grad_norm": 0.3769383132457733, + "learning_rate": 2.8995666666666667e-05, + "loss": 0.0226, + "step": 13017 + }, + { + "epoch": 10.994932432432432, + "grad_norm": 0.43193334341049194, + "learning_rate": 2.8995333333333333e-05, + "loss": 0.0312, + "step": 13018 + }, + { + "epoch": 10.995777027027026, + "grad_norm": 0.2396862506866455, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0122, + "step": 13019 + }, + { + "epoch": 10.996621621621621, + "grad_norm": 0.4375292658805847, + "learning_rate": 2.8994666666666665e-05, + "loss": 0.0294, + "step": 13020 + }, + { + "epoch": 10.997466216216216, + "grad_norm": 0.6287668943405151, + "learning_rate": 2.8994333333333335e-05, + "loss": 0.0355, + "step": 13021 + }, + { + "epoch": 10.99831081081081, + "grad_norm": 0.6668413877487183, + "learning_rate": 2.8994e-05, + "loss": 0.0329, + "step": 13022 + }, + { + "epoch": 10.999155405405405, + "grad_norm": 0.7063038349151611, + "learning_rate": 2.8993666666666667e-05, + "loss": 0.0524, + "step": 13023 + }, + { + "epoch": 11.0, + "grad_norm": 0.4728873074054718, + "learning_rate": 2.8993333333333332e-05, + "loss": 0.0348, + "step": 13024 + }, + { + "epoch": 11.000844594594595, + "grad_norm": 0.8206998109817505, + "learning_rate": 2.8993000000000002e-05, + "loss": 0.2306, + "step": 13025 + }, + { + "epoch": 11.00168918918919, + "grad_norm": 0.4112188220024109, + "learning_rate": 2.8992666666666668e-05, + "loss": 0.1368, + "step": 13026 + }, + { + "epoch": 11.002533783783784, + "grad_norm": 0.34764280915260315, + "learning_rate": 2.8992333333333334e-05, + "loss": 0.073, + "step": 13027 + }, + { + "epoch": 11.003378378378379, + "grad_norm": 0.35122010111808777, + "learning_rate": 2.8992000000000003e-05, + "loss": 0.0435, + "step": 13028 + }, + { + "epoch": 11.004222972972974, + "grad_norm": 0.3996942937374115, + "learning_rate": 2.8991666666666666e-05, + "loss": 0.0369, + "step": 13029 + }, + { + "epoch": 11.005067567567568, + "grad_norm": 0.35660549998283386, + "learning_rate": 2.8991333333333335e-05, + "loss": 0.019, + "step": 13030 + }, + { + "epoch": 11.005912162162161, + "grad_norm": 0.20216071605682373, + "learning_rate": 2.8991e-05, + "loss": 0.0309, + "step": 13031 + }, + { + "epoch": 11.006756756756756, + "grad_norm": 0.21390579640865326, + "learning_rate": 2.8990666666666667e-05, + "loss": 0.0202, + "step": 13032 + }, + { + "epoch": 11.00760135135135, + "grad_norm": 0.27049487829208374, + "learning_rate": 2.8990333333333333e-05, + "loss": 0.0176, + "step": 13033 + }, + { + "epoch": 11.008445945945946, + "grad_norm": 0.27660825848579407, + "learning_rate": 2.8990000000000002e-05, + "loss": 0.0294, + "step": 13034 + }, + { + "epoch": 11.00929054054054, + "grad_norm": 0.19729627668857574, + "learning_rate": 2.8989666666666668e-05, + "loss": 0.0258, + "step": 13035 + }, + { + "epoch": 11.010135135135135, + "grad_norm": 0.2448125183582306, + "learning_rate": 2.8989333333333334e-05, + "loss": 0.0191, + "step": 13036 + }, + { + "epoch": 11.01097972972973, + "grad_norm": 0.18902646005153656, + "learning_rate": 2.8989000000000003e-05, + "loss": 0.0159, + "step": 13037 + }, + { + "epoch": 11.011824324324325, + "grad_norm": 0.3622312545776367, + "learning_rate": 2.8988666666666666e-05, + "loss": 0.0246, + "step": 13038 + }, + { + "epoch": 11.01266891891892, + "grad_norm": 0.2716869115829468, + "learning_rate": 2.8988333333333335e-05, + "loss": 0.035, + "step": 13039 + }, + { + "epoch": 11.013513513513514, + "grad_norm": 0.274217426776886, + "learning_rate": 2.8988e-05, + "loss": 0.0154, + "step": 13040 + }, + { + "epoch": 11.014358108108109, + "grad_norm": 0.28907567262649536, + "learning_rate": 2.8987666666666667e-05, + "loss": 0.0211, + "step": 13041 + }, + { + "epoch": 11.015202702702704, + "grad_norm": 0.19959616661071777, + "learning_rate": 2.8987333333333336e-05, + "loss": 0.0188, + "step": 13042 + }, + { + "epoch": 11.016047297297296, + "grad_norm": 0.3720344603061676, + "learning_rate": 2.8987000000000002e-05, + "loss": 0.0204, + "step": 13043 + }, + { + "epoch": 11.016891891891891, + "grad_norm": 0.22467553615570068, + "learning_rate": 2.8986666666666668e-05, + "loss": 0.0153, + "step": 13044 + }, + { + "epoch": 11.017736486486486, + "grad_norm": 0.25467649102211, + "learning_rate": 2.8986333333333334e-05, + "loss": 0.0167, + "step": 13045 + }, + { + "epoch": 11.01858108108108, + "grad_norm": 0.16498024761676788, + "learning_rate": 2.8986e-05, + "loss": 0.0104, + "step": 13046 + }, + { + "epoch": 11.019425675675675, + "grad_norm": 0.2197941094636917, + "learning_rate": 2.8985666666666666e-05, + "loss": 0.0182, + "step": 13047 + }, + { + "epoch": 11.02027027027027, + "grad_norm": 0.11598673462867737, + "learning_rate": 2.8985333333333335e-05, + "loss": 0.0081, + "step": 13048 + }, + { + "epoch": 11.021114864864865, + "grad_norm": 0.22217901051044464, + "learning_rate": 2.8984999999999998e-05, + "loss": 0.0178, + "step": 13049 + }, + { + "epoch": 11.02195945945946, + "grad_norm": 0.3891279399394989, + "learning_rate": 2.8984666666666667e-05, + "loss": 0.0162, + "step": 13050 + }, + { + "epoch": 11.022804054054054, + "grad_norm": 0.4678122103214264, + "learning_rate": 2.8984333333333336e-05, + "loss": 0.0279, + "step": 13051 + }, + { + "epoch": 11.02364864864865, + "grad_norm": 0.32499396800994873, + "learning_rate": 2.8984e-05, + "loss": 0.0197, + "step": 13052 + }, + { + "epoch": 11.024493243243244, + "grad_norm": 0.3669408857822418, + "learning_rate": 2.8983666666666668e-05, + "loss": 0.0269, + "step": 13053 + }, + { + "epoch": 11.025337837837839, + "grad_norm": 0.3288588225841522, + "learning_rate": 2.8983333333333334e-05, + "loss": 0.0221, + "step": 13054 + }, + { + "epoch": 11.026182432432432, + "grad_norm": 0.33171316981315613, + "learning_rate": 2.8983e-05, + "loss": 0.0161, + "step": 13055 + }, + { + "epoch": 11.027027027027026, + "grad_norm": 0.2623145878314972, + "learning_rate": 2.8982666666666666e-05, + "loss": 0.0225, + "step": 13056 + }, + { + "epoch": 11.027871621621621, + "grad_norm": 0.21896684169769287, + "learning_rate": 2.8982333333333335e-05, + "loss": 0.0193, + "step": 13057 + }, + { + "epoch": 11.028716216216216, + "grad_norm": 1.104914665222168, + "learning_rate": 2.8981999999999998e-05, + "loss": 0.0284, + "step": 13058 + }, + { + "epoch": 11.02956081081081, + "grad_norm": 0.30104026198387146, + "learning_rate": 2.8981666666666667e-05, + "loss": 0.0758, + "step": 13059 + }, + { + "epoch": 11.030405405405405, + "grad_norm": 0.22669166326522827, + "learning_rate": 2.8981333333333337e-05, + "loss": 0.0086, + "step": 13060 + }, + { + "epoch": 11.03125, + "grad_norm": 0.6289067268371582, + "learning_rate": 2.8981e-05, + "loss": 0.0106, + "step": 13061 + }, + { + "epoch": 11.032094594594595, + "grad_norm": 0.24595166742801666, + "learning_rate": 2.898066666666667e-05, + "loss": 0.0149, + "step": 13062 + }, + { + "epoch": 11.03293918918919, + "grad_norm": 0.3919391632080078, + "learning_rate": 2.8980333333333334e-05, + "loss": 0.0266, + "step": 13063 + }, + { + "epoch": 11.033783783783784, + "grad_norm": 0.27026036381721497, + "learning_rate": 2.898e-05, + "loss": 0.012, + "step": 13064 + }, + { + "epoch": 11.034628378378379, + "grad_norm": 0.2775316536426544, + "learning_rate": 2.8979666666666666e-05, + "loss": 0.0186, + "step": 13065 + }, + { + "epoch": 11.035472972972974, + "grad_norm": 0.27944642305374146, + "learning_rate": 2.8979333333333336e-05, + "loss": 0.0215, + "step": 13066 + }, + { + "epoch": 11.036317567567568, + "grad_norm": 0.27606579661369324, + "learning_rate": 2.8979e-05, + "loss": 0.028, + "step": 13067 + }, + { + "epoch": 11.037162162162161, + "grad_norm": 0.5063722729682922, + "learning_rate": 2.8978666666666667e-05, + "loss": 0.0184, + "step": 13068 + }, + { + "epoch": 11.038006756756756, + "grad_norm": 0.3194851875305176, + "learning_rate": 2.8978333333333337e-05, + "loss": 0.0101, + "step": 13069 + }, + { + "epoch": 11.03885135135135, + "grad_norm": 1.3386019468307495, + "learning_rate": 2.8978e-05, + "loss": 0.0427, + "step": 13070 + }, + { + "epoch": 11.039695945945946, + "grad_norm": 0.4686077833175659, + "learning_rate": 2.897766666666667e-05, + "loss": 0.0225, + "step": 13071 + }, + { + "epoch": 11.04054054054054, + "grad_norm": 0.5034211874008179, + "learning_rate": 2.8977333333333335e-05, + "loss": 0.0159, + "step": 13072 + }, + { + "epoch": 11.041385135135135, + "grad_norm": 1.1613564491271973, + "learning_rate": 2.8977e-05, + "loss": 0.0523, + "step": 13073 + }, + { + "epoch": 11.04222972972973, + "grad_norm": 4.728181838989258, + "learning_rate": 2.8976666666666666e-05, + "loss": 0.0273, + "step": 13074 + }, + { + "epoch": 11.043074324324325, + "grad_norm": 0.6401835680007935, + "learning_rate": 2.8976333333333332e-05, + "loss": 0.2218, + "step": 13075 + }, + { + "epoch": 11.04391891891892, + "grad_norm": 0.36767518520355225, + "learning_rate": 2.8976000000000002e-05, + "loss": 0.1191, + "step": 13076 + }, + { + "epoch": 11.044763513513514, + "grad_norm": 0.37062570452690125, + "learning_rate": 2.8975666666666668e-05, + "loss": 0.0633, + "step": 13077 + }, + { + "epoch": 11.045608108108109, + "grad_norm": 0.5015704035758972, + "learning_rate": 2.8975333333333334e-05, + "loss": 0.0492, + "step": 13078 + }, + { + "epoch": 11.046452702702704, + "grad_norm": 0.3018023371696472, + "learning_rate": 2.8975e-05, + "loss": 0.0275, + "step": 13079 + }, + { + "epoch": 11.047297297297296, + "grad_norm": 0.3260855972766876, + "learning_rate": 2.897466666666667e-05, + "loss": 0.0283, + "step": 13080 + }, + { + "epoch": 11.048141891891891, + "grad_norm": 0.3540445864200592, + "learning_rate": 2.897433333333333e-05, + "loss": 0.0431, + "step": 13081 + }, + { + "epoch": 11.048986486486486, + "grad_norm": 0.257798433303833, + "learning_rate": 2.8974e-05, + "loss": 0.0113, + "step": 13082 + }, + { + "epoch": 11.04983108108108, + "grad_norm": 0.2005210518836975, + "learning_rate": 2.8973666666666667e-05, + "loss": 0.0194, + "step": 13083 + }, + { + "epoch": 11.050675675675675, + "grad_norm": 0.31737297773361206, + "learning_rate": 2.8973333333333333e-05, + "loss": 0.0275, + "step": 13084 + }, + { + "epoch": 11.05152027027027, + "grad_norm": 0.2265988439321518, + "learning_rate": 2.8973000000000002e-05, + "loss": 0.014, + "step": 13085 + }, + { + "epoch": 11.052364864864865, + "grad_norm": 0.22200024127960205, + "learning_rate": 2.8972666666666668e-05, + "loss": 0.0149, + "step": 13086 + }, + { + "epoch": 11.05320945945946, + "grad_norm": 0.4462668299674988, + "learning_rate": 2.8972333333333334e-05, + "loss": 0.0298, + "step": 13087 + }, + { + "epoch": 11.054054054054054, + "grad_norm": 0.35795775055885315, + "learning_rate": 2.8972e-05, + "loss": 0.0291, + "step": 13088 + }, + { + "epoch": 11.05489864864865, + "grad_norm": 0.18022702634334564, + "learning_rate": 2.897166666666667e-05, + "loss": 0.0201, + "step": 13089 + }, + { + "epoch": 11.055743243243244, + "grad_norm": 0.19151952862739563, + "learning_rate": 2.897133333333333e-05, + "loss": 0.0201, + "step": 13090 + }, + { + "epoch": 11.056587837837839, + "grad_norm": 0.25240829586982727, + "learning_rate": 2.8971e-05, + "loss": 0.0276, + "step": 13091 + }, + { + "epoch": 11.057432432432432, + "grad_norm": 0.21674397587776184, + "learning_rate": 2.897066666666667e-05, + "loss": 0.0212, + "step": 13092 + }, + { + "epoch": 11.058277027027026, + "grad_norm": 0.19063976407051086, + "learning_rate": 2.8970333333333333e-05, + "loss": 0.0122, + "step": 13093 + }, + { + "epoch": 11.059121621621621, + "grad_norm": 0.40862569212913513, + "learning_rate": 2.8970000000000002e-05, + "loss": 0.0281, + "step": 13094 + }, + { + "epoch": 11.059966216216216, + "grad_norm": 0.7099766135215759, + "learning_rate": 2.8969666666666668e-05, + "loss": 0.0233, + "step": 13095 + }, + { + "epoch": 11.06081081081081, + "grad_norm": 0.46318504214286804, + "learning_rate": 2.8969333333333334e-05, + "loss": 0.0211, + "step": 13096 + }, + { + "epoch": 11.061655405405405, + "grad_norm": 0.19414934515953064, + "learning_rate": 2.8969e-05, + "loss": 0.02, + "step": 13097 + }, + { + "epoch": 11.0625, + "grad_norm": 0.16527721285820007, + "learning_rate": 2.896866666666667e-05, + "loss": 0.015, + "step": 13098 + }, + { + "epoch": 11.063344594594595, + "grad_norm": 0.4462612569332123, + "learning_rate": 2.8968333333333332e-05, + "loss": 0.0142, + "step": 13099 + }, + { + "epoch": 11.06418918918919, + "grad_norm": 0.32744696736335754, + "learning_rate": 2.8968e-05, + "loss": 0.0171, + "step": 13100 + }, + { + "epoch": 11.065033783783784, + "grad_norm": 0.36895257234573364, + "learning_rate": 2.896766666666667e-05, + "loss": 0.0217, + "step": 13101 + }, + { + "epoch": 11.065878378378379, + "grad_norm": 0.37511709332466125, + "learning_rate": 2.8967333333333333e-05, + "loss": 0.0136, + "step": 13102 + }, + { + "epoch": 11.066722972972974, + "grad_norm": 0.3927084803581238, + "learning_rate": 2.8967000000000002e-05, + "loss": 0.0198, + "step": 13103 + }, + { + "epoch": 11.067567567567568, + "grad_norm": 0.4729727506637573, + "learning_rate": 2.8966666666666665e-05, + "loss": 0.0324, + "step": 13104 + }, + { + "epoch": 11.068412162162161, + "grad_norm": 0.4223637282848358, + "learning_rate": 2.8966333333333334e-05, + "loss": 0.0261, + "step": 13105 + }, + { + "epoch": 11.069256756756756, + "grad_norm": 0.4450232684612274, + "learning_rate": 2.8966e-05, + "loss": 0.0348, + "step": 13106 + }, + { + "epoch": 11.07010135135135, + "grad_norm": 0.25384068489074707, + "learning_rate": 2.8965666666666666e-05, + "loss": 0.0235, + "step": 13107 + }, + { + "epoch": 11.070945945945946, + "grad_norm": 0.6387161016464233, + "learning_rate": 2.8965333333333332e-05, + "loss": 0.0199, + "step": 13108 + }, + { + "epoch": 11.07179054054054, + "grad_norm": 0.5949127078056335, + "learning_rate": 2.8965e-05, + "loss": 0.0264, + "step": 13109 + }, + { + "epoch": 11.072635135135135, + "grad_norm": 0.2899635434150696, + "learning_rate": 2.8964666666666667e-05, + "loss": 0.0171, + "step": 13110 + }, + { + "epoch": 11.07347972972973, + "grad_norm": 0.33717355132102966, + "learning_rate": 2.8964333333333333e-05, + "loss": 0.0265, + "step": 13111 + }, + { + "epoch": 11.074324324324325, + "grad_norm": 0.5981661677360535, + "learning_rate": 2.8964000000000003e-05, + "loss": 0.0301, + "step": 13112 + }, + { + "epoch": 11.07516891891892, + "grad_norm": 0.5142730474472046, + "learning_rate": 2.8963666666666665e-05, + "loss": 0.0303, + "step": 13113 + }, + { + "epoch": 11.076013513513514, + "grad_norm": 0.39685341715812683, + "learning_rate": 2.8963333333333334e-05, + "loss": 0.0338, + "step": 13114 + }, + { + "epoch": 11.076858108108109, + "grad_norm": 0.9923896789550781, + "learning_rate": 2.8963e-05, + "loss": 0.0236, + "step": 13115 + }, + { + "epoch": 11.077702702702704, + "grad_norm": 0.35903602838516235, + "learning_rate": 2.8962666666666666e-05, + "loss": 0.0242, + "step": 13116 + }, + { + "epoch": 11.078547297297296, + "grad_norm": 0.34879064559936523, + "learning_rate": 2.8962333333333336e-05, + "loss": 0.0348, + "step": 13117 + }, + { + "epoch": 11.079391891891891, + "grad_norm": 0.46386051177978516, + "learning_rate": 2.8962e-05, + "loss": 0.015, + "step": 13118 + }, + { + "epoch": 11.080236486486486, + "grad_norm": 0.6351109743118286, + "learning_rate": 2.8961666666666667e-05, + "loss": 0.0422, + "step": 13119 + }, + { + "epoch": 11.08108108108108, + "grad_norm": 0.5890246033668518, + "learning_rate": 2.8961333333333333e-05, + "loss": 0.0211, + "step": 13120 + }, + { + "epoch": 11.081925675675675, + "grad_norm": 0.28302764892578125, + "learning_rate": 2.8961000000000003e-05, + "loss": 0.0225, + "step": 13121 + }, + { + "epoch": 11.08277027027027, + "grad_norm": 0.19336187839508057, + "learning_rate": 2.8960666666666665e-05, + "loss": 0.0137, + "step": 13122 + }, + { + "epoch": 11.083614864864865, + "grad_norm": 1.7451894283294678, + "learning_rate": 2.8960333333333335e-05, + "loss": 0.0936, + "step": 13123 + }, + { + "epoch": 11.08445945945946, + "grad_norm": 0.5280264019966125, + "learning_rate": 2.896e-05, + "loss": 0.0567, + "step": 13124 + }, + { + "epoch": 11.085304054054054, + "grad_norm": 0.5488444566726685, + "learning_rate": 2.8959666666666666e-05, + "loss": 0.2247, + "step": 13125 + }, + { + "epoch": 11.08614864864865, + "grad_norm": 0.4668688476085663, + "learning_rate": 2.8959333333333336e-05, + "loss": 0.134, + "step": 13126 + }, + { + "epoch": 11.086993243243244, + "grad_norm": 0.48444026708602905, + "learning_rate": 2.8959000000000002e-05, + "loss": 0.0803, + "step": 13127 + }, + { + "epoch": 11.087837837837839, + "grad_norm": 0.24850842356681824, + "learning_rate": 2.8958666666666668e-05, + "loss": 0.0521, + "step": 13128 + }, + { + "epoch": 11.088682432432432, + "grad_norm": 0.236888587474823, + "learning_rate": 2.8958333333333334e-05, + "loss": 0.0344, + "step": 13129 + }, + { + "epoch": 11.089527027027026, + "grad_norm": 0.4146687388420105, + "learning_rate": 2.8958000000000003e-05, + "loss": 0.0674, + "step": 13130 + }, + { + "epoch": 11.090371621621621, + "grad_norm": 0.3084748089313507, + "learning_rate": 2.8957666666666665e-05, + "loss": 0.0223, + "step": 13131 + }, + { + "epoch": 11.091216216216216, + "grad_norm": 0.24742475152015686, + "learning_rate": 2.8957333333333335e-05, + "loss": 0.0153, + "step": 13132 + }, + { + "epoch": 11.09206081081081, + "grad_norm": 0.25432583689689636, + "learning_rate": 2.8957e-05, + "loss": 0.0196, + "step": 13133 + }, + { + "epoch": 11.092905405405405, + "grad_norm": 0.4209749698638916, + "learning_rate": 2.8956666666666667e-05, + "loss": 0.0291, + "step": 13134 + }, + { + "epoch": 11.09375, + "grad_norm": 0.25628069043159485, + "learning_rate": 2.8956333333333336e-05, + "loss": 0.0245, + "step": 13135 + }, + { + "epoch": 11.094594594594595, + "grad_norm": 0.5024426579475403, + "learning_rate": 2.8956e-05, + "loss": 0.0307, + "step": 13136 + }, + { + "epoch": 11.09543918918919, + "grad_norm": 0.4132632315158844, + "learning_rate": 2.8955666666666668e-05, + "loss": 0.0246, + "step": 13137 + }, + { + "epoch": 11.096283783783784, + "grad_norm": 0.2963131070137024, + "learning_rate": 2.8955333333333334e-05, + "loss": 0.0252, + "step": 13138 + }, + { + "epoch": 11.097128378378379, + "grad_norm": 0.2694861888885498, + "learning_rate": 2.8955e-05, + "loss": 0.0256, + "step": 13139 + }, + { + "epoch": 11.097972972972974, + "grad_norm": 0.334175705909729, + "learning_rate": 2.8954666666666666e-05, + "loss": 0.0207, + "step": 13140 + }, + { + "epoch": 11.098817567567568, + "grad_norm": 0.27350518107414246, + "learning_rate": 2.8954333333333335e-05, + "loss": 0.0185, + "step": 13141 + }, + { + "epoch": 11.099662162162161, + "grad_norm": 0.21673211455345154, + "learning_rate": 2.8954e-05, + "loss": 0.0152, + "step": 13142 + }, + { + "epoch": 11.100506756756756, + "grad_norm": 0.3151170611381531, + "learning_rate": 2.8953666666666667e-05, + "loss": 0.0202, + "step": 13143 + }, + { + "epoch": 11.10135135135135, + "grad_norm": 0.3272322714328766, + "learning_rate": 2.8953333333333336e-05, + "loss": 0.022, + "step": 13144 + }, + { + "epoch": 11.102195945945946, + "grad_norm": 0.2327454537153244, + "learning_rate": 2.8953e-05, + "loss": 0.0168, + "step": 13145 + }, + { + "epoch": 11.10304054054054, + "grad_norm": 0.2699129283428192, + "learning_rate": 2.8952666666666668e-05, + "loss": 0.0132, + "step": 13146 + }, + { + "epoch": 11.103885135135135, + "grad_norm": 0.24824991822242737, + "learning_rate": 2.8952333333333334e-05, + "loss": 0.0196, + "step": 13147 + }, + { + "epoch": 11.10472972972973, + "grad_norm": 0.5986071825027466, + "learning_rate": 2.8952e-05, + "loss": 0.0927, + "step": 13148 + }, + { + "epoch": 11.105574324324325, + "grad_norm": 0.24640750885009766, + "learning_rate": 2.8951666666666666e-05, + "loss": 0.0191, + "step": 13149 + }, + { + "epoch": 11.10641891891892, + "grad_norm": 0.3353199064731598, + "learning_rate": 2.8951333333333335e-05, + "loss": 0.0208, + "step": 13150 + }, + { + "epoch": 11.107263513513514, + "grad_norm": 0.645599901676178, + "learning_rate": 2.8951e-05, + "loss": 0.0345, + "step": 13151 + }, + { + "epoch": 11.108108108108109, + "grad_norm": 0.2505562901496887, + "learning_rate": 2.8950666666666667e-05, + "loss": 0.0241, + "step": 13152 + }, + { + "epoch": 11.108952702702704, + "grad_norm": 0.5409688949584961, + "learning_rate": 2.8950333333333336e-05, + "loss": 0.0267, + "step": 13153 + }, + { + "epoch": 11.109797297297296, + "grad_norm": 0.2638753354549408, + "learning_rate": 2.895e-05, + "loss": 0.0246, + "step": 13154 + }, + { + "epoch": 11.110641891891891, + "grad_norm": 0.21556873619556427, + "learning_rate": 2.8949666666666668e-05, + "loss": 0.0219, + "step": 13155 + }, + { + "epoch": 11.111486486486486, + "grad_norm": 0.3626895844936371, + "learning_rate": 2.8949333333333334e-05, + "loss": 0.0132, + "step": 13156 + }, + { + "epoch": 11.11233108108108, + "grad_norm": 0.23602887988090515, + "learning_rate": 2.8949e-05, + "loss": 0.0217, + "step": 13157 + }, + { + "epoch": 11.113175675675675, + "grad_norm": 0.7708430886268616, + "learning_rate": 2.8948666666666666e-05, + "loss": 0.0257, + "step": 13158 + }, + { + "epoch": 11.11402027027027, + "grad_norm": 0.5378855466842651, + "learning_rate": 2.8948333333333335e-05, + "loss": 0.0219, + "step": 13159 + }, + { + "epoch": 11.114864864864865, + "grad_norm": 0.35599377751350403, + "learning_rate": 2.8948e-05, + "loss": 0.0313, + "step": 13160 + }, + { + "epoch": 11.11570945945946, + "grad_norm": 1.898960828781128, + "learning_rate": 2.8947666666666667e-05, + "loss": 0.0187, + "step": 13161 + }, + { + "epoch": 11.116554054054054, + "grad_norm": 0.37468022108078003, + "learning_rate": 2.8947333333333337e-05, + "loss": 0.029, + "step": 13162 + }, + { + "epoch": 11.11739864864865, + "grad_norm": 0.3299810290336609, + "learning_rate": 2.8947e-05, + "loss": 0.0215, + "step": 13163 + }, + { + "epoch": 11.118243243243244, + "grad_norm": 0.2741609513759613, + "learning_rate": 2.894666666666667e-05, + "loss": 0.0251, + "step": 13164 + }, + { + "epoch": 11.119087837837839, + "grad_norm": 0.27528348565101624, + "learning_rate": 2.894633333333333e-05, + "loss": 0.0279, + "step": 13165 + }, + { + "epoch": 11.119932432432432, + "grad_norm": 0.2619735598564148, + "learning_rate": 2.8946e-05, + "loss": 0.0168, + "step": 13166 + }, + { + "epoch": 11.120777027027026, + "grad_norm": 0.16688357293605804, + "learning_rate": 2.894566666666667e-05, + "loss": 0.0158, + "step": 13167 + }, + { + "epoch": 11.121621621621621, + "grad_norm": 0.3251689076423645, + "learning_rate": 2.8945333333333332e-05, + "loss": 0.0188, + "step": 13168 + }, + { + "epoch": 11.122466216216216, + "grad_norm": 0.5653364062309265, + "learning_rate": 2.8945e-05, + "loss": 0.0735, + "step": 13169 + }, + { + "epoch": 11.12331081081081, + "grad_norm": 0.6996309161186218, + "learning_rate": 2.8944666666666667e-05, + "loss": 0.0244, + "step": 13170 + }, + { + "epoch": 11.124155405405405, + "grad_norm": 1.162320613861084, + "learning_rate": 2.8944333333333333e-05, + "loss": 0.0293, + "step": 13171 + }, + { + "epoch": 11.125, + "grad_norm": 0.9634835720062256, + "learning_rate": 2.8944e-05, + "loss": 0.0426, + "step": 13172 + }, + { + "epoch": 11.125844594594595, + "grad_norm": 0.6759665012359619, + "learning_rate": 2.894366666666667e-05, + "loss": 0.0309, + "step": 13173 + }, + { + "epoch": 11.12668918918919, + "grad_norm": 0.6490093469619751, + "learning_rate": 2.894333333333333e-05, + "loss": 0.0377, + "step": 13174 + }, + { + "epoch": 11.127533783783784, + "grad_norm": 0.4606296122074127, + "learning_rate": 2.8943e-05, + "loss": 0.1859, + "step": 13175 + }, + { + "epoch": 11.128378378378379, + "grad_norm": 0.4785768687725067, + "learning_rate": 2.894266666666667e-05, + "loss": 0.1207, + "step": 13176 + }, + { + "epoch": 11.129222972972974, + "grad_norm": 0.526800274848938, + "learning_rate": 2.8942333333333332e-05, + "loss": 0.0853, + "step": 13177 + }, + { + "epoch": 11.130067567567568, + "grad_norm": 0.33195436000823975, + "learning_rate": 2.8942000000000002e-05, + "loss": 0.035, + "step": 13178 + }, + { + "epoch": 11.130912162162161, + "grad_norm": 0.416847288608551, + "learning_rate": 2.8941666666666668e-05, + "loss": 0.022, + "step": 13179 + }, + { + "epoch": 11.131756756756756, + "grad_norm": 0.4094191789627075, + "learning_rate": 2.8941333333333334e-05, + "loss": 0.0373, + "step": 13180 + }, + { + "epoch": 11.13260135135135, + "grad_norm": 0.440845787525177, + "learning_rate": 2.8941e-05, + "loss": 0.0292, + "step": 13181 + }, + { + "epoch": 11.133445945945946, + "grad_norm": 0.2109796553850174, + "learning_rate": 2.894066666666667e-05, + "loss": 0.0209, + "step": 13182 + }, + { + "epoch": 11.13429054054054, + "grad_norm": 0.3730315864086151, + "learning_rate": 2.894033333333333e-05, + "loss": 0.0304, + "step": 13183 + }, + { + "epoch": 11.135135135135135, + "grad_norm": 0.3483772873878479, + "learning_rate": 2.894e-05, + "loss": 0.0311, + "step": 13184 + }, + { + "epoch": 11.13597972972973, + "grad_norm": 0.492523193359375, + "learning_rate": 2.893966666666667e-05, + "loss": 0.0217, + "step": 13185 + }, + { + "epoch": 11.136824324324325, + "grad_norm": 0.26318806409835815, + "learning_rate": 2.8939333333333333e-05, + "loss": 0.0217, + "step": 13186 + }, + { + "epoch": 11.13766891891892, + "grad_norm": 0.33199435472488403, + "learning_rate": 2.8939000000000002e-05, + "loss": 0.0264, + "step": 13187 + }, + { + "epoch": 11.138513513513514, + "grad_norm": 0.27787795662879944, + "learning_rate": 2.8938666666666668e-05, + "loss": 0.0312, + "step": 13188 + }, + { + "epoch": 11.139358108108109, + "grad_norm": 0.33325013518333435, + "learning_rate": 2.8938333333333334e-05, + "loss": 0.0215, + "step": 13189 + }, + { + "epoch": 11.140202702702704, + "grad_norm": 0.27151021361351013, + "learning_rate": 2.8938e-05, + "loss": 0.0191, + "step": 13190 + }, + { + "epoch": 11.141047297297296, + "grad_norm": 0.31158384680747986, + "learning_rate": 2.893766666666667e-05, + "loss": 0.021, + "step": 13191 + }, + { + "epoch": 11.141891891891891, + "grad_norm": 0.14881101250648499, + "learning_rate": 2.8937333333333335e-05, + "loss": 0.012, + "step": 13192 + }, + { + "epoch": 11.142736486486486, + "grad_norm": 0.4276638329029083, + "learning_rate": 2.8937e-05, + "loss": 0.0268, + "step": 13193 + }, + { + "epoch": 11.14358108108108, + "grad_norm": 0.225076824426651, + "learning_rate": 2.893666666666667e-05, + "loss": 0.0281, + "step": 13194 + }, + { + "epoch": 11.144425675675675, + "grad_norm": 2.6633782386779785, + "learning_rate": 2.8936333333333333e-05, + "loss": 0.027, + "step": 13195 + }, + { + "epoch": 11.14527027027027, + "grad_norm": 0.19960172474384308, + "learning_rate": 2.8936000000000002e-05, + "loss": 0.0195, + "step": 13196 + }, + { + "epoch": 11.146114864864865, + "grad_norm": 0.3818069398403168, + "learning_rate": 2.8935666666666665e-05, + "loss": 0.024, + "step": 13197 + }, + { + "epoch": 11.14695945945946, + "grad_norm": 0.40806737542152405, + "learning_rate": 2.8935333333333334e-05, + "loss": 0.0349, + "step": 13198 + }, + { + "epoch": 11.147804054054054, + "grad_norm": 0.5738896131515503, + "learning_rate": 2.8935e-05, + "loss": 0.0214, + "step": 13199 + }, + { + "epoch": 11.14864864864865, + "grad_norm": 0.23021897673606873, + "learning_rate": 2.8934666666666666e-05, + "loss": 0.0172, + "step": 13200 + }, + { + "epoch": 11.149493243243244, + "grad_norm": 0.30731266736984253, + "learning_rate": 2.8934333333333335e-05, + "loss": 0.0136, + "step": 13201 + }, + { + "epoch": 11.150337837837839, + "grad_norm": 0.2513158917427063, + "learning_rate": 2.8934e-05, + "loss": 0.0223, + "step": 13202 + }, + { + "epoch": 11.151182432432432, + "grad_norm": 0.23287732899188995, + "learning_rate": 2.8933666666666667e-05, + "loss": 0.0143, + "step": 13203 + }, + { + "epoch": 11.152027027027026, + "grad_norm": 0.4396311044692993, + "learning_rate": 2.8933333333333333e-05, + "loss": 0.0245, + "step": 13204 + }, + { + "epoch": 11.152871621621621, + "grad_norm": 0.23163369297981262, + "learning_rate": 2.8933000000000002e-05, + "loss": 0.0124, + "step": 13205 + }, + { + "epoch": 11.153716216216216, + "grad_norm": 0.4875125586986542, + "learning_rate": 2.8932666666666665e-05, + "loss": 0.0242, + "step": 13206 + }, + { + "epoch": 11.15456081081081, + "grad_norm": 0.38863855600357056, + "learning_rate": 2.8932333333333334e-05, + "loss": 0.031, + "step": 13207 + }, + { + "epoch": 11.155405405405405, + "grad_norm": 0.6028311252593994, + "learning_rate": 2.8932e-05, + "loss": 0.0302, + "step": 13208 + }, + { + "epoch": 11.15625, + "grad_norm": 0.3805679380893707, + "learning_rate": 2.8931666666666666e-05, + "loss": 0.0208, + "step": 13209 + }, + { + "epoch": 11.157094594594595, + "grad_norm": 0.31961947679519653, + "learning_rate": 2.8931333333333335e-05, + "loss": 0.0209, + "step": 13210 + }, + { + "epoch": 11.15793918918919, + "grad_norm": 0.23877401649951935, + "learning_rate": 2.8931e-05, + "loss": 0.0177, + "step": 13211 + }, + { + "epoch": 11.158783783783784, + "grad_norm": 0.4114755094051361, + "learning_rate": 2.8930666666666667e-05, + "loss": 0.0236, + "step": 13212 + }, + { + "epoch": 11.159628378378379, + "grad_norm": 0.42043963074684143, + "learning_rate": 2.8930333333333333e-05, + "loss": 0.0178, + "step": 13213 + }, + { + "epoch": 11.160472972972974, + "grad_norm": 0.19499464333057404, + "learning_rate": 2.8930000000000003e-05, + "loss": 0.0144, + "step": 13214 + }, + { + "epoch": 11.161317567567568, + "grad_norm": 0.23069797456264496, + "learning_rate": 2.8929666666666665e-05, + "loss": 0.0152, + "step": 13215 + }, + { + "epoch": 11.162162162162161, + "grad_norm": 0.2821025252342224, + "learning_rate": 2.8929333333333334e-05, + "loss": 0.019, + "step": 13216 + }, + { + "epoch": 11.163006756756756, + "grad_norm": 1.8149100542068481, + "learning_rate": 2.8929000000000004e-05, + "loss": 0.0414, + "step": 13217 + }, + { + "epoch": 11.16385135135135, + "grad_norm": 0.38168391585350037, + "learning_rate": 2.8928666666666666e-05, + "loss": 0.0416, + "step": 13218 + }, + { + "epoch": 11.164695945945946, + "grad_norm": 1.3059308528900146, + "learning_rate": 2.8928333333333336e-05, + "loss": 0.0154, + "step": 13219 + }, + { + "epoch": 11.16554054054054, + "grad_norm": 0.36763662099838257, + "learning_rate": 2.8928e-05, + "loss": 0.0222, + "step": 13220 + }, + { + "epoch": 11.166385135135135, + "grad_norm": 0.4098605215549469, + "learning_rate": 2.8927666666666667e-05, + "loss": 0.0245, + "step": 13221 + }, + { + "epoch": 11.16722972972973, + "grad_norm": 0.5807985663414001, + "learning_rate": 2.8927333333333333e-05, + "loss": 0.0412, + "step": 13222 + }, + { + "epoch": 11.168074324324325, + "grad_norm": 0.6480174660682678, + "learning_rate": 2.8927000000000003e-05, + "loss": 0.0397, + "step": 13223 + }, + { + "epoch": 11.16891891891892, + "grad_norm": 1.3071975708007812, + "learning_rate": 2.8926666666666665e-05, + "loss": 0.0759, + "step": 13224 + }, + { + "epoch": 11.169763513513514, + "grad_norm": 0.754775881767273, + "learning_rate": 2.8926333333333335e-05, + "loss": 0.2237, + "step": 13225 + }, + { + "epoch": 11.170608108108109, + "grad_norm": 0.762479841709137, + "learning_rate": 2.8926e-05, + "loss": 0.1434, + "step": 13226 + }, + { + "epoch": 11.171452702702704, + "grad_norm": 0.3912084698677063, + "learning_rate": 2.8925666666666666e-05, + "loss": 0.0641, + "step": 13227 + }, + { + "epoch": 11.172297297297296, + "grad_norm": 0.33482593297958374, + "learning_rate": 2.8925333333333336e-05, + "loss": 0.051, + "step": 13228 + }, + { + "epoch": 11.173141891891891, + "grad_norm": 0.15677499771118164, + "learning_rate": 2.8925e-05, + "loss": 0.021, + "step": 13229 + }, + { + "epoch": 11.173986486486486, + "grad_norm": 0.34715163707733154, + "learning_rate": 2.8924666666666668e-05, + "loss": 0.0451, + "step": 13230 + }, + { + "epoch": 11.17483108108108, + "grad_norm": 0.3268284499645233, + "learning_rate": 2.8924333333333334e-05, + "loss": 0.0317, + "step": 13231 + }, + { + "epoch": 11.175675675675675, + "grad_norm": 0.46208205819129944, + "learning_rate": 2.8924e-05, + "loss": 0.0301, + "step": 13232 + }, + { + "epoch": 11.17652027027027, + "grad_norm": 0.344820499420166, + "learning_rate": 2.8923666666666665e-05, + "loss": 0.0316, + "step": 13233 + }, + { + "epoch": 11.177364864864865, + "grad_norm": 0.24868379533290863, + "learning_rate": 2.8923333333333335e-05, + "loss": 0.0298, + "step": 13234 + }, + { + "epoch": 11.17820945945946, + "grad_norm": 0.43892380595207214, + "learning_rate": 2.8923e-05, + "loss": 0.0378, + "step": 13235 + }, + { + "epoch": 11.179054054054054, + "grad_norm": 0.4490687847137451, + "learning_rate": 2.8922666666666667e-05, + "loss": 0.0344, + "step": 13236 + }, + { + "epoch": 11.17989864864865, + "grad_norm": 0.308015912771225, + "learning_rate": 2.8922333333333336e-05, + "loss": 0.0301, + "step": 13237 + }, + { + "epoch": 11.180743243243244, + "grad_norm": 0.17095354199409485, + "learning_rate": 2.8922e-05, + "loss": 0.0185, + "step": 13238 + }, + { + "epoch": 11.181587837837839, + "grad_norm": 0.3495922386646271, + "learning_rate": 2.8921666666666668e-05, + "loss": 0.0293, + "step": 13239 + }, + { + "epoch": 11.182432432432432, + "grad_norm": 0.274347186088562, + "learning_rate": 2.8921333333333334e-05, + "loss": 0.019, + "step": 13240 + }, + { + "epoch": 11.183277027027026, + "grad_norm": 0.42120060324668884, + "learning_rate": 2.8921e-05, + "loss": 0.0258, + "step": 13241 + }, + { + "epoch": 11.184121621621621, + "grad_norm": 0.23439398407936096, + "learning_rate": 2.892066666666667e-05, + "loss": 0.0241, + "step": 13242 + }, + { + "epoch": 11.184966216216216, + "grad_norm": 0.29519030451774597, + "learning_rate": 2.8920333333333335e-05, + "loss": 0.0307, + "step": 13243 + }, + { + "epoch": 11.18581081081081, + "grad_norm": 0.17324283719062805, + "learning_rate": 2.892e-05, + "loss": 0.0156, + "step": 13244 + }, + { + "epoch": 11.186655405405405, + "grad_norm": 0.5679822564125061, + "learning_rate": 2.8919666666666667e-05, + "loss": 0.0272, + "step": 13245 + }, + { + "epoch": 11.1875, + "grad_norm": 0.43053263425827026, + "learning_rate": 2.8919333333333336e-05, + "loss": 0.0309, + "step": 13246 + }, + { + "epoch": 11.188344594594595, + "grad_norm": 0.29057472944259644, + "learning_rate": 2.8919e-05, + "loss": 0.0245, + "step": 13247 + }, + { + "epoch": 11.18918918918919, + "grad_norm": 0.22088806331157684, + "learning_rate": 2.8918666666666668e-05, + "loss": 0.0118, + "step": 13248 + }, + { + "epoch": 11.190033783783784, + "grad_norm": 0.5719829201698303, + "learning_rate": 2.8918333333333334e-05, + "loss": 0.0237, + "step": 13249 + }, + { + "epoch": 11.190878378378379, + "grad_norm": 0.2413702756166458, + "learning_rate": 2.8918e-05, + "loss": 0.0211, + "step": 13250 + }, + { + "epoch": 11.191722972972974, + "grad_norm": 0.20415395498275757, + "learning_rate": 2.891766666666667e-05, + "loss": 0.021, + "step": 13251 + }, + { + "epoch": 11.192567567567568, + "grad_norm": 0.28793102502822876, + "learning_rate": 2.8917333333333335e-05, + "loss": 0.0182, + "step": 13252 + }, + { + "epoch": 11.193412162162161, + "grad_norm": 0.3078107237815857, + "learning_rate": 2.8917e-05, + "loss": 0.0214, + "step": 13253 + }, + { + "epoch": 11.194256756756756, + "grad_norm": 0.3673016130924225, + "learning_rate": 2.8916666666666667e-05, + "loss": 0.0229, + "step": 13254 + }, + { + "epoch": 11.19510135135135, + "grad_norm": 0.2379383146762848, + "learning_rate": 2.8916333333333336e-05, + "loss": 0.019, + "step": 13255 + }, + { + "epoch": 11.195945945945946, + "grad_norm": 0.35107681155204773, + "learning_rate": 2.8916e-05, + "loss": 0.0283, + "step": 13256 + }, + { + "epoch": 11.19679054054054, + "grad_norm": 0.21708612143993378, + "learning_rate": 2.8915666666666668e-05, + "loss": 0.0157, + "step": 13257 + }, + { + "epoch": 11.197635135135135, + "grad_norm": 0.18295179307460785, + "learning_rate": 2.8915333333333334e-05, + "loss": 0.0113, + "step": 13258 + }, + { + "epoch": 11.19847972972973, + "grad_norm": 0.21108223497867584, + "learning_rate": 2.8915e-05, + "loss": 0.0196, + "step": 13259 + }, + { + "epoch": 11.199324324324325, + "grad_norm": 0.5252999067306519, + "learning_rate": 2.891466666666667e-05, + "loss": 0.0181, + "step": 13260 + }, + { + "epoch": 11.20016891891892, + "grad_norm": 0.9493067860603333, + "learning_rate": 2.8914333333333332e-05, + "loss": 0.0443, + "step": 13261 + }, + { + "epoch": 11.201013513513514, + "grad_norm": 0.5777038335800171, + "learning_rate": 2.8914e-05, + "loss": 0.0247, + "step": 13262 + }, + { + "epoch": 11.201858108108109, + "grad_norm": 0.3192337155342102, + "learning_rate": 2.8913666666666667e-05, + "loss": 0.0247, + "step": 13263 + }, + { + "epoch": 11.202702702702704, + "grad_norm": 0.8578418493270874, + "learning_rate": 2.8913333333333333e-05, + "loss": 0.0419, + "step": 13264 + }, + { + "epoch": 11.203547297297296, + "grad_norm": 0.3734293580055237, + "learning_rate": 2.8913e-05, + "loss": 0.0178, + "step": 13265 + }, + { + "epoch": 11.204391891891891, + "grad_norm": 0.20607200264930725, + "learning_rate": 2.891266666666667e-05, + "loss": 0.0196, + "step": 13266 + }, + { + "epoch": 11.205236486486486, + "grad_norm": 0.6002614498138428, + "learning_rate": 2.8912333333333334e-05, + "loss": 0.027, + "step": 13267 + }, + { + "epoch": 11.20608108108108, + "grad_norm": 0.42234858870506287, + "learning_rate": 2.8912e-05, + "loss": 0.0305, + "step": 13268 + }, + { + "epoch": 11.206925675675675, + "grad_norm": 0.4570094645023346, + "learning_rate": 2.891166666666667e-05, + "loss": 0.0204, + "step": 13269 + }, + { + "epoch": 11.20777027027027, + "grad_norm": 0.3681126832962036, + "learning_rate": 2.8911333333333332e-05, + "loss": 0.0191, + "step": 13270 + }, + { + "epoch": 11.208614864864865, + "grad_norm": 0.5357041358947754, + "learning_rate": 2.8911e-05, + "loss": 0.0398, + "step": 13271 + }, + { + "epoch": 11.20945945945946, + "grad_norm": 0.3891794681549072, + "learning_rate": 2.8910666666666667e-05, + "loss": 0.0314, + "step": 13272 + }, + { + "epoch": 11.210304054054054, + "grad_norm": 0.634811282157898, + "learning_rate": 2.8910333333333333e-05, + "loss": 0.0371, + "step": 13273 + }, + { + "epoch": 11.21114864864865, + "grad_norm": 0.8334209322929382, + "learning_rate": 2.891e-05, + "loss": 0.0396, + "step": 13274 + }, + { + "epoch": 11.211993243243244, + "grad_norm": 0.6452298164367676, + "learning_rate": 2.890966666666667e-05, + "loss": 0.2028, + "step": 13275 + }, + { + "epoch": 11.212837837837839, + "grad_norm": 0.9776099920272827, + "learning_rate": 2.8909333333333335e-05, + "loss": 0.1553, + "step": 13276 + }, + { + "epoch": 11.213682432432432, + "grad_norm": 0.6752314567565918, + "learning_rate": 2.8909e-05, + "loss": 0.0553, + "step": 13277 + }, + { + "epoch": 11.214527027027026, + "grad_norm": 0.4024762511253357, + "learning_rate": 2.890866666666667e-05, + "loss": 0.0513, + "step": 13278 + }, + { + "epoch": 11.215371621621621, + "grad_norm": 0.6183620691299438, + "learning_rate": 2.8908333333333332e-05, + "loss": 0.0521, + "step": 13279 + }, + { + "epoch": 11.216216216216216, + "grad_norm": 0.39287069439888, + "learning_rate": 2.8908000000000002e-05, + "loss": 0.0221, + "step": 13280 + }, + { + "epoch": 11.21706081081081, + "grad_norm": 0.3952770531177521, + "learning_rate": 2.8907666666666668e-05, + "loss": 0.0264, + "step": 13281 + }, + { + "epoch": 11.217905405405405, + "grad_norm": 0.2141428142786026, + "learning_rate": 2.8907333333333334e-05, + "loss": 0.022, + "step": 13282 + }, + { + "epoch": 11.21875, + "grad_norm": 0.39474159479141235, + "learning_rate": 2.8907e-05, + "loss": 0.029, + "step": 13283 + }, + { + "epoch": 11.219594594594595, + "grad_norm": 0.3958672285079956, + "learning_rate": 2.890666666666667e-05, + "loss": 0.0324, + "step": 13284 + }, + { + "epoch": 11.22043918918919, + "grad_norm": 0.3286426067352295, + "learning_rate": 2.8906333333333335e-05, + "loss": 0.0355, + "step": 13285 + }, + { + "epoch": 11.221283783783784, + "grad_norm": 0.47025492787361145, + "learning_rate": 2.8906e-05, + "loss": 0.0405, + "step": 13286 + }, + { + "epoch": 11.222128378378379, + "grad_norm": 0.4082956314086914, + "learning_rate": 2.8905666666666667e-05, + "loss": 0.0274, + "step": 13287 + }, + { + "epoch": 11.222972972972974, + "grad_norm": 0.34145835041999817, + "learning_rate": 2.8905333333333333e-05, + "loss": 0.0289, + "step": 13288 + }, + { + "epoch": 11.223817567567568, + "grad_norm": 0.4821031093597412, + "learning_rate": 2.8905000000000002e-05, + "loss": 0.0382, + "step": 13289 + }, + { + "epoch": 11.224662162162161, + "grad_norm": 0.3709290325641632, + "learning_rate": 2.8904666666666664e-05, + "loss": 0.0356, + "step": 13290 + }, + { + "epoch": 11.225506756756756, + "grad_norm": 0.22337749600410461, + "learning_rate": 2.8904333333333334e-05, + "loss": 0.0204, + "step": 13291 + }, + { + "epoch": 11.22635135135135, + "grad_norm": 0.3269781470298767, + "learning_rate": 2.8904000000000003e-05, + "loss": 0.019, + "step": 13292 + }, + { + "epoch": 11.227195945945946, + "grad_norm": 0.3952428996562958, + "learning_rate": 2.8903666666666666e-05, + "loss": 0.0319, + "step": 13293 + }, + { + "epoch": 11.22804054054054, + "grad_norm": 0.22134961187839508, + "learning_rate": 2.8903333333333335e-05, + "loss": 0.0202, + "step": 13294 + }, + { + "epoch": 11.228885135135135, + "grad_norm": 0.2391107976436615, + "learning_rate": 2.8903e-05, + "loss": 0.0169, + "step": 13295 + }, + { + "epoch": 11.22972972972973, + "grad_norm": 0.28643807768821716, + "learning_rate": 2.8902666666666667e-05, + "loss": 0.0248, + "step": 13296 + }, + { + "epoch": 11.230574324324325, + "grad_norm": 0.2599354684352875, + "learning_rate": 2.8902333333333333e-05, + "loss": 0.0178, + "step": 13297 + }, + { + "epoch": 11.23141891891892, + "grad_norm": 0.48062944412231445, + "learning_rate": 2.8902000000000002e-05, + "loss": 0.03, + "step": 13298 + }, + { + "epoch": 11.232263513513514, + "grad_norm": 0.4941083490848541, + "learning_rate": 2.8901666666666665e-05, + "loss": 0.0205, + "step": 13299 + }, + { + "epoch": 11.233108108108109, + "grad_norm": 0.2669568359851837, + "learning_rate": 2.8901333333333334e-05, + "loss": 0.0182, + "step": 13300 + }, + { + "epoch": 11.233952702702704, + "grad_norm": 0.11465996503829956, + "learning_rate": 2.8901000000000003e-05, + "loss": 0.0093, + "step": 13301 + }, + { + "epoch": 11.234797297297296, + "grad_norm": 0.2697223722934723, + "learning_rate": 2.8900666666666666e-05, + "loss": 0.0228, + "step": 13302 + }, + { + "epoch": 11.235641891891891, + "grad_norm": 0.29020312428474426, + "learning_rate": 2.8900333333333335e-05, + "loss": 0.0254, + "step": 13303 + }, + { + "epoch": 11.236486486486486, + "grad_norm": 0.22125886380672455, + "learning_rate": 2.89e-05, + "loss": 0.0176, + "step": 13304 + }, + { + "epoch": 11.23733108108108, + "grad_norm": 0.3666611313819885, + "learning_rate": 2.8899666666666667e-05, + "loss": 0.0298, + "step": 13305 + }, + { + "epoch": 11.238175675675675, + "grad_norm": 0.3753068149089813, + "learning_rate": 2.8899333333333333e-05, + "loss": 0.0245, + "step": 13306 + }, + { + "epoch": 11.23902027027027, + "grad_norm": 0.5129044651985168, + "learning_rate": 2.8899000000000002e-05, + "loss": 0.0343, + "step": 13307 + }, + { + "epoch": 11.239864864864865, + "grad_norm": 0.3518127202987671, + "learning_rate": 2.8898666666666668e-05, + "loss": 0.02, + "step": 13308 + }, + { + "epoch": 11.24070945945946, + "grad_norm": 0.21002864837646484, + "learning_rate": 2.8898333333333334e-05, + "loss": 0.0157, + "step": 13309 + }, + { + "epoch": 11.241554054054054, + "grad_norm": 0.36892321705818176, + "learning_rate": 2.8898000000000004e-05, + "loss": 0.0147, + "step": 13310 + }, + { + "epoch": 11.24239864864865, + "grad_norm": 0.507050633430481, + "learning_rate": 2.8897666666666666e-05, + "loss": 0.0331, + "step": 13311 + }, + { + "epoch": 11.243243243243244, + "grad_norm": 0.4299495816230774, + "learning_rate": 2.8897333333333335e-05, + "loss": 0.0299, + "step": 13312 + }, + { + "epoch": 11.244087837837839, + "grad_norm": 0.26560935378074646, + "learning_rate": 2.8897e-05, + "loss": 0.016, + "step": 13313 + }, + { + "epoch": 11.244932432432432, + "grad_norm": 0.403227299451828, + "learning_rate": 2.8896666666666667e-05, + "loss": 0.0348, + "step": 13314 + }, + { + "epoch": 11.245777027027026, + "grad_norm": 0.3773462772369385, + "learning_rate": 2.8896333333333333e-05, + "loss": 0.0277, + "step": 13315 + }, + { + "epoch": 11.246621621621621, + "grad_norm": 0.5546189546585083, + "learning_rate": 2.8896e-05, + "loss": 0.0209, + "step": 13316 + }, + { + "epoch": 11.247466216216216, + "grad_norm": 0.18876461684703827, + "learning_rate": 2.889566666666667e-05, + "loss": 0.0153, + "step": 13317 + }, + { + "epoch": 11.24831081081081, + "grad_norm": 0.27813461422920227, + "learning_rate": 2.8895333333333334e-05, + "loss": 0.016, + "step": 13318 + }, + { + "epoch": 11.249155405405405, + "grad_norm": 0.34169405698776245, + "learning_rate": 2.8895e-05, + "loss": 0.0144, + "step": 13319 + }, + { + "epoch": 11.25, + "grad_norm": 0.287861168384552, + "learning_rate": 2.8894666666666666e-05, + "loss": 0.0238, + "step": 13320 + }, + { + "epoch": 11.250844594594595, + "grad_norm": 0.3912266492843628, + "learning_rate": 2.8894333333333336e-05, + "loss": 0.0278, + "step": 13321 + }, + { + "epoch": 11.25168918918919, + "grad_norm": 0.4569343030452728, + "learning_rate": 2.8893999999999998e-05, + "loss": 0.0373, + "step": 13322 + }, + { + "epoch": 11.252533783783784, + "grad_norm": 0.4668809771537781, + "learning_rate": 2.8893666666666667e-05, + "loss": 0.0262, + "step": 13323 + }, + { + "epoch": 11.253378378378379, + "grad_norm": 0.9178073406219482, + "learning_rate": 2.8893333333333333e-05, + "loss": 0.0299, + "step": 13324 + }, + { + "epoch": 11.254222972972974, + "grad_norm": 1.2366129159927368, + "learning_rate": 2.8893e-05, + "loss": 0.3138, + "step": 13325 + }, + { + "epoch": 11.255067567567568, + "grad_norm": 0.6799986362457275, + "learning_rate": 2.889266666666667e-05, + "loss": 0.1627, + "step": 13326 + }, + { + "epoch": 11.255912162162161, + "grad_norm": 0.4048082232475281, + "learning_rate": 2.8892333333333335e-05, + "loss": 0.071, + "step": 13327 + }, + { + "epoch": 11.256756756756756, + "grad_norm": 0.2681369185447693, + "learning_rate": 2.8892e-05, + "loss": 0.0387, + "step": 13328 + }, + { + "epoch": 11.25760135135135, + "grad_norm": 0.2987842261791229, + "learning_rate": 2.8891666666666666e-05, + "loss": 0.029, + "step": 13329 + }, + { + "epoch": 11.258445945945946, + "grad_norm": 0.3866059482097626, + "learning_rate": 2.8891333333333336e-05, + "loss": 0.0433, + "step": 13330 + }, + { + "epoch": 11.25929054054054, + "grad_norm": 0.30199798941612244, + "learning_rate": 2.8891e-05, + "loss": 0.0256, + "step": 13331 + }, + { + "epoch": 11.260135135135135, + "grad_norm": 0.23308072984218597, + "learning_rate": 2.8890666666666668e-05, + "loss": 0.0273, + "step": 13332 + }, + { + "epoch": 11.26097972972973, + "grad_norm": 0.42955872416496277, + "learning_rate": 2.8890333333333334e-05, + "loss": 0.0341, + "step": 13333 + }, + { + "epoch": 11.261824324324325, + "grad_norm": 0.30649176239967346, + "learning_rate": 2.889e-05, + "loss": 0.0257, + "step": 13334 + }, + { + "epoch": 11.26266891891892, + "grad_norm": 0.30981358885765076, + "learning_rate": 2.888966666666667e-05, + "loss": 0.0337, + "step": 13335 + }, + { + "epoch": 11.263513513513514, + "grad_norm": 0.2916530668735504, + "learning_rate": 2.8889333333333335e-05, + "loss": 0.0215, + "step": 13336 + }, + { + "epoch": 11.264358108108109, + "grad_norm": 0.24196037650108337, + "learning_rate": 2.8889e-05, + "loss": 0.0218, + "step": 13337 + }, + { + "epoch": 11.265202702702704, + "grad_norm": 0.4039018452167511, + "learning_rate": 2.8888666666666667e-05, + "loss": 0.0313, + "step": 13338 + }, + { + "epoch": 11.266047297297296, + "grad_norm": 0.3209850490093231, + "learning_rate": 2.8888333333333336e-05, + "loss": 0.0319, + "step": 13339 + }, + { + "epoch": 11.266891891891891, + "grad_norm": 0.35540771484375, + "learning_rate": 2.8888e-05, + "loss": 0.0228, + "step": 13340 + }, + { + "epoch": 11.267736486486486, + "grad_norm": 0.33648255467414856, + "learning_rate": 2.8887666666666668e-05, + "loss": 0.0265, + "step": 13341 + }, + { + "epoch": 11.26858108108108, + "grad_norm": 0.23084965348243713, + "learning_rate": 2.8887333333333337e-05, + "loss": 0.0153, + "step": 13342 + }, + { + "epoch": 11.269425675675675, + "grad_norm": 0.24179592728614807, + "learning_rate": 2.8887e-05, + "loss": 0.0246, + "step": 13343 + }, + { + "epoch": 11.27027027027027, + "grad_norm": 0.2792541980743408, + "learning_rate": 2.888666666666667e-05, + "loss": 0.017, + "step": 13344 + }, + { + "epoch": 11.271114864864865, + "grad_norm": 0.20389649271965027, + "learning_rate": 2.8886333333333335e-05, + "loss": 0.0212, + "step": 13345 + }, + { + "epoch": 11.27195945945946, + "grad_norm": 0.4236728549003601, + "learning_rate": 2.8886e-05, + "loss": 0.0169, + "step": 13346 + }, + { + "epoch": 11.272804054054054, + "grad_norm": 0.28855597972869873, + "learning_rate": 2.8885666666666667e-05, + "loss": 0.0188, + "step": 13347 + }, + { + "epoch": 11.27364864864865, + "grad_norm": 0.6422390937805176, + "learning_rate": 2.8885333333333333e-05, + "loss": 0.025, + "step": 13348 + }, + { + "epoch": 11.274493243243244, + "grad_norm": 1.2564144134521484, + "learning_rate": 2.8885e-05, + "loss": 0.0203, + "step": 13349 + }, + { + "epoch": 11.275337837837839, + "grad_norm": 0.5350531339645386, + "learning_rate": 2.8884666666666668e-05, + "loss": 0.0184, + "step": 13350 + }, + { + "epoch": 11.276182432432432, + "grad_norm": 0.4160880744457245, + "learning_rate": 2.8884333333333334e-05, + "loss": 0.0176, + "step": 13351 + }, + { + "epoch": 11.277027027027026, + "grad_norm": 0.23404213786125183, + "learning_rate": 2.8884e-05, + "loss": 0.0162, + "step": 13352 + }, + { + "epoch": 11.277871621621621, + "grad_norm": 0.4672091603279114, + "learning_rate": 2.888366666666667e-05, + "loss": 0.0299, + "step": 13353 + }, + { + "epoch": 11.278716216216216, + "grad_norm": 0.46650242805480957, + "learning_rate": 2.8883333333333332e-05, + "loss": 0.0394, + "step": 13354 + }, + { + "epoch": 11.27956081081081, + "grad_norm": 0.31795722246170044, + "learning_rate": 2.8883e-05, + "loss": 0.0205, + "step": 13355 + }, + { + "epoch": 11.280405405405405, + "grad_norm": 0.8952333331108093, + "learning_rate": 2.8882666666666667e-05, + "loss": 0.0351, + "step": 13356 + }, + { + "epoch": 11.28125, + "grad_norm": 0.23535799980163574, + "learning_rate": 2.8882333333333333e-05, + "loss": 0.0238, + "step": 13357 + }, + { + "epoch": 11.282094594594595, + "grad_norm": 0.27340346574783325, + "learning_rate": 2.8882000000000002e-05, + "loss": 0.0234, + "step": 13358 + }, + { + "epoch": 11.28293918918919, + "grad_norm": 0.6633935570716858, + "learning_rate": 2.8881666666666668e-05, + "loss": 0.0215, + "step": 13359 + }, + { + "epoch": 11.283783783783784, + "grad_norm": 0.20966877043247223, + "learning_rate": 2.8881333333333334e-05, + "loss": 0.0139, + "step": 13360 + }, + { + "epoch": 11.284628378378379, + "grad_norm": 0.4705295264720917, + "learning_rate": 2.8881e-05, + "loss": 0.0336, + "step": 13361 + }, + { + "epoch": 11.285472972972974, + "grad_norm": 0.39971858263015747, + "learning_rate": 2.888066666666667e-05, + "loss": 0.0276, + "step": 13362 + }, + { + "epoch": 11.286317567567568, + "grad_norm": 0.6346693634986877, + "learning_rate": 2.8880333333333332e-05, + "loss": 0.0204, + "step": 13363 + }, + { + "epoch": 11.287162162162161, + "grad_norm": 0.3543180525302887, + "learning_rate": 2.888e-05, + "loss": 0.0197, + "step": 13364 + }, + { + "epoch": 11.288006756756756, + "grad_norm": 0.37722793221473694, + "learning_rate": 2.8879666666666667e-05, + "loss": 0.0262, + "step": 13365 + }, + { + "epoch": 11.28885135135135, + "grad_norm": 0.26735246181488037, + "learning_rate": 2.8879333333333333e-05, + "loss": 0.0172, + "step": 13366 + }, + { + "epoch": 11.289695945945946, + "grad_norm": 0.43146300315856934, + "learning_rate": 2.8879000000000003e-05, + "loss": 0.0356, + "step": 13367 + }, + { + "epoch": 11.29054054054054, + "grad_norm": 0.251391738653183, + "learning_rate": 2.887866666666667e-05, + "loss": 0.0209, + "step": 13368 + }, + { + "epoch": 11.291385135135135, + "grad_norm": 0.5030809640884399, + "learning_rate": 2.8878333333333334e-05, + "loss": 0.03, + "step": 13369 + }, + { + "epoch": 11.29222972972973, + "grad_norm": 0.2964487373828888, + "learning_rate": 2.8878e-05, + "loss": 0.0205, + "step": 13370 + }, + { + "epoch": 11.293074324324325, + "grad_norm": 0.6391454935073853, + "learning_rate": 2.887766666666667e-05, + "loss": 0.022, + "step": 13371 + }, + { + "epoch": 11.29391891891892, + "grad_norm": 0.4802909791469574, + "learning_rate": 2.8877333333333332e-05, + "loss": 0.0276, + "step": 13372 + }, + { + "epoch": 11.294763513513514, + "grad_norm": 0.6302375793457031, + "learning_rate": 2.8877e-05, + "loss": 0.0404, + "step": 13373 + }, + { + "epoch": 11.295608108108109, + "grad_norm": 0.5916298031806946, + "learning_rate": 2.8876666666666667e-05, + "loss": 0.0351, + "step": 13374 + }, + { + "epoch": 11.296452702702704, + "grad_norm": 1.63316810131073, + "learning_rate": 2.8876333333333333e-05, + "loss": 0.2704, + "step": 13375 + }, + { + "epoch": 11.297297297297296, + "grad_norm": 0.5783657431602478, + "learning_rate": 2.8876000000000003e-05, + "loss": 0.1351, + "step": 13376 + }, + { + "epoch": 11.298141891891891, + "grad_norm": 0.3574660122394562, + "learning_rate": 2.8875666666666665e-05, + "loss": 0.0709, + "step": 13377 + }, + { + "epoch": 11.298986486486486, + "grad_norm": 0.6193667054176331, + "learning_rate": 2.8875333333333335e-05, + "loss": 0.0378, + "step": 13378 + }, + { + "epoch": 11.29983108108108, + "grad_norm": 0.4081268310546875, + "learning_rate": 2.8875e-05, + "loss": 0.0344, + "step": 13379 + }, + { + "epoch": 11.300675675675675, + "grad_norm": 0.41153484582901, + "learning_rate": 2.8874666666666666e-05, + "loss": 0.0326, + "step": 13380 + }, + { + "epoch": 11.30152027027027, + "grad_norm": 0.6017024517059326, + "learning_rate": 2.8874333333333332e-05, + "loss": 0.0257, + "step": 13381 + }, + { + "epoch": 11.302364864864865, + "grad_norm": 0.9411514401435852, + "learning_rate": 2.8874000000000002e-05, + "loss": 0.0429, + "step": 13382 + }, + { + "epoch": 11.30320945945946, + "grad_norm": 0.3820509910583496, + "learning_rate": 2.8873666666666668e-05, + "loss": 0.0273, + "step": 13383 + }, + { + "epoch": 11.304054054054054, + "grad_norm": 0.5539248585700989, + "learning_rate": 2.8873333333333334e-05, + "loss": 0.0349, + "step": 13384 + }, + { + "epoch": 11.30489864864865, + "grad_norm": 0.42844489216804504, + "learning_rate": 2.8873000000000003e-05, + "loss": 0.0272, + "step": 13385 + }, + { + "epoch": 11.305743243243244, + "grad_norm": 0.4374629557132721, + "learning_rate": 2.8872666666666665e-05, + "loss": 0.0268, + "step": 13386 + }, + { + "epoch": 11.306587837837839, + "grad_norm": 0.17088209092617035, + "learning_rate": 2.8872333333333335e-05, + "loss": 0.0178, + "step": 13387 + }, + { + "epoch": 11.307432432432432, + "grad_norm": 0.32115134596824646, + "learning_rate": 2.8872e-05, + "loss": 0.0287, + "step": 13388 + }, + { + "epoch": 11.308277027027026, + "grad_norm": 0.3470774292945862, + "learning_rate": 2.8871666666666667e-05, + "loss": 0.0226, + "step": 13389 + }, + { + "epoch": 11.309121621621621, + "grad_norm": 0.2239147275686264, + "learning_rate": 2.8871333333333333e-05, + "loss": 0.019, + "step": 13390 + }, + { + "epoch": 11.309966216216216, + "grad_norm": 0.30839285254478455, + "learning_rate": 2.8871000000000002e-05, + "loss": 0.0319, + "step": 13391 + }, + { + "epoch": 11.31081081081081, + "grad_norm": 0.21353290975093842, + "learning_rate": 2.8870666666666668e-05, + "loss": 0.0181, + "step": 13392 + }, + { + "epoch": 11.311655405405405, + "grad_norm": 0.24709109961986542, + "learning_rate": 2.8870333333333334e-05, + "loss": 0.0164, + "step": 13393 + }, + { + "epoch": 11.3125, + "grad_norm": 0.3286270201206207, + "learning_rate": 2.8870000000000003e-05, + "loss": 0.0265, + "step": 13394 + }, + { + "epoch": 11.313344594594595, + "grad_norm": 0.4174235463142395, + "learning_rate": 2.8869666666666666e-05, + "loss": 0.0274, + "step": 13395 + }, + { + "epoch": 11.31418918918919, + "grad_norm": 1.031136393547058, + "learning_rate": 2.8869333333333335e-05, + "loss": 0.0712, + "step": 13396 + }, + { + "epoch": 11.315033783783784, + "grad_norm": 0.4685761630535126, + "learning_rate": 2.8869e-05, + "loss": 0.0266, + "step": 13397 + }, + { + "epoch": 11.315878378378379, + "grad_norm": 0.318936824798584, + "learning_rate": 2.8868666666666667e-05, + "loss": 0.0211, + "step": 13398 + }, + { + "epoch": 11.316722972972974, + "grad_norm": 0.2431543916463852, + "learning_rate": 2.8868333333333333e-05, + "loss": 0.024, + "step": 13399 + }, + { + "epoch": 11.317567567567568, + "grad_norm": 0.272859662771225, + "learning_rate": 2.8868000000000002e-05, + "loss": 0.0279, + "step": 13400 + }, + { + "epoch": 11.318412162162161, + "grad_norm": 0.2544752359390259, + "learning_rate": 2.8867666666666668e-05, + "loss": 0.0196, + "step": 13401 + }, + { + "epoch": 11.319256756756756, + "grad_norm": 0.3112501800060272, + "learning_rate": 2.8867333333333334e-05, + "loss": 0.0224, + "step": 13402 + }, + { + "epoch": 11.32010135135135, + "grad_norm": 0.2615077793598175, + "learning_rate": 2.8867000000000003e-05, + "loss": 0.0138, + "step": 13403 + }, + { + "epoch": 11.320945945945946, + "grad_norm": 0.33421435952186584, + "learning_rate": 2.8866666666666666e-05, + "loss": 0.0146, + "step": 13404 + }, + { + "epoch": 11.32179054054054, + "grad_norm": 0.29426947236061096, + "learning_rate": 2.8866333333333335e-05, + "loss": 0.0202, + "step": 13405 + }, + { + "epoch": 11.322635135135135, + "grad_norm": 0.296769917011261, + "learning_rate": 2.8866e-05, + "loss": 0.0195, + "step": 13406 + }, + { + "epoch": 11.32347972972973, + "grad_norm": 0.36254626512527466, + "learning_rate": 2.8865666666666667e-05, + "loss": 0.0297, + "step": 13407 + }, + { + "epoch": 11.324324324324325, + "grad_norm": 0.5423793196678162, + "learning_rate": 2.8865333333333336e-05, + "loss": 0.0233, + "step": 13408 + }, + { + "epoch": 11.32516891891892, + "grad_norm": 0.4468822777271271, + "learning_rate": 2.8865e-05, + "loss": 0.0134, + "step": 13409 + }, + { + "epoch": 11.326013513513514, + "grad_norm": 0.5670009851455688, + "learning_rate": 2.8864666666666668e-05, + "loss": 0.0229, + "step": 13410 + }, + { + "epoch": 11.326858108108109, + "grad_norm": 0.42124807834625244, + "learning_rate": 2.8864333333333334e-05, + "loss": 0.0287, + "step": 13411 + }, + { + "epoch": 11.327702702702704, + "grad_norm": 0.6409432888031006, + "learning_rate": 2.8864e-05, + "loss": 0.0148, + "step": 13412 + }, + { + "epoch": 11.328547297297296, + "grad_norm": 0.33915382623672485, + "learning_rate": 2.8863666666666666e-05, + "loss": 0.0241, + "step": 13413 + }, + { + "epoch": 11.329391891891891, + "grad_norm": 0.314426988363266, + "learning_rate": 2.8863333333333335e-05, + "loss": 0.0247, + "step": 13414 + }, + { + "epoch": 11.330236486486486, + "grad_norm": 0.586858332157135, + "learning_rate": 2.8862999999999998e-05, + "loss": 0.0225, + "step": 13415 + }, + { + "epoch": 11.33108108108108, + "grad_norm": 0.5115801095962524, + "learning_rate": 2.8862666666666667e-05, + "loss": 0.0201, + "step": 13416 + }, + { + "epoch": 11.331925675675675, + "grad_norm": 0.6158075332641602, + "learning_rate": 2.8862333333333337e-05, + "loss": 0.0276, + "step": 13417 + }, + { + "epoch": 11.33277027027027, + "grad_norm": 0.2908308207988739, + "learning_rate": 2.8862e-05, + "loss": 0.0251, + "step": 13418 + }, + { + "epoch": 11.333614864864865, + "grad_norm": 0.5373671650886536, + "learning_rate": 2.886166666666667e-05, + "loss": 0.0248, + "step": 13419 + }, + { + "epoch": 11.33445945945946, + "grad_norm": 0.3128868341445923, + "learning_rate": 2.8861333333333334e-05, + "loss": 0.0299, + "step": 13420 + }, + { + "epoch": 11.335304054054054, + "grad_norm": 1.0448366403579712, + "learning_rate": 2.8861e-05, + "loss": 0.06, + "step": 13421 + }, + { + "epoch": 11.33614864864865, + "grad_norm": 0.43270590901374817, + "learning_rate": 2.8860666666666666e-05, + "loss": 0.0352, + "step": 13422 + }, + { + "epoch": 11.336993243243244, + "grad_norm": 0.9266477823257446, + "learning_rate": 2.8860333333333336e-05, + "loss": 0.0336, + "step": 13423 + }, + { + "epoch": 11.337837837837839, + "grad_norm": 1.1304726600646973, + "learning_rate": 2.8859999999999998e-05, + "loss": 0.044, + "step": 13424 + }, + { + "epoch": 11.338682432432432, + "grad_norm": 0.8589267730712891, + "learning_rate": 2.8859666666666667e-05, + "loss": 0.2748, + "step": 13425 + }, + { + "epoch": 11.339527027027026, + "grad_norm": 0.6567010283470154, + "learning_rate": 2.8859333333333337e-05, + "loss": 0.1337, + "step": 13426 + }, + { + "epoch": 11.340371621621621, + "grad_norm": 0.4419073760509491, + "learning_rate": 2.8859e-05, + "loss": 0.0609, + "step": 13427 + }, + { + "epoch": 11.341216216216216, + "grad_norm": 0.4116525650024414, + "learning_rate": 2.885866666666667e-05, + "loss": 0.0561, + "step": 13428 + }, + { + "epoch": 11.34206081081081, + "grad_norm": 0.40698280930519104, + "learning_rate": 2.8858333333333335e-05, + "loss": 0.0354, + "step": 13429 + }, + { + "epoch": 11.342905405405405, + "grad_norm": 0.41091012954711914, + "learning_rate": 2.8858e-05, + "loss": 0.0389, + "step": 13430 + }, + { + "epoch": 11.34375, + "grad_norm": 0.5889737010002136, + "learning_rate": 2.8857666666666666e-05, + "loss": 0.0435, + "step": 13431 + }, + { + "epoch": 11.344594594594595, + "grad_norm": 0.3464842736721039, + "learning_rate": 2.8857333333333336e-05, + "loss": 0.0314, + "step": 13432 + }, + { + "epoch": 11.34543918918919, + "grad_norm": 0.26707467436790466, + "learning_rate": 2.8857000000000002e-05, + "loss": 0.0214, + "step": 13433 + }, + { + "epoch": 11.346283783783784, + "grad_norm": 0.42085686326026917, + "learning_rate": 2.8856666666666668e-05, + "loss": 0.0316, + "step": 13434 + }, + { + "epoch": 11.347128378378379, + "grad_norm": 0.2637348473072052, + "learning_rate": 2.8856333333333337e-05, + "loss": 0.02, + "step": 13435 + }, + { + "epoch": 11.347972972972974, + "grad_norm": 0.25764843821525574, + "learning_rate": 2.8856e-05, + "loss": 0.0278, + "step": 13436 + }, + { + "epoch": 11.348817567567568, + "grad_norm": 0.27024319767951965, + "learning_rate": 2.885566666666667e-05, + "loss": 0.0198, + "step": 13437 + }, + { + "epoch": 11.349662162162161, + "grad_norm": 0.6962035298347473, + "learning_rate": 2.885533333333333e-05, + "loss": 0.0346, + "step": 13438 + }, + { + "epoch": 11.350506756756756, + "grad_norm": 0.4141395390033722, + "learning_rate": 2.8855e-05, + "loss": 0.0269, + "step": 13439 + }, + { + "epoch": 11.35135135135135, + "grad_norm": 0.31405311822891235, + "learning_rate": 2.8854666666666667e-05, + "loss": 0.0234, + "step": 13440 + }, + { + "epoch": 11.352195945945946, + "grad_norm": 0.511419415473938, + "learning_rate": 2.8854333333333333e-05, + "loss": 0.033, + "step": 13441 + }, + { + "epoch": 11.35304054054054, + "grad_norm": 0.43538424372673035, + "learning_rate": 2.8854000000000002e-05, + "loss": 0.0193, + "step": 13442 + }, + { + "epoch": 11.353885135135135, + "grad_norm": 0.9378079771995544, + "learning_rate": 2.8853666666666668e-05, + "loss": 0.0234, + "step": 13443 + }, + { + "epoch": 11.35472972972973, + "grad_norm": 0.312630832195282, + "learning_rate": 2.8853333333333334e-05, + "loss": 0.0296, + "step": 13444 + }, + { + "epoch": 11.355574324324325, + "grad_norm": 0.3513604700565338, + "learning_rate": 2.8853e-05, + "loss": 0.0256, + "step": 13445 + }, + { + "epoch": 11.35641891891892, + "grad_norm": 0.1850769817829132, + "learning_rate": 2.885266666666667e-05, + "loss": 0.0148, + "step": 13446 + }, + { + "epoch": 11.357263513513514, + "grad_norm": 0.5418533682823181, + "learning_rate": 2.885233333333333e-05, + "loss": 0.0325, + "step": 13447 + }, + { + "epoch": 11.358108108108109, + "grad_norm": 0.4278711974620819, + "learning_rate": 2.8852e-05, + "loss": 0.0296, + "step": 13448 + }, + { + "epoch": 11.358952702702704, + "grad_norm": 0.4529189467430115, + "learning_rate": 2.8851666666666667e-05, + "loss": 0.0414, + "step": 13449 + }, + { + "epoch": 11.359797297297296, + "grad_norm": 0.4008634686470032, + "learning_rate": 2.8851333333333333e-05, + "loss": 0.0291, + "step": 13450 + }, + { + "epoch": 11.360641891891891, + "grad_norm": 0.1801183521747589, + "learning_rate": 2.8851000000000002e-05, + "loss": 0.0186, + "step": 13451 + }, + { + "epoch": 11.361486486486486, + "grad_norm": 0.32650354504585266, + "learning_rate": 2.8850666666666668e-05, + "loss": 0.0252, + "step": 13452 + }, + { + "epoch": 11.36233108108108, + "grad_norm": 0.8447557687759399, + "learning_rate": 2.8850333333333334e-05, + "loss": 0.0211, + "step": 13453 + }, + { + "epoch": 11.363175675675675, + "grad_norm": 0.3128970265388489, + "learning_rate": 2.885e-05, + "loss": 0.025, + "step": 13454 + }, + { + "epoch": 11.36402027027027, + "grad_norm": 0.11389606446027756, + "learning_rate": 2.884966666666667e-05, + "loss": 0.0097, + "step": 13455 + }, + { + "epoch": 11.364864864864865, + "grad_norm": 0.29847997426986694, + "learning_rate": 2.8849333333333332e-05, + "loss": 0.0173, + "step": 13456 + }, + { + "epoch": 11.36570945945946, + "grad_norm": 0.210740327835083, + "learning_rate": 2.8849e-05, + "loss": 0.0166, + "step": 13457 + }, + { + "epoch": 11.366554054054054, + "grad_norm": 0.45365968346595764, + "learning_rate": 2.884866666666667e-05, + "loss": 0.0301, + "step": 13458 + }, + { + "epoch": 11.36739864864865, + "grad_norm": 0.6107977032661438, + "learning_rate": 2.8848333333333333e-05, + "loss": 0.0287, + "step": 13459 + }, + { + "epoch": 11.368243243243244, + "grad_norm": 0.3293609321117401, + "learning_rate": 2.8848000000000002e-05, + "loss": 0.0158, + "step": 13460 + }, + { + "epoch": 11.369087837837839, + "grad_norm": 0.629101037979126, + "learning_rate": 2.8847666666666668e-05, + "loss": 0.0295, + "step": 13461 + }, + { + "epoch": 11.369932432432432, + "grad_norm": 0.41253259778022766, + "learning_rate": 2.8847333333333334e-05, + "loss": 0.0381, + "step": 13462 + }, + { + "epoch": 11.370777027027026, + "grad_norm": 0.7230366468429565, + "learning_rate": 2.8847e-05, + "loss": 0.0298, + "step": 13463 + }, + { + "epoch": 11.371621621621621, + "grad_norm": 0.43229517340660095, + "learning_rate": 2.884666666666667e-05, + "loss": 0.0322, + "step": 13464 + }, + { + "epoch": 11.372466216216216, + "grad_norm": 0.2799398601055145, + "learning_rate": 2.8846333333333332e-05, + "loss": 0.0272, + "step": 13465 + }, + { + "epoch": 11.37331081081081, + "grad_norm": 0.294293612241745, + "learning_rate": 2.8846e-05, + "loss": 0.0237, + "step": 13466 + }, + { + "epoch": 11.374155405405405, + "grad_norm": 0.8150919079780579, + "learning_rate": 2.8845666666666667e-05, + "loss": 0.0224, + "step": 13467 + }, + { + "epoch": 11.375, + "grad_norm": 0.34684014320373535, + "learning_rate": 2.8845333333333333e-05, + "loss": 0.0227, + "step": 13468 + }, + { + "epoch": 11.375844594594595, + "grad_norm": 0.8295963406562805, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0296, + "step": 13469 + }, + { + "epoch": 11.37668918918919, + "grad_norm": 0.5102479457855225, + "learning_rate": 2.8844666666666665e-05, + "loss": 0.0292, + "step": 13470 + }, + { + "epoch": 11.377533783783784, + "grad_norm": 0.285756915807724, + "learning_rate": 2.8844333333333334e-05, + "loss": 0.0266, + "step": 13471 + }, + { + "epoch": 11.378378378378379, + "grad_norm": 1.4051127433776855, + "learning_rate": 2.8844e-05, + "loss": 0.0428, + "step": 13472 + }, + { + "epoch": 11.379222972972974, + "grad_norm": 0.9182218909263611, + "learning_rate": 2.8843666666666666e-05, + "loss": 0.0356, + "step": 13473 + }, + { + "epoch": 11.380067567567568, + "grad_norm": 0.4268559515476227, + "learning_rate": 2.8843333333333332e-05, + "loss": 0.029, + "step": 13474 + }, + { + "epoch": 11.380912162162161, + "grad_norm": 0.617394208908081, + "learning_rate": 2.8843e-05, + "loss": 0.209, + "step": 13475 + }, + { + "epoch": 11.381756756756756, + "grad_norm": 0.7270370125770569, + "learning_rate": 2.8842666666666667e-05, + "loss": 0.1599, + "step": 13476 + }, + { + "epoch": 11.38260135135135, + "grad_norm": 0.36972281336784363, + "learning_rate": 2.8842333333333333e-05, + "loss": 0.0776, + "step": 13477 + }, + { + "epoch": 11.383445945945946, + "grad_norm": 0.378614217042923, + "learning_rate": 2.8842000000000003e-05, + "loss": 0.056, + "step": 13478 + }, + { + "epoch": 11.38429054054054, + "grad_norm": 0.2452418953180313, + "learning_rate": 2.8841666666666665e-05, + "loss": 0.0253, + "step": 13479 + }, + { + "epoch": 11.385135135135135, + "grad_norm": 0.3114946186542511, + "learning_rate": 2.8841333333333335e-05, + "loss": 0.0438, + "step": 13480 + }, + { + "epoch": 11.38597972972973, + "grad_norm": 0.2784431576728821, + "learning_rate": 2.8841e-05, + "loss": 0.0199, + "step": 13481 + }, + { + "epoch": 11.386824324324325, + "grad_norm": 0.3742481470108032, + "learning_rate": 2.8840666666666666e-05, + "loss": 0.0417, + "step": 13482 + }, + { + "epoch": 11.38766891891892, + "grad_norm": 0.37853744626045227, + "learning_rate": 2.8840333333333336e-05, + "loss": 0.0314, + "step": 13483 + }, + { + "epoch": 11.388513513513514, + "grad_norm": 0.2741019129753113, + "learning_rate": 2.8840000000000002e-05, + "loss": 0.0197, + "step": 13484 + }, + { + "epoch": 11.389358108108109, + "grad_norm": 0.2700972855091095, + "learning_rate": 2.8839666666666668e-05, + "loss": 0.0235, + "step": 13485 + }, + { + "epoch": 11.390202702702704, + "grad_norm": 0.24149107933044434, + "learning_rate": 2.8839333333333334e-05, + "loss": 0.03, + "step": 13486 + }, + { + "epoch": 11.391047297297296, + "grad_norm": 0.33767932653427124, + "learning_rate": 2.8839000000000003e-05, + "loss": 0.0299, + "step": 13487 + }, + { + "epoch": 11.391891891891891, + "grad_norm": 0.5320448279380798, + "learning_rate": 2.8838666666666665e-05, + "loss": 0.0297, + "step": 13488 + }, + { + "epoch": 11.392736486486486, + "grad_norm": 0.33649420738220215, + "learning_rate": 2.8838333333333335e-05, + "loss": 0.0446, + "step": 13489 + }, + { + "epoch": 11.39358108108108, + "grad_norm": 0.4889771640300751, + "learning_rate": 2.8838e-05, + "loss": 0.0295, + "step": 13490 + }, + { + "epoch": 11.394425675675675, + "grad_norm": 0.24430713057518005, + "learning_rate": 2.8837666666666667e-05, + "loss": 0.0256, + "step": 13491 + }, + { + "epoch": 11.39527027027027, + "grad_norm": 0.4141314625740051, + "learning_rate": 2.8837333333333336e-05, + "loss": 0.0267, + "step": 13492 + }, + { + "epoch": 11.396114864864865, + "grad_norm": 0.2737921476364136, + "learning_rate": 2.8837000000000002e-05, + "loss": 0.0159, + "step": 13493 + }, + { + "epoch": 11.39695945945946, + "grad_norm": 0.35471826791763306, + "learning_rate": 2.8836666666666668e-05, + "loss": 0.0217, + "step": 13494 + }, + { + "epoch": 11.397804054054054, + "grad_norm": 0.24215254187583923, + "learning_rate": 2.8836333333333334e-05, + "loss": 0.0259, + "step": 13495 + }, + { + "epoch": 11.39864864864865, + "grad_norm": 0.17165718972682953, + "learning_rate": 2.8836000000000003e-05, + "loss": 0.0155, + "step": 13496 + }, + { + "epoch": 11.399493243243244, + "grad_norm": 0.31254076957702637, + "learning_rate": 2.8835666666666666e-05, + "loss": 0.0292, + "step": 13497 + }, + { + "epoch": 11.400337837837839, + "grad_norm": 0.3506532907485962, + "learning_rate": 2.8835333333333335e-05, + "loss": 0.0213, + "step": 13498 + }, + { + "epoch": 11.401182432432432, + "grad_norm": 0.399107426404953, + "learning_rate": 2.8834999999999998e-05, + "loss": 0.0129, + "step": 13499 + }, + { + "epoch": 11.402027027027026, + "grad_norm": 0.2600437104701996, + "learning_rate": 2.8834666666666667e-05, + "loss": 0.0196, + "step": 13500 + }, + { + "epoch": 11.402871621621621, + "grad_norm": 0.4799364507198334, + "learning_rate": 2.8834333333333336e-05, + "loss": 0.0274, + "step": 13501 + }, + { + "epoch": 11.403716216216216, + "grad_norm": 0.3503527343273163, + "learning_rate": 2.8834e-05, + "loss": 0.0174, + "step": 13502 + }, + { + "epoch": 11.40456081081081, + "grad_norm": 0.15146222710609436, + "learning_rate": 2.8833666666666668e-05, + "loss": 0.0112, + "step": 13503 + }, + { + "epoch": 11.405405405405405, + "grad_norm": 0.3694586753845215, + "learning_rate": 2.8833333333333334e-05, + "loss": 0.0233, + "step": 13504 + }, + { + "epoch": 11.40625, + "grad_norm": 0.337032288312912, + "learning_rate": 2.8833e-05, + "loss": 0.0303, + "step": 13505 + }, + { + "epoch": 11.407094594594595, + "grad_norm": 0.26016733050346375, + "learning_rate": 2.8832666666666666e-05, + "loss": 0.0275, + "step": 13506 + }, + { + "epoch": 11.40793918918919, + "grad_norm": 0.36784589290618896, + "learning_rate": 2.8832333333333335e-05, + "loss": 0.025, + "step": 13507 + }, + { + "epoch": 11.408783783783784, + "grad_norm": 0.24308271706104279, + "learning_rate": 2.8832e-05, + "loss": 0.0253, + "step": 13508 + }, + { + "epoch": 11.409628378378379, + "grad_norm": 0.2883460819721222, + "learning_rate": 2.8831666666666667e-05, + "loss": 0.0359, + "step": 13509 + }, + { + "epoch": 11.410472972972974, + "grad_norm": 0.28979048132896423, + "learning_rate": 2.8831333333333336e-05, + "loss": 0.0178, + "step": 13510 + }, + { + "epoch": 11.411317567567568, + "grad_norm": 0.3335135281085968, + "learning_rate": 2.8831e-05, + "loss": 0.0198, + "step": 13511 + }, + { + "epoch": 11.412162162162161, + "grad_norm": 0.3920573890209198, + "learning_rate": 2.8830666666666668e-05, + "loss": 0.0247, + "step": 13512 + }, + { + "epoch": 11.413006756756756, + "grad_norm": 0.2814907431602478, + "learning_rate": 2.8830333333333334e-05, + "loss": 0.023, + "step": 13513 + }, + { + "epoch": 11.41385135135135, + "grad_norm": 0.468325138092041, + "learning_rate": 2.883e-05, + "loss": 0.0281, + "step": 13514 + }, + { + "epoch": 11.414695945945946, + "grad_norm": 0.327646404504776, + "learning_rate": 2.8829666666666666e-05, + "loss": 0.0192, + "step": 13515 + }, + { + "epoch": 11.41554054054054, + "grad_norm": 0.4817902743816376, + "learning_rate": 2.8829333333333335e-05, + "loss": 0.0276, + "step": 13516 + }, + { + "epoch": 11.416385135135135, + "grad_norm": 0.35432833433151245, + "learning_rate": 2.8829e-05, + "loss": 0.0211, + "step": 13517 + }, + { + "epoch": 11.41722972972973, + "grad_norm": 0.31672003865242004, + "learning_rate": 2.8828666666666667e-05, + "loss": 0.0249, + "step": 13518 + }, + { + "epoch": 11.418074324324325, + "grad_norm": 0.16917763650417328, + "learning_rate": 2.8828333333333337e-05, + "loss": 0.012, + "step": 13519 + }, + { + "epoch": 11.41891891891892, + "grad_norm": 0.29449737071990967, + "learning_rate": 2.8828e-05, + "loss": 0.0148, + "step": 13520 + }, + { + "epoch": 11.419763513513514, + "grad_norm": 0.9960865378379822, + "learning_rate": 2.882766666666667e-05, + "loss": 0.0363, + "step": 13521 + }, + { + "epoch": 11.420608108108109, + "grad_norm": 0.2152402400970459, + "learning_rate": 2.8827333333333334e-05, + "loss": 0.0216, + "step": 13522 + }, + { + "epoch": 11.421452702702704, + "grad_norm": 0.5345959663391113, + "learning_rate": 2.8827e-05, + "loss": 0.0296, + "step": 13523 + }, + { + "epoch": 11.422297297297296, + "grad_norm": 0.49557510018348694, + "learning_rate": 2.8826666666666666e-05, + "loss": 0.0226, + "step": 13524 + }, + { + "epoch": 11.423141891891891, + "grad_norm": 0.5220391154289246, + "learning_rate": 2.8826333333333336e-05, + "loss": 0.1936, + "step": 13525 + }, + { + "epoch": 11.423986486486486, + "grad_norm": 0.47518661618232727, + "learning_rate": 2.8826e-05, + "loss": 0.0942, + "step": 13526 + }, + { + "epoch": 11.42483108108108, + "grad_norm": 0.3627246916294098, + "learning_rate": 2.8825666666666667e-05, + "loss": 0.065, + "step": 13527 + }, + { + "epoch": 11.425675675675675, + "grad_norm": 0.4585052728652954, + "learning_rate": 2.8825333333333333e-05, + "loss": 0.0412, + "step": 13528 + }, + { + "epoch": 11.42652027027027, + "grad_norm": 0.3345335125923157, + "learning_rate": 2.8825e-05, + "loss": 0.0378, + "step": 13529 + }, + { + "epoch": 11.427364864864865, + "grad_norm": 0.3242034316062927, + "learning_rate": 2.882466666666667e-05, + "loss": 0.0252, + "step": 13530 + }, + { + "epoch": 11.42820945945946, + "grad_norm": 0.2726854383945465, + "learning_rate": 2.882433333333333e-05, + "loss": 0.0273, + "step": 13531 + }, + { + "epoch": 11.429054054054054, + "grad_norm": 0.4775456190109253, + "learning_rate": 2.8824e-05, + "loss": 0.0325, + "step": 13532 + }, + { + "epoch": 11.42989864864865, + "grad_norm": 0.2526550889015198, + "learning_rate": 2.882366666666667e-05, + "loss": 0.0178, + "step": 13533 + }, + { + "epoch": 11.430743243243244, + "grad_norm": 0.2973991334438324, + "learning_rate": 2.8823333333333332e-05, + "loss": 0.0261, + "step": 13534 + }, + { + "epoch": 11.431587837837839, + "grad_norm": 0.31252187490463257, + "learning_rate": 2.8823000000000002e-05, + "loss": 0.0224, + "step": 13535 + }, + { + "epoch": 11.432432432432432, + "grad_norm": 0.5864198803901672, + "learning_rate": 2.8822666666666668e-05, + "loss": 0.0303, + "step": 13536 + }, + { + "epoch": 11.433277027027026, + "grad_norm": 0.22236618399620056, + "learning_rate": 2.8822333333333334e-05, + "loss": 0.0231, + "step": 13537 + }, + { + "epoch": 11.434121621621621, + "grad_norm": 0.258003830909729, + "learning_rate": 2.8822e-05, + "loss": 0.0222, + "step": 13538 + }, + { + "epoch": 11.434966216216216, + "grad_norm": 0.279030442237854, + "learning_rate": 2.882166666666667e-05, + "loss": 0.0189, + "step": 13539 + }, + { + "epoch": 11.43581081081081, + "grad_norm": 0.3123462200164795, + "learning_rate": 2.882133333333333e-05, + "loss": 0.0193, + "step": 13540 + }, + { + "epoch": 11.436655405405405, + "grad_norm": 0.2831032872200012, + "learning_rate": 2.8821e-05, + "loss": 0.0315, + "step": 13541 + }, + { + "epoch": 11.4375, + "grad_norm": 0.2396755814552307, + "learning_rate": 2.882066666666667e-05, + "loss": 0.0245, + "step": 13542 + }, + { + "epoch": 11.438344594594595, + "grad_norm": 0.523515522480011, + "learning_rate": 2.8820333333333333e-05, + "loss": 0.0414, + "step": 13543 + }, + { + "epoch": 11.43918918918919, + "grad_norm": 0.33846619725227356, + "learning_rate": 2.8820000000000002e-05, + "loss": 0.0313, + "step": 13544 + }, + { + "epoch": 11.440033783783784, + "grad_norm": 0.5576091408729553, + "learning_rate": 2.8819666666666668e-05, + "loss": 0.0263, + "step": 13545 + }, + { + "epoch": 11.440878378378379, + "grad_norm": 0.21720600128173828, + "learning_rate": 2.8819333333333334e-05, + "loss": 0.0142, + "step": 13546 + }, + { + "epoch": 11.441722972972974, + "grad_norm": 0.37199410796165466, + "learning_rate": 2.8819e-05, + "loss": 0.0261, + "step": 13547 + }, + { + "epoch": 11.442567567567568, + "grad_norm": 0.2431042641401291, + "learning_rate": 2.881866666666667e-05, + "loss": 0.023, + "step": 13548 + }, + { + "epoch": 11.443412162162161, + "grad_norm": 0.2216186374425888, + "learning_rate": 2.881833333333333e-05, + "loss": 0.0172, + "step": 13549 + }, + { + "epoch": 11.444256756756756, + "grad_norm": 0.26732850074768066, + "learning_rate": 2.8818e-05, + "loss": 0.024, + "step": 13550 + }, + { + "epoch": 11.44510135135135, + "grad_norm": 0.1999468356370926, + "learning_rate": 2.881766666666667e-05, + "loss": 0.0164, + "step": 13551 + }, + { + "epoch": 11.445945945945946, + "grad_norm": 0.20077773928642273, + "learning_rate": 2.8817333333333333e-05, + "loss": 0.0134, + "step": 13552 + }, + { + "epoch": 11.44679054054054, + "grad_norm": 0.2484414428472519, + "learning_rate": 2.8817000000000002e-05, + "loss": 0.0153, + "step": 13553 + }, + { + "epoch": 11.447635135135135, + "grad_norm": 1.6798944473266602, + "learning_rate": 2.8816666666666668e-05, + "loss": 0.0462, + "step": 13554 + }, + { + "epoch": 11.44847972972973, + "grad_norm": 0.322839617729187, + "learning_rate": 2.8816333333333334e-05, + "loss": 0.0198, + "step": 13555 + }, + { + "epoch": 11.449324324324325, + "grad_norm": 0.3374478220939636, + "learning_rate": 2.8816e-05, + "loss": 0.0267, + "step": 13556 + }, + { + "epoch": 11.45016891891892, + "grad_norm": 0.4568428695201874, + "learning_rate": 2.881566666666667e-05, + "loss": 0.0375, + "step": 13557 + }, + { + "epoch": 11.451013513513514, + "grad_norm": 0.38225057721138, + "learning_rate": 2.8815333333333335e-05, + "loss": 0.0221, + "step": 13558 + }, + { + "epoch": 11.451858108108109, + "grad_norm": 0.6893587112426758, + "learning_rate": 2.8815e-05, + "loss": 0.0337, + "step": 13559 + }, + { + "epoch": 11.452702702702704, + "grad_norm": 0.6149942278862, + "learning_rate": 2.8814666666666667e-05, + "loss": 0.0238, + "step": 13560 + }, + { + "epoch": 11.453547297297296, + "grad_norm": 0.55075603723526, + "learning_rate": 2.8814333333333333e-05, + "loss": 0.0255, + "step": 13561 + }, + { + "epoch": 11.454391891891891, + "grad_norm": 0.3818422555923462, + "learning_rate": 2.8814000000000002e-05, + "loss": 0.0603, + "step": 13562 + }, + { + "epoch": 11.455236486486486, + "grad_norm": 0.3249177932739258, + "learning_rate": 2.8813666666666665e-05, + "loss": 0.0178, + "step": 13563 + }, + { + "epoch": 11.45608108108108, + "grad_norm": 0.3920181691646576, + "learning_rate": 2.8813333333333334e-05, + "loss": 0.0268, + "step": 13564 + }, + { + "epoch": 11.456925675675675, + "grad_norm": 0.448263943195343, + "learning_rate": 2.8813e-05, + "loss": 0.0269, + "step": 13565 + }, + { + "epoch": 11.45777027027027, + "grad_norm": 0.3481191396713257, + "learning_rate": 2.8812666666666666e-05, + "loss": 0.0195, + "step": 13566 + }, + { + "epoch": 11.458614864864865, + "grad_norm": 0.20308363437652588, + "learning_rate": 2.8812333333333335e-05, + "loss": 0.0143, + "step": 13567 + }, + { + "epoch": 11.45945945945946, + "grad_norm": 0.278501033782959, + "learning_rate": 2.8812e-05, + "loss": 0.0249, + "step": 13568 + }, + { + "epoch": 11.460304054054054, + "grad_norm": 0.33208468556404114, + "learning_rate": 2.8811666666666667e-05, + "loss": 0.0275, + "step": 13569 + }, + { + "epoch": 11.46114864864865, + "grad_norm": 0.3918728232383728, + "learning_rate": 2.8811333333333333e-05, + "loss": 0.0208, + "step": 13570 + }, + { + "epoch": 11.461993243243244, + "grad_norm": 0.36438852548599243, + "learning_rate": 2.8811000000000002e-05, + "loss": 0.0275, + "step": 13571 + }, + { + "epoch": 11.462837837837839, + "grad_norm": 0.4104701578617096, + "learning_rate": 2.8810666666666665e-05, + "loss": 0.0181, + "step": 13572 + }, + { + "epoch": 11.463682432432432, + "grad_norm": 0.8396336436271667, + "learning_rate": 2.8810333333333334e-05, + "loss": 0.0525, + "step": 13573 + }, + { + "epoch": 11.464527027027026, + "grad_norm": 0.8270746469497681, + "learning_rate": 2.881e-05, + "loss": 0.105, + "step": 13574 + }, + { + "epoch": 11.465371621621621, + "grad_norm": 0.6395666003227234, + "learning_rate": 2.8809666666666666e-05, + "loss": 0.1968, + "step": 13575 + }, + { + "epoch": 11.466216216216216, + "grad_norm": 0.7750589847564697, + "learning_rate": 2.8809333333333336e-05, + "loss": 0.0981, + "step": 13576 + }, + { + "epoch": 11.46706081081081, + "grad_norm": 0.47813931107521057, + "learning_rate": 2.8809e-05, + "loss": 0.0652, + "step": 13577 + }, + { + "epoch": 11.467905405405405, + "grad_norm": 0.2417578101158142, + "learning_rate": 2.8808666666666667e-05, + "loss": 0.0314, + "step": 13578 + }, + { + "epoch": 11.46875, + "grad_norm": 0.33409976959228516, + "learning_rate": 2.8808333333333333e-05, + "loss": 0.0476, + "step": 13579 + }, + { + "epoch": 11.469594594594595, + "grad_norm": 0.7999302744865417, + "learning_rate": 2.8808000000000003e-05, + "loss": 0.061, + "step": 13580 + }, + { + "epoch": 11.47043918918919, + "grad_norm": 0.33886590600013733, + "learning_rate": 2.8807666666666665e-05, + "loss": 0.0182, + "step": 13581 + }, + { + "epoch": 11.471283783783784, + "grad_norm": 1.2611913681030273, + "learning_rate": 2.8807333333333335e-05, + "loss": 0.0261, + "step": 13582 + }, + { + "epoch": 11.472128378378379, + "grad_norm": 0.22558213770389557, + "learning_rate": 2.8807000000000004e-05, + "loss": 0.0315, + "step": 13583 + }, + { + "epoch": 11.472972972972974, + "grad_norm": 0.362215131521225, + "learning_rate": 2.8806666666666666e-05, + "loss": 0.0396, + "step": 13584 + }, + { + "epoch": 11.473817567567568, + "grad_norm": 0.26424047350883484, + "learning_rate": 2.8806333333333336e-05, + "loss": 0.0188, + "step": 13585 + }, + { + "epoch": 11.474662162162161, + "grad_norm": 0.3422916531562805, + "learning_rate": 2.8806e-05, + "loss": 0.0231, + "step": 13586 + }, + { + "epoch": 11.475506756756756, + "grad_norm": 0.3548143804073334, + "learning_rate": 2.8805666666666668e-05, + "loss": 0.048, + "step": 13587 + }, + { + "epoch": 11.47635135135135, + "grad_norm": 0.3209940195083618, + "learning_rate": 2.8805333333333334e-05, + "loss": 0.0237, + "step": 13588 + }, + { + "epoch": 11.477195945945946, + "grad_norm": 0.41238972544670105, + "learning_rate": 2.8805e-05, + "loss": 0.0258, + "step": 13589 + }, + { + "epoch": 11.47804054054054, + "grad_norm": 0.3355490565299988, + "learning_rate": 2.8804666666666665e-05, + "loss": 0.0227, + "step": 13590 + }, + { + "epoch": 11.478885135135135, + "grad_norm": 0.3590989112854004, + "learning_rate": 2.8804333333333335e-05, + "loss": 0.0327, + "step": 13591 + }, + { + "epoch": 11.47972972972973, + "grad_norm": 0.2878223657608032, + "learning_rate": 2.8804e-05, + "loss": 0.0237, + "step": 13592 + }, + { + "epoch": 11.480574324324325, + "grad_norm": 0.3814324140548706, + "learning_rate": 2.8803666666666667e-05, + "loss": 0.0306, + "step": 13593 + }, + { + "epoch": 11.48141891891892, + "grad_norm": 0.19279910624027252, + "learning_rate": 2.8803333333333336e-05, + "loss": 0.0219, + "step": 13594 + }, + { + "epoch": 11.482263513513514, + "grad_norm": 0.20941248536109924, + "learning_rate": 2.8803e-05, + "loss": 0.0167, + "step": 13595 + }, + { + "epoch": 11.483108108108109, + "grad_norm": 0.5013051629066467, + "learning_rate": 2.8802666666666668e-05, + "loss": 0.0277, + "step": 13596 + }, + { + "epoch": 11.483952702702704, + "grad_norm": 0.41034430265426636, + "learning_rate": 2.8802333333333334e-05, + "loss": 0.0278, + "step": 13597 + }, + { + "epoch": 11.484797297297296, + "grad_norm": 0.2548760175704956, + "learning_rate": 2.8802e-05, + "loss": 0.0217, + "step": 13598 + }, + { + "epoch": 11.485641891891891, + "grad_norm": 0.4238179922103882, + "learning_rate": 2.8801666666666666e-05, + "loss": 0.0467, + "step": 13599 + }, + { + "epoch": 11.486486486486486, + "grad_norm": 0.29049330949783325, + "learning_rate": 2.8801333333333335e-05, + "loss": 0.0185, + "step": 13600 + }, + { + "epoch": 11.48733108108108, + "grad_norm": 0.4254271686077118, + "learning_rate": 2.8801e-05, + "loss": 0.0295, + "step": 13601 + }, + { + "epoch": 11.488175675675675, + "grad_norm": 0.4025759696960449, + "learning_rate": 2.8800666666666667e-05, + "loss": 0.0196, + "step": 13602 + }, + { + "epoch": 11.48902027027027, + "grad_norm": 0.33187946677207947, + "learning_rate": 2.8800333333333336e-05, + "loss": 0.0316, + "step": 13603 + }, + { + "epoch": 11.489864864864865, + "grad_norm": 0.26855477690696716, + "learning_rate": 2.88e-05, + "loss": 0.0238, + "step": 13604 + }, + { + "epoch": 11.49070945945946, + "grad_norm": 0.34282463788986206, + "learning_rate": 2.8799666666666668e-05, + "loss": 0.0159, + "step": 13605 + }, + { + "epoch": 11.491554054054054, + "grad_norm": 0.44711512327194214, + "learning_rate": 2.8799333333333334e-05, + "loss": 0.0191, + "step": 13606 + }, + { + "epoch": 11.49239864864865, + "grad_norm": 0.20596739649772644, + "learning_rate": 2.8799e-05, + "loss": 0.0257, + "step": 13607 + }, + { + "epoch": 11.493243243243244, + "grad_norm": 0.48141488432884216, + "learning_rate": 2.879866666666667e-05, + "loss": 0.0395, + "step": 13608 + }, + { + "epoch": 11.494087837837839, + "grad_norm": 0.43543824553489685, + "learning_rate": 2.8798333333333335e-05, + "loss": 0.0423, + "step": 13609 + }, + { + "epoch": 11.494932432432432, + "grad_norm": 0.2954997718334198, + "learning_rate": 2.8798e-05, + "loss": 0.0281, + "step": 13610 + }, + { + "epoch": 11.495777027027026, + "grad_norm": 0.2731856405735016, + "learning_rate": 2.8797666666666667e-05, + "loss": 0.0226, + "step": 13611 + }, + { + "epoch": 11.496621621621621, + "grad_norm": 0.17516763508319855, + "learning_rate": 2.8797333333333336e-05, + "loss": 0.0197, + "step": 13612 + }, + { + "epoch": 11.497466216216216, + "grad_norm": 0.2959500849246979, + "learning_rate": 2.8797e-05, + "loss": 0.0167, + "step": 13613 + }, + { + "epoch": 11.49831081081081, + "grad_norm": 0.4664461016654968, + "learning_rate": 2.8796666666666668e-05, + "loss": 0.0265, + "step": 13614 + }, + { + "epoch": 11.499155405405405, + "grad_norm": 0.3439592719078064, + "learning_rate": 2.8796333333333334e-05, + "loss": 0.0304, + "step": 13615 + }, + { + "epoch": 11.5, + "grad_norm": 0.20778866112232208, + "learning_rate": 2.8796e-05, + "loss": 0.0144, + "step": 13616 + }, + { + "epoch": 11.500844594594595, + "grad_norm": 0.32025179266929626, + "learning_rate": 2.879566666666667e-05, + "loss": 0.0231, + "step": 13617 + }, + { + "epoch": 11.50168918918919, + "grad_norm": 0.21295997500419617, + "learning_rate": 2.8795333333333332e-05, + "loss": 0.0198, + "step": 13618 + }, + { + "epoch": 11.502533783783784, + "grad_norm": 0.6278092265129089, + "learning_rate": 2.8795e-05, + "loss": 0.0313, + "step": 13619 + }, + { + "epoch": 11.503378378378379, + "grad_norm": 0.6439099311828613, + "learning_rate": 2.8794666666666667e-05, + "loss": 0.0426, + "step": 13620 + }, + { + "epoch": 11.504222972972974, + "grad_norm": 0.3631150424480438, + "learning_rate": 2.8794333333333333e-05, + "loss": 0.032, + "step": 13621 + }, + { + "epoch": 11.505067567567568, + "grad_norm": 1.1315277814865112, + "learning_rate": 2.8794e-05, + "loss": 0.0321, + "step": 13622 + }, + { + "epoch": 11.505912162162161, + "grad_norm": 1.0340323448181152, + "learning_rate": 2.879366666666667e-05, + "loss": 0.0556, + "step": 13623 + }, + { + "epoch": 11.506756756756756, + "grad_norm": 1.1579055786132812, + "learning_rate": 2.879333333333333e-05, + "loss": 0.0613, + "step": 13624 + }, + { + "epoch": 11.50760135135135, + "grad_norm": 0.7918294072151184, + "learning_rate": 2.8793e-05, + "loss": 0.198, + "step": 13625 + }, + { + "epoch": 11.508445945945946, + "grad_norm": 0.3678589165210724, + "learning_rate": 2.879266666666667e-05, + "loss": 0.0934, + "step": 13626 + }, + { + "epoch": 11.50929054054054, + "grad_norm": 0.5512824654579163, + "learning_rate": 2.8792333333333332e-05, + "loss": 0.0473, + "step": 13627 + }, + { + "epoch": 11.510135135135135, + "grad_norm": 0.2118225395679474, + "learning_rate": 2.8792e-05, + "loss": 0.0262, + "step": 13628 + }, + { + "epoch": 11.51097972972973, + "grad_norm": 0.31482622027397156, + "learning_rate": 2.8791666666666667e-05, + "loss": 0.0477, + "step": 13629 + }, + { + "epoch": 11.511824324324325, + "grad_norm": 0.46805012226104736, + "learning_rate": 2.8791333333333333e-05, + "loss": 0.0452, + "step": 13630 + }, + { + "epoch": 11.51266891891892, + "grad_norm": 0.6306435465812683, + "learning_rate": 2.8791e-05, + "loss": 0.0274, + "step": 13631 + }, + { + "epoch": 11.513513513513514, + "grad_norm": 0.3375716507434845, + "learning_rate": 2.879066666666667e-05, + "loss": 0.0276, + "step": 13632 + }, + { + "epoch": 11.514358108108109, + "grad_norm": 0.45380643010139465, + "learning_rate": 2.8790333333333335e-05, + "loss": 0.0309, + "step": 13633 + }, + { + "epoch": 11.515202702702704, + "grad_norm": 0.40070122480392456, + "learning_rate": 2.879e-05, + "loss": 0.03, + "step": 13634 + }, + { + "epoch": 11.516047297297296, + "grad_norm": 0.25571852922439575, + "learning_rate": 2.878966666666667e-05, + "loss": 0.0151, + "step": 13635 + }, + { + "epoch": 11.516891891891891, + "grad_norm": 0.337115079164505, + "learning_rate": 2.8789333333333332e-05, + "loss": 0.035, + "step": 13636 + }, + { + "epoch": 11.517736486486486, + "grad_norm": 0.47598496079444885, + "learning_rate": 2.8789e-05, + "loss": 0.0279, + "step": 13637 + }, + { + "epoch": 11.51858108108108, + "grad_norm": 0.27178770303726196, + "learning_rate": 2.8788666666666668e-05, + "loss": 0.0307, + "step": 13638 + }, + { + "epoch": 11.519425675675675, + "grad_norm": 0.2478650063276291, + "learning_rate": 2.8788333333333334e-05, + "loss": 0.0254, + "step": 13639 + }, + { + "epoch": 11.52027027027027, + "grad_norm": 0.25798293948173523, + "learning_rate": 2.8788e-05, + "loss": 0.028, + "step": 13640 + }, + { + "epoch": 11.521114864864865, + "grad_norm": 0.26836690306663513, + "learning_rate": 2.878766666666667e-05, + "loss": 0.0205, + "step": 13641 + }, + { + "epoch": 11.52195945945946, + "grad_norm": 0.23713107407093048, + "learning_rate": 2.8787333333333335e-05, + "loss": 0.0227, + "step": 13642 + }, + { + "epoch": 11.522804054054054, + "grad_norm": 0.2444659322500229, + "learning_rate": 2.8787e-05, + "loss": 0.0205, + "step": 13643 + }, + { + "epoch": 11.52364864864865, + "grad_norm": 0.7330899238586426, + "learning_rate": 2.878666666666667e-05, + "loss": 0.0242, + "step": 13644 + }, + { + "epoch": 11.524493243243244, + "grad_norm": 0.37575384974479675, + "learning_rate": 2.8786333333333333e-05, + "loss": 0.0176, + "step": 13645 + }, + { + "epoch": 11.525337837837839, + "grad_norm": 0.29699674248695374, + "learning_rate": 2.8786000000000002e-05, + "loss": 0.0319, + "step": 13646 + }, + { + "epoch": 11.526182432432432, + "grad_norm": 0.23884417116641998, + "learning_rate": 2.8785666666666668e-05, + "loss": 0.0213, + "step": 13647 + }, + { + "epoch": 11.527027027027026, + "grad_norm": 0.3170488476753235, + "learning_rate": 2.8785333333333334e-05, + "loss": 0.0292, + "step": 13648 + }, + { + "epoch": 11.527871621621621, + "grad_norm": 0.6899104714393616, + "learning_rate": 2.8785e-05, + "loss": 0.03, + "step": 13649 + }, + { + "epoch": 11.528716216216216, + "grad_norm": 0.33873090147972107, + "learning_rate": 2.8784666666666666e-05, + "loss": 0.0131, + "step": 13650 + }, + { + "epoch": 11.52956081081081, + "grad_norm": 0.4495965242385864, + "learning_rate": 2.8784333333333335e-05, + "loss": 0.039, + "step": 13651 + }, + { + "epoch": 11.530405405405405, + "grad_norm": 0.2806569039821625, + "learning_rate": 2.8784e-05, + "loss": 0.0119, + "step": 13652 + }, + { + "epoch": 11.53125, + "grad_norm": 0.36799055337905884, + "learning_rate": 2.8783666666666667e-05, + "loss": 0.0235, + "step": 13653 + }, + { + "epoch": 11.532094594594595, + "grad_norm": 0.3447432219982147, + "learning_rate": 2.8783333333333333e-05, + "loss": 0.0289, + "step": 13654 + }, + { + "epoch": 11.53293918918919, + "grad_norm": 0.36367344856262207, + "learning_rate": 2.8783000000000002e-05, + "loss": 0.028, + "step": 13655 + }, + { + "epoch": 11.533783783783784, + "grad_norm": 0.30374690890312195, + "learning_rate": 2.8782666666666665e-05, + "loss": 0.0232, + "step": 13656 + }, + { + "epoch": 11.534628378378379, + "grad_norm": 0.23870769143104553, + "learning_rate": 2.8782333333333334e-05, + "loss": 0.0212, + "step": 13657 + }, + { + "epoch": 11.535472972972974, + "grad_norm": 0.49551922082901, + "learning_rate": 2.8782000000000003e-05, + "loss": 0.0282, + "step": 13658 + }, + { + "epoch": 11.536317567567568, + "grad_norm": 0.25334542989730835, + "learning_rate": 2.8781666666666666e-05, + "loss": 0.0111, + "step": 13659 + }, + { + "epoch": 11.537162162162161, + "grad_norm": 0.25567764043807983, + "learning_rate": 2.8781333333333335e-05, + "loss": 0.0213, + "step": 13660 + }, + { + "epoch": 11.538006756756756, + "grad_norm": 0.2759258449077606, + "learning_rate": 2.8781e-05, + "loss": 0.0202, + "step": 13661 + }, + { + "epoch": 11.53885135135135, + "grad_norm": 0.5298434495925903, + "learning_rate": 2.8780666666666667e-05, + "loss": 0.0231, + "step": 13662 + }, + { + "epoch": 11.539695945945946, + "grad_norm": 0.30068644881248474, + "learning_rate": 2.8780333333333333e-05, + "loss": 0.0218, + "step": 13663 + }, + { + "epoch": 11.54054054054054, + "grad_norm": 0.531242847442627, + "learning_rate": 2.8780000000000002e-05, + "loss": 0.017, + "step": 13664 + }, + { + "epoch": 11.541385135135135, + "grad_norm": 0.2183656543493271, + "learning_rate": 2.8779666666666665e-05, + "loss": 0.0173, + "step": 13665 + }, + { + "epoch": 11.54222972972973, + "grad_norm": 0.2758581340312958, + "learning_rate": 2.8779333333333334e-05, + "loss": 0.0128, + "step": 13666 + }, + { + "epoch": 11.543074324324325, + "grad_norm": 0.49302583932876587, + "learning_rate": 2.8779000000000003e-05, + "loss": 0.0322, + "step": 13667 + }, + { + "epoch": 11.54391891891892, + "grad_norm": 0.2805187702178955, + "learning_rate": 2.8778666666666666e-05, + "loss": 0.0219, + "step": 13668 + }, + { + "epoch": 11.544763513513514, + "grad_norm": 0.29009243845939636, + "learning_rate": 2.8778333333333335e-05, + "loss": 0.0239, + "step": 13669 + }, + { + "epoch": 11.545608108108109, + "grad_norm": 0.6491265296936035, + "learning_rate": 2.8778e-05, + "loss": 0.0417, + "step": 13670 + }, + { + "epoch": 11.546452702702704, + "grad_norm": 1.341991662979126, + "learning_rate": 2.8777666666666667e-05, + "loss": 0.022, + "step": 13671 + }, + { + "epoch": 11.547297297297296, + "grad_norm": 0.4814881980419159, + "learning_rate": 2.8777333333333333e-05, + "loss": 0.039, + "step": 13672 + }, + { + "epoch": 11.548141891891891, + "grad_norm": 0.3525877296924591, + "learning_rate": 2.8777000000000002e-05, + "loss": 0.0219, + "step": 13673 + }, + { + "epoch": 11.548986486486486, + "grad_norm": 1.130208969116211, + "learning_rate": 2.8776666666666665e-05, + "loss": 0.0396, + "step": 13674 + }, + { + "epoch": 11.54983108108108, + "grad_norm": 0.4910792112350464, + "learning_rate": 2.8776333333333334e-05, + "loss": 0.211, + "step": 13675 + }, + { + "epoch": 11.550675675675675, + "grad_norm": 0.5633717179298401, + "learning_rate": 2.8776000000000004e-05, + "loss": 0.1706, + "step": 13676 + }, + { + "epoch": 11.55152027027027, + "grad_norm": 0.4284374713897705, + "learning_rate": 2.8775666666666666e-05, + "loss": 0.0732, + "step": 13677 + }, + { + "epoch": 11.552364864864865, + "grad_norm": 0.6311453580856323, + "learning_rate": 2.8775333333333336e-05, + "loss": 0.0627, + "step": 13678 + }, + { + "epoch": 11.55320945945946, + "grad_norm": 0.463954895734787, + "learning_rate": 2.8774999999999998e-05, + "loss": 0.0372, + "step": 13679 + }, + { + "epoch": 11.554054054054054, + "grad_norm": 0.3665878176689148, + "learning_rate": 2.8774666666666667e-05, + "loss": 0.0273, + "step": 13680 + }, + { + "epoch": 11.55489864864865, + "grad_norm": 0.2524253726005554, + "learning_rate": 2.8774333333333333e-05, + "loss": 0.0273, + "step": 13681 + }, + { + "epoch": 11.555743243243244, + "grad_norm": 0.22148072719573975, + "learning_rate": 2.8774e-05, + "loss": 0.0205, + "step": 13682 + }, + { + "epoch": 11.556587837837839, + "grad_norm": 0.28074049949645996, + "learning_rate": 2.877366666666667e-05, + "loss": 0.0304, + "step": 13683 + }, + { + "epoch": 11.557432432432432, + "grad_norm": 0.2614453434944153, + "learning_rate": 2.8773333333333335e-05, + "loss": 0.0227, + "step": 13684 + }, + { + "epoch": 11.558277027027026, + "grad_norm": 0.2715836763381958, + "learning_rate": 2.8773e-05, + "loss": 0.0248, + "step": 13685 + }, + { + "epoch": 11.559121621621621, + "grad_norm": 0.32747358083724976, + "learning_rate": 2.8772666666666666e-05, + "loss": 0.0325, + "step": 13686 + }, + { + "epoch": 11.559966216216216, + "grad_norm": 0.30502820014953613, + "learning_rate": 2.8772333333333336e-05, + "loss": 0.0343, + "step": 13687 + }, + { + "epoch": 11.56081081081081, + "grad_norm": 0.23988886177539825, + "learning_rate": 2.8771999999999998e-05, + "loss": 0.0161, + "step": 13688 + }, + { + "epoch": 11.561655405405405, + "grad_norm": 0.20716869831085205, + "learning_rate": 2.8771666666666668e-05, + "loss": 0.0154, + "step": 13689 + }, + { + "epoch": 11.5625, + "grad_norm": 0.2617085576057434, + "learning_rate": 2.8771333333333334e-05, + "loss": 0.0383, + "step": 13690 + }, + { + "epoch": 11.563344594594595, + "grad_norm": 0.4268733859062195, + "learning_rate": 2.8771e-05, + "loss": 0.0296, + "step": 13691 + }, + { + "epoch": 11.56418918918919, + "grad_norm": 0.2123112678527832, + "learning_rate": 2.877066666666667e-05, + "loss": 0.0218, + "step": 13692 + }, + { + "epoch": 11.565033783783784, + "grad_norm": 0.11097699403762817, + "learning_rate": 2.8770333333333335e-05, + "loss": 0.0158, + "step": 13693 + }, + { + "epoch": 11.565878378378379, + "grad_norm": 0.15120829641819, + "learning_rate": 2.877e-05, + "loss": 0.0144, + "step": 13694 + }, + { + "epoch": 11.566722972972974, + "grad_norm": 0.37396231293678284, + "learning_rate": 2.8769666666666667e-05, + "loss": 0.0269, + "step": 13695 + }, + { + "epoch": 11.567567567567568, + "grad_norm": 0.3999028503894806, + "learning_rate": 2.8769333333333336e-05, + "loss": 0.0262, + "step": 13696 + }, + { + "epoch": 11.568412162162161, + "grad_norm": 0.38771095871925354, + "learning_rate": 2.8769e-05, + "loss": 0.022, + "step": 13697 + }, + { + "epoch": 11.569256756756756, + "grad_norm": 0.6471091508865356, + "learning_rate": 2.8768666666666668e-05, + "loss": 0.0261, + "step": 13698 + }, + { + "epoch": 11.57010135135135, + "grad_norm": 0.502131998538971, + "learning_rate": 2.8768333333333334e-05, + "loss": 0.0351, + "step": 13699 + }, + { + "epoch": 11.570945945945946, + "grad_norm": 0.5360338687896729, + "learning_rate": 2.8768e-05, + "loss": 0.0344, + "step": 13700 + }, + { + "epoch": 11.57179054054054, + "grad_norm": 0.26280477643013, + "learning_rate": 2.876766666666667e-05, + "loss": 0.0176, + "step": 13701 + }, + { + "epoch": 11.572635135135135, + "grad_norm": 0.283238023519516, + "learning_rate": 2.8767333333333335e-05, + "loss": 0.0153, + "step": 13702 + }, + { + "epoch": 11.57347972972973, + "grad_norm": 0.2732756435871124, + "learning_rate": 2.8767e-05, + "loss": 0.0231, + "step": 13703 + }, + { + "epoch": 11.574324324324325, + "grad_norm": 0.6135159730911255, + "learning_rate": 2.8766666666666667e-05, + "loss": 0.0365, + "step": 13704 + }, + { + "epoch": 11.57516891891892, + "grad_norm": 0.4173235595226288, + "learning_rate": 2.8766333333333336e-05, + "loss": 0.0283, + "step": 13705 + }, + { + "epoch": 11.576013513513514, + "grad_norm": 0.3728609085083008, + "learning_rate": 2.8766e-05, + "loss": 0.0389, + "step": 13706 + }, + { + "epoch": 11.576858108108109, + "grad_norm": 0.17257781326770782, + "learning_rate": 2.8765666666666668e-05, + "loss": 0.0172, + "step": 13707 + }, + { + "epoch": 11.577702702702704, + "grad_norm": 0.4616944193840027, + "learning_rate": 2.8765333333333337e-05, + "loss": 0.0301, + "step": 13708 + }, + { + "epoch": 11.578547297297296, + "grad_norm": 0.6203587055206299, + "learning_rate": 2.8765e-05, + "loss": 0.03, + "step": 13709 + }, + { + "epoch": 11.579391891891891, + "grad_norm": 0.3521270751953125, + "learning_rate": 2.876466666666667e-05, + "loss": 0.0166, + "step": 13710 + }, + { + "epoch": 11.580236486486486, + "grad_norm": 0.4294292628765106, + "learning_rate": 2.8764333333333332e-05, + "loss": 0.0308, + "step": 13711 + }, + { + "epoch": 11.58108108108108, + "grad_norm": 0.3224467635154724, + "learning_rate": 2.8764e-05, + "loss": 0.0209, + "step": 13712 + }, + { + "epoch": 11.581925675675675, + "grad_norm": 0.2997576892375946, + "learning_rate": 2.8763666666666667e-05, + "loss": 0.0224, + "step": 13713 + }, + { + "epoch": 11.58277027027027, + "grad_norm": 0.35305118560791016, + "learning_rate": 2.8763333333333333e-05, + "loss": 0.0236, + "step": 13714 + }, + { + "epoch": 11.583614864864865, + "grad_norm": 0.4124198257923126, + "learning_rate": 2.8763e-05, + "loss": 0.0313, + "step": 13715 + }, + { + "epoch": 11.58445945945946, + "grad_norm": 0.35394543409347534, + "learning_rate": 2.8762666666666668e-05, + "loss": 0.0176, + "step": 13716 + }, + { + "epoch": 11.585304054054054, + "grad_norm": 0.5688841342926025, + "learning_rate": 2.8762333333333334e-05, + "loss": 0.0318, + "step": 13717 + }, + { + "epoch": 11.58614864864865, + "grad_norm": 0.3786866366863251, + "learning_rate": 2.8762e-05, + "loss": 0.0195, + "step": 13718 + }, + { + "epoch": 11.586993243243244, + "grad_norm": 0.3278697729110718, + "learning_rate": 2.876166666666667e-05, + "loss": 0.0232, + "step": 13719 + }, + { + "epoch": 11.587837837837839, + "grad_norm": 0.2638886272907257, + "learning_rate": 2.8761333333333332e-05, + "loss": 0.0219, + "step": 13720 + }, + { + "epoch": 11.588682432432432, + "grad_norm": 0.21993261575698853, + "learning_rate": 2.8761e-05, + "loss": 0.0147, + "step": 13721 + }, + { + "epoch": 11.589527027027026, + "grad_norm": 0.6928108334541321, + "learning_rate": 2.8760666666666667e-05, + "loss": 0.0253, + "step": 13722 + }, + { + "epoch": 11.590371621621621, + "grad_norm": 1.9212005138397217, + "learning_rate": 2.8760333333333333e-05, + "loss": 0.0879, + "step": 13723 + }, + { + "epoch": 11.591216216216216, + "grad_norm": 0.4612758457660675, + "learning_rate": 2.876e-05, + "loss": 0.0287, + "step": 13724 + }, + { + "epoch": 11.59206081081081, + "grad_norm": 0.5565750002861023, + "learning_rate": 2.875966666666667e-05, + "loss": 0.2344, + "step": 13725 + }, + { + "epoch": 11.592905405405405, + "grad_norm": 0.6976006031036377, + "learning_rate": 2.8759333333333334e-05, + "loss": 0.1267, + "step": 13726 + }, + { + "epoch": 11.59375, + "grad_norm": 0.488300621509552, + "learning_rate": 2.8759e-05, + "loss": 0.0814, + "step": 13727 + }, + { + "epoch": 11.594594594594595, + "grad_norm": 0.22508738934993744, + "learning_rate": 2.875866666666667e-05, + "loss": 0.0271, + "step": 13728 + }, + { + "epoch": 11.59543918918919, + "grad_norm": 0.43909209966659546, + "learning_rate": 2.8758333333333332e-05, + "loss": 0.0335, + "step": 13729 + }, + { + "epoch": 11.596283783783784, + "grad_norm": 0.27178582549095154, + "learning_rate": 2.8758e-05, + "loss": 0.0261, + "step": 13730 + }, + { + "epoch": 11.597128378378379, + "grad_norm": 0.39525723457336426, + "learning_rate": 2.8757666666666667e-05, + "loss": 0.0241, + "step": 13731 + }, + { + "epoch": 11.597972972972974, + "grad_norm": 1.1374846696853638, + "learning_rate": 2.8757333333333333e-05, + "loss": 0.0531, + "step": 13732 + }, + { + "epoch": 11.598817567567568, + "grad_norm": 0.33274954557418823, + "learning_rate": 2.8757000000000003e-05, + "loss": 0.0294, + "step": 13733 + }, + { + "epoch": 11.599662162162161, + "grad_norm": 0.27636557817459106, + "learning_rate": 2.875666666666667e-05, + "loss": 0.0235, + "step": 13734 + }, + { + "epoch": 11.600506756756756, + "grad_norm": 0.30873653292655945, + "learning_rate": 2.8756333333333335e-05, + "loss": 0.0303, + "step": 13735 + }, + { + "epoch": 11.60135135135135, + "grad_norm": 0.363850861787796, + "learning_rate": 2.8756e-05, + "loss": 0.0464, + "step": 13736 + }, + { + "epoch": 11.602195945945946, + "grad_norm": 0.45625072717666626, + "learning_rate": 2.875566666666667e-05, + "loss": 0.0181, + "step": 13737 + }, + { + "epoch": 11.60304054054054, + "grad_norm": 0.3806172013282776, + "learning_rate": 2.8755333333333332e-05, + "loss": 0.0214, + "step": 13738 + }, + { + "epoch": 11.603885135135135, + "grad_norm": 0.3951556384563446, + "learning_rate": 2.8755e-05, + "loss": 0.0455, + "step": 13739 + }, + { + "epoch": 11.60472972972973, + "grad_norm": 0.3470630943775177, + "learning_rate": 2.8754666666666664e-05, + "loss": 0.0162, + "step": 13740 + }, + { + "epoch": 11.605574324324325, + "grad_norm": 0.40702927112579346, + "learning_rate": 2.8754333333333334e-05, + "loss": 0.0509, + "step": 13741 + }, + { + "epoch": 11.60641891891892, + "grad_norm": 0.5094730854034424, + "learning_rate": 2.8754000000000003e-05, + "loss": 0.0259, + "step": 13742 + }, + { + "epoch": 11.607263513513514, + "grad_norm": 0.28448691964149475, + "learning_rate": 2.8753666666666665e-05, + "loss": 0.0247, + "step": 13743 + }, + { + "epoch": 11.608108108108109, + "grad_norm": 0.3955211937427521, + "learning_rate": 2.8753333333333335e-05, + "loss": 0.0344, + "step": 13744 + }, + { + "epoch": 11.608952702702704, + "grad_norm": 0.2133610099554062, + "learning_rate": 2.8753e-05, + "loss": 0.0174, + "step": 13745 + }, + { + "epoch": 11.609797297297296, + "grad_norm": 0.2726026177406311, + "learning_rate": 2.8752666666666667e-05, + "loss": 0.0184, + "step": 13746 + }, + { + "epoch": 11.610641891891891, + "grad_norm": 0.34016990661621094, + "learning_rate": 2.8752333333333333e-05, + "loss": 0.0193, + "step": 13747 + }, + { + "epoch": 11.611486486486486, + "grad_norm": 0.44431793689727783, + "learning_rate": 2.8752000000000002e-05, + "loss": 0.0307, + "step": 13748 + }, + { + "epoch": 11.61233108108108, + "grad_norm": 0.3447590470314026, + "learning_rate": 2.8751666666666664e-05, + "loss": 0.0235, + "step": 13749 + }, + { + "epoch": 11.613175675675675, + "grad_norm": 0.3530270457267761, + "learning_rate": 2.8751333333333334e-05, + "loss": 0.0279, + "step": 13750 + }, + { + "epoch": 11.61402027027027, + "grad_norm": 0.37323030829429626, + "learning_rate": 2.8751000000000003e-05, + "loss": 0.0231, + "step": 13751 + }, + { + "epoch": 11.614864864864865, + "grad_norm": 0.23229266703128815, + "learning_rate": 2.8750666666666666e-05, + "loss": 0.015, + "step": 13752 + }, + { + "epoch": 11.61570945945946, + "grad_norm": 0.30373093485832214, + "learning_rate": 2.8750333333333335e-05, + "loss": 0.0194, + "step": 13753 + }, + { + "epoch": 11.616554054054054, + "grad_norm": 1.142599105834961, + "learning_rate": 2.875e-05, + "loss": 0.0252, + "step": 13754 + }, + { + "epoch": 11.61739864864865, + "grad_norm": 0.37066709995269775, + "learning_rate": 2.8749666666666667e-05, + "loss": 0.0299, + "step": 13755 + }, + { + "epoch": 11.618243243243244, + "grad_norm": 0.20960764586925507, + "learning_rate": 2.8749333333333333e-05, + "loss": 0.0187, + "step": 13756 + }, + { + "epoch": 11.619087837837839, + "grad_norm": 0.26778388023376465, + "learning_rate": 2.8749000000000002e-05, + "loss": 0.0113, + "step": 13757 + }, + { + "epoch": 11.619932432432432, + "grad_norm": 0.4618911147117615, + "learning_rate": 2.8748666666666668e-05, + "loss": 0.0251, + "step": 13758 + }, + { + "epoch": 11.620777027027026, + "grad_norm": 0.30566203594207764, + "learning_rate": 2.8748333333333334e-05, + "loss": 0.019, + "step": 13759 + }, + { + "epoch": 11.621621621621621, + "grad_norm": 0.47457945346832275, + "learning_rate": 2.8748000000000003e-05, + "loss": 0.0358, + "step": 13760 + }, + { + "epoch": 11.622466216216216, + "grad_norm": 0.31017735600471497, + "learning_rate": 2.8747666666666666e-05, + "loss": 0.0221, + "step": 13761 + }, + { + "epoch": 11.62331081081081, + "grad_norm": 0.32433855533599854, + "learning_rate": 2.8747333333333335e-05, + "loss": 0.0115, + "step": 13762 + }, + { + "epoch": 11.624155405405405, + "grad_norm": 0.5599989295005798, + "learning_rate": 2.8747e-05, + "loss": 0.0207, + "step": 13763 + }, + { + "epoch": 11.625, + "grad_norm": 0.5071843266487122, + "learning_rate": 2.8746666666666667e-05, + "loss": 0.0298, + "step": 13764 + }, + { + "epoch": 11.625844594594595, + "grad_norm": 0.3551429212093353, + "learning_rate": 2.8746333333333333e-05, + "loss": 0.0216, + "step": 13765 + }, + { + "epoch": 11.62668918918919, + "grad_norm": 0.2754417955875397, + "learning_rate": 2.8746000000000002e-05, + "loss": 0.017, + "step": 13766 + }, + { + "epoch": 11.627533783783784, + "grad_norm": 0.3114033639431, + "learning_rate": 2.8745666666666668e-05, + "loss": 0.0244, + "step": 13767 + }, + { + "epoch": 11.628378378378379, + "grad_norm": 0.22930631041526794, + "learning_rate": 2.8745333333333334e-05, + "loss": 0.0135, + "step": 13768 + }, + { + "epoch": 11.629222972972974, + "grad_norm": 0.6307181119918823, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0609, + "step": 13769 + }, + { + "epoch": 11.630067567567568, + "grad_norm": 0.5595002770423889, + "learning_rate": 2.8744666666666666e-05, + "loss": 0.0347, + "step": 13770 + }, + { + "epoch": 11.630912162162161, + "grad_norm": 0.2946513593196869, + "learning_rate": 2.8744333333333335e-05, + "loss": 0.0232, + "step": 13771 + }, + { + "epoch": 11.631756756756756, + "grad_norm": 0.5373883247375488, + "learning_rate": 2.8743999999999998e-05, + "loss": 0.0563, + "step": 13772 + }, + { + "epoch": 11.63260135135135, + "grad_norm": 0.8739615678787231, + "learning_rate": 2.8743666666666667e-05, + "loss": 0.0297, + "step": 13773 + }, + { + "epoch": 11.633445945945946, + "grad_norm": 0.6214396357536316, + "learning_rate": 2.8743333333333333e-05, + "loss": 0.0443, + "step": 13774 + }, + { + "epoch": 11.63429054054054, + "grad_norm": 0.5974547863006592, + "learning_rate": 2.8743e-05, + "loss": 0.2003, + "step": 13775 + }, + { + "epoch": 11.635135135135135, + "grad_norm": 0.7613686323165894, + "learning_rate": 2.874266666666667e-05, + "loss": 0.1586, + "step": 13776 + }, + { + "epoch": 11.63597972972973, + "grad_norm": 0.33249449729919434, + "learning_rate": 2.8742333333333334e-05, + "loss": 0.0613, + "step": 13777 + }, + { + "epoch": 11.636824324324325, + "grad_norm": 0.31242185831069946, + "learning_rate": 2.8742e-05, + "loss": 0.0507, + "step": 13778 + }, + { + "epoch": 11.63766891891892, + "grad_norm": 0.3738376498222351, + "learning_rate": 2.8741666666666666e-05, + "loss": 0.0388, + "step": 13779 + }, + { + "epoch": 11.638513513513514, + "grad_norm": 0.25303950905799866, + "learning_rate": 2.8741333333333336e-05, + "loss": 0.0313, + "step": 13780 + }, + { + "epoch": 11.639358108108109, + "grad_norm": 0.35218873620033264, + "learning_rate": 2.8740999999999998e-05, + "loss": 0.0362, + "step": 13781 + }, + { + "epoch": 11.640202702702704, + "grad_norm": 0.24106787145137787, + "learning_rate": 2.8740666666666667e-05, + "loss": 0.0256, + "step": 13782 + }, + { + "epoch": 11.641047297297296, + "grad_norm": 0.24456173181533813, + "learning_rate": 2.8740333333333337e-05, + "loss": 0.0325, + "step": 13783 + }, + { + "epoch": 11.641891891891891, + "grad_norm": 0.30514636635780334, + "learning_rate": 2.874e-05, + "loss": 0.0263, + "step": 13784 + }, + { + "epoch": 11.642736486486486, + "grad_norm": 0.2343491166830063, + "learning_rate": 2.873966666666667e-05, + "loss": 0.0179, + "step": 13785 + }, + { + "epoch": 11.64358108108108, + "grad_norm": 0.26094427704811096, + "learning_rate": 2.8739333333333335e-05, + "loss": 0.0224, + "step": 13786 + }, + { + "epoch": 11.644425675675675, + "grad_norm": 0.4748925268650055, + "learning_rate": 2.8739e-05, + "loss": 0.0321, + "step": 13787 + }, + { + "epoch": 11.64527027027027, + "grad_norm": 0.25398772954940796, + "learning_rate": 2.8738666666666666e-05, + "loss": 0.0325, + "step": 13788 + }, + { + "epoch": 11.646114864864865, + "grad_norm": 0.3233223557472229, + "learning_rate": 2.8738333333333336e-05, + "loss": 0.0232, + "step": 13789 + }, + { + "epoch": 11.64695945945946, + "grad_norm": 0.35592353343963623, + "learning_rate": 2.8737999999999998e-05, + "loss": 0.0228, + "step": 13790 + }, + { + "epoch": 11.647804054054054, + "grad_norm": 0.2147139459848404, + "learning_rate": 2.8737666666666668e-05, + "loss": 0.0206, + "step": 13791 + }, + { + "epoch": 11.64864864864865, + "grad_norm": 0.3655080199241638, + "learning_rate": 2.8737333333333337e-05, + "loss": 0.0249, + "step": 13792 + }, + { + "epoch": 11.649493243243244, + "grad_norm": 0.3408186733722687, + "learning_rate": 2.8737e-05, + "loss": 0.0298, + "step": 13793 + }, + { + "epoch": 11.650337837837839, + "grad_norm": 0.6810666918754578, + "learning_rate": 2.873666666666667e-05, + "loss": 0.064, + "step": 13794 + }, + { + "epoch": 11.651182432432432, + "grad_norm": 0.3282298445701599, + "learning_rate": 2.8736333333333335e-05, + "loss": 0.0278, + "step": 13795 + }, + { + "epoch": 11.652027027027026, + "grad_norm": 0.2513687312602997, + "learning_rate": 2.8736e-05, + "loss": 0.0191, + "step": 13796 + }, + { + "epoch": 11.652871621621621, + "grad_norm": 0.3289521336555481, + "learning_rate": 2.8735666666666667e-05, + "loss": 0.0201, + "step": 13797 + }, + { + "epoch": 11.653716216216216, + "grad_norm": 0.4041377604007721, + "learning_rate": 2.8735333333333336e-05, + "loss": 0.0244, + "step": 13798 + }, + { + "epoch": 11.65456081081081, + "grad_norm": 0.4181070625782013, + "learning_rate": 2.8735e-05, + "loss": 0.054, + "step": 13799 + }, + { + "epoch": 11.655405405405405, + "grad_norm": 0.38195639848709106, + "learning_rate": 2.8734666666666668e-05, + "loss": 0.0375, + "step": 13800 + }, + { + "epoch": 11.65625, + "grad_norm": 0.4097534716129303, + "learning_rate": 2.8734333333333334e-05, + "loss": 0.0441, + "step": 13801 + }, + { + "epoch": 11.657094594594595, + "grad_norm": 0.24792355298995972, + "learning_rate": 2.8734e-05, + "loss": 0.02, + "step": 13802 + }, + { + "epoch": 11.65793918918919, + "grad_norm": 0.46182000637054443, + "learning_rate": 2.873366666666667e-05, + "loss": 0.0309, + "step": 13803 + }, + { + "epoch": 11.658783783783784, + "grad_norm": 0.2848120927810669, + "learning_rate": 2.873333333333333e-05, + "loss": 0.0268, + "step": 13804 + }, + { + "epoch": 11.659628378378379, + "grad_norm": 0.3739336133003235, + "learning_rate": 2.8733e-05, + "loss": 0.0222, + "step": 13805 + }, + { + "epoch": 11.660472972972974, + "grad_norm": 0.315173476934433, + "learning_rate": 2.8732666666666667e-05, + "loss": 0.0294, + "step": 13806 + }, + { + "epoch": 11.661317567567568, + "grad_norm": 0.2924747169017792, + "learning_rate": 2.8732333333333333e-05, + "loss": 0.0192, + "step": 13807 + }, + { + "epoch": 11.662162162162161, + "grad_norm": 0.4090851843357086, + "learning_rate": 2.8732000000000002e-05, + "loss": 0.0226, + "step": 13808 + }, + { + "epoch": 11.663006756756756, + "grad_norm": 0.4473981261253357, + "learning_rate": 2.8731666666666668e-05, + "loss": 0.0146, + "step": 13809 + }, + { + "epoch": 11.66385135135135, + "grad_norm": 0.45193150639533997, + "learning_rate": 2.8731333333333334e-05, + "loss": 0.0258, + "step": 13810 + }, + { + "epoch": 11.664695945945946, + "grad_norm": 0.2639828622341156, + "learning_rate": 2.8731e-05, + "loss": 0.0188, + "step": 13811 + }, + { + "epoch": 11.66554054054054, + "grad_norm": 0.9272861480712891, + "learning_rate": 2.873066666666667e-05, + "loss": 0.0204, + "step": 13812 + }, + { + "epoch": 11.666385135135135, + "grad_norm": 0.3022681474685669, + "learning_rate": 2.8730333333333332e-05, + "loss": 0.0259, + "step": 13813 + }, + { + "epoch": 11.66722972972973, + "grad_norm": 0.25463685393333435, + "learning_rate": 2.873e-05, + "loss": 0.0161, + "step": 13814 + }, + { + "epoch": 11.668074324324325, + "grad_norm": 0.3919992744922638, + "learning_rate": 2.8729666666666667e-05, + "loss": 0.0218, + "step": 13815 + }, + { + "epoch": 11.66891891891892, + "grad_norm": 0.2849421203136444, + "learning_rate": 2.8729333333333333e-05, + "loss": 0.0217, + "step": 13816 + }, + { + "epoch": 11.669763513513514, + "grad_norm": 0.3851924538612366, + "learning_rate": 2.8729000000000002e-05, + "loss": 0.02, + "step": 13817 + }, + { + "epoch": 11.670608108108109, + "grad_norm": 0.4899166226387024, + "learning_rate": 2.8728666666666668e-05, + "loss": 0.0266, + "step": 13818 + }, + { + "epoch": 11.671452702702704, + "grad_norm": 0.9011946320533752, + "learning_rate": 2.8728333333333334e-05, + "loss": 0.0274, + "step": 13819 + }, + { + "epoch": 11.672297297297296, + "grad_norm": 0.3076525330543518, + "learning_rate": 2.8728e-05, + "loss": 0.0273, + "step": 13820 + }, + { + "epoch": 11.673141891891891, + "grad_norm": 1.0515977144241333, + "learning_rate": 2.872766666666667e-05, + "loss": 0.0321, + "step": 13821 + }, + { + "epoch": 11.673986486486486, + "grad_norm": 0.479239284992218, + "learning_rate": 2.8727333333333332e-05, + "loss": 0.0311, + "step": 13822 + }, + { + "epoch": 11.67483108108108, + "grad_norm": 1.033499836921692, + "learning_rate": 2.8727e-05, + "loss": 0.0537, + "step": 13823 + }, + { + "epoch": 11.675675675675675, + "grad_norm": 0.7260008454322815, + "learning_rate": 2.8726666666666667e-05, + "loss": 0.0433, + "step": 13824 + }, + { + "epoch": 11.67652027027027, + "grad_norm": 0.6207340955734253, + "learning_rate": 2.8726333333333333e-05, + "loss": 0.2153, + "step": 13825 + }, + { + "epoch": 11.677364864864865, + "grad_norm": 0.47997763752937317, + "learning_rate": 2.8726000000000002e-05, + "loss": 0.1326, + "step": 13826 + }, + { + "epoch": 11.67820945945946, + "grad_norm": 0.5075411796569824, + "learning_rate": 2.872566666666667e-05, + "loss": 0.0866, + "step": 13827 + }, + { + "epoch": 11.679054054054054, + "grad_norm": 0.2903025448322296, + "learning_rate": 2.8725333333333334e-05, + "loss": 0.042, + "step": 13828 + }, + { + "epoch": 11.67989864864865, + "grad_norm": 1.3329360485076904, + "learning_rate": 2.8725e-05, + "loss": 0.0778, + "step": 13829 + }, + { + "epoch": 11.680743243243244, + "grad_norm": 0.33494260907173157, + "learning_rate": 2.8724666666666666e-05, + "loss": 0.0408, + "step": 13830 + }, + { + "epoch": 11.681587837837839, + "grad_norm": 0.2595856487751007, + "learning_rate": 2.8724333333333332e-05, + "loss": 0.0187, + "step": 13831 + }, + { + "epoch": 11.682432432432432, + "grad_norm": 0.409651517868042, + "learning_rate": 2.8724e-05, + "loss": 0.0289, + "step": 13832 + }, + { + "epoch": 11.683277027027026, + "grad_norm": 0.3884858787059784, + "learning_rate": 2.8723666666666667e-05, + "loss": 0.0352, + "step": 13833 + }, + { + "epoch": 11.684121621621621, + "grad_norm": 0.6137455105781555, + "learning_rate": 2.8723333333333333e-05, + "loss": 0.0322, + "step": 13834 + }, + { + "epoch": 11.684966216216216, + "grad_norm": 0.2902837097644806, + "learning_rate": 2.8723000000000003e-05, + "loss": 0.0283, + "step": 13835 + }, + { + "epoch": 11.68581081081081, + "grad_norm": 0.4010300040245056, + "learning_rate": 2.8722666666666665e-05, + "loss": 0.0181, + "step": 13836 + }, + { + "epoch": 11.686655405405405, + "grad_norm": 0.24743519723415375, + "learning_rate": 2.8722333333333335e-05, + "loss": 0.0227, + "step": 13837 + }, + { + "epoch": 11.6875, + "grad_norm": 0.3974035382270813, + "learning_rate": 2.8722e-05, + "loss": 0.0419, + "step": 13838 + }, + { + "epoch": 11.688344594594595, + "grad_norm": 0.18103784322738647, + "learning_rate": 2.8721666666666666e-05, + "loss": 0.0115, + "step": 13839 + }, + { + "epoch": 11.68918918918919, + "grad_norm": 0.40607884526252747, + "learning_rate": 2.8721333333333332e-05, + "loss": 0.0221, + "step": 13840 + }, + { + "epoch": 11.690033783783784, + "grad_norm": 0.2939049303531647, + "learning_rate": 2.8721e-05, + "loss": 0.0293, + "step": 13841 + }, + { + "epoch": 11.690878378378379, + "grad_norm": 0.24795493483543396, + "learning_rate": 2.8720666666666668e-05, + "loss": 0.0204, + "step": 13842 + }, + { + "epoch": 11.691722972972974, + "grad_norm": 0.2889518439769745, + "learning_rate": 2.8720333333333334e-05, + "loss": 0.0217, + "step": 13843 + }, + { + "epoch": 11.692567567567568, + "grad_norm": 0.2529059648513794, + "learning_rate": 2.8720000000000003e-05, + "loss": 0.0192, + "step": 13844 + }, + { + "epoch": 11.693412162162161, + "grad_norm": 0.4727870225906372, + "learning_rate": 2.8719666666666665e-05, + "loss": 0.0245, + "step": 13845 + }, + { + "epoch": 11.694256756756756, + "grad_norm": 0.30395516753196716, + "learning_rate": 2.8719333333333335e-05, + "loss": 0.0183, + "step": 13846 + }, + { + "epoch": 11.69510135135135, + "grad_norm": 0.353228360414505, + "learning_rate": 2.8719e-05, + "loss": 0.036, + "step": 13847 + }, + { + "epoch": 11.695945945945946, + "grad_norm": 0.2947676181793213, + "learning_rate": 2.8718666666666667e-05, + "loss": 0.0248, + "step": 13848 + }, + { + "epoch": 11.69679054054054, + "grad_norm": 0.2639087736606598, + "learning_rate": 2.8718333333333333e-05, + "loss": 0.02, + "step": 13849 + }, + { + "epoch": 11.697635135135135, + "grad_norm": 0.6180364489555359, + "learning_rate": 2.8718000000000002e-05, + "loss": 0.0411, + "step": 13850 + }, + { + "epoch": 11.69847972972973, + "grad_norm": 0.22254550457000732, + "learning_rate": 2.8717666666666668e-05, + "loss": 0.0214, + "step": 13851 + }, + { + "epoch": 11.699324324324325, + "grad_norm": 0.19113792479038239, + "learning_rate": 2.8717333333333334e-05, + "loss": 0.0223, + "step": 13852 + }, + { + "epoch": 11.70016891891892, + "grad_norm": 0.62351393699646, + "learning_rate": 2.8717000000000003e-05, + "loss": 0.0179, + "step": 13853 + }, + { + "epoch": 11.701013513513514, + "grad_norm": 0.3039683401584625, + "learning_rate": 2.8716666666666666e-05, + "loss": 0.0215, + "step": 13854 + }, + { + "epoch": 11.701858108108109, + "grad_norm": 0.28689834475517273, + "learning_rate": 2.8716333333333335e-05, + "loss": 0.0187, + "step": 13855 + }, + { + "epoch": 11.702702702702704, + "grad_norm": 0.26977407932281494, + "learning_rate": 2.8716e-05, + "loss": 0.0191, + "step": 13856 + }, + { + "epoch": 11.703547297297296, + "grad_norm": 0.23441599309444427, + "learning_rate": 2.8715666666666667e-05, + "loss": 0.0191, + "step": 13857 + }, + { + "epoch": 11.704391891891891, + "grad_norm": 0.2510550916194916, + "learning_rate": 2.8715333333333336e-05, + "loss": 0.0625, + "step": 13858 + }, + { + "epoch": 11.705236486486486, + "grad_norm": 0.38026052713394165, + "learning_rate": 2.8715000000000002e-05, + "loss": 0.0335, + "step": 13859 + }, + { + "epoch": 11.70608108108108, + "grad_norm": 0.21693569421768188, + "learning_rate": 2.8714666666666668e-05, + "loss": 0.0153, + "step": 13860 + }, + { + "epoch": 11.706925675675675, + "grad_norm": 0.31847113370895386, + "learning_rate": 2.8714333333333334e-05, + "loss": 0.0268, + "step": 13861 + }, + { + "epoch": 11.70777027027027, + "grad_norm": 0.302998423576355, + "learning_rate": 2.8714e-05, + "loss": 0.0174, + "step": 13862 + }, + { + "epoch": 11.708614864864865, + "grad_norm": 0.5578328967094421, + "learning_rate": 2.8713666666666666e-05, + "loss": 0.0224, + "step": 13863 + }, + { + "epoch": 11.70945945945946, + "grad_norm": 0.2432406097650528, + "learning_rate": 2.8713333333333335e-05, + "loss": 0.0137, + "step": 13864 + }, + { + "epoch": 11.710304054054054, + "grad_norm": 0.31624382734298706, + "learning_rate": 2.8712999999999998e-05, + "loss": 0.0231, + "step": 13865 + }, + { + "epoch": 11.71114864864865, + "grad_norm": 0.303323894739151, + "learning_rate": 2.8712666666666667e-05, + "loss": 0.0167, + "step": 13866 + }, + { + "epoch": 11.711993243243244, + "grad_norm": 0.42283686995506287, + "learning_rate": 2.8712333333333336e-05, + "loss": 0.0308, + "step": 13867 + }, + { + "epoch": 11.712837837837839, + "grad_norm": 0.2789173126220703, + "learning_rate": 2.8712e-05, + "loss": 0.0158, + "step": 13868 + }, + { + "epoch": 11.713682432432432, + "grad_norm": 0.4326708912849426, + "learning_rate": 2.8711666666666668e-05, + "loss": 0.0215, + "step": 13869 + }, + { + "epoch": 11.714527027027026, + "grad_norm": 0.46062374114990234, + "learning_rate": 2.8711333333333334e-05, + "loss": 0.025, + "step": 13870 + }, + { + "epoch": 11.715371621621621, + "grad_norm": 0.26092469692230225, + "learning_rate": 2.8711e-05, + "loss": 0.0171, + "step": 13871 + }, + { + "epoch": 11.716216216216216, + "grad_norm": 0.48943883180618286, + "learning_rate": 2.8710666666666666e-05, + "loss": 0.0258, + "step": 13872 + }, + { + "epoch": 11.71706081081081, + "grad_norm": 0.526214599609375, + "learning_rate": 2.8710333333333335e-05, + "loss": 0.0285, + "step": 13873 + }, + { + "epoch": 11.717905405405405, + "grad_norm": 1.0155187845230103, + "learning_rate": 2.871e-05, + "loss": 0.0328, + "step": 13874 + }, + { + "epoch": 11.71875, + "grad_norm": 0.8932115435600281, + "learning_rate": 2.8709666666666667e-05, + "loss": 0.1931, + "step": 13875 + }, + { + "epoch": 11.719594594594595, + "grad_norm": 0.7996664047241211, + "learning_rate": 2.8709333333333337e-05, + "loss": 0.2503, + "step": 13876 + }, + { + "epoch": 11.72043918918919, + "grad_norm": 0.2994701862335205, + "learning_rate": 2.8709e-05, + "loss": 0.0512, + "step": 13877 + }, + { + "epoch": 11.721283783783784, + "grad_norm": 0.34299030900001526, + "learning_rate": 2.870866666666667e-05, + "loss": 0.0315, + "step": 13878 + }, + { + "epoch": 11.722128378378379, + "grad_norm": 0.32444238662719727, + "learning_rate": 2.8708333333333334e-05, + "loss": 0.0249, + "step": 13879 + }, + { + "epoch": 11.722972972972974, + "grad_norm": 0.3590782582759857, + "learning_rate": 2.8708e-05, + "loss": 0.0368, + "step": 13880 + }, + { + "epoch": 11.723817567567568, + "grad_norm": 0.3692159354686737, + "learning_rate": 2.8707666666666666e-05, + "loss": 0.0328, + "step": 13881 + }, + { + "epoch": 11.724662162162161, + "grad_norm": 0.3107603192329407, + "learning_rate": 2.8707333333333336e-05, + "loss": 0.03, + "step": 13882 + }, + { + "epoch": 11.725506756756756, + "grad_norm": 0.3641768991947174, + "learning_rate": 2.8707e-05, + "loss": 0.0297, + "step": 13883 + }, + { + "epoch": 11.72635135135135, + "grad_norm": 0.3329491913318634, + "learning_rate": 2.8706666666666667e-05, + "loss": 0.0159, + "step": 13884 + }, + { + "epoch": 11.727195945945946, + "grad_norm": 0.3052823543548584, + "learning_rate": 2.8706333333333337e-05, + "loss": 0.0254, + "step": 13885 + }, + { + "epoch": 11.72804054054054, + "grad_norm": 0.27319884300231934, + "learning_rate": 2.8706e-05, + "loss": 0.0266, + "step": 13886 + }, + { + "epoch": 11.728885135135135, + "grad_norm": 0.31062763929367065, + "learning_rate": 2.870566666666667e-05, + "loss": 0.0165, + "step": 13887 + }, + { + "epoch": 11.72972972972973, + "grad_norm": 0.6651361584663391, + "learning_rate": 2.8705333333333335e-05, + "loss": 0.0273, + "step": 13888 + }, + { + "epoch": 11.730574324324325, + "grad_norm": 0.15032505989074707, + "learning_rate": 2.8705e-05, + "loss": 0.015, + "step": 13889 + }, + { + "epoch": 11.73141891891892, + "grad_norm": 0.1723906248807907, + "learning_rate": 2.8704666666666666e-05, + "loss": 0.0195, + "step": 13890 + }, + { + "epoch": 11.732263513513514, + "grad_norm": 0.24067267775535583, + "learning_rate": 2.8704333333333332e-05, + "loss": 0.0298, + "step": 13891 + }, + { + "epoch": 11.733108108108109, + "grad_norm": 0.20486165583133698, + "learning_rate": 2.8704e-05, + "loss": 0.0127, + "step": 13892 + }, + { + "epoch": 11.733952702702704, + "grad_norm": 0.8764453530311584, + "learning_rate": 2.8703666666666668e-05, + "loss": 0.0257, + "step": 13893 + }, + { + "epoch": 11.734797297297296, + "grad_norm": 0.2674832344055176, + "learning_rate": 2.8703333333333334e-05, + "loss": 0.0183, + "step": 13894 + }, + { + "epoch": 11.735641891891891, + "grad_norm": 0.36233407258987427, + "learning_rate": 2.8703e-05, + "loss": 0.0255, + "step": 13895 + }, + { + "epoch": 11.736486486486486, + "grad_norm": 0.1936245858669281, + "learning_rate": 2.870266666666667e-05, + "loss": 0.025, + "step": 13896 + }, + { + "epoch": 11.73733108108108, + "grad_norm": 0.4468633234500885, + "learning_rate": 2.870233333333333e-05, + "loss": 0.0306, + "step": 13897 + }, + { + "epoch": 11.738175675675675, + "grad_norm": 0.4150993525981903, + "learning_rate": 2.8702e-05, + "loss": 0.0336, + "step": 13898 + }, + { + "epoch": 11.73902027027027, + "grad_norm": 0.4284968078136444, + "learning_rate": 2.8701666666666667e-05, + "loss": 0.0228, + "step": 13899 + }, + { + "epoch": 11.739864864864865, + "grad_norm": 0.3617030084133148, + "learning_rate": 2.8701333333333333e-05, + "loss": 0.0195, + "step": 13900 + }, + { + "epoch": 11.74070945945946, + "grad_norm": 0.3857932388782501, + "learning_rate": 2.8701000000000002e-05, + "loss": 0.0292, + "step": 13901 + }, + { + "epoch": 11.741554054054054, + "grad_norm": 0.29500511288642883, + "learning_rate": 2.8700666666666668e-05, + "loss": 0.0139, + "step": 13902 + }, + { + "epoch": 11.74239864864865, + "grad_norm": 0.2380410134792328, + "learning_rate": 2.8700333333333334e-05, + "loss": 0.021, + "step": 13903 + }, + { + "epoch": 11.743243243243244, + "grad_norm": 0.39507728815078735, + "learning_rate": 2.87e-05, + "loss": 0.0268, + "step": 13904 + }, + { + "epoch": 11.744087837837839, + "grad_norm": 0.22620128095149994, + "learning_rate": 2.869966666666667e-05, + "loss": 0.0155, + "step": 13905 + }, + { + "epoch": 11.744932432432432, + "grad_norm": 0.3248273432254791, + "learning_rate": 2.869933333333333e-05, + "loss": 0.0179, + "step": 13906 + }, + { + "epoch": 11.745777027027026, + "grad_norm": 0.17151443660259247, + "learning_rate": 2.8699e-05, + "loss": 0.0105, + "step": 13907 + }, + { + "epoch": 11.746621621621621, + "grad_norm": 0.3293960988521576, + "learning_rate": 2.869866666666667e-05, + "loss": 0.0372, + "step": 13908 + }, + { + "epoch": 11.747466216216216, + "grad_norm": 0.2474668025970459, + "learning_rate": 2.8698333333333333e-05, + "loss": 0.0157, + "step": 13909 + }, + { + "epoch": 11.74831081081081, + "grad_norm": 0.3694739043712616, + "learning_rate": 2.8698000000000002e-05, + "loss": 0.0274, + "step": 13910 + }, + { + "epoch": 11.749155405405405, + "grad_norm": 0.3552241027355194, + "learning_rate": 2.8697666666666668e-05, + "loss": 0.0259, + "step": 13911 + }, + { + "epoch": 11.75, + "grad_norm": 0.4258910119533539, + "learning_rate": 2.8697333333333334e-05, + "loss": 0.0295, + "step": 13912 + }, + { + "epoch": 11.750844594594595, + "grad_norm": 0.44545647501945496, + "learning_rate": 2.8697e-05, + "loss": 0.0348, + "step": 13913 + }, + { + "epoch": 11.75168918918919, + "grad_norm": 0.37278348207473755, + "learning_rate": 2.869666666666667e-05, + "loss": 0.0288, + "step": 13914 + }, + { + "epoch": 11.752533783783784, + "grad_norm": 0.4782062768936157, + "learning_rate": 2.8696333333333332e-05, + "loss": 0.0261, + "step": 13915 + }, + { + "epoch": 11.753378378378379, + "grad_norm": 0.39590513706207275, + "learning_rate": 2.8696e-05, + "loss": 0.0323, + "step": 13916 + }, + { + "epoch": 11.754222972972974, + "grad_norm": 0.2871064841747284, + "learning_rate": 2.869566666666667e-05, + "loss": 0.0268, + "step": 13917 + }, + { + "epoch": 11.755067567567568, + "grad_norm": 0.3381471633911133, + "learning_rate": 2.8695333333333333e-05, + "loss": 0.0266, + "step": 13918 + }, + { + "epoch": 11.755912162162161, + "grad_norm": 0.3236404359340668, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0242, + "step": 13919 + }, + { + "epoch": 11.756756756756756, + "grad_norm": 0.32967400550842285, + "learning_rate": 2.8694666666666668e-05, + "loss": 0.0333, + "step": 13920 + }, + { + "epoch": 11.75760135135135, + "grad_norm": 0.6134118437767029, + "learning_rate": 2.8694333333333334e-05, + "loss": 0.0241, + "step": 13921 + }, + { + "epoch": 11.758445945945946, + "grad_norm": 0.5577414035797119, + "learning_rate": 2.8694e-05, + "loss": 0.0347, + "step": 13922 + }, + { + "epoch": 11.75929054054054, + "grad_norm": 0.5891189575195312, + "learning_rate": 2.8693666666666666e-05, + "loss": 0.0454, + "step": 13923 + }, + { + "epoch": 11.760135135135135, + "grad_norm": 0.6829861998558044, + "learning_rate": 2.8693333333333335e-05, + "loss": 0.0488, + "step": 13924 + }, + { + "epoch": 11.76097972972973, + "grad_norm": 0.6294547319412231, + "learning_rate": 2.8693e-05, + "loss": 0.2168, + "step": 13925 + }, + { + "epoch": 11.761824324324325, + "grad_norm": 0.7766996026039124, + "learning_rate": 2.8692666666666667e-05, + "loss": 0.1509, + "step": 13926 + }, + { + "epoch": 11.76266891891892, + "grad_norm": 0.47961941361427307, + "learning_rate": 2.8692333333333333e-05, + "loss": 0.0864, + "step": 13927 + }, + { + "epoch": 11.763513513513514, + "grad_norm": 0.3663056194782257, + "learning_rate": 2.8692000000000002e-05, + "loss": 0.0509, + "step": 13928 + }, + { + "epoch": 11.764358108108109, + "grad_norm": 0.261339008808136, + "learning_rate": 2.8691666666666665e-05, + "loss": 0.0354, + "step": 13929 + }, + { + "epoch": 11.765202702702704, + "grad_norm": 0.2923562824726105, + "learning_rate": 2.8691333333333334e-05, + "loss": 0.0365, + "step": 13930 + }, + { + "epoch": 11.766047297297296, + "grad_norm": 0.2907135784626007, + "learning_rate": 2.8691e-05, + "loss": 0.0258, + "step": 13931 + }, + { + "epoch": 11.766891891891891, + "grad_norm": 0.2653803527355194, + "learning_rate": 2.8690666666666666e-05, + "loss": 0.0341, + "step": 13932 + }, + { + "epoch": 11.767736486486486, + "grad_norm": 0.4911186695098877, + "learning_rate": 2.8690333333333336e-05, + "loss": 0.0395, + "step": 13933 + }, + { + "epoch": 11.76858108108108, + "grad_norm": 0.38901352882385254, + "learning_rate": 2.869e-05, + "loss": 0.0296, + "step": 13934 + }, + { + "epoch": 11.769425675675675, + "grad_norm": 0.43531277775764465, + "learning_rate": 2.8689666666666667e-05, + "loss": 0.0229, + "step": 13935 + }, + { + "epoch": 11.77027027027027, + "grad_norm": 0.15820547938346863, + "learning_rate": 2.8689333333333333e-05, + "loss": 0.0187, + "step": 13936 + }, + { + "epoch": 11.771114864864865, + "grad_norm": 0.31771987676620483, + "learning_rate": 2.8689000000000003e-05, + "loss": 0.0262, + "step": 13937 + }, + { + "epoch": 11.77195945945946, + "grad_norm": 0.22024111449718475, + "learning_rate": 2.8688666666666665e-05, + "loss": 0.016, + "step": 13938 + }, + { + "epoch": 11.772804054054054, + "grad_norm": 0.2499646097421646, + "learning_rate": 2.8688333333333335e-05, + "loss": 0.0278, + "step": 13939 + }, + { + "epoch": 11.77364864864865, + "grad_norm": 0.2590027153491974, + "learning_rate": 2.8688e-05, + "loss": 0.008, + "step": 13940 + }, + { + "epoch": 11.774493243243244, + "grad_norm": 0.5659560561180115, + "learning_rate": 2.8687666666666666e-05, + "loss": 0.0312, + "step": 13941 + }, + { + "epoch": 11.775337837837839, + "grad_norm": 0.24380765855312347, + "learning_rate": 2.8687333333333336e-05, + "loss": 0.0255, + "step": 13942 + }, + { + "epoch": 11.776182432432432, + "grad_norm": 0.19694246351718903, + "learning_rate": 2.8687e-05, + "loss": 0.0143, + "step": 13943 + }, + { + "epoch": 11.777027027027026, + "grad_norm": 0.23336409032344818, + "learning_rate": 2.8686666666666668e-05, + "loss": 0.0204, + "step": 13944 + }, + { + "epoch": 11.777871621621621, + "grad_norm": 0.24150046706199646, + "learning_rate": 2.8686333333333334e-05, + "loss": 0.0185, + "step": 13945 + }, + { + "epoch": 11.778716216216216, + "grad_norm": 0.2859288454055786, + "learning_rate": 2.8686000000000003e-05, + "loss": 0.0287, + "step": 13946 + }, + { + "epoch": 11.77956081081081, + "grad_norm": 0.37676993012428284, + "learning_rate": 2.8685666666666665e-05, + "loss": 0.0264, + "step": 13947 + }, + { + "epoch": 11.780405405405405, + "grad_norm": 0.280521422624588, + "learning_rate": 2.8685333333333335e-05, + "loss": 0.0205, + "step": 13948 + }, + { + "epoch": 11.78125, + "grad_norm": 0.2914467453956604, + "learning_rate": 2.8685e-05, + "loss": 0.0226, + "step": 13949 + }, + { + "epoch": 11.782094594594595, + "grad_norm": 0.331711083650589, + "learning_rate": 2.8684666666666667e-05, + "loss": 0.0317, + "step": 13950 + }, + { + "epoch": 11.78293918918919, + "grad_norm": 0.2450009435415268, + "learning_rate": 2.8684333333333336e-05, + "loss": 0.0236, + "step": 13951 + }, + { + "epoch": 11.783783783783784, + "grad_norm": 0.21598586440086365, + "learning_rate": 2.8684e-05, + "loss": 0.0142, + "step": 13952 + }, + { + "epoch": 11.784628378378379, + "grad_norm": 0.21018679440021515, + "learning_rate": 2.8683666666666668e-05, + "loss": 0.0164, + "step": 13953 + }, + { + "epoch": 11.785472972972974, + "grad_norm": 0.35047629475593567, + "learning_rate": 2.8683333333333334e-05, + "loss": 0.0377, + "step": 13954 + }, + { + "epoch": 11.786317567567568, + "grad_norm": 0.36739403009414673, + "learning_rate": 2.8683e-05, + "loss": 0.019, + "step": 13955 + }, + { + "epoch": 11.787162162162161, + "grad_norm": 0.31190726161003113, + "learning_rate": 2.8682666666666666e-05, + "loss": 0.0203, + "step": 13956 + }, + { + "epoch": 11.788006756756756, + "grad_norm": 0.2802690863609314, + "learning_rate": 2.8682333333333335e-05, + "loss": 0.016, + "step": 13957 + }, + { + "epoch": 11.78885135135135, + "grad_norm": 0.36358749866485596, + "learning_rate": 2.8682e-05, + "loss": 0.0257, + "step": 13958 + }, + { + "epoch": 11.789695945945946, + "grad_norm": 0.3154979944229126, + "learning_rate": 2.8681666666666667e-05, + "loss": 0.0207, + "step": 13959 + }, + { + "epoch": 11.79054054054054, + "grad_norm": 0.28854241967201233, + "learning_rate": 2.8681333333333336e-05, + "loss": 0.0133, + "step": 13960 + }, + { + "epoch": 11.791385135135135, + "grad_norm": 0.27949458360671997, + "learning_rate": 2.8681e-05, + "loss": 0.0183, + "step": 13961 + }, + { + "epoch": 11.79222972972973, + "grad_norm": 0.17790938913822174, + "learning_rate": 2.8680666666666668e-05, + "loss": 0.0156, + "step": 13962 + }, + { + "epoch": 11.793074324324325, + "grad_norm": 0.24163131415843964, + "learning_rate": 2.8680333333333334e-05, + "loss": 0.0162, + "step": 13963 + }, + { + "epoch": 11.79391891891892, + "grad_norm": 1.1508851051330566, + "learning_rate": 2.868e-05, + "loss": 0.0448, + "step": 13964 + }, + { + "epoch": 11.794763513513514, + "grad_norm": 0.3196278214454651, + "learning_rate": 2.8679666666666666e-05, + "loss": 0.022, + "step": 13965 + }, + { + "epoch": 11.795608108108109, + "grad_norm": 0.3489202857017517, + "learning_rate": 2.8679333333333335e-05, + "loss": 0.031, + "step": 13966 + }, + { + "epoch": 11.796452702702704, + "grad_norm": 0.6122796535491943, + "learning_rate": 2.8679e-05, + "loss": 0.0249, + "step": 13967 + }, + { + "epoch": 11.797297297297296, + "grad_norm": 0.41612282395362854, + "learning_rate": 2.8678666666666667e-05, + "loss": 0.0242, + "step": 13968 + }, + { + "epoch": 11.798141891891891, + "grad_norm": 0.3433351218700409, + "learning_rate": 2.8678333333333336e-05, + "loss": 0.0245, + "step": 13969 + }, + { + "epoch": 11.798986486486486, + "grad_norm": 0.6064863204956055, + "learning_rate": 2.8678e-05, + "loss": 0.0217, + "step": 13970 + }, + { + "epoch": 11.79983108108108, + "grad_norm": 0.36473989486694336, + "learning_rate": 2.8677666666666668e-05, + "loss": 0.0246, + "step": 13971 + }, + { + "epoch": 11.800675675675675, + "grad_norm": 0.553395688533783, + "learning_rate": 2.8677333333333334e-05, + "loss": 0.0374, + "step": 13972 + }, + { + "epoch": 11.80152027027027, + "grad_norm": 1.2872211933135986, + "learning_rate": 2.8677e-05, + "loss": 0.0433, + "step": 13973 + }, + { + "epoch": 11.802364864864865, + "grad_norm": 0.5405806303024292, + "learning_rate": 2.867666666666667e-05, + "loss": 0.028, + "step": 13974 + }, + { + "epoch": 11.80320945945946, + "grad_norm": 0.655630350112915, + "learning_rate": 2.8676333333333335e-05, + "loss": 0.2104, + "step": 13975 + }, + { + "epoch": 11.804054054054054, + "grad_norm": 0.3133805990219116, + "learning_rate": 2.8676e-05, + "loss": 0.096, + "step": 13976 + }, + { + "epoch": 11.80489864864865, + "grad_norm": 0.3347418010234833, + "learning_rate": 2.8675666666666667e-05, + "loss": 0.0496, + "step": 13977 + }, + { + "epoch": 11.805743243243244, + "grad_norm": 0.27074146270751953, + "learning_rate": 2.8675333333333336e-05, + "loss": 0.0271, + "step": 13978 + }, + { + "epoch": 11.806587837837839, + "grad_norm": 0.32743072509765625, + "learning_rate": 2.8675e-05, + "loss": 0.041, + "step": 13979 + }, + { + "epoch": 11.807432432432432, + "grad_norm": 0.5288968682289124, + "learning_rate": 2.867466666666667e-05, + "loss": 0.034, + "step": 13980 + }, + { + "epoch": 11.808277027027026, + "grad_norm": 0.3094894289970398, + "learning_rate": 2.867433333333333e-05, + "loss": 0.0251, + "step": 13981 + }, + { + "epoch": 11.809121621621621, + "grad_norm": 0.20911644399166107, + "learning_rate": 2.8674e-05, + "loss": 0.0169, + "step": 13982 + }, + { + "epoch": 11.809966216216216, + "grad_norm": 0.2729310989379883, + "learning_rate": 2.867366666666667e-05, + "loss": 0.0321, + "step": 13983 + }, + { + "epoch": 11.81081081081081, + "grad_norm": 0.49859321117401123, + "learning_rate": 2.8673333333333332e-05, + "loss": 0.0215, + "step": 13984 + }, + { + "epoch": 11.811655405405405, + "grad_norm": 0.2374233901500702, + "learning_rate": 2.8673e-05, + "loss": 0.0236, + "step": 13985 + }, + { + "epoch": 11.8125, + "grad_norm": 0.3693476915359497, + "learning_rate": 2.8672666666666667e-05, + "loss": 0.0257, + "step": 13986 + }, + { + "epoch": 11.813344594594595, + "grad_norm": 0.2878570556640625, + "learning_rate": 2.8672333333333333e-05, + "loss": 0.0274, + "step": 13987 + }, + { + "epoch": 11.81418918918919, + "grad_norm": 0.23694568872451782, + "learning_rate": 2.8672e-05, + "loss": 0.021, + "step": 13988 + }, + { + "epoch": 11.815033783783784, + "grad_norm": 0.34891703724861145, + "learning_rate": 2.867166666666667e-05, + "loss": 0.0225, + "step": 13989 + }, + { + "epoch": 11.815878378378379, + "grad_norm": 0.23725968599319458, + "learning_rate": 2.867133333333333e-05, + "loss": 0.0168, + "step": 13990 + }, + { + "epoch": 11.816722972972974, + "grad_norm": 0.3906271755695343, + "learning_rate": 2.8671e-05, + "loss": 0.0246, + "step": 13991 + }, + { + "epoch": 11.817567567567568, + "grad_norm": 0.4005364179611206, + "learning_rate": 2.867066666666667e-05, + "loss": 0.0326, + "step": 13992 + }, + { + "epoch": 11.818412162162161, + "grad_norm": 0.28680214285850525, + "learning_rate": 2.8670333333333332e-05, + "loss": 0.0155, + "step": 13993 + }, + { + "epoch": 11.819256756756756, + "grad_norm": 0.2763347923755646, + "learning_rate": 2.867e-05, + "loss": 0.0299, + "step": 13994 + }, + { + "epoch": 11.82010135135135, + "grad_norm": 0.2764371335506439, + "learning_rate": 2.8669666666666668e-05, + "loss": 0.0189, + "step": 13995 + }, + { + "epoch": 11.820945945945946, + "grad_norm": 0.4921996593475342, + "learning_rate": 2.8669333333333334e-05, + "loss": 0.0195, + "step": 13996 + }, + { + "epoch": 11.82179054054054, + "grad_norm": 0.23946136236190796, + "learning_rate": 2.8669e-05, + "loss": 0.0191, + "step": 13997 + }, + { + "epoch": 11.822635135135135, + "grad_norm": 0.17326292395591736, + "learning_rate": 2.866866666666667e-05, + "loss": 0.0154, + "step": 13998 + }, + { + "epoch": 11.82347972972973, + "grad_norm": 0.3565545380115509, + "learning_rate": 2.8668333333333335e-05, + "loss": 0.0285, + "step": 13999 + }, + { + "epoch": 11.824324324324325, + "grad_norm": 0.35956403613090515, + "learning_rate": 2.8668e-05, + "loss": 0.0218, + "step": 14000 + }, + { + "epoch": 11.824324324324325, + "eval_cer": 0.09043450325275078, + "eval_loss": 0.2694067358970642, + "eval_runtime": 8.0227, + "eval_samples_per_second": 12.589, + "eval_steps_per_second": 0.499, + "eval_wer": 0.26455026455026454, + "step": 14000 + }, + { + "epoch": 11.82516891891892, + "grad_norm": 0.538373589515686, + "learning_rate": 2.866766666666667e-05, + "loss": 0.0273, + "step": 14001 + }, + { + "epoch": 11.826013513513514, + "grad_norm": 0.410717636346817, + "learning_rate": 2.8667333333333333e-05, + "loss": 0.0308, + "step": 14002 + }, + { + "epoch": 11.826858108108109, + "grad_norm": 0.3081144392490387, + "learning_rate": 2.8667000000000002e-05, + "loss": 0.0263, + "step": 14003 + }, + { + "epoch": 11.827702702702704, + "grad_norm": 0.3600448966026306, + "learning_rate": 2.8666666666666668e-05, + "loss": 0.0375, + "step": 14004 + }, + { + "epoch": 11.828547297297296, + "grad_norm": 0.31716474890708923, + "learning_rate": 2.8666333333333334e-05, + "loss": 0.0192, + "step": 14005 + }, + { + "epoch": 11.829391891891891, + "grad_norm": 0.2252165973186493, + "learning_rate": 2.8666e-05, + "loss": 0.012, + "step": 14006 + }, + { + "epoch": 11.830236486486486, + "grad_norm": 0.24529175460338593, + "learning_rate": 2.866566666666667e-05, + "loss": 0.0186, + "step": 14007 + }, + { + "epoch": 11.83108108108108, + "grad_norm": 0.446125328540802, + "learning_rate": 2.8665333333333335e-05, + "loss": 0.0288, + "step": 14008 + }, + { + "epoch": 11.831925675675675, + "grad_norm": 0.22180193662643433, + "learning_rate": 2.8665e-05, + "loss": 0.0203, + "step": 14009 + }, + { + "epoch": 11.83277027027027, + "grad_norm": 0.36729559302330017, + "learning_rate": 2.866466666666667e-05, + "loss": 0.0337, + "step": 14010 + }, + { + "epoch": 11.833614864864865, + "grad_norm": 0.40151098370552063, + "learning_rate": 2.8664333333333333e-05, + "loss": 0.0236, + "step": 14011 + }, + { + "epoch": 11.83445945945946, + "grad_norm": 0.2750368118286133, + "learning_rate": 2.8664000000000002e-05, + "loss": 0.0231, + "step": 14012 + }, + { + "epoch": 11.835304054054054, + "grad_norm": 0.3308071792125702, + "learning_rate": 2.8663666666666665e-05, + "loss": 0.0201, + "step": 14013 + }, + { + "epoch": 11.83614864864865, + "grad_norm": 0.20678779482841492, + "learning_rate": 2.8663333333333334e-05, + "loss": 0.0217, + "step": 14014 + }, + { + "epoch": 11.836993243243244, + "grad_norm": 0.4449287950992584, + "learning_rate": 2.8663e-05, + "loss": 0.0432, + "step": 14015 + }, + { + "epoch": 11.837837837837839, + "grad_norm": 0.24395962059497833, + "learning_rate": 2.8662666666666666e-05, + "loss": 0.0199, + "step": 14016 + }, + { + "epoch": 11.838682432432432, + "grad_norm": 0.38942790031433105, + "learning_rate": 2.8662333333333335e-05, + "loss": 0.0244, + "step": 14017 + }, + { + "epoch": 11.839527027027026, + "grad_norm": 0.2558569312095642, + "learning_rate": 2.8662e-05, + "loss": 0.0172, + "step": 14018 + }, + { + "epoch": 11.840371621621621, + "grad_norm": 0.30507680773735046, + "learning_rate": 2.8661666666666667e-05, + "loss": 0.0186, + "step": 14019 + }, + { + "epoch": 11.841216216216216, + "grad_norm": 0.5758376121520996, + "learning_rate": 2.8661333333333333e-05, + "loss": 0.0181, + "step": 14020 + }, + { + "epoch": 11.84206081081081, + "grad_norm": 0.5925665497779846, + "learning_rate": 2.8661000000000002e-05, + "loss": 0.034, + "step": 14021 + }, + { + "epoch": 11.842905405405405, + "grad_norm": 0.4490446150302887, + "learning_rate": 2.8660666666666665e-05, + "loss": 0.0231, + "step": 14022 + }, + { + "epoch": 11.84375, + "grad_norm": 0.9952989220619202, + "learning_rate": 2.8660333333333334e-05, + "loss": 0.0406, + "step": 14023 + }, + { + "epoch": 11.844594594594595, + "grad_norm": 0.8985205292701721, + "learning_rate": 2.8660000000000003e-05, + "loss": 0.0458, + "step": 14024 + }, + { + "epoch": 11.84543918918919, + "grad_norm": 0.7442185878753662, + "learning_rate": 2.8659666666666666e-05, + "loss": 0.2206, + "step": 14025 + }, + { + "epoch": 11.846283783783784, + "grad_norm": 0.5345494747161865, + "learning_rate": 2.8659333333333335e-05, + "loss": 0.118, + "step": 14026 + }, + { + "epoch": 11.847128378378379, + "grad_norm": 0.46838441491127014, + "learning_rate": 2.8659e-05, + "loss": 0.1185, + "step": 14027 + }, + { + "epoch": 11.847972972972974, + "grad_norm": 0.43293240666389465, + "learning_rate": 2.8658666666666667e-05, + "loss": 0.0454, + "step": 14028 + }, + { + "epoch": 11.848817567567568, + "grad_norm": 0.3221350610256195, + "learning_rate": 2.8658333333333333e-05, + "loss": 0.0287, + "step": 14029 + }, + { + "epoch": 11.849662162162161, + "grad_norm": 0.26161080598831177, + "learning_rate": 2.8658000000000002e-05, + "loss": 0.0352, + "step": 14030 + }, + { + "epoch": 11.850506756756756, + "grad_norm": 0.2231685072183609, + "learning_rate": 2.8657666666666665e-05, + "loss": 0.015, + "step": 14031 + }, + { + "epoch": 11.85135135135135, + "grad_norm": 0.6526025533676147, + "learning_rate": 2.8657333333333334e-05, + "loss": 0.0233, + "step": 14032 + }, + { + "epoch": 11.852195945945946, + "grad_norm": 0.23720481991767883, + "learning_rate": 2.8657000000000004e-05, + "loss": 0.0178, + "step": 14033 + }, + { + "epoch": 11.85304054054054, + "grad_norm": 0.23653939366340637, + "learning_rate": 2.8656666666666666e-05, + "loss": 0.0213, + "step": 14034 + }, + { + "epoch": 11.853885135135135, + "grad_norm": 0.24570240080356598, + "learning_rate": 2.8656333333333335e-05, + "loss": 0.0329, + "step": 14035 + }, + { + "epoch": 11.85472972972973, + "grad_norm": 0.2463359832763672, + "learning_rate": 2.8656e-05, + "loss": 0.0341, + "step": 14036 + }, + { + "epoch": 11.855574324324325, + "grad_norm": 0.12690337002277374, + "learning_rate": 2.8655666666666667e-05, + "loss": 0.0125, + "step": 14037 + }, + { + "epoch": 11.85641891891892, + "grad_norm": 0.8483830690383911, + "learning_rate": 2.8655333333333333e-05, + "loss": 0.0367, + "step": 14038 + }, + { + "epoch": 11.857263513513514, + "grad_norm": 0.16836023330688477, + "learning_rate": 2.8655000000000003e-05, + "loss": 0.0168, + "step": 14039 + }, + { + "epoch": 11.858108108108109, + "grad_norm": 0.25909361243247986, + "learning_rate": 2.8654666666666665e-05, + "loss": 0.024, + "step": 14040 + }, + { + "epoch": 11.858952702702704, + "grad_norm": 0.24841855466365814, + "learning_rate": 2.8654333333333334e-05, + "loss": 0.0211, + "step": 14041 + }, + { + "epoch": 11.859797297297296, + "grad_norm": 0.4346330165863037, + "learning_rate": 2.8654e-05, + "loss": 0.0243, + "step": 14042 + }, + { + "epoch": 11.860641891891891, + "grad_norm": 0.24782694876194, + "learning_rate": 2.8653666666666666e-05, + "loss": 0.0177, + "step": 14043 + }, + { + "epoch": 11.861486486486486, + "grad_norm": 0.5478566288948059, + "learning_rate": 2.8653333333333336e-05, + "loss": 0.0512, + "step": 14044 + }, + { + "epoch": 11.86233108108108, + "grad_norm": 0.4233212172985077, + "learning_rate": 2.8652999999999998e-05, + "loss": 0.0184, + "step": 14045 + }, + { + "epoch": 11.863175675675675, + "grad_norm": 0.20579861104488373, + "learning_rate": 2.8652666666666668e-05, + "loss": 0.0203, + "step": 14046 + }, + { + "epoch": 11.86402027027027, + "grad_norm": 0.34763357043266296, + "learning_rate": 2.8652333333333334e-05, + "loss": 0.0206, + "step": 14047 + }, + { + "epoch": 11.864864864864865, + "grad_norm": 0.2838532328605652, + "learning_rate": 2.8652e-05, + "loss": 0.0296, + "step": 14048 + }, + { + "epoch": 11.86570945945946, + "grad_norm": 0.48796799778938293, + "learning_rate": 2.865166666666667e-05, + "loss": 0.0231, + "step": 14049 + }, + { + "epoch": 11.866554054054054, + "grad_norm": 0.26850828528404236, + "learning_rate": 2.8651333333333335e-05, + "loss": 0.0285, + "step": 14050 + }, + { + "epoch": 11.86739864864865, + "grad_norm": 0.37979385256767273, + "learning_rate": 2.8651e-05, + "loss": 0.0202, + "step": 14051 + }, + { + "epoch": 11.868243243243244, + "grad_norm": 0.6361658573150635, + "learning_rate": 2.8650666666666667e-05, + "loss": 0.0272, + "step": 14052 + }, + { + "epoch": 11.869087837837839, + "grad_norm": 0.1877208948135376, + "learning_rate": 2.8650333333333336e-05, + "loss": 0.0147, + "step": 14053 + }, + { + "epoch": 11.869932432432432, + "grad_norm": 0.179025799036026, + "learning_rate": 2.865e-05, + "loss": 0.0183, + "step": 14054 + }, + { + "epoch": 11.870777027027026, + "grad_norm": 0.23676814138889313, + "learning_rate": 2.8649666666666668e-05, + "loss": 0.017, + "step": 14055 + }, + { + "epoch": 11.871621621621621, + "grad_norm": 0.3976914584636688, + "learning_rate": 2.8649333333333334e-05, + "loss": 0.0204, + "step": 14056 + }, + { + "epoch": 11.872466216216216, + "grad_norm": 0.3478899300098419, + "learning_rate": 2.8649e-05, + "loss": 0.0267, + "step": 14057 + }, + { + "epoch": 11.87331081081081, + "grad_norm": 0.14376525580883026, + "learning_rate": 2.864866666666667e-05, + "loss": 0.0096, + "step": 14058 + }, + { + "epoch": 11.874155405405405, + "grad_norm": 0.34111008048057556, + "learning_rate": 2.8648333333333335e-05, + "loss": 0.0151, + "step": 14059 + }, + { + "epoch": 11.875, + "grad_norm": 0.40799006819725037, + "learning_rate": 2.8648e-05, + "loss": 0.0288, + "step": 14060 + }, + { + "epoch": 11.875844594594595, + "grad_norm": 0.1898212432861328, + "learning_rate": 2.8647666666666667e-05, + "loss": 0.0117, + "step": 14061 + }, + { + "epoch": 11.87668918918919, + "grad_norm": 0.46513450145721436, + "learning_rate": 2.8647333333333336e-05, + "loss": 0.022, + "step": 14062 + }, + { + "epoch": 11.877533783783784, + "grad_norm": 0.39983296394348145, + "learning_rate": 2.8647e-05, + "loss": 0.0241, + "step": 14063 + }, + { + "epoch": 11.878378378378379, + "grad_norm": 0.2835431694984436, + "learning_rate": 2.8646666666666668e-05, + "loss": 0.0164, + "step": 14064 + }, + { + "epoch": 11.879222972972974, + "grad_norm": 0.48361286520957947, + "learning_rate": 2.8646333333333334e-05, + "loss": 0.0233, + "step": 14065 + }, + { + "epoch": 11.880067567567568, + "grad_norm": 0.43889009952545166, + "learning_rate": 2.8646e-05, + "loss": 0.0305, + "step": 14066 + }, + { + "epoch": 11.880912162162161, + "grad_norm": 0.2532615661621094, + "learning_rate": 2.864566666666667e-05, + "loss": 0.014, + "step": 14067 + }, + { + "epoch": 11.881756756756756, + "grad_norm": 0.5206319689750671, + "learning_rate": 2.8645333333333335e-05, + "loss": 0.031, + "step": 14068 + }, + { + "epoch": 11.88260135135135, + "grad_norm": 0.35985326766967773, + "learning_rate": 2.8645e-05, + "loss": 0.0271, + "step": 14069 + }, + { + "epoch": 11.883445945945946, + "grad_norm": 0.3312321603298187, + "learning_rate": 2.8644666666666667e-05, + "loss": 0.0248, + "step": 14070 + }, + { + "epoch": 11.88429054054054, + "grad_norm": 0.48149627447128296, + "learning_rate": 2.8644333333333336e-05, + "loss": 0.0189, + "step": 14071 + }, + { + "epoch": 11.885135135135135, + "grad_norm": 0.5878880023956299, + "learning_rate": 2.8644e-05, + "loss": 0.0472, + "step": 14072 + }, + { + "epoch": 11.88597972972973, + "grad_norm": 0.3842317461967468, + "learning_rate": 2.8643666666666668e-05, + "loss": 0.0378, + "step": 14073 + }, + { + "epoch": 11.886824324324325, + "grad_norm": 0.6984393000602722, + "learning_rate": 2.8643333333333334e-05, + "loss": 0.0385, + "step": 14074 + }, + { + "epoch": 11.88766891891892, + "grad_norm": 0.6724914908409119, + "learning_rate": 2.8643e-05, + "loss": 0.1901, + "step": 14075 + }, + { + "epoch": 11.888513513513514, + "grad_norm": 0.5622327327728271, + "learning_rate": 2.864266666666667e-05, + "loss": 0.1471, + "step": 14076 + }, + { + "epoch": 11.889358108108109, + "grad_norm": 0.4766501784324646, + "learning_rate": 2.8642333333333332e-05, + "loss": 0.0605, + "step": 14077 + }, + { + "epoch": 11.890202702702704, + "grad_norm": 0.39829888939857483, + "learning_rate": 2.8642e-05, + "loss": 0.055, + "step": 14078 + }, + { + "epoch": 11.891047297297296, + "grad_norm": 0.718622624874115, + "learning_rate": 2.8641666666666667e-05, + "loss": 0.0372, + "step": 14079 + }, + { + "epoch": 11.891891891891891, + "grad_norm": 0.24966531991958618, + "learning_rate": 2.8641333333333333e-05, + "loss": 0.0229, + "step": 14080 + }, + { + "epoch": 11.892736486486486, + "grad_norm": 0.1998220980167389, + "learning_rate": 2.8641e-05, + "loss": 0.0178, + "step": 14081 + }, + { + "epoch": 11.89358108108108, + "grad_norm": 0.17840756475925446, + "learning_rate": 2.864066666666667e-05, + "loss": 0.0163, + "step": 14082 + }, + { + "epoch": 11.894425675675675, + "grad_norm": 0.27907153964042664, + "learning_rate": 2.8640333333333334e-05, + "loss": 0.0221, + "step": 14083 + }, + { + "epoch": 11.89527027027027, + "grad_norm": 0.3579525649547577, + "learning_rate": 2.864e-05, + "loss": 0.0153, + "step": 14084 + }, + { + "epoch": 11.896114864864865, + "grad_norm": 0.32408642768859863, + "learning_rate": 2.863966666666667e-05, + "loss": 0.0369, + "step": 14085 + }, + { + "epoch": 11.89695945945946, + "grad_norm": 0.32146501541137695, + "learning_rate": 2.8639333333333332e-05, + "loss": 0.0423, + "step": 14086 + }, + { + "epoch": 11.897804054054054, + "grad_norm": 0.2881489396095276, + "learning_rate": 2.8639e-05, + "loss": 0.027, + "step": 14087 + }, + { + "epoch": 11.89864864864865, + "grad_norm": 0.47792261838912964, + "learning_rate": 2.8638666666666667e-05, + "loss": 0.0245, + "step": 14088 + }, + { + "epoch": 11.899493243243244, + "grad_norm": 0.3147623538970947, + "learning_rate": 2.8638333333333333e-05, + "loss": 0.0282, + "step": 14089 + }, + { + "epoch": 11.900337837837839, + "grad_norm": 0.2668652832508087, + "learning_rate": 2.8638e-05, + "loss": 0.0269, + "step": 14090 + }, + { + "epoch": 11.901182432432432, + "grad_norm": 0.3014220595359802, + "learning_rate": 2.863766666666667e-05, + "loss": 0.0259, + "step": 14091 + }, + { + "epoch": 11.902027027027026, + "grad_norm": 0.5646927356719971, + "learning_rate": 2.8637333333333334e-05, + "loss": 0.0229, + "step": 14092 + }, + { + "epoch": 11.902871621621621, + "grad_norm": 0.2044973224401474, + "learning_rate": 2.8637e-05, + "loss": 0.0167, + "step": 14093 + }, + { + "epoch": 11.903716216216216, + "grad_norm": 0.17513127624988556, + "learning_rate": 2.863666666666667e-05, + "loss": 0.0129, + "step": 14094 + }, + { + "epoch": 11.90456081081081, + "grad_norm": 0.22154600918293, + "learning_rate": 2.8636333333333332e-05, + "loss": 0.0158, + "step": 14095 + }, + { + "epoch": 11.905405405405405, + "grad_norm": 0.18605570495128632, + "learning_rate": 2.8636e-05, + "loss": 0.0153, + "step": 14096 + }, + { + "epoch": 11.90625, + "grad_norm": 0.2174096554517746, + "learning_rate": 2.8635666666666668e-05, + "loss": 0.0226, + "step": 14097 + }, + { + "epoch": 11.907094594594595, + "grad_norm": 0.15766040980815887, + "learning_rate": 2.8635333333333333e-05, + "loss": 0.0126, + "step": 14098 + }, + { + "epoch": 11.90793918918919, + "grad_norm": 0.49092188477516174, + "learning_rate": 2.8635000000000003e-05, + "loss": 0.0257, + "step": 14099 + }, + { + "epoch": 11.908783783783784, + "grad_norm": 0.20225584506988525, + "learning_rate": 2.863466666666667e-05, + "loss": 0.0139, + "step": 14100 + }, + { + "epoch": 11.909628378378379, + "grad_norm": 0.37570720911026, + "learning_rate": 2.8634333333333335e-05, + "loss": 0.0342, + "step": 14101 + }, + { + "epoch": 11.910472972972974, + "grad_norm": 0.5451659560203552, + "learning_rate": 2.8634e-05, + "loss": 0.0292, + "step": 14102 + }, + { + "epoch": 11.911317567567568, + "grad_norm": 0.6551663875579834, + "learning_rate": 2.8633666666666667e-05, + "loss": 0.0273, + "step": 14103 + }, + { + "epoch": 11.912162162162161, + "grad_norm": 0.34880968928337097, + "learning_rate": 2.8633333333333332e-05, + "loss": 0.0221, + "step": 14104 + }, + { + "epoch": 11.913006756756756, + "grad_norm": 0.41292065382003784, + "learning_rate": 2.8633000000000002e-05, + "loss": 0.0227, + "step": 14105 + }, + { + "epoch": 11.91385135135135, + "grad_norm": 0.5904343724250793, + "learning_rate": 2.8632666666666664e-05, + "loss": 0.0232, + "step": 14106 + }, + { + "epoch": 11.914695945945946, + "grad_norm": 0.3581533432006836, + "learning_rate": 2.8632333333333334e-05, + "loss": 0.0214, + "step": 14107 + }, + { + "epoch": 11.91554054054054, + "grad_norm": 0.34676554799079895, + "learning_rate": 2.8632000000000003e-05, + "loss": 0.0226, + "step": 14108 + }, + { + "epoch": 11.916385135135135, + "grad_norm": 0.3769499957561493, + "learning_rate": 2.8631666666666666e-05, + "loss": 0.0297, + "step": 14109 + }, + { + "epoch": 11.91722972972973, + "grad_norm": 0.3176209330558777, + "learning_rate": 2.8631333333333335e-05, + "loss": 0.0174, + "step": 14110 + }, + { + "epoch": 11.918074324324325, + "grad_norm": 0.5308434367179871, + "learning_rate": 2.8631e-05, + "loss": 0.0916, + "step": 14111 + }, + { + "epoch": 11.91891891891892, + "grad_norm": 0.5034207701683044, + "learning_rate": 2.8630666666666667e-05, + "loss": 0.0277, + "step": 14112 + }, + { + "epoch": 11.919763513513514, + "grad_norm": 0.7375887632369995, + "learning_rate": 2.8630333333333333e-05, + "loss": 0.0387, + "step": 14113 + }, + { + "epoch": 11.920608108108109, + "grad_norm": 0.2904530167579651, + "learning_rate": 2.8630000000000002e-05, + "loss": 0.0231, + "step": 14114 + }, + { + "epoch": 11.921452702702704, + "grad_norm": 0.629758358001709, + "learning_rate": 2.8629666666666665e-05, + "loss": 0.0328, + "step": 14115 + }, + { + "epoch": 11.922297297297296, + "grad_norm": 0.4465383291244507, + "learning_rate": 2.8629333333333334e-05, + "loss": 0.0291, + "step": 14116 + }, + { + "epoch": 11.923141891891891, + "grad_norm": 0.7726544141769409, + "learning_rate": 2.8629000000000003e-05, + "loss": 0.0377, + "step": 14117 + }, + { + "epoch": 11.923986486486486, + "grad_norm": 0.29861921072006226, + "learning_rate": 2.8628666666666666e-05, + "loss": 0.0204, + "step": 14118 + }, + { + "epoch": 11.92483108108108, + "grad_norm": 0.2848009765148163, + "learning_rate": 2.8628333333333335e-05, + "loss": 0.0285, + "step": 14119 + }, + { + "epoch": 11.925675675675675, + "grad_norm": 0.6830737590789795, + "learning_rate": 2.8628e-05, + "loss": 0.0232, + "step": 14120 + }, + { + "epoch": 11.92652027027027, + "grad_norm": 0.25739410519599915, + "learning_rate": 2.8627666666666667e-05, + "loss": 0.0192, + "step": 14121 + }, + { + "epoch": 11.927364864864865, + "grad_norm": 0.6081969141960144, + "learning_rate": 2.8627333333333333e-05, + "loss": 0.0376, + "step": 14122 + }, + { + "epoch": 11.92820945945946, + "grad_norm": 0.7733201384544373, + "learning_rate": 2.8627000000000002e-05, + "loss": 0.0361, + "step": 14123 + }, + { + "epoch": 11.929054054054054, + "grad_norm": 1.0659983158111572, + "learning_rate": 2.8626666666666668e-05, + "loss": 0.0759, + "step": 14124 + }, + { + "epoch": 11.92989864864865, + "grad_norm": 0.7986548542976379, + "learning_rate": 2.8626333333333334e-05, + "loss": 0.2622, + "step": 14125 + }, + { + "epoch": 11.930743243243244, + "grad_norm": 0.48477572202682495, + "learning_rate": 2.8626000000000003e-05, + "loss": 0.152, + "step": 14126 + }, + { + "epoch": 11.931587837837839, + "grad_norm": 0.2808171212673187, + "learning_rate": 2.8625666666666666e-05, + "loss": 0.0685, + "step": 14127 + }, + { + "epoch": 11.932432432432432, + "grad_norm": 0.5804234743118286, + "learning_rate": 2.8625333333333335e-05, + "loss": 0.0654, + "step": 14128 + }, + { + "epoch": 11.933277027027026, + "grad_norm": 0.3186016082763672, + "learning_rate": 2.8625e-05, + "loss": 0.0423, + "step": 14129 + }, + { + "epoch": 11.934121621621621, + "grad_norm": 0.3026050329208374, + "learning_rate": 2.8624666666666667e-05, + "loss": 0.0246, + "step": 14130 + }, + { + "epoch": 11.934966216216216, + "grad_norm": 0.32979142665863037, + "learning_rate": 2.8624333333333333e-05, + "loss": 0.025, + "step": 14131 + }, + { + "epoch": 11.93581081081081, + "grad_norm": 0.29641401767730713, + "learning_rate": 2.8624000000000002e-05, + "loss": 0.0217, + "step": 14132 + }, + { + "epoch": 11.936655405405405, + "grad_norm": 0.24468708038330078, + "learning_rate": 2.862366666666667e-05, + "loss": 0.0282, + "step": 14133 + }, + { + "epoch": 11.9375, + "grad_norm": 0.39193567633628845, + "learning_rate": 2.8623333333333334e-05, + "loss": 0.0216, + "step": 14134 + }, + { + "epoch": 11.938344594594595, + "grad_norm": 0.37263014912605286, + "learning_rate": 2.8623e-05, + "loss": 0.0258, + "step": 14135 + }, + { + "epoch": 11.93918918918919, + "grad_norm": 0.14327441155910492, + "learning_rate": 2.8622666666666666e-05, + "loss": 0.0124, + "step": 14136 + }, + { + "epoch": 11.940033783783784, + "grad_norm": 0.6497406959533691, + "learning_rate": 2.8622333333333335e-05, + "loss": 0.0382, + "step": 14137 + }, + { + "epoch": 11.940878378378379, + "grad_norm": 0.2831668257713318, + "learning_rate": 2.8621999999999998e-05, + "loss": 0.0249, + "step": 14138 + }, + { + "epoch": 11.941722972972974, + "grad_norm": 0.2645612061023712, + "learning_rate": 2.8621666666666667e-05, + "loss": 0.0226, + "step": 14139 + }, + { + "epoch": 11.942567567567568, + "grad_norm": 0.2770535945892334, + "learning_rate": 2.8621333333333333e-05, + "loss": 0.0146, + "step": 14140 + }, + { + "epoch": 11.943412162162161, + "grad_norm": 0.3253582715988159, + "learning_rate": 2.8621e-05, + "loss": 0.0268, + "step": 14141 + }, + { + "epoch": 11.944256756756756, + "grad_norm": 0.38236770033836365, + "learning_rate": 2.862066666666667e-05, + "loss": 0.0221, + "step": 14142 + }, + { + "epoch": 11.94510135135135, + "grad_norm": 0.46198758482933044, + "learning_rate": 2.8620333333333334e-05, + "loss": 0.0186, + "step": 14143 + }, + { + "epoch": 11.945945945945946, + "grad_norm": 0.40779033303260803, + "learning_rate": 2.862e-05, + "loss": 0.1042, + "step": 14144 + }, + { + "epoch": 11.94679054054054, + "grad_norm": 0.32032981514930725, + "learning_rate": 2.8619666666666666e-05, + "loss": 0.0253, + "step": 14145 + }, + { + "epoch": 11.947635135135135, + "grad_norm": 0.3046322762966156, + "learning_rate": 2.8619333333333336e-05, + "loss": 0.0292, + "step": 14146 + }, + { + "epoch": 11.94847972972973, + "grad_norm": 0.2052968144416809, + "learning_rate": 2.8618999999999998e-05, + "loss": 0.015, + "step": 14147 + }, + { + "epoch": 11.949324324324325, + "grad_norm": 0.2834811806678772, + "learning_rate": 2.8618666666666668e-05, + "loss": 0.0211, + "step": 14148 + }, + { + "epoch": 11.95016891891892, + "grad_norm": 0.29210764169692993, + "learning_rate": 2.8618333333333337e-05, + "loss": 0.0194, + "step": 14149 + }, + { + "epoch": 11.951013513513514, + "grad_norm": 0.49124646186828613, + "learning_rate": 2.8618e-05, + "loss": 0.0249, + "step": 14150 + }, + { + "epoch": 11.951858108108109, + "grad_norm": 0.26520559191703796, + "learning_rate": 2.861766666666667e-05, + "loss": 0.0214, + "step": 14151 + }, + { + "epoch": 11.952702702702704, + "grad_norm": 0.4253673255443573, + "learning_rate": 2.8617333333333335e-05, + "loss": 0.021, + "step": 14152 + }, + { + "epoch": 11.953547297297296, + "grad_norm": 0.16250643134117126, + "learning_rate": 2.8617e-05, + "loss": 0.0173, + "step": 14153 + }, + { + "epoch": 11.954391891891891, + "grad_norm": 0.1859256476163864, + "learning_rate": 2.8616666666666667e-05, + "loss": 0.0114, + "step": 14154 + }, + { + "epoch": 11.955236486486486, + "grad_norm": 0.25798073410987854, + "learning_rate": 2.8616333333333336e-05, + "loss": 0.0182, + "step": 14155 + }, + { + "epoch": 11.95608108108108, + "grad_norm": 0.5070990324020386, + "learning_rate": 2.8616e-05, + "loss": 0.0259, + "step": 14156 + }, + { + "epoch": 11.956925675675675, + "grad_norm": 0.500828742980957, + "learning_rate": 2.8615666666666668e-05, + "loss": 0.0425, + "step": 14157 + }, + { + "epoch": 11.95777027027027, + "grad_norm": 0.30007821321487427, + "learning_rate": 2.8615333333333337e-05, + "loss": 0.0255, + "step": 14158 + }, + { + "epoch": 11.958614864864865, + "grad_norm": 0.34460777044296265, + "learning_rate": 2.8615e-05, + "loss": 0.0257, + "step": 14159 + }, + { + "epoch": 11.95945945945946, + "grad_norm": 0.5663167238235474, + "learning_rate": 2.861466666666667e-05, + "loss": 0.03, + "step": 14160 + }, + { + "epoch": 11.960304054054054, + "grad_norm": 0.9668729901313782, + "learning_rate": 2.8614333333333335e-05, + "loss": 0.0211, + "step": 14161 + }, + { + "epoch": 11.96114864864865, + "grad_norm": 0.26827478408813477, + "learning_rate": 2.8614e-05, + "loss": 0.0192, + "step": 14162 + }, + { + "epoch": 11.961993243243244, + "grad_norm": 0.24604713916778564, + "learning_rate": 2.8613666666666667e-05, + "loss": 0.0213, + "step": 14163 + }, + { + "epoch": 11.962837837837839, + "grad_norm": 0.7708138227462769, + "learning_rate": 2.8613333333333333e-05, + "loss": 0.024, + "step": 14164 + }, + { + "epoch": 11.963682432432432, + "grad_norm": 0.2698531746864319, + "learning_rate": 2.8613e-05, + "loss": 0.0207, + "step": 14165 + }, + { + "epoch": 11.964527027027026, + "grad_norm": 0.4183349907398224, + "learning_rate": 2.8612666666666668e-05, + "loss": 0.0232, + "step": 14166 + }, + { + "epoch": 11.965371621621621, + "grad_norm": 0.37901508808135986, + "learning_rate": 2.8612333333333334e-05, + "loss": 0.0265, + "step": 14167 + }, + { + "epoch": 11.966216216216216, + "grad_norm": 0.28112247586250305, + "learning_rate": 2.8612e-05, + "loss": 0.0167, + "step": 14168 + }, + { + "epoch": 11.96706081081081, + "grad_norm": 0.24971289932727814, + "learning_rate": 2.861166666666667e-05, + "loss": 0.0158, + "step": 14169 + }, + { + "epoch": 11.967905405405405, + "grad_norm": 0.39698711037635803, + "learning_rate": 2.861133333333333e-05, + "loss": 0.0289, + "step": 14170 + }, + { + "epoch": 11.96875, + "grad_norm": 1.4614381790161133, + "learning_rate": 2.8611e-05, + "loss": 0.0398, + "step": 14171 + }, + { + "epoch": 11.969594594594595, + "grad_norm": 0.5245623588562012, + "learning_rate": 2.8610666666666667e-05, + "loss": 0.025, + "step": 14172 + }, + { + "epoch": 11.97043918918919, + "grad_norm": 0.905626654624939, + "learning_rate": 2.8610333333333333e-05, + "loss": 0.0501, + "step": 14173 + }, + { + "epoch": 11.971283783783784, + "grad_norm": 0.8275316953659058, + "learning_rate": 2.8610000000000002e-05, + "loss": 0.0338, + "step": 14174 + }, + { + "epoch": 11.972128378378379, + "grad_norm": 0.7438571453094482, + "learning_rate": 2.8609666666666668e-05, + "loss": 0.2072, + "step": 14175 + }, + { + "epoch": 11.972972972972974, + "grad_norm": 0.3034786880016327, + "learning_rate": 2.8609333333333334e-05, + "loss": 0.0597, + "step": 14176 + }, + { + "epoch": 11.973817567567568, + "grad_norm": 0.41806885600090027, + "learning_rate": 2.8609e-05, + "loss": 0.0449, + "step": 14177 + }, + { + "epoch": 11.974662162162161, + "grad_norm": 0.33154264092445374, + "learning_rate": 2.860866666666667e-05, + "loss": 0.0407, + "step": 14178 + }, + { + "epoch": 11.975506756756756, + "grad_norm": 0.3812901973724365, + "learning_rate": 2.8608333333333332e-05, + "loss": 0.0301, + "step": 14179 + }, + { + "epoch": 11.97635135135135, + "grad_norm": 0.1828855723142624, + "learning_rate": 2.8608e-05, + "loss": 0.018, + "step": 14180 + }, + { + "epoch": 11.977195945945946, + "grad_norm": 0.39686861634254456, + "learning_rate": 2.8607666666666667e-05, + "loss": 0.0381, + "step": 14181 + }, + { + "epoch": 11.97804054054054, + "grad_norm": 0.2832949161529541, + "learning_rate": 2.8607333333333333e-05, + "loss": 0.018, + "step": 14182 + }, + { + "epoch": 11.978885135135135, + "grad_norm": 0.28900033235549927, + "learning_rate": 2.8607000000000002e-05, + "loss": 0.0248, + "step": 14183 + }, + { + "epoch": 11.97972972972973, + "grad_norm": 0.3304891884326935, + "learning_rate": 2.860666666666667e-05, + "loss": 0.0256, + "step": 14184 + }, + { + "epoch": 11.980574324324325, + "grad_norm": 0.3236885964870453, + "learning_rate": 2.8606333333333334e-05, + "loss": 0.0158, + "step": 14185 + }, + { + "epoch": 11.98141891891892, + "grad_norm": 0.1756795346736908, + "learning_rate": 2.8606e-05, + "loss": 0.0152, + "step": 14186 + }, + { + "epoch": 11.982263513513514, + "grad_norm": 0.3285803198814392, + "learning_rate": 2.860566666666667e-05, + "loss": 0.0453, + "step": 14187 + }, + { + "epoch": 11.983108108108109, + "grad_norm": 0.35700303316116333, + "learning_rate": 2.8605333333333332e-05, + "loss": 0.0289, + "step": 14188 + }, + { + "epoch": 11.983952702702704, + "grad_norm": 0.21650251746177673, + "learning_rate": 2.8605e-05, + "loss": 0.0161, + "step": 14189 + }, + { + "epoch": 11.984797297297296, + "grad_norm": 0.4653680622577667, + "learning_rate": 2.8604666666666667e-05, + "loss": 0.0202, + "step": 14190 + }, + { + "epoch": 11.985641891891891, + "grad_norm": 0.3271961510181427, + "learning_rate": 2.8604333333333333e-05, + "loss": 0.0229, + "step": 14191 + }, + { + "epoch": 11.986486486486486, + "grad_norm": 0.369990736246109, + "learning_rate": 2.8604000000000003e-05, + "loss": 0.0282, + "step": 14192 + }, + { + "epoch": 11.98733108108108, + "grad_norm": 0.192704975605011, + "learning_rate": 2.8603666666666665e-05, + "loss": 0.0127, + "step": 14193 + }, + { + "epoch": 11.988175675675675, + "grad_norm": 0.2684963047504425, + "learning_rate": 2.8603333333333334e-05, + "loss": 0.0214, + "step": 14194 + }, + { + "epoch": 11.98902027027027, + "grad_norm": 0.3265538513660431, + "learning_rate": 2.8603e-05, + "loss": 0.0357, + "step": 14195 + }, + { + "epoch": 11.989864864864865, + "grad_norm": 0.46396544575691223, + "learning_rate": 2.8602666666666666e-05, + "loss": 0.0257, + "step": 14196 + }, + { + "epoch": 11.99070945945946, + "grad_norm": 0.3753075897693634, + "learning_rate": 2.8602333333333332e-05, + "loss": 0.0215, + "step": 14197 + }, + { + "epoch": 11.991554054054054, + "grad_norm": 0.24823899567127228, + "learning_rate": 2.8602e-05, + "loss": 0.0189, + "step": 14198 + }, + { + "epoch": 11.99239864864865, + "grad_norm": 0.5698351263999939, + "learning_rate": 2.8601666666666668e-05, + "loss": 0.0238, + "step": 14199 + }, + { + "epoch": 11.993243243243244, + "grad_norm": 0.43737736344337463, + "learning_rate": 2.8601333333333333e-05, + "loss": 0.0299, + "step": 14200 + }, + { + "epoch": 11.994087837837839, + "grad_norm": 0.3594253957271576, + "learning_rate": 2.8601000000000003e-05, + "loss": 0.0283, + "step": 14201 + }, + { + "epoch": 11.994932432432432, + "grad_norm": 0.4951845407485962, + "learning_rate": 2.8600666666666665e-05, + "loss": 0.0197, + "step": 14202 + }, + { + "epoch": 11.995777027027026, + "grad_norm": 0.2605746388435364, + "learning_rate": 2.8600333333333335e-05, + "loss": 0.0193, + "step": 14203 + }, + { + "epoch": 11.996621621621621, + "grad_norm": 0.42141327261924744, + "learning_rate": 2.86e-05, + "loss": 0.0269, + "step": 14204 + }, + { + "epoch": 11.997466216216216, + "grad_norm": 0.3951706886291504, + "learning_rate": 2.8599666666666667e-05, + "loss": 0.0278, + "step": 14205 + }, + { + "epoch": 11.99831081081081, + "grad_norm": 0.5495996475219727, + "learning_rate": 2.8599333333333332e-05, + "loss": 0.028, + "step": 14206 + }, + { + "epoch": 11.999155405405405, + "grad_norm": 0.39280715584754944, + "learning_rate": 2.8599000000000002e-05, + "loss": 0.033, + "step": 14207 + }, + { + "epoch": 12.0, + "grad_norm": 2.6494479179382324, + "learning_rate": 2.8598666666666668e-05, + "loss": 0.0445, + "step": 14208 + }, + { + "epoch": 12.000844594594595, + "grad_norm": 0.8629067540168762, + "learning_rate": 2.8598333333333334e-05, + "loss": 0.1827, + "step": 14209 + }, + { + "epoch": 12.00168918918919, + "grad_norm": 0.4807661175727844, + "learning_rate": 2.8598000000000003e-05, + "loss": 0.1069, + "step": 14210 + }, + { + "epoch": 12.002533783783784, + "grad_norm": 0.7641931772232056, + "learning_rate": 2.8597666666666666e-05, + "loss": 0.1003, + "step": 14211 + }, + { + "epoch": 12.003378378378379, + "grad_norm": 0.37342506647109985, + "learning_rate": 2.8597333333333335e-05, + "loss": 0.0385, + "step": 14212 + }, + { + "epoch": 12.004222972972974, + "grad_norm": 0.2431814968585968, + "learning_rate": 2.8597e-05, + "loss": 0.0273, + "step": 14213 + }, + { + "epoch": 12.005067567567568, + "grad_norm": 0.38156774640083313, + "learning_rate": 2.8596666666666667e-05, + "loss": 0.0296, + "step": 14214 + }, + { + "epoch": 12.005912162162161, + "grad_norm": 0.5361897349357605, + "learning_rate": 2.8596333333333333e-05, + "loss": 0.0202, + "step": 14215 + }, + { + "epoch": 12.006756756756756, + "grad_norm": 0.24882015585899353, + "learning_rate": 2.8596000000000002e-05, + "loss": 0.019, + "step": 14216 + }, + { + "epoch": 12.00760135135135, + "grad_norm": 0.31731775403022766, + "learning_rate": 2.8595666666666668e-05, + "loss": 0.0285, + "step": 14217 + }, + { + "epoch": 12.008445945945946, + "grad_norm": 0.229139044880867, + "learning_rate": 2.8595333333333334e-05, + "loss": 0.0174, + "step": 14218 + }, + { + "epoch": 12.00929054054054, + "grad_norm": 0.2713060677051544, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.029, + "step": 14219 + }, + { + "epoch": 12.010135135135135, + "grad_norm": 0.4445675015449524, + "learning_rate": 2.8594666666666666e-05, + "loss": 0.0267, + "step": 14220 + }, + { + "epoch": 12.01097972972973, + "grad_norm": 0.25197842717170715, + "learning_rate": 2.8594333333333335e-05, + "loss": 0.0246, + "step": 14221 + }, + { + "epoch": 12.011824324324325, + "grad_norm": 0.2700243294239044, + "learning_rate": 2.8594e-05, + "loss": 0.0218, + "step": 14222 + }, + { + "epoch": 12.01266891891892, + "grad_norm": 0.3519600033760071, + "learning_rate": 2.8593666666666667e-05, + "loss": 0.0254, + "step": 14223 + }, + { + "epoch": 12.013513513513514, + "grad_norm": 0.3434760272502899, + "learning_rate": 2.8593333333333336e-05, + "loss": 0.0234, + "step": 14224 + }, + { + "epoch": 12.014358108108109, + "grad_norm": 0.1980137825012207, + "learning_rate": 2.8593e-05, + "loss": 0.0167, + "step": 14225 + }, + { + "epoch": 12.015202702702704, + "grad_norm": 0.239728644490242, + "learning_rate": 2.8592666666666668e-05, + "loss": 0.0157, + "step": 14226 + }, + { + "epoch": 12.016047297297296, + "grad_norm": 0.36627569794654846, + "learning_rate": 2.8592333333333334e-05, + "loss": 0.0217, + "step": 14227 + }, + { + "epoch": 12.016891891891891, + "grad_norm": 0.5625806450843811, + "learning_rate": 2.8592e-05, + "loss": 0.0361, + "step": 14228 + }, + { + "epoch": 12.017736486486486, + "grad_norm": 0.2485320270061493, + "learning_rate": 2.8591666666666666e-05, + "loss": 0.0159, + "step": 14229 + }, + { + "epoch": 12.01858108108108, + "grad_norm": 0.3947502374649048, + "learning_rate": 2.8591333333333335e-05, + "loss": 0.0176, + "step": 14230 + }, + { + "epoch": 12.019425675675675, + "grad_norm": 0.38370591402053833, + "learning_rate": 2.8590999999999998e-05, + "loss": 0.0369, + "step": 14231 + }, + { + "epoch": 12.02027027027027, + "grad_norm": 0.28068602085113525, + "learning_rate": 2.8590666666666667e-05, + "loss": 0.0141, + "step": 14232 + }, + { + "epoch": 12.021114864864865, + "grad_norm": 0.22821980714797974, + "learning_rate": 2.8590333333333336e-05, + "loss": 0.0176, + "step": 14233 + }, + { + "epoch": 12.02195945945946, + "grad_norm": 0.31573987007141113, + "learning_rate": 2.859e-05, + "loss": 0.0287, + "step": 14234 + }, + { + "epoch": 12.022804054054054, + "grad_norm": 0.1743588149547577, + "learning_rate": 2.858966666666667e-05, + "loss": 0.0186, + "step": 14235 + }, + { + "epoch": 12.02364864864865, + "grad_norm": 0.23819014430046082, + "learning_rate": 2.8589333333333334e-05, + "loss": 0.0151, + "step": 14236 + }, + { + "epoch": 12.024493243243244, + "grad_norm": 0.21787339448928833, + "learning_rate": 2.8589e-05, + "loss": 0.0197, + "step": 14237 + }, + { + "epoch": 12.025337837837839, + "grad_norm": 0.38751810789108276, + "learning_rate": 2.8588666666666666e-05, + "loss": 0.0229, + "step": 14238 + }, + { + "epoch": 12.026182432432432, + "grad_norm": 0.44712427258491516, + "learning_rate": 2.8588333333333335e-05, + "loss": 0.0209, + "step": 14239 + }, + { + "epoch": 12.027027027027026, + "grad_norm": 1.0933061838150024, + "learning_rate": 2.8587999999999998e-05, + "loss": 0.0212, + "step": 14240 + }, + { + "epoch": 12.027871621621621, + "grad_norm": 0.39960426092147827, + "learning_rate": 2.8587666666666667e-05, + "loss": 0.0337, + "step": 14241 + }, + { + "epoch": 12.028716216216216, + "grad_norm": 0.7815983891487122, + "learning_rate": 2.8587333333333337e-05, + "loss": 0.0114, + "step": 14242 + }, + { + "epoch": 12.02956081081081, + "grad_norm": 1.1466535329818726, + "learning_rate": 2.8587e-05, + "loss": 0.0338, + "step": 14243 + }, + { + "epoch": 12.030405405405405, + "grad_norm": 0.7795867919921875, + "learning_rate": 2.858666666666667e-05, + "loss": 0.0181, + "step": 14244 + }, + { + "epoch": 12.03125, + "grad_norm": 0.3149898648262024, + "learning_rate": 2.8586333333333334e-05, + "loss": 0.0112, + "step": 14245 + }, + { + "epoch": 12.032094594594595, + "grad_norm": 0.23577021062374115, + "learning_rate": 2.8586e-05, + "loss": 0.014, + "step": 14246 + }, + { + "epoch": 12.03293918918919, + "grad_norm": 0.2778014838695526, + "learning_rate": 2.8585666666666666e-05, + "loss": 0.0218, + "step": 14247 + }, + { + "epoch": 12.033783783783784, + "grad_norm": 0.26905348896980286, + "learning_rate": 2.8585333333333336e-05, + "loss": 0.0159, + "step": 14248 + }, + { + "epoch": 12.034628378378379, + "grad_norm": 0.3806495666503906, + "learning_rate": 2.8585e-05, + "loss": 0.0333, + "step": 14249 + }, + { + "epoch": 12.035472972972974, + "grad_norm": 0.3848082721233368, + "learning_rate": 2.8584666666666668e-05, + "loss": 0.0299, + "step": 14250 + }, + { + "epoch": 12.036317567567568, + "grad_norm": 0.6047413945198059, + "learning_rate": 2.8584333333333337e-05, + "loss": 0.0299, + "step": 14251 + }, + { + "epoch": 12.037162162162161, + "grad_norm": 0.3651602268218994, + "learning_rate": 2.8584e-05, + "loss": 0.0276, + "step": 14252 + }, + { + "epoch": 12.038006756756756, + "grad_norm": 0.4335003197193146, + "learning_rate": 2.858366666666667e-05, + "loss": 0.0239, + "step": 14253 + }, + { + "epoch": 12.03885135135135, + "grad_norm": 0.7443845272064209, + "learning_rate": 2.858333333333333e-05, + "loss": 0.0223, + "step": 14254 + }, + { + "epoch": 12.039695945945946, + "grad_norm": 0.45775628089904785, + "learning_rate": 2.8583e-05, + "loss": 0.026, + "step": 14255 + }, + { + "epoch": 12.04054054054054, + "grad_norm": 0.5589524507522583, + "learning_rate": 2.8582666666666667e-05, + "loss": 0.0576, + "step": 14256 + }, + { + "epoch": 12.041385135135135, + "grad_norm": 0.31058716773986816, + "learning_rate": 2.8582333333333332e-05, + "loss": 0.0165, + "step": 14257 + }, + { + "epoch": 12.04222972972973, + "grad_norm": 0.925097644329071, + "learning_rate": 2.8582000000000002e-05, + "loss": 0.0386, + "step": 14258 + }, + { + "epoch": 12.043074324324325, + "grad_norm": 1.4076651334762573, + "learning_rate": 2.8581666666666668e-05, + "loss": 0.2756, + "step": 14259 + }, + { + "epoch": 12.04391891891892, + "grad_norm": 0.38489848375320435, + "learning_rate": 2.8581333333333334e-05, + "loss": 0.1502, + "step": 14260 + }, + { + "epoch": 12.044763513513514, + "grad_norm": 0.40760135650634766, + "learning_rate": 2.8581e-05, + "loss": 0.0705, + "step": 14261 + }, + { + "epoch": 12.045608108108109, + "grad_norm": 0.2603314220905304, + "learning_rate": 2.858066666666667e-05, + "loss": 0.02, + "step": 14262 + }, + { + "epoch": 12.046452702702704, + "grad_norm": 0.5805001854896545, + "learning_rate": 2.858033333333333e-05, + "loss": 0.0512, + "step": 14263 + }, + { + "epoch": 12.047297297297296, + "grad_norm": 1.1098285913467407, + "learning_rate": 2.858e-05, + "loss": 0.0414, + "step": 14264 + }, + { + "epoch": 12.048141891891891, + "grad_norm": 0.38924458622932434, + "learning_rate": 2.8579666666666667e-05, + "loss": 0.0382, + "step": 14265 + }, + { + "epoch": 12.048986486486486, + "grad_norm": 0.2501746118068695, + "learning_rate": 2.8579333333333333e-05, + "loss": 0.02, + "step": 14266 + }, + { + "epoch": 12.04983108108108, + "grad_norm": 0.2785361409187317, + "learning_rate": 2.8579000000000002e-05, + "loss": 0.0234, + "step": 14267 + }, + { + "epoch": 12.050675675675675, + "grad_norm": 0.24907338619232178, + "learning_rate": 2.8578666666666668e-05, + "loss": 0.0153, + "step": 14268 + }, + { + "epoch": 12.05152027027027, + "grad_norm": 0.29247361421585083, + "learning_rate": 2.8578333333333334e-05, + "loss": 0.028, + "step": 14269 + }, + { + "epoch": 12.052364864864865, + "grad_norm": 0.2488372027873993, + "learning_rate": 2.8578e-05, + "loss": 0.0246, + "step": 14270 + }, + { + "epoch": 12.05320945945946, + "grad_norm": 0.40993279218673706, + "learning_rate": 2.857766666666667e-05, + "loss": 0.0181, + "step": 14271 + }, + { + "epoch": 12.054054054054054, + "grad_norm": 0.33167141675949097, + "learning_rate": 2.857733333333333e-05, + "loss": 0.0217, + "step": 14272 + }, + { + "epoch": 12.05489864864865, + "grad_norm": 0.5470162034034729, + "learning_rate": 2.8577e-05, + "loss": 0.0316, + "step": 14273 + }, + { + "epoch": 12.055743243243244, + "grad_norm": 0.2630906105041504, + "learning_rate": 2.857666666666667e-05, + "loss": 0.0196, + "step": 14274 + }, + { + "epoch": 12.056587837837839, + "grad_norm": 0.2790750563144684, + "learning_rate": 2.8576333333333333e-05, + "loss": 0.0216, + "step": 14275 + }, + { + "epoch": 12.057432432432432, + "grad_norm": 0.3223074674606323, + "learning_rate": 2.8576000000000002e-05, + "loss": 0.0199, + "step": 14276 + }, + { + "epoch": 12.058277027027026, + "grad_norm": 0.7601632475852966, + "learning_rate": 2.8575666666666668e-05, + "loss": 0.0227, + "step": 14277 + }, + { + "epoch": 12.059121621621621, + "grad_norm": 0.2855890095233917, + "learning_rate": 2.8575333333333334e-05, + "loss": 0.0173, + "step": 14278 + }, + { + "epoch": 12.059966216216216, + "grad_norm": 0.291706383228302, + "learning_rate": 2.8575e-05, + "loss": 0.0205, + "step": 14279 + }, + { + "epoch": 12.06081081081081, + "grad_norm": 0.2407863289117813, + "learning_rate": 2.857466666666667e-05, + "loss": 0.0117, + "step": 14280 + }, + { + "epoch": 12.061655405405405, + "grad_norm": 0.4498355984687805, + "learning_rate": 2.8574333333333332e-05, + "loss": 0.0725, + "step": 14281 + }, + { + "epoch": 12.0625, + "grad_norm": 0.2856082618236542, + "learning_rate": 2.8574e-05, + "loss": 0.0175, + "step": 14282 + }, + { + "epoch": 12.063344594594595, + "grad_norm": 0.47063547372817993, + "learning_rate": 2.857366666666667e-05, + "loss": 0.0194, + "step": 14283 + }, + { + "epoch": 12.06418918918919, + "grad_norm": 0.3765352964401245, + "learning_rate": 2.8573333333333333e-05, + "loss": 0.0222, + "step": 14284 + }, + { + "epoch": 12.065033783783784, + "grad_norm": 0.2275628298521042, + "learning_rate": 2.8573000000000002e-05, + "loss": 0.0109, + "step": 14285 + }, + { + "epoch": 12.065878378378379, + "grad_norm": 0.14068515598773956, + "learning_rate": 2.8572666666666665e-05, + "loss": 0.0102, + "step": 14286 + }, + { + "epoch": 12.066722972972974, + "grad_norm": 0.317962646484375, + "learning_rate": 2.8572333333333334e-05, + "loss": 0.0114, + "step": 14287 + }, + { + "epoch": 12.067567567567568, + "grad_norm": 0.542002260684967, + "learning_rate": 2.8572e-05, + "loss": 0.0418, + "step": 14288 + }, + { + "epoch": 12.068412162162161, + "grad_norm": 0.3401987850666046, + "learning_rate": 2.8571666666666666e-05, + "loss": 0.0114, + "step": 14289 + }, + { + "epoch": 12.069256756756756, + "grad_norm": 0.594541072845459, + "learning_rate": 2.8571333333333332e-05, + "loss": 0.0335, + "step": 14290 + }, + { + "epoch": 12.07010135135135, + "grad_norm": 0.2793961763381958, + "learning_rate": 2.8571e-05, + "loss": 0.0251, + "step": 14291 + }, + { + "epoch": 12.070945945945946, + "grad_norm": 0.22216129302978516, + "learning_rate": 2.8570666666666667e-05, + "loss": 0.0124, + "step": 14292 + }, + { + "epoch": 12.07179054054054, + "grad_norm": 0.34566691517829895, + "learning_rate": 2.8570333333333333e-05, + "loss": 0.024, + "step": 14293 + }, + { + "epoch": 12.072635135135135, + "grad_norm": 0.3619843125343323, + "learning_rate": 2.8570000000000003e-05, + "loss": 0.0219, + "step": 14294 + }, + { + "epoch": 12.07347972972973, + "grad_norm": 0.3484742343425751, + "learning_rate": 2.8569666666666665e-05, + "loss": 0.0197, + "step": 14295 + }, + { + "epoch": 12.074324324324325, + "grad_norm": 0.5316829681396484, + "learning_rate": 2.8569333333333334e-05, + "loss": 0.0189, + "step": 14296 + }, + { + "epoch": 12.07516891891892, + "grad_norm": 0.422658234834671, + "learning_rate": 2.8569e-05, + "loss": 0.014, + "step": 14297 + }, + { + "epoch": 12.076013513513514, + "grad_norm": 0.20011278986930847, + "learning_rate": 2.8568666666666666e-05, + "loss": 0.0126, + "step": 14298 + }, + { + "epoch": 12.076858108108109, + "grad_norm": 0.23210009932518005, + "learning_rate": 2.8568333333333336e-05, + "loss": 0.0155, + "step": 14299 + }, + { + "epoch": 12.077702702702704, + "grad_norm": 0.2749461829662323, + "learning_rate": 2.8568e-05, + "loss": 0.012, + "step": 14300 + }, + { + "epoch": 12.078547297297296, + "grad_norm": 0.29191261529922485, + "learning_rate": 2.8567666666666668e-05, + "loss": 0.0221, + "step": 14301 + }, + { + "epoch": 12.079391891891891, + "grad_norm": 1.214272379875183, + "learning_rate": 2.8567333333333333e-05, + "loss": 0.0263, + "step": 14302 + }, + { + "epoch": 12.080236486486486, + "grad_norm": 0.8623942136764526, + "learning_rate": 2.8567000000000003e-05, + "loss": 0.0524, + "step": 14303 + }, + { + "epoch": 12.08108108108108, + "grad_norm": 0.41424086689949036, + "learning_rate": 2.8566666666666665e-05, + "loss": 0.0222, + "step": 14304 + }, + { + "epoch": 12.081925675675675, + "grad_norm": 0.24560849368572235, + "learning_rate": 2.8566333333333335e-05, + "loss": 0.0189, + "step": 14305 + }, + { + "epoch": 12.08277027027027, + "grad_norm": 0.2188667207956314, + "learning_rate": 2.8566e-05, + "loss": 0.0188, + "step": 14306 + }, + { + "epoch": 12.083614864864865, + "grad_norm": 0.5923473834991455, + "learning_rate": 2.8565666666666667e-05, + "loss": 0.0383, + "step": 14307 + }, + { + "epoch": 12.08445945945946, + "grad_norm": 0.35467275977134705, + "learning_rate": 2.8565333333333336e-05, + "loss": 0.0371, + "step": 14308 + }, + { + "epoch": 12.085304054054054, + "grad_norm": 1.3218083381652832, + "learning_rate": 2.8565000000000002e-05, + "loss": 0.2279, + "step": 14309 + }, + { + "epoch": 12.08614864864865, + "grad_norm": 0.8028162717819214, + "learning_rate": 2.8564666666666668e-05, + "loss": 0.1427, + "step": 14310 + }, + { + "epoch": 12.086993243243244, + "grad_norm": 0.4981938600540161, + "learning_rate": 2.8564333333333334e-05, + "loss": 0.067, + "step": 14311 + }, + { + "epoch": 12.087837837837839, + "grad_norm": 0.36980530619621277, + "learning_rate": 2.8564000000000003e-05, + "loss": 0.0484, + "step": 14312 + }, + { + "epoch": 12.088682432432432, + "grad_norm": 0.3724154531955719, + "learning_rate": 2.8563666666666666e-05, + "loss": 0.035, + "step": 14313 + }, + { + "epoch": 12.089527027027026, + "grad_norm": 0.28864601254463196, + "learning_rate": 2.8563333333333335e-05, + "loss": 0.0251, + "step": 14314 + }, + { + "epoch": 12.090371621621621, + "grad_norm": 0.23434728384017944, + "learning_rate": 2.8563e-05, + "loss": 0.0168, + "step": 14315 + }, + { + "epoch": 12.091216216216216, + "grad_norm": 0.4106171727180481, + "learning_rate": 2.8562666666666667e-05, + "loss": 0.0266, + "step": 14316 + }, + { + "epoch": 12.09206081081081, + "grad_norm": 0.2608390152454376, + "learning_rate": 2.8562333333333336e-05, + "loss": 0.0168, + "step": 14317 + }, + { + "epoch": 12.092905405405405, + "grad_norm": 0.3633280098438263, + "learning_rate": 2.8562e-05, + "loss": 0.03, + "step": 14318 + }, + { + "epoch": 12.09375, + "grad_norm": 0.2768166661262512, + "learning_rate": 2.8561666666666668e-05, + "loss": 0.0201, + "step": 14319 + }, + { + "epoch": 12.094594594594595, + "grad_norm": 0.407368540763855, + "learning_rate": 2.8561333333333334e-05, + "loss": 0.0431, + "step": 14320 + }, + { + "epoch": 12.09543918918919, + "grad_norm": 0.2532147169113159, + "learning_rate": 2.8561e-05, + "loss": 0.0212, + "step": 14321 + }, + { + "epoch": 12.096283783783784, + "grad_norm": 0.46027469635009766, + "learning_rate": 2.8560666666666666e-05, + "loss": 0.022, + "step": 14322 + }, + { + "epoch": 12.097128378378379, + "grad_norm": 0.42203542590141296, + "learning_rate": 2.8560333333333335e-05, + "loss": 0.0173, + "step": 14323 + }, + { + "epoch": 12.097972972972974, + "grad_norm": 0.3962097764015198, + "learning_rate": 2.856e-05, + "loss": 0.0264, + "step": 14324 + }, + { + "epoch": 12.098817567567568, + "grad_norm": 0.3135410249233246, + "learning_rate": 2.8559666666666667e-05, + "loss": 0.0297, + "step": 14325 + }, + { + "epoch": 12.099662162162161, + "grad_norm": 0.38496196269989014, + "learning_rate": 2.8559333333333336e-05, + "loss": 0.0129, + "step": 14326 + }, + { + "epoch": 12.100506756756756, + "grad_norm": 0.319898396730423, + "learning_rate": 2.8559e-05, + "loss": 0.0302, + "step": 14327 + }, + { + "epoch": 12.10135135135135, + "grad_norm": 0.32069945335388184, + "learning_rate": 2.8558666666666668e-05, + "loss": 0.0189, + "step": 14328 + }, + { + "epoch": 12.102195945945946, + "grad_norm": 0.5991360545158386, + "learning_rate": 2.8558333333333334e-05, + "loss": 0.0226, + "step": 14329 + }, + { + "epoch": 12.10304054054054, + "grad_norm": 0.4235656261444092, + "learning_rate": 2.8558e-05, + "loss": 0.0352, + "step": 14330 + }, + { + "epoch": 12.103885135135135, + "grad_norm": 0.4459403157234192, + "learning_rate": 2.8557666666666666e-05, + "loss": 0.0306, + "step": 14331 + }, + { + "epoch": 12.10472972972973, + "grad_norm": 0.35552021861076355, + "learning_rate": 2.8557333333333335e-05, + "loss": 0.0483, + "step": 14332 + }, + { + "epoch": 12.105574324324325, + "grad_norm": 0.28817883133888245, + "learning_rate": 2.8557e-05, + "loss": 0.0221, + "step": 14333 + }, + { + "epoch": 12.10641891891892, + "grad_norm": 0.2592022716999054, + "learning_rate": 2.8556666666666667e-05, + "loss": 0.0418, + "step": 14334 + }, + { + "epoch": 12.107263513513514, + "grad_norm": 0.3904736638069153, + "learning_rate": 2.8556333333333336e-05, + "loss": 0.0151, + "step": 14335 + }, + { + "epoch": 12.108108108108109, + "grad_norm": 0.4588797986507416, + "learning_rate": 2.8556e-05, + "loss": 0.0298, + "step": 14336 + }, + { + "epoch": 12.108952702702704, + "grad_norm": 0.452217698097229, + "learning_rate": 2.855566666666667e-05, + "loss": 0.0219, + "step": 14337 + }, + { + "epoch": 12.109797297297296, + "grad_norm": 0.28362134099006653, + "learning_rate": 2.8555333333333334e-05, + "loss": 0.0192, + "step": 14338 + }, + { + "epoch": 12.110641891891891, + "grad_norm": 0.4313940703868866, + "learning_rate": 2.8555e-05, + "loss": 0.0373, + "step": 14339 + }, + { + "epoch": 12.111486486486486, + "grad_norm": 0.48439162969589233, + "learning_rate": 2.8554666666666666e-05, + "loss": 0.0186, + "step": 14340 + }, + { + "epoch": 12.11233108108108, + "grad_norm": 0.09731058031320572, + "learning_rate": 2.8554333333333335e-05, + "loss": 0.0079, + "step": 14341 + }, + { + "epoch": 12.113175675675675, + "grad_norm": 0.3330135643482208, + "learning_rate": 2.8554e-05, + "loss": 0.0158, + "step": 14342 + }, + { + "epoch": 12.11402027027027, + "grad_norm": 0.3411490321159363, + "learning_rate": 2.8553666666666667e-05, + "loss": 0.0194, + "step": 14343 + }, + { + "epoch": 12.114864864864865, + "grad_norm": 0.3045237362384796, + "learning_rate": 2.8553333333333333e-05, + "loss": 0.0154, + "step": 14344 + }, + { + "epoch": 12.11570945945946, + "grad_norm": 0.2751268446445465, + "learning_rate": 2.8553e-05, + "loss": 0.023, + "step": 14345 + }, + { + "epoch": 12.116554054054054, + "grad_norm": 0.3422355055809021, + "learning_rate": 2.855266666666667e-05, + "loss": 0.0118, + "step": 14346 + }, + { + "epoch": 12.11739864864865, + "grad_norm": 0.29264920949935913, + "learning_rate": 2.855233333333333e-05, + "loss": 0.018, + "step": 14347 + }, + { + "epoch": 12.118243243243244, + "grad_norm": 0.30645278096199036, + "learning_rate": 2.8552e-05, + "loss": 0.0125, + "step": 14348 + }, + { + "epoch": 12.119087837837839, + "grad_norm": 0.36557072401046753, + "learning_rate": 2.855166666666667e-05, + "loss": 0.0173, + "step": 14349 + }, + { + "epoch": 12.119932432432432, + "grad_norm": 0.8415243625640869, + "learning_rate": 2.8551333333333332e-05, + "loss": 0.0201, + "step": 14350 + }, + { + "epoch": 12.120777027027026, + "grad_norm": 0.641435980796814, + "learning_rate": 2.8551e-05, + "loss": 0.0272, + "step": 14351 + }, + { + "epoch": 12.121621621621621, + "grad_norm": 0.38621965050697327, + "learning_rate": 2.8550666666666668e-05, + "loss": 0.015, + "step": 14352 + }, + { + "epoch": 12.122466216216216, + "grad_norm": 0.2322784662246704, + "learning_rate": 2.8550333333333333e-05, + "loss": 0.009, + "step": 14353 + }, + { + "epoch": 12.12331081081081, + "grad_norm": 0.17661741375923157, + "learning_rate": 2.855e-05, + "loss": 0.0146, + "step": 14354 + }, + { + "epoch": 12.124155405405405, + "grad_norm": 1.1559395790100098, + "learning_rate": 2.854966666666667e-05, + "loss": 0.0211, + "step": 14355 + }, + { + "epoch": 12.125, + "grad_norm": 0.341596782207489, + "learning_rate": 2.854933333333333e-05, + "loss": 0.0211, + "step": 14356 + }, + { + "epoch": 12.125844594594595, + "grad_norm": 0.43858495354652405, + "learning_rate": 2.8549e-05, + "loss": 0.0264, + "step": 14357 + }, + { + "epoch": 12.12668918918919, + "grad_norm": 0.744966447353363, + "learning_rate": 2.854866666666667e-05, + "loss": 0.0281, + "step": 14358 + }, + { + "epoch": 12.127533783783784, + "grad_norm": 0.7600439786911011, + "learning_rate": 2.8548333333333332e-05, + "loss": 0.2403, + "step": 14359 + }, + { + "epoch": 12.128378378378379, + "grad_norm": 0.7064679861068726, + "learning_rate": 2.8548000000000002e-05, + "loss": 0.139, + "step": 14360 + }, + { + "epoch": 12.129222972972974, + "grad_norm": 0.5773196816444397, + "learning_rate": 2.8547666666666668e-05, + "loss": 0.0816, + "step": 14361 + }, + { + "epoch": 12.130067567567568, + "grad_norm": 0.484454870223999, + "learning_rate": 2.8547333333333334e-05, + "loss": 0.0417, + "step": 14362 + }, + { + "epoch": 12.130912162162161, + "grad_norm": 0.642062246799469, + "learning_rate": 2.8547e-05, + "loss": 0.021, + "step": 14363 + }, + { + "epoch": 12.131756756756756, + "grad_norm": 0.23268739879131317, + "learning_rate": 2.854666666666667e-05, + "loss": 0.0226, + "step": 14364 + }, + { + "epoch": 12.13260135135135, + "grad_norm": 0.2605620324611664, + "learning_rate": 2.8546333333333335e-05, + "loss": 0.0192, + "step": 14365 + }, + { + "epoch": 12.133445945945946, + "grad_norm": 0.2997807264328003, + "learning_rate": 2.8546e-05, + "loss": 0.0207, + "step": 14366 + }, + { + "epoch": 12.13429054054054, + "grad_norm": 0.20992232859134674, + "learning_rate": 2.854566666666667e-05, + "loss": 0.0208, + "step": 14367 + }, + { + "epoch": 12.135135135135135, + "grad_norm": 0.31302496790885925, + "learning_rate": 2.8545333333333333e-05, + "loss": 0.0308, + "step": 14368 + }, + { + "epoch": 12.13597972972973, + "grad_norm": 0.2984400987625122, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0215, + "step": 14369 + }, + { + "epoch": 12.136824324324325, + "grad_norm": 0.1907411515712738, + "learning_rate": 2.8544666666666668e-05, + "loss": 0.0178, + "step": 14370 + }, + { + "epoch": 12.13766891891892, + "grad_norm": 0.37322545051574707, + "learning_rate": 2.8544333333333334e-05, + "loss": 0.026, + "step": 14371 + }, + { + "epoch": 12.138513513513514, + "grad_norm": 0.20898620784282684, + "learning_rate": 2.8544e-05, + "loss": 0.0151, + "step": 14372 + }, + { + "epoch": 12.139358108108109, + "grad_norm": 0.24393302202224731, + "learning_rate": 2.854366666666667e-05, + "loss": 0.0313, + "step": 14373 + }, + { + "epoch": 12.140202702702704, + "grad_norm": 0.30936646461486816, + "learning_rate": 2.8543333333333335e-05, + "loss": 0.0335, + "step": 14374 + }, + { + "epoch": 12.141047297297296, + "grad_norm": 0.27720949053764343, + "learning_rate": 2.8543e-05, + "loss": 0.0249, + "step": 14375 + }, + { + "epoch": 12.141891891891891, + "grad_norm": 0.2861649990081787, + "learning_rate": 2.8542666666666667e-05, + "loss": 0.0246, + "step": 14376 + }, + { + "epoch": 12.142736486486486, + "grad_norm": 0.335174560546875, + "learning_rate": 2.8542333333333333e-05, + "loss": 0.0254, + "step": 14377 + }, + { + "epoch": 12.14358108108108, + "grad_norm": 0.6636695861816406, + "learning_rate": 2.8542000000000002e-05, + "loss": 0.0295, + "step": 14378 + }, + { + "epoch": 12.144425675675675, + "grad_norm": 0.28632038831710815, + "learning_rate": 2.8541666666666665e-05, + "loss": 0.0181, + "step": 14379 + }, + { + "epoch": 12.14527027027027, + "grad_norm": 0.2449706494808197, + "learning_rate": 2.8541333333333334e-05, + "loss": 0.0207, + "step": 14380 + }, + { + "epoch": 12.146114864864865, + "grad_norm": 0.24591529369354248, + "learning_rate": 2.8541e-05, + "loss": 0.016, + "step": 14381 + }, + { + "epoch": 12.14695945945946, + "grad_norm": 0.4497009217739105, + "learning_rate": 2.8540666666666666e-05, + "loss": 0.0216, + "step": 14382 + }, + { + "epoch": 12.147804054054054, + "grad_norm": 0.17493686079978943, + "learning_rate": 2.8540333333333335e-05, + "loss": 0.0127, + "step": 14383 + }, + { + "epoch": 12.14864864864865, + "grad_norm": 0.2498861700296402, + "learning_rate": 2.854e-05, + "loss": 0.0178, + "step": 14384 + }, + { + "epoch": 12.149493243243244, + "grad_norm": 0.2349196821451187, + "learning_rate": 2.8539666666666667e-05, + "loss": 0.021, + "step": 14385 + }, + { + "epoch": 12.150337837837839, + "grad_norm": 0.4625854194164276, + "learning_rate": 2.8539333333333333e-05, + "loss": 0.0222, + "step": 14386 + }, + { + "epoch": 12.151182432432432, + "grad_norm": 0.21148669719696045, + "learning_rate": 2.8539000000000002e-05, + "loss": 0.0132, + "step": 14387 + }, + { + "epoch": 12.152027027027026, + "grad_norm": 0.3462897539138794, + "learning_rate": 2.8538666666666665e-05, + "loss": 0.0192, + "step": 14388 + }, + { + "epoch": 12.152871621621621, + "grad_norm": 0.6829212307929993, + "learning_rate": 2.8538333333333334e-05, + "loss": 0.0158, + "step": 14389 + }, + { + "epoch": 12.153716216216216, + "grad_norm": 0.2936362624168396, + "learning_rate": 2.8538e-05, + "loss": 0.0221, + "step": 14390 + }, + { + "epoch": 12.15456081081081, + "grad_norm": 0.2396087646484375, + "learning_rate": 2.8537666666666666e-05, + "loss": 0.0116, + "step": 14391 + }, + { + "epoch": 12.155405405405405, + "grad_norm": 0.2588749825954437, + "learning_rate": 2.8537333333333335e-05, + "loss": 0.0206, + "step": 14392 + }, + { + "epoch": 12.15625, + "grad_norm": 0.3894292712211609, + "learning_rate": 2.8537e-05, + "loss": 0.017, + "step": 14393 + }, + { + "epoch": 12.157094594594595, + "grad_norm": 0.40890854597091675, + "learning_rate": 2.8536666666666667e-05, + "loss": 0.0214, + "step": 14394 + }, + { + "epoch": 12.15793918918919, + "grad_norm": 0.32907935976982117, + "learning_rate": 2.8536333333333333e-05, + "loss": 0.0172, + "step": 14395 + }, + { + "epoch": 12.158783783783784, + "grad_norm": 0.42854076623916626, + "learning_rate": 2.8536000000000003e-05, + "loss": 0.0211, + "step": 14396 + }, + { + "epoch": 12.159628378378379, + "grad_norm": 0.5927668809890747, + "learning_rate": 2.8535666666666665e-05, + "loss": 0.0161, + "step": 14397 + }, + { + "epoch": 12.160472972972974, + "grad_norm": 0.3663746416568756, + "learning_rate": 2.8535333333333334e-05, + "loss": 0.0212, + "step": 14398 + }, + { + "epoch": 12.161317567567568, + "grad_norm": 0.2960839569568634, + "learning_rate": 2.8535000000000004e-05, + "loss": 0.016, + "step": 14399 + }, + { + "epoch": 12.162162162162161, + "grad_norm": 0.4784718155860901, + "learning_rate": 2.8534666666666666e-05, + "loss": 0.0244, + "step": 14400 + }, + { + "epoch": 12.163006756756756, + "grad_norm": 0.9097440838813782, + "learning_rate": 2.8534333333333336e-05, + "loss": 0.0286, + "step": 14401 + }, + { + "epoch": 12.16385135135135, + "grad_norm": 0.36783701181411743, + "learning_rate": 2.8534e-05, + "loss": 0.0213, + "step": 14402 + }, + { + "epoch": 12.164695945945946, + "grad_norm": 0.5142865180969238, + "learning_rate": 2.8533666666666668e-05, + "loss": 0.0399, + "step": 14403 + }, + { + "epoch": 12.16554054054054, + "grad_norm": 0.23484709858894348, + "learning_rate": 2.8533333333333333e-05, + "loss": 0.0278, + "step": 14404 + }, + { + "epoch": 12.166385135135135, + "grad_norm": 0.35064688324928284, + "learning_rate": 2.8533e-05, + "loss": 0.0189, + "step": 14405 + }, + { + "epoch": 12.16722972972973, + "grad_norm": 0.6447892785072327, + "learning_rate": 2.8532666666666665e-05, + "loss": 0.022, + "step": 14406 + }, + { + "epoch": 12.168074324324325, + "grad_norm": 1.4652271270751953, + "learning_rate": 2.8532333333333335e-05, + "loss": 0.0483, + "step": 14407 + }, + { + "epoch": 12.16891891891892, + "grad_norm": 0.45684823393821716, + "learning_rate": 2.8532e-05, + "loss": 0.0234, + "step": 14408 + }, + { + "epoch": 12.169763513513514, + "grad_norm": 0.623930037021637, + "learning_rate": 2.8531666666666667e-05, + "loss": 0.2089, + "step": 14409 + }, + { + "epoch": 12.170608108108109, + "grad_norm": 0.3987964391708374, + "learning_rate": 2.8531333333333336e-05, + "loss": 0.096, + "step": 14410 + }, + { + "epoch": 12.171452702702704, + "grad_norm": 0.24145425856113434, + "learning_rate": 2.8531e-05, + "loss": 0.0382, + "step": 14411 + }, + { + "epoch": 12.172297297297296, + "grad_norm": 0.34570834040641785, + "learning_rate": 2.8530666666666668e-05, + "loss": 0.0846, + "step": 14412 + }, + { + "epoch": 12.173141891891891, + "grad_norm": 0.3306005001068115, + "learning_rate": 2.8530333333333334e-05, + "loss": 0.026, + "step": 14413 + }, + { + "epoch": 12.173986486486486, + "grad_norm": 0.5336334109306335, + "learning_rate": 2.853e-05, + "loss": 0.0481, + "step": 14414 + }, + { + "epoch": 12.17483108108108, + "grad_norm": 0.21633794903755188, + "learning_rate": 2.852966666666667e-05, + "loss": 0.0228, + "step": 14415 + }, + { + "epoch": 12.175675675675675, + "grad_norm": 0.394395649433136, + "learning_rate": 2.8529333333333335e-05, + "loss": 0.0294, + "step": 14416 + }, + { + "epoch": 12.17652027027027, + "grad_norm": 0.42275354266166687, + "learning_rate": 2.8529e-05, + "loss": 0.0323, + "step": 14417 + }, + { + "epoch": 12.177364864864865, + "grad_norm": 0.40877681970596313, + "learning_rate": 2.8528666666666667e-05, + "loss": 0.0227, + "step": 14418 + }, + { + "epoch": 12.17820945945946, + "grad_norm": 0.24926970899105072, + "learning_rate": 2.8528333333333336e-05, + "loss": 0.0143, + "step": 14419 + }, + { + "epoch": 12.179054054054054, + "grad_norm": 0.2037222534418106, + "learning_rate": 2.8528e-05, + "loss": 0.0167, + "step": 14420 + }, + { + "epoch": 12.17989864864865, + "grad_norm": 0.3359799087047577, + "learning_rate": 2.8527666666666668e-05, + "loss": 0.0262, + "step": 14421 + }, + { + "epoch": 12.180743243243244, + "grad_norm": 0.23773634433746338, + "learning_rate": 2.8527333333333334e-05, + "loss": 0.024, + "step": 14422 + }, + { + "epoch": 12.181587837837839, + "grad_norm": 0.4316944479942322, + "learning_rate": 2.8527e-05, + "loss": 0.0235, + "step": 14423 + }, + { + "epoch": 12.182432432432432, + "grad_norm": 0.3033808469772339, + "learning_rate": 2.852666666666667e-05, + "loss": 0.018, + "step": 14424 + }, + { + "epoch": 12.183277027027026, + "grad_norm": 0.38573792576789856, + "learning_rate": 2.8526333333333335e-05, + "loss": 0.0132, + "step": 14425 + }, + { + "epoch": 12.184121621621621, + "grad_norm": 0.2353583127260208, + "learning_rate": 2.8526e-05, + "loss": 0.02, + "step": 14426 + }, + { + "epoch": 12.184966216216216, + "grad_norm": 0.41435468196868896, + "learning_rate": 2.8525666666666667e-05, + "loss": 0.0193, + "step": 14427 + }, + { + "epoch": 12.18581081081081, + "grad_norm": 0.1693248748779297, + "learning_rate": 2.8525333333333336e-05, + "loss": 0.0171, + "step": 14428 + }, + { + "epoch": 12.186655405405405, + "grad_norm": 0.24613270163536072, + "learning_rate": 2.8525e-05, + "loss": 0.0202, + "step": 14429 + }, + { + "epoch": 12.1875, + "grad_norm": 0.297497421503067, + "learning_rate": 2.8524666666666668e-05, + "loss": 0.0247, + "step": 14430 + }, + { + "epoch": 12.188344594594595, + "grad_norm": 0.23886801302433014, + "learning_rate": 2.8524333333333334e-05, + "loss": 0.0148, + "step": 14431 + }, + { + "epoch": 12.18918918918919, + "grad_norm": 0.46906253695487976, + "learning_rate": 2.8524e-05, + "loss": 0.0215, + "step": 14432 + }, + { + "epoch": 12.190033783783784, + "grad_norm": 0.26248669624328613, + "learning_rate": 2.852366666666667e-05, + "loss": 0.0171, + "step": 14433 + }, + { + "epoch": 12.190878378378379, + "grad_norm": 0.22219252586364746, + "learning_rate": 2.8523333333333335e-05, + "loss": 0.0161, + "step": 14434 + }, + { + "epoch": 12.191722972972974, + "grad_norm": 0.20952804386615753, + "learning_rate": 2.8523e-05, + "loss": 0.0212, + "step": 14435 + }, + { + "epoch": 12.192567567567568, + "grad_norm": 0.27840641140937805, + "learning_rate": 2.8522666666666667e-05, + "loss": 0.0175, + "step": 14436 + }, + { + "epoch": 12.193412162162161, + "grad_norm": 0.23914293944835663, + "learning_rate": 2.8522333333333333e-05, + "loss": 0.0172, + "step": 14437 + }, + { + "epoch": 12.194256756756756, + "grad_norm": 0.1710580587387085, + "learning_rate": 2.8522e-05, + "loss": 0.0114, + "step": 14438 + }, + { + "epoch": 12.19510135135135, + "grad_norm": 0.8229402899742126, + "learning_rate": 2.852166666666667e-05, + "loss": 0.0404, + "step": 14439 + }, + { + "epoch": 12.195945945945946, + "grad_norm": 0.536679208278656, + "learning_rate": 2.8521333333333334e-05, + "loss": 0.0236, + "step": 14440 + }, + { + "epoch": 12.19679054054054, + "grad_norm": 0.5787582397460938, + "learning_rate": 2.8521e-05, + "loss": 0.0689, + "step": 14441 + }, + { + "epoch": 12.197635135135135, + "grad_norm": 0.16629810631275177, + "learning_rate": 2.852066666666667e-05, + "loss": 0.0094, + "step": 14442 + }, + { + "epoch": 12.19847972972973, + "grad_norm": 0.4979732632637024, + "learning_rate": 2.8520333333333332e-05, + "loss": 0.0334, + "step": 14443 + }, + { + "epoch": 12.199324324324325, + "grad_norm": 0.3520967662334442, + "learning_rate": 2.852e-05, + "loss": 0.0272, + "step": 14444 + }, + { + "epoch": 12.20016891891892, + "grad_norm": 0.3031921684741974, + "learning_rate": 2.8519666666666667e-05, + "loss": 0.0184, + "step": 14445 + }, + { + "epoch": 12.201013513513514, + "grad_norm": 0.3328046202659607, + "learning_rate": 2.8519333333333333e-05, + "loss": 0.024, + "step": 14446 + }, + { + "epoch": 12.201858108108109, + "grad_norm": 0.4772268533706665, + "learning_rate": 2.8519e-05, + "loss": 0.0151, + "step": 14447 + }, + { + "epoch": 12.202702702702704, + "grad_norm": 0.3297695517539978, + "learning_rate": 2.851866666666667e-05, + "loss": 0.0266, + "step": 14448 + }, + { + "epoch": 12.203547297297296, + "grad_norm": 0.25031185150146484, + "learning_rate": 2.8518333333333334e-05, + "loss": 0.0137, + "step": 14449 + }, + { + "epoch": 12.204391891891891, + "grad_norm": 0.5200149416923523, + "learning_rate": 2.8518e-05, + "loss": 0.0207, + "step": 14450 + }, + { + "epoch": 12.205236486486486, + "grad_norm": 0.3001696765422821, + "learning_rate": 2.851766666666667e-05, + "loss": 0.0234, + "step": 14451 + }, + { + "epoch": 12.20608108108108, + "grad_norm": 0.34436917304992676, + "learning_rate": 2.8517333333333332e-05, + "loss": 0.0173, + "step": 14452 + }, + { + "epoch": 12.206925675675675, + "grad_norm": 0.2732868492603302, + "learning_rate": 2.8517e-05, + "loss": 0.023, + "step": 14453 + }, + { + "epoch": 12.20777027027027, + "grad_norm": 0.4070006310939789, + "learning_rate": 2.8516666666666668e-05, + "loss": 0.0208, + "step": 14454 + }, + { + "epoch": 12.208614864864865, + "grad_norm": 0.4121190905570984, + "learning_rate": 2.8516333333333333e-05, + "loss": 0.0222, + "step": 14455 + }, + { + "epoch": 12.20945945945946, + "grad_norm": 0.46038565039634705, + "learning_rate": 2.8516e-05, + "loss": 0.0302, + "step": 14456 + }, + { + "epoch": 12.210304054054054, + "grad_norm": 0.6788070201873779, + "learning_rate": 2.851566666666667e-05, + "loss": 0.0312, + "step": 14457 + }, + { + "epoch": 12.21114864864865, + "grad_norm": 2.1550047397613525, + "learning_rate": 2.8515333333333335e-05, + "loss": 0.0689, + "step": 14458 + }, + { + "epoch": 12.211993243243244, + "grad_norm": 0.45407262444496155, + "learning_rate": 2.8515e-05, + "loss": 0.1774, + "step": 14459 + }, + { + "epoch": 12.212837837837839, + "grad_norm": 0.492054283618927, + "learning_rate": 2.851466666666667e-05, + "loss": 0.086, + "step": 14460 + }, + { + "epoch": 12.213682432432432, + "grad_norm": 0.2637867331504822, + "learning_rate": 2.8514333333333332e-05, + "loss": 0.044, + "step": 14461 + }, + { + "epoch": 12.214527027027026, + "grad_norm": 0.2921176254749298, + "learning_rate": 2.8514000000000002e-05, + "loss": 0.0426, + "step": 14462 + }, + { + "epoch": 12.215371621621621, + "grad_norm": 0.20732854306697845, + "learning_rate": 2.8513666666666668e-05, + "loss": 0.0257, + "step": 14463 + }, + { + "epoch": 12.216216216216216, + "grad_norm": 0.23703984916210175, + "learning_rate": 2.8513333333333334e-05, + "loss": 0.0282, + "step": 14464 + }, + { + "epoch": 12.21706081081081, + "grad_norm": 0.42260071635246277, + "learning_rate": 2.8513000000000003e-05, + "loss": 0.052, + "step": 14465 + }, + { + "epoch": 12.217905405405405, + "grad_norm": 0.239018514752388, + "learning_rate": 2.8512666666666666e-05, + "loss": 0.0181, + "step": 14466 + }, + { + "epoch": 12.21875, + "grad_norm": 0.2663419246673584, + "learning_rate": 2.8512333333333335e-05, + "loss": 0.0273, + "step": 14467 + }, + { + "epoch": 12.219594594594595, + "grad_norm": 0.19876021146774292, + "learning_rate": 2.8512e-05, + "loss": 0.0233, + "step": 14468 + }, + { + "epoch": 12.22043918918919, + "grad_norm": 0.280319482088089, + "learning_rate": 2.8511666666666667e-05, + "loss": 0.0123, + "step": 14469 + }, + { + "epoch": 12.221283783783784, + "grad_norm": 0.33140528202056885, + "learning_rate": 2.8511333333333333e-05, + "loss": 0.026, + "step": 14470 + }, + { + "epoch": 12.222128378378379, + "grad_norm": 0.2239091843366623, + "learning_rate": 2.8511000000000002e-05, + "loss": 0.0199, + "step": 14471 + }, + { + "epoch": 12.222972972972974, + "grad_norm": 0.24956020712852478, + "learning_rate": 2.8510666666666665e-05, + "loss": 0.0278, + "step": 14472 + }, + { + "epoch": 12.223817567567568, + "grad_norm": 0.22672539949417114, + "learning_rate": 2.8510333333333334e-05, + "loss": 0.0377, + "step": 14473 + }, + { + "epoch": 12.224662162162161, + "grad_norm": 0.24113476276397705, + "learning_rate": 2.8510000000000003e-05, + "loss": 0.0235, + "step": 14474 + }, + { + "epoch": 12.225506756756756, + "grad_norm": 0.2572963535785675, + "learning_rate": 2.8509666666666666e-05, + "loss": 0.0183, + "step": 14475 + }, + { + "epoch": 12.22635135135135, + "grad_norm": 0.2257220596075058, + "learning_rate": 2.8509333333333335e-05, + "loss": 0.0141, + "step": 14476 + }, + { + "epoch": 12.227195945945946, + "grad_norm": 0.17972882091999054, + "learning_rate": 2.8509e-05, + "loss": 0.0181, + "step": 14477 + }, + { + "epoch": 12.22804054054054, + "grad_norm": 0.37387338280677795, + "learning_rate": 2.8508666666666667e-05, + "loss": 0.017, + "step": 14478 + }, + { + "epoch": 12.228885135135135, + "grad_norm": 0.24213413894176483, + "learning_rate": 2.8508333333333333e-05, + "loss": 0.0211, + "step": 14479 + }, + { + "epoch": 12.22972972972973, + "grad_norm": 0.401499480009079, + "learning_rate": 2.8508000000000002e-05, + "loss": 0.0147, + "step": 14480 + }, + { + "epoch": 12.230574324324325, + "grad_norm": 0.2420925348997116, + "learning_rate": 2.8507666666666665e-05, + "loss": 0.0286, + "step": 14481 + }, + { + "epoch": 12.23141891891892, + "grad_norm": 0.44811245799064636, + "learning_rate": 2.8507333333333334e-05, + "loss": 0.0178, + "step": 14482 + }, + { + "epoch": 12.232263513513514, + "grad_norm": 0.28852784633636475, + "learning_rate": 2.8507000000000003e-05, + "loss": 0.0188, + "step": 14483 + }, + { + "epoch": 12.233108108108109, + "grad_norm": 0.3208601176738739, + "learning_rate": 2.8506666666666666e-05, + "loss": 0.022, + "step": 14484 + }, + { + "epoch": 12.233952702702704, + "grad_norm": 0.20564472675323486, + "learning_rate": 2.8506333333333335e-05, + "loss": 0.0099, + "step": 14485 + }, + { + "epoch": 12.234797297297296, + "grad_norm": 0.26423123478889465, + "learning_rate": 2.8506e-05, + "loss": 0.0206, + "step": 14486 + }, + { + "epoch": 12.235641891891891, + "grad_norm": 0.5247923731803894, + "learning_rate": 2.8505666666666667e-05, + "loss": 0.0204, + "step": 14487 + }, + { + "epoch": 12.236486486486486, + "grad_norm": 0.21818062663078308, + "learning_rate": 2.8505333333333333e-05, + "loss": 0.012, + "step": 14488 + }, + { + "epoch": 12.23733108108108, + "grad_norm": 0.16976988315582275, + "learning_rate": 2.8505000000000002e-05, + "loss": 0.0134, + "step": 14489 + }, + { + "epoch": 12.238175675675675, + "grad_norm": 0.3244425356388092, + "learning_rate": 2.8504666666666668e-05, + "loss": 0.0293, + "step": 14490 + }, + { + "epoch": 12.23902027027027, + "grad_norm": 0.2270132601261139, + "learning_rate": 2.8504333333333334e-05, + "loss": 0.0138, + "step": 14491 + }, + { + "epoch": 12.239864864864865, + "grad_norm": 0.47697246074676514, + "learning_rate": 2.8504000000000004e-05, + "loss": 0.0187, + "step": 14492 + }, + { + "epoch": 12.24070945945946, + "grad_norm": 0.3218161463737488, + "learning_rate": 2.8503666666666666e-05, + "loss": 0.0165, + "step": 14493 + }, + { + "epoch": 12.241554054054054, + "grad_norm": 0.387844443321228, + "learning_rate": 2.8503333333333335e-05, + "loss": 0.0191, + "step": 14494 + }, + { + "epoch": 12.24239864864865, + "grad_norm": 0.4465315341949463, + "learning_rate": 2.8502999999999998e-05, + "loss": 0.0172, + "step": 14495 + }, + { + "epoch": 12.243243243243244, + "grad_norm": 0.5282753705978394, + "learning_rate": 2.8502666666666667e-05, + "loss": 0.0809, + "step": 14496 + }, + { + "epoch": 12.244087837837839, + "grad_norm": 0.38395369052886963, + "learning_rate": 2.8502333333333333e-05, + "loss": 0.0195, + "step": 14497 + }, + { + "epoch": 12.244932432432432, + "grad_norm": 0.17470704019069672, + "learning_rate": 2.8502e-05, + "loss": 0.0092, + "step": 14498 + }, + { + "epoch": 12.245777027027026, + "grad_norm": 0.3644154667854309, + "learning_rate": 2.850166666666667e-05, + "loss": 0.0207, + "step": 14499 + }, + { + "epoch": 12.246621621621621, + "grad_norm": 0.46150171756744385, + "learning_rate": 2.8501333333333334e-05, + "loss": 0.0168, + "step": 14500 + }, + { + "epoch": 12.247466216216216, + "grad_norm": 0.25324904918670654, + "learning_rate": 2.8501e-05, + "loss": 0.0157, + "step": 14501 + }, + { + "epoch": 12.24831081081081, + "grad_norm": 0.3716530501842499, + "learning_rate": 2.8500666666666666e-05, + "loss": 0.0111, + "step": 14502 + }, + { + "epoch": 12.249155405405405, + "grad_norm": 0.32819294929504395, + "learning_rate": 2.8500333333333336e-05, + "loss": 0.0197, + "step": 14503 + }, + { + "epoch": 12.25, + "grad_norm": 0.2997418940067291, + "learning_rate": 2.8499999999999998e-05, + "loss": 0.0232, + "step": 14504 + }, + { + "epoch": 12.250844594594595, + "grad_norm": 0.4901242256164551, + "learning_rate": 2.8499666666666668e-05, + "loss": 0.0198, + "step": 14505 + }, + { + "epoch": 12.25168918918919, + "grad_norm": 0.5407283902168274, + "learning_rate": 2.8499333333333333e-05, + "loss": 0.0206, + "step": 14506 + }, + { + "epoch": 12.252533783783784, + "grad_norm": 0.4850989580154419, + "learning_rate": 2.8499e-05, + "loss": 0.0343, + "step": 14507 + }, + { + "epoch": 12.253378378378379, + "grad_norm": 0.575343668460846, + "learning_rate": 2.849866666666667e-05, + "loss": 0.0316, + "step": 14508 + }, + { + "epoch": 12.254222972972974, + "grad_norm": 0.7254335284233093, + "learning_rate": 2.8498333333333335e-05, + "loss": 0.2425, + "step": 14509 + }, + { + "epoch": 12.255067567567568, + "grad_norm": 0.5207211971282959, + "learning_rate": 2.8498e-05, + "loss": 0.1152, + "step": 14510 + }, + { + "epoch": 12.255912162162161, + "grad_norm": 0.3439999520778656, + "learning_rate": 2.8497666666666667e-05, + "loss": 0.0493, + "step": 14511 + }, + { + "epoch": 12.256756756756756, + "grad_norm": 0.36219751834869385, + "learning_rate": 2.8497333333333336e-05, + "loss": 0.0507, + "step": 14512 + }, + { + "epoch": 12.25760135135135, + "grad_norm": 0.37118658423423767, + "learning_rate": 2.8497e-05, + "loss": 0.0313, + "step": 14513 + }, + { + "epoch": 12.258445945945946, + "grad_norm": 0.2219746857881546, + "learning_rate": 2.8496666666666668e-05, + "loss": 0.0192, + "step": 14514 + }, + { + "epoch": 12.25929054054054, + "grad_norm": 0.8363267779350281, + "learning_rate": 2.8496333333333337e-05, + "loss": 0.0265, + "step": 14515 + }, + { + "epoch": 12.260135135135135, + "grad_norm": 0.2026151716709137, + "learning_rate": 2.8496e-05, + "loss": 0.0195, + "step": 14516 + }, + { + "epoch": 12.26097972972973, + "grad_norm": 0.4186707139015198, + "learning_rate": 2.849566666666667e-05, + "loss": 0.0237, + "step": 14517 + }, + { + "epoch": 12.261824324324325, + "grad_norm": 0.20585595071315765, + "learning_rate": 2.8495333333333335e-05, + "loss": 0.0182, + "step": 14518 + }, + { + "epoch": 12.26266891891892, + "grad_norm": 0.3175782561302185, + "learning_rate": 2.8495e-05, + "loss": 0.0247, + "step": 14519 + }, + { + "epoch": 12.263513513513514, + "grad_norm": 0.20383071899414062, + "learning_rate": 2.8494666666666667e-05, + "loss": 0.0144, + "step": 14520 + }, + { + "epoch": 12.264358108108109, + "grad_norm": 0.30210763216018677, + "learning_rate": 2.8494333333333336e-05, + "loss": 0.0193, + "step": 14521 + }, + { + "epoch": 12.265202702702704, + "grad_norm": 0.23196426033973694, + "learning_rate": 2.8494e-05, + "loss": 0.0176, + "step": 14522 + }, + { + "epoch": 12.266047297297296, + "grad_norm": 0.28667253255844116, + "learning_rate": 2.8493666666666668e-05, + "loss": 0.0157, + "step": 14523 + }, + { + "epoch": 12.266891891891891, + "grad_norm": 0.23349736630916595, + "learning_rate": 2.8493333333333337e-05, + "loss": 0.02, + "step": 14524 + }, + { + "epoch": 12.267736486486486, + "grad_norm": 0.172597274184227, + "learning_rate": 2.8493e-05, + "loss": 0.0097, + "step": 14525 + }, + { + "epoch": 12.26858108108108, + "grad_norm": 0.3873598575592041, + "learning_rate": 2.849266666666667e-05, + "loss": 0.0302, + "step": 14526 + }, + { + "epoch": 12.269425675675675, + "grad_norm": 0.38571545481681824, + "learning_rate": 2.849233333333333e-05, + "loss": 0.0208, + "step": 14527 + }, + { + "epoch": 12.27027027027027, + "grad_norm": 0.24993900954723358, + "learning_rate": 2.8492e-05, + "loss": 0.022, + "step": 14528 + }, + { + "epoch": 12.271114864864865, + "grad_norm": 1.184820294380188, + "learning_rate": 2.8491666666666667e-05, + "loss": 0.0215, + "step": 14529 + }, + { + "epoch": 12.27195945945946, + "grad_norm": 0.5520285964012146, + "learning_rate": 2.8491333333333333e-05, + "loss": 0.029, + "step": 14530 + }, + { + "epoch": 12.272804054054054, + "grad_norm": 0.3321287930011749, + "learning_rate": 2.8491e-05, + "loss": 0.0388, + "step": 14531 + }, + { + "epoch": 12.27364864864865, + "grad_norm": 0.4048806428909302, + "learning_rate": 2.8490666666666668e-05, + "loss": 0.0299, + "step": 14532 + }, + { + "epoch": 12.274493243243244, + "grad_norm": 0.24803151190280914, + "learning_rate": 2.8490333333333334e-05, + "loss": 0.0242, + "step": 14533 + }, + { + "epoch": 12.275337837837839, + "grad_norm": 0.2690000832080841, + "learning_rate": 2.849e-05, + "loss": 0.0189, + "step": 14534 + }, + { + "epoch": 12.276182432432432, + "grad_norm": 0.1998991072177887, + "learning_rate": 2.848966666666667e-05, + "loss": 0.0163, + "step": 14535 + }, + { + "epoch": 12.277027027027026, + "grad_norm": 0.4339063763618469, + "learning_rate": 2.8489333333333332e-05, + "loss": 0.0251, + "step": 14536 + }, + { + "epoch": 12.277871621621621, + "grad_norm": 1.1063495874404907, + "learning_rate": 2.8489e-05, + "loss": 0.0312, + "step": 14537 + }, + { + "epoch": 12.278716216216216, + "grad_norm": 0.17656433582305908, + "learning_rate": 2.8488666666666667e-05, + "loss": 0.0135, + "step": 14538 + }, + { + "epoch": 12.27956081081081, + "grad_norm": 0.25560054183006287, + "learning_rate": 2.8488333333333333e-05, + "loss": 0.0219, + "step": 14539 + }, + { + "epoch": 12.280405405405405, + "grad_norm": 0.27793681621551514, + "learning_rate": 2.8488000000000002e-05, + "loss": 0.018, + "step": 14540 + }, + { + "epoch": 12.28125, + "grad_norm": 0.4825434684753418, + "learning_rate": 2.8487666666666668e-05, + "loss": 0.0183, + "step": 14541 + }, + { + "epoch": 12.282094594594595, + "grad_norm": 0.5109711289405823, + "learning_rate": 2.8487333333333334e-05, + "loss": 0.0151, + "step": 14542 + }, + { + "epoch": 12.28293918918919, + "grad_norm": 0.20029275119304657, + "learning_rate": 2.8487e-05, + "loss": 0.0165, + "step": 14543 + }, + { + "epoch": 12.283783783783784, + "grad_norm": 0.5149775743484497, + "learning_rate": 2.848666666666667e-05, + "loss": 0.0215, + "step": 14544 + }, + { + "epoch": 12.284628378378379, + "grad_norm": 0.5369022488594055, + "learning_rate": 2.8486333333333332e-05, + "loss": 0.0219, + "step": 14545 + }, + { + "epoch": 12.285472972972974, + "grad_norm": 0.3411664664745331, + "learning_rate": 2.8486e-05, + "loss": 0.0231, + "step": 14546 + }, + { + "epoch": 12.286317567567568, + "grad_norm": 0.5738191604614258, + "learning_rate": 2.8485666666666667e-05, + "loss": 0.0223, + "step": 14547 + }, + { + "epoch": 12.287162162162161, + "grad_norm": 0.23676031827926636, + "learning_rate": 2.8485333333333333e-05, + "loss": 0.0191, + "step": 14548 + }, + { + "epoch": 12.288006756756756, + "grad_norm": 0.258929044008255, + "learning_rate": 2.8485000000000003e-05, + "loss": 0.0204, + "step": 14549 + }, + { + "epoch": 12.28885135135135, + "grad_norm": 0.42956477403640747, + "learning_rate": 2.848466666666667e-05, + "loss": 0.0214, + "step": 14550 + }, + { + "epoch": 12.289695945945946, + "grad_norm": 0.28503426909446716, + "learning_rate": 2.8484333333333334e-05, + "loss": 0.022, + "step": 14551 + }, + { + "epoch": 12.29054054054054, + "grad_norm": 0.30444109439849854, + "learning_rate": 2.8484e-05, + "loss": 0.0084, + "step": 14552 + }, + { + "epoch": 12.291385135135135, + "grad_norm": 0.8349081873893738, + "learning_rate": 2.848366666666667e-05, + "loss": 0.0355, + "step": 14553 + }, + { + "epoch": 12.29222972972973, + "grad_norm": 1.6700788736343384, + "learning_rate": 2.8483333333333332e-05, + "loss": 0.0355, + "step": 14554 + }, + { + "epoch": 12.293074324324325, + "grad_norm": 0.6037343144416809, + "learning_rate": 2.8483e-05, + "loss": 0.0286, + "step": 14555 + }, + { + "epoch": 12.29391891891892, + "grad_norm": 0.6371506452560425, + "learning_rate": 2.8482666666666664e-05, + "loss": 0.0348, + "step": 14556 + }, + { + "epoch": 12.294763513513514, + "grad_norm": 0.39724791049957275, + "learning_rate": 2.8482333333333333e-05, + "loss": 0.0189, + "step": 14557 + }, + { + "epoch": 12.295608108108109, + "grad_norm": 1.2751173973083496, + "learning_rate": 2.8482000000000003e-05, + "loss": 0.0665, + "step": 14558 + }, + { + "epoch": 12.296452702702704, + "grad_norm": 0.6349449157714844, + "learning_rate": 2.8481666666666665e-05, + "loss": 0.1838, + "step": 14559 + }, + { + "epoch": 12.297297297297296, + "grad_norm": 0.4897264242172241, + "learning_rate": 2.8481333333333335e-05, + "loss": 0.1154, + "step": 14560 + }, + { + "epoch": 12.298141891891891, + "grad_norm": 0.3848241865634918, + "learning_rate": 2.8481e-05, + "loss": 0.061, + "step": 14561 + }, + { + "epoch": 12.298986486486486, + "grad_norm": 0.6607927680015564, + "learning_rate": 2.8480666666666667e-05, + "loss": 0.0588, + "step": 14562 + }, + { + "epoch": 12.29983108108108, + "grad_norm": 0.4833548665046692, + "learning_rate": 2.8480333333333332e-05, + "loss": 0.0479, + "step": 14563 + }, + { + "epoch": 12.300675675675675, + "grad_norm": 0.29603004455566406, + "learning_rate": 2.8480000000000002e-05, + "loss": 0.0273, + "step": 14564 + }, + { + "epoch": 12.30152027027027, + "grad_norm": 0.2310435026884079, + "learning_rate": 2.8479666666666668e-05, + "loss": 0.0281, + "step": 14565 + }, + { + "epoch": 12.302364864864865, + "grad_norm": 0.29279497265815735, + "learning_rate": 2.8479333333333334e-05, + "loss": 0.0304, + "step": 14566 + }, + { + "epoch": 12.30320945945946, + "grad_norm": 0.33505645394325256, + "learning_rate": 2.8479000000000003e-05, + "loss": 0.0161, + "step": 14567 + }, + { + "epoch": 12.304054054054054, + "grad_norm": 0.2500156760215759, + "learning_rate": 2.8478666666666666e-05, + "loss": 0.0179, + "step": 14568 + }, + { + "epoch": 12.30489864864865, + "grad_norm": 0.49397793412208557, + "learning_rate": 2.8478333333333335e-05, + "loss": 0.0154, + "step": 14569 + }, + { + "epoch": 12.305743243243244, + "grad_norm": 0.2990396320819855, + "learning_rate": 2.8478e-05, + "loss": 0.0309, + "step": 14570 + }, + { + "epoch": 12.306587837837839, + "grad_norm": 0.33129847049713135, + "learning_rate": 2.8477666666666667e-05, + "loss": 0.0158, + "step": 14571 + }, + { + "epoch": 12.307432432432432, + "grad_norm": 0.3091185390949249, + "learning_rate": 2.8477333333333333e-05, + "loss": 0.0258, + "step": 14572 + }, + { + "epoch": 12.308277027027026, + "grad_norm": 0.3200698792934418, + "learning_rate": 2.8477000000000002e-05, + "loss": 0.027, + "step": 14573 + }, + { + "epoch": 12.309121621621621, + "grad_norm": 0.18855223059654236, + "learning_rate": 2.8476666666666668e-05, + "loss": 0.0179, + "step": 14574 + }, + { + "epoch": 12.309966216216216, + "grad_norm": 0.2851942479610443, + "learning_rate": 2.8476333333333334e-05, + "loss": 0.0213, + "step": 14575 + }, + { + "epoch": 12.31081081081081, + "grad_norm": 0.21169127523899078, + "learning_rate": 2.8476000000000003e-05, + "loss": 0.0235, + "step": 14576 + }, + { + "epoch": 12.311655405405405, + "grad_norm": 0.1601058542728424, + "learning_rate": 2.8475666666666666e-05, + "loss": 0.0132, + "step": 14577 + }, + { + "epoch": 12.3125, + "grad_norm": 0.9887166023254395, + "learning_rate": 2.8475333333333335e-05, + "loss": 0.0301, + "step": 14578 + }, + { + "epoch": 12.313344594594595, + "grad_norm": 0.6580596566200256, + "learning_rate": 2.8475e-05, + "loss": 0.0267, + "step": 14579 + }, + { + "epoch": 12.31418918918919, + "grad_norm": 0.2836493253707886, + "learning_rate": 2.8474666666666667e-05, + "loss": 0.0278, + "step": 14580 + }, + { + "epoch": 12.315033783783784, + "grad_norm": 0.5062731504440308, + "learning_rate": 2.8474333333333333e-05, + "loss": 0.032, + "step": 14581 + }, + { + "epoch": 12.315878378378379, + "grad_norm": 0.2239212840795517, + "learning_rate": 2.8474000000000002e-05, + "loss": 0.017, + "step": 14582 + }, + { + "epoch": 12.316722972972974, + "grad_norm": 0.24102622270584106, + "learning_rate": 2.8473666666666668e-05, + "loss": 0.0221, + "step": 14583 + }, + { + "epoch": 12.317567567567568, + "grad_norm": 0.38254842162132263, + "learning_rate": 2.8473333333333334e-05, + "loss": 0.02, + "step": 14584 + }, + { + "epoch": 12.318412162162161, + "grad_norm": 0.28766822814941406, + "learning_rate": 2.8473000000000003e-05, + "loss": 0.023, + "step": 14585 + }, + { + "epoch": 12.319256756756756, + "grad_norm": 0.3750613033771515, + "learning_rate": 2.8472666666666666e-05, + "loss": 0.0227, + "step": 14586 + }, + { + "epoch": 12.32010135135135, + "grad_norm": 0.25464728474617004, + "learning_rate": 2.8472333333333335e-05, + "loss": 0.0151, + "step": 14587 + }, + { + "epoch": 12.320945945945946, + "grad_norm": 0.1990385353565216, + "learning_rate": 2.8471999999999998e-05, + "loss": 0.0168, + "step": 14588 + }, + { + "epoch": 12.32179054054054, + "grad_norm": 0.2582668662071228, + "learning_rate": 2.8471666666666667e-05, + "loss": 0.0152, + "step": 14589 + }, + { + "epoch": 12.322635135135135, + "grad_norm": 0.2606586813926697, + "learning_rate": 2.8471333333333336e-05, + "loss": 0.0229, + "step": 14590 + }, + { + "epoch": 12.32347972972973, + "grad_norm": 0.24020442366600037, + "learning_rate": 2.8471e-05, + "loss": 0.0109, + "step": 14591 + }, + { + "epoch": 12.324324324324325, + "grad_norm": 0.24420596659183502, + "learning_rate": 2.8470666666666668e-05, + "loss": 0.0239, + "step": 14592 + }, + { + "epoch": 12.32516891891892, + "grad_norm": 0.4709406793117523, + "learning_rate": 2.8470333333333334e-05, + "loss": 0.0679, + "step": 14593 + }, + { + "epoch": 12.326013513513514, + "grad_norm": 0.2477836310863495, + "learning_rate": 2.847e-05, + "loss": 0.0098, + "step": 14594 + }, + { + "epoch": 12.326858108108109, + "grad_norm": 0.4020811915397644, + "learning_rate": 2.8469666666666666e-05, + "loss": 0.0191, + "step": 14595 + }, + { + "epoch": 12.327702702702704, + "grad_norm": 0.567537248134613, + "learning_rate": 2.8469333333333335e-05, + "loss": 0.0329, + "step": 14596 + }, + { + "epoch": 12.328547297297296, + "grad_norm": 0.3389328420162201, + "learning_rate": 2.8468999999999998e-05, + "loss": 0.0161, + "step": 14597 + }, + { + "epoch": 12.329391891891891, + "grad_norm": 0.1916974037885666, + "learning_rate": 2.8468666666666667e-05, + "loss": 0.0133, + "step": 14598 + }, + { + "epoch": 12.330236486486486, + "grad_norm": 0.35793501138687134, + "learning_rate": 2.8468333333333337e-05, + "loss": 0.0178, + "step": 14599 + }, + { + "epoch": 12.33108108108108, + "grad_norm": 0.43108296394348145, + "learning_rate": 2.8468e-05, + "loss": 0.0285, + "step": 14600 + }, + { + "epoch": 12.331925675675675, + "grad_norm": 0.3006189167499542, + "learning_rate": 2.846766666666667e-05, + "loss": 0.0252, + "step": 14601 + }, + { + "epoch": 12.33277027027027, + "grad_norm": 0.46748867630958557, + "learning_rate": 2.8467333333333334e-05, + "loss": 0.0205, + "step": 14602 + }, + { + "epoch": 12.333614864864865, + "grad_norm": 0.24035878479480743, + "learning_rate": 2.8467e-05, + "loss": 0.014, + "step": 14603 + }, + { + "epoch": 12.33445945945946, + "grad_norm": 0.4616272747516632, + "learning_rate": 2.8466666666666666e-05, + "loss": 0.0278, + "step": 14604 + }, + { + "epoch": 12.335304054054054, + "grad_norm": 0.4165104031562805, + "learning_rate": 2.8466333333333336e-05, + "loss": 0.0156, + "step": 14605 + }, + { + "epoch": 12.33614864864865, + "grad_norm": 0.38804465532302856, + "learning_rate": 2.8465999999999998e-05, + "loss": 0.0325, + "step": 14606 + }, + { + "epoch": 12.336993243243244, + "grad_norm": 0.7269166707992554, + "learning_rate": 2.8465666666666667e-05, + "loss": 0.0332, + "step": 14607 + }, + { + "epoch": 12.337837837837839, + "grad_norm": 0.5561630129814148, + "learning_rate": 2.8465333333333337e-05, + "loss": 0.0237, + "step": 14608 + }, + { + "epoch": 12.338682432432432, + "grad_norm": 0.7953087687492371, + "learning_rate": 2.8465e-05, + "loss": 0.199, + "step": 14609 + }, + { + "epoch": 12.339527027027026, + "grad_norm": 0.6942903995513916, + "learning_rate": 2.846466666666667e-05, + "loss": 0.122, + "step": 14610 + }, + { + "epoch": 12.340371621621621, + "grad_norm": 0.3961504101753235, + "learning_rate": 2.8464333333333335e-05, + "loss": 0.0821, + "step": 14611 + }, + { + "epoch": 12.341216216216216, + "grad_norm": 0.26473715901374817, + "learning_rate": 2.8464e-05, + "loss": 0.0272, + "step": 14612 + }, + { + "epoch": 12.34206081081081, + "grad_norm": 0.4304613173007965, + "learning_rate": 2.8463666666666666e-05, + "loss": 0.0398, + "step": 14613 + }, + { + "epoch": 12.342905405405405, + "grad_norm": 0.27029454708099365, + "learning_rate": 2.8463333333333336e-05, + "loss": 0.0245, + "step": 14614 + }, + { + "epoch": 12.34375, + "grad_norm": 0.3027857840061188, + "learning_rate": 2.8463000000000002e-05, + "loss": 0.0162, + "step": 14615 + }, + { + "epoch": 12.344594594594595, + "grad_norm": 0.2503134608268738, + "learning_rate": 2.8462666666666668e-05, + "loss": 0.0244, + "step": 14616 + }, + { + "epoch": 12.34543918918919, + "grad_norm": 0.45462456345558167, + "learning_rate": 2.8462333333333334e-05, + "loss": 0.0258, + "step": 14617 + }, + { + "epoch": 12.346283783783784, + "grad_norm": 0.20808672904968262, + "learning_rate": 2.8462e-05, + "loss": 0.0174, + "step": 14618 + }, + { + "epoch": 12.347128378378379, + "grad_norm": 0.7127395272254944, + "learning_rate": 2.846166666666667e-05, + "loss": 0.0264, + "step": 14619 + }, + { + "epoch": 12.347972972972974, + "grad_norm": 0.26699188351631165, + "learning_rate": 2.846133333333333e-05, + "loss": 0.0174, + "step": 14620 + }, + { + "epoch": 12.348817567567568, + "grad_norm": 0.40177664160728455, + "learning_rate": 2.8461e-05, + "loss": 0.0242, + "step": 14621 + }, + { + "epoch": 12.349662162162161, + "grad_norm": 0.22611117362976074, + "learning_rate": 2.8460666666666667e-05, + "loss": 0.0164, + "step": 14622 + }, + { + "epoch": 12.350506756756756, + "grad_norm": 0.3467330038547516, + "learning_rate": 2.8460333333333333e-05, + "loss": 0.028, + "step": 14623 + }, + { + "epoch": 12.35135135135135, + "grad_norm": 0.37090224027633667, + "learning_rate": 2.8460000000000002e-05, + "loss": 0.0191, + "step": 14624 + }, + { + "epoch": 12.352195945945946, + "grad_norm": 0.3148168921470642, + "learning_rate": 2.8459666666666668e-05, + "loss": 0.0238, + "step": 14625 + }, + { + "epoch": 12.35304054054054, + "grad_norm": 0.2369658350944519, + "learning_rate": 2.8459333333333334e-05, + "loss": 0.0192, + "step": 14626 + }, + { + "epoch": 12.353885135135135, + "grad_norm": 0.41049158573150635, + "learning_rate": 2.8459e-05, + "loss": 0.0482, + "step": 14627 + }, + { + "epoch": 12.35472972972973, + "grad_norm": 0.19173862040042877, + "learning_rate": 2.845866666666667e-05, + "loss": 0.0184, + "step": 14628 + }, + { + "epoch": 12.355574324324325, + "grad_norm": 0.27451416850090027, + "learning_rate": 2.845833333333333e-05, + "loss": 0.0157, + "step": 14629 + }, + { + "epoch": 12.35641891891892, + "grad_norm": 0.2129741758108139, + "learning_rate": 2.8458e-05, + "loss": 0.0249, + "step": 14630 + }, + { + "epoch": 12.357263513513514, + "grad_norm": 0.17975960671901703, + "learning_rate": 2.8457666666666667e-05, + "loss": 0.0152, + "step": 14631 + }, + { + "epoch": 12.358108108108109, + "grad_norm": 0.2877718508243561, + "learning_rate": 2.8457333333333333e-05, + "loss": 0.021, + "step": 14632 + }, + { + "epoch": 12.358952702702704, + "grad_norm": 0.4016983211040497, + "learning_rate": 2.8457000000000002e-05, + "loss": 0.0204, + "step": 14633 + }, + { + "epoch": 12.359797297297296, + "grad_norm": 0.4024302661418915, + "learning_rate": 2.8456666666666668e-05, + "loss": 0.0174, + "step": 14634 + }, + { + "epoch": 12.360641891891891, + "grad_norm": 0.16377274692058563, + "learning_rate": 2.8456333333333334e-05, + "loss": 0.0122, + "step": 14635 + }, + { + "epoch": 12.361486486486486, + "grad_norm": 0.22511790692806244, + "learning_rate": 2.8456e-05, + "loss": 0.0166, + "step": 14636 + }, + { + "epoch": 12.36233108108108, + "grad_norm": 0.2573513686656952, + "learning_rate": 2.845566666666667e-05, + "loss": 0.0293, + "step": 14637 + }, + { + "epoch": 12.363175675675675, + "grad_norm": 0.2662251889705658, + "learning_rate": 2.8455333333333332e-05, + "loss": 0.0231, + "step": 14638 + }, + { + "epoch": 12.36402027027027, + "grad_norm": 0.8085383772850037, + "learning_rate": 2.8455e-05, + "loss": 0.0248, + "step": 14639 + }, + { + "epoch": 12.364864864864865, + "grad_norm": 0.27098366618156433, + "learning_rate": 2.845466666666667e-05, + "loss": 0.0124, + "step": 14640 + }, + { + "epoch": 12.36570945945946, + "grad_norm": 0.384355753660202, + "learning_rate": 2.8454333333333333e-05, + "loss": 0.0245, + "step": 14641 + }, + { + "epoch": 12.366554054054054, + "grad_norm": 0.15167832374572754, + "learning_rate": 2.8454000000000002e-05, + "loss": 0.0074, + "step": 14642 + }, + { + "epoch": 12.36739864864865, + "grad_norm": 1.1477224826812744, + "learning_rate": 2.8453666666666668e-05, + "loss": 0.028, + "step": 14643 + }, + { + "epoch": 12.368243243243244, + "grad_norm": 0.14833973348140717, + "learning_rate": 2.8453333333333334e-05, + "loss": 0.0098, + "step": 14644 + }, + { + "epoch": 12.369087837837839, + "grad_norm": 0.49782121181488037, + "learning_rate": 2.8453e-05, + "loss": 0.072, + "step": 14645 + }, + { + "epoch": 12.369932432432432, + "grad_norm": 0.5076579451560974, + "learning_rate": 2.845266666666667e-05, + "loss": 0.0324, + "step": 14646 + }, + { + "epoch": 12.370777027027026, + "grad_norm": 0.2545683681964874, + "learning_rate": 2.8452333333333332e-05, + "loss": 0.0107, + "step": 14647 + }, + { + "epoch": 12.371621621621621, + "grad_norm": 0.2782152593135834, + "learning_rate": 2.8452e-05, + "loss": 0.0276, + "step": 14648 + }, + { + "epoch": 12.372466216216216, + "grad_norm": 0.19221216440200806, + "learning_rate": 2.8451666666666667e-05, + "loss": 0.0109, + "step": 14649 + }, + { + "epoch": 12.37331081081081, + "grad_norm": 0.5139702558517456, + "learning_rate": 2.8451333333333333e-05, + "loss": 0.0206, + "step": 14650 + }, + { + "epoch": 12.374155405405405, + "grad_norm": 0.523102879524231, + "learning_rate": 2.8451000000000003e-05, + "loss": 0.0159, + "step": 14651 + }, + { + "epoch": 12.375, + "grad_norm": 0.2790570855140686, + "learning_rate": 2.8450666666666665e-05, + "loss": 0.0191, + "step": 14652 + }, + { + "epoch": 12.375844594594595, + "grad_norm": 0.3381100594997406, + "learning_rate": 2.8450333333333334e-05, + "loss": 0.0188, + "step": 14653 + }, + { + "epoch": 12.37668918918919, + "grad_norm": 0.37306052446365356, + "learning_rate": 2.845e-05, + "loss": 0.0285, + "step": 14654 + }, + { + "epoch": 12.377533783783784, + "grad_norm": 0.2580490708351135, + "learning_rate": 2.8449666666666666e-05, + "loss": 0.0124, + "step": 14655 + }, + { + "epoch": 12.378378378378379, + "grad_norm": 0.9596828818321228, + "learning_rate": 2.8449333333333332e-05, + "loss": 0.0654, + "step": 14656 + }, + { + "epoch": 12.379222972972974, + "grad_norm": 0.6508536338806152, + "learning_rate": 2.8449e-05, + "loss": 0.0307, + "step": 14657 + }, + { + "epoch": 12.380067567567568, + "grad_norm": 1.6032030582427979, + "learning_rate": 2.8448666666666667e-05, + "loss": 0.1412, + "step": 14658 + }, + { + "epoch": 12.380912162162161, + "grad_norm": 0.9022103548049927, + "learning_rate": 2.8448333333333333e-05, + "loss": 0.1918, + "step": 14659 + }, + { + "epoch": 12.381756756756756, + "grad_norm": 0.5871270895004272, + "learning_rate": 2.8448000000000003e-05, + "loss": 0.139, + "step": 14660 + }, + { + "epoch": 12.38260135135135, + "grad_norm": 0.35619378089904785, + "learning_rate": 2.8447666666666665e-05, + "loss": 0.0707, + "step": 14661 + }, + { + "epoch": 12.383445945945946, + "grad_norm": 0.3183538615703583, + "learning_rate": 2.8447333333333335e-05, + "loss": 0.0407, + "step": 14662 + }, + { + "epoch": 12.38429054054054, + "grad_norm": 0.23301522433757782, + "learning_rate": 2.8447e-05, + "loss": 0.023, + "step": 14663 + }, + { + "epoch": 12.385135135135135, + "grad_norm": 0.26531893014907837, + "learning_rate": 2.8446666666666666e-05, + "loss": 0.0265, + "step": 14664 + }, + { + "epoch": 12.38597972972973, + "grad_norm": 0.3095424771308899, + "learning_rate": 2.8446333333333336e-05, + "loss": 0.0206, + "step": 14665 + }, + { + "epoch": 12.386824324324325, + "grad_norm": 0.3582712411880493, + "learning_rate": 2.8446000000000002e-05, + "loss": 0.0325, + "step": 14666 + }, + { + "epoch": 12.38766891891892, + "grad_norm": 0.276973158121109, + "learning_rate": 2.8445666666666668e-05, + "loss": 0.0308, + "step": 14667 + }, + { + "epoch": 12.388513513513514, + "grad_norm": 0.23073451220989227, + "learning_rate": 2.8445333333333334e-05, + "loss": 0.0203, + "step": 14668 + }, + { + "epoch": 12.389358108108109, + "grad_norm": 0.30708616971969604, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0233, + "step": 14669 + }, + { + "epoch": 12.390202702702704, + "grad_norm": 0.2171623706817627, + "learning_rate": 2.8444666666666665e-05, + "loss": 0.0191, + "step": 14670 + }, + { + "epoch": 12.391047297297296, + "grad_norm": 0.16552862524986267, + "learning_rate": 2.8444333333333335e-05, + "loss": 0.0165, + "step": 14671 + }, + { + "epoch": 12.391891891891891, + "grad_norm": 0.2501842677593231, + "learning_rate": 2.8444e-05, + "loss": 0.0213, + "step": 14672 + }, + { + "epoch": 12.392736486486486, + "grad_norm": 0.7529529333114624, + "learning_rate": 2.8443666666666667e-05, + "loss": 0.0214, + "step": 14673 + }, + { + "epoch": 12.39358108108108, + "grad_norm": 0.1818118840456009, + "learning_rate": 2.8443333333333336e-05, + "loss": 0.0167, + "step": 14674 + }, + { + "epoch": 12.394425675675675, + "grad_norm": 0.2787616550922394, + "learning_rate": 2.8443000000000002e-05, + "loss": 0.0236, + "step": 14675 + }, + { + "epoch": 12.39527027027027, + "grad_norm": 0.30337023735046387, + "learning_rate": 2.8442666666666668e-05, + "loss": 0.0293, + "step": 14676 + }, + { + "epoch": 12.396114864864865, + "grad_norm": 0.18306393921375275, + "learning_rate": 2.8442333333333334e-05, + "loss": 0.0155, + "step": 14677 + }, + { + "epoch": 12.39695945945946, + "grad_norm": 0.7256408929824829, + "learning_rate": 2.8442e-05, + "loss": 0.0959, + "step": 14678 + }, + { + "epoch": 12.397804054054054, + "grad_norm": 0.280879408121109, + "learning_rate": 2.8441666666666666e-05, + "loss": 0.0236, + "step": 14679 + }, + { + "epoch": 12.39864864864865, + "grad_norm": 0.26987046003341675, + "learning_rate": 2.8441333333333335e-05, + "loss": 0.0193, + "step": 14680 + }, + { + "epoch": 12.399493243243244, + "grad_norm": 0.16596640646457672, + "learning_rate": 2.8440999999999998e-05, + "loss": 0.0132, + "step": 14681 + }, + { + "epoch": 12.400337837837839, + "grad_norm": 0.24165864288806915, + "learning_rate": 2.8440666666666667e-05, + "loss": 0.0178, + "step": 14682 + }, + { + "epoch": 12.401182432432432, + "grad_norm": 0.17353732883930206, + "learning_rate": 2.8440333333333336e-05, + "loss": 0.0093, + "step": 14683 + }, + { + "epoch": 12.402027027027026, + "grad_norm": 0.3030380308628082, + "learning_rate": 2.844e-05, + "loss": 0.0183, + "step": 14684 + }, + { + "epoch": 12.402871621621621, + "grad_norm": 0.30274492502212524, + "learning_rate": 2.8439666666666668e-05, + "loss": 0.0362, + "step": 14685 + }, + { + "epoch": 12.403716216216216, + "grad_norm": 0.277746319770813, + "learning_rate": 2.8439333333333334e-05, + "loss": 0.0157, + "step": 14686 + }, + { + "epoch": 12.40456081081081, + "grad_norm": 0.43724507093429565, + "learning_rate": 2.8439e-05, + "loss": 0.0146, + "step": 14687 + }, + { + "epoch": 12.405405405405405, + "grad_norm": 0.2936532199382782, + "learning_rate": 2.8438666666666666e-05, + "loss": 0.0264, + "step": 14688 + }, + { + "epoch": 12.40625, + "grad_norm": 0.43989309668540955, + "learning_rate": 2.8438333333333335e-05, + "loss": 0.0238, + "step": 14689 + }, + { + "epoch": 12.407094594594595, + "grad_norm": 0.2895200550556183, + "learning_rate": 2.8438e-05, + "loss": 0.0179, + "step": 14690 + }, + { + "epoch": 12.40793918918919, + "grad_norm": 0.2161063849925995, + "learning_rate": 2.8437666666666667e-05, + "loss": 0.0177, + "step": 14691 + }, + { + "epoch": 12.408783783783784, + "grad_norm": 0.6475774049758911, + "learning_rate": 2.8437333333333336e-05, + "loss": 0.0299, + "step": 14692 + }, + { + "epoch": 12.409628378378379, + "grad_norm": 0.3971114754676819, + "learning_rate": 2.8437e-05, + "loss": 0.0282, + "step": 14693 + }, + { + "epoch": 12.410472972972974, + "grad_norm": 0.29026147723197937, + "learning_rate": 2.8436666666666668e-05, + "loss": 0.0219, + "step": 14694 + }, + { + "epoch": 12.411317567567568, + "grad_norm": 0.28736695647239685, + "learning_rate": 2.8436333333333334e-05, + "loss": 0.0156, + "step": 14695 + }, + { + "epoch": 12.412162162162161, + "grad_norm": 0.304841548204422, + "learning_rate": 2.8436e-05, + "loss": 0.0198, + "step": 14696 + }, + { + "epoch": 12.413006756756756, + "grad_norm": 0.6523225903511047, + "learning_rate": 2.8435666666666666e-05, + "loss": 0.0322, + "step": 14697 + }, + { + "epoch": 12.41385135135135, + "grad_norm": 0.7699766755104065, + "learning_rate": 2.8435333333333335e-05, + "loss": 0.035, + "step": 14698 + }, + { + "epoch": 12.414695945945946, + "grad_norm": 0.16614122688770294, + "learning_rate": 2.8435e-05, + "loss": 0.0114, + "step": 14699 + }, + { + "epoch": 12.41554054054054, + "grad_norm": 0.17303644120693207, + "learning_rate": 2.8434666666666667e-05, + "loss": 0.016, + "step": 14700 + }, + { + "epoch": 12.416385135135135, + "grad_norm": 0.49021294713020325, + "learning_rate": 2.8434333333333337e-05, + "loss": 0.0239, + "step": 14701 + }, + { + "epoch": 12.41722972972973, + "grad_norm": 0.36615628004074097, + "learning_rate": 2.8434e-05, + "loss": 0.0232, + "step": 14702 + }, + { + "epoch": 12.418074324324325, + "grad_norm": 0.331154465675354, + "learning_rate": 2.843366666666667e-05, + "loss": 0.022, + "step": 14703 + }, + { + "epoch": 12.41891891891892, + "grad_norm": 0.5060903429985046, + "learning_rate": 2.8433333333333334e-05, + "loss": 0.0161, + "step": 14704 + }, + { + "epoch": 12.419763513513514, + "grad_norm": 0.2635253667831421, + "learning_rate": 2.8433e-05, + "loss": 0.0164, + "step": 14705 + }, + { + "epoch": 12.420608108108109, + "grad_norm": 0.8879970908164978, + "learning_rate": 2.8432666666666666e-05, + "loss": 0.0298, + "step": 14706 + }, + { + "epoch": 12.421452702702704, + "grad_norm": 0.629880964756012, + "learning_rate": 2.8432333333333332e-05, + "loss": 0.0613, + "step": 14707 + }, + { + "epoch": 12.422297297297296, + "grad_norm": 0.6545870304107666, + "learning_rate": 2.8432e-05, + "loss": 0.0344, + "step": 14708 + }, + { + "epoch": 12.423141891891891, + "grad_norm": 0.6569843888282776, + "learning_rate": 2.8431666666666667e-05, + "loss": 0.189, + "step": 14709 + }, + { + "epoch": 12.423986486486486, + "grad_norm": 0.8085178136825562, + "learning_rate": 2.8431333333333333e-05, + "loss": 0.1149, + "step": 14710 + }, + { + "epoch": 12.42483108108108, + "grad_norm": 0.5664506554603577, + "learning_rate": 2.8431e-05, + "loss": 0.0759, + "step": 14711 + }, + { + "epoch": 12.425675675675675, + "grad_norm": 0.3709944188594818, + "learning_rate": 2.843066666666667e-05, + "loss": 0.0404, + "step": 14712 + }, + { + "epoch": 12.42652027027027, + "grad_norm": 0.2917416989803314, + "learning_rate": 2.843033333333333e-05, + "loss": 0.0301, + "step": 14713 + }, + { + "epoch": 12.427364864864865, + "grad_norm": 0.49026763439178467, + "learning_rate": 2.843e-05, + "loss": 0.0275, + "step": 14714 + }, + { + "epoch": 12.42820945945946, + "grad_norm": 0.25523141026496887, + "learning_rate": 2.842966666666667e-05, + "loss": 0.0186, + "step": 14715 + }, + { + "epoch": 12.429054054054054, + "grad_norm": 0.3542098104953766, + "learning_rate": 2.8429333333333332e-05, + "loss": 0.0234, + "step": 14716 + }, + { + "epoch": 12.42989864864865, + "grad_norm": 0.26954737305641174, + "learning_rate": 2.8429000000000002e-05, + "loss": 0.0278, + "step": 14717 + }, + { + "epoch": 12.430743243243244, + "grad_norm": 0.5818050503730774, + "learning_rate": 2.8428666666666668e-05, + "loss": 0.0439, + "step": 14718 + }, + { + "epoch": 12.431587837837839, + "grad_norm": 0.3858194649219513, + "learning_rate": 2.8428333333333334e-05, + "loss": 0.0339, + "step": 14719 + }, + { + "epoch": 12.432432432432432, + "grad_norm": 0.35383597016334534, + "learning_rate": 2.8428e-05, + "loss": 0.0299, + "step": 14720 + }, + { + "epoch": 12.433277027027026, + "grad_norm": 0.598964512348175, + "learning_rate": 2.842766666666667e-05, + "loss": 0.0266, + "step": 14721 + }, + { + "epoch": 12.434121621621621, + "grad_norm": 0.2440526932477951, + "learning_rate": 2.842733333333333e-05, + "loss": 0.0186, + "step": 14722 + }, + { + "epoch": 12.434966216216216, + "grad_norm": 0.36622533202171326, + "learning_rate": 2.8427e-05, + "loss": 0.0318, + "step": 14723 + }, + { + "epoch": 12.43581081081081, + "grad_norm": 0.18722423911094666, + "learning_rate": 2.842666666666667e-05, + "loss": 0.0134, + "step": 14724 + }, + { + "epoch": 12.436655405405405, + "grad_norm": 0.3042807877063751, + "learning_rate": 2.8426333333333333e-05, + "loss": 0.0289, + "step": 14725 + }, + { + "epoch": 12.4375, + "grad_norm": 0.3502228856086731, + "learning_rate": 2.8426000000000002e-05, + "loss": 0.0219, + "step": 14726 + }, + { + "epoch": 12.438344594594595, + "grad_norm": 0.1656685620546341, + "learning_rate": 2.8425666666666668e-05, + "loss": 0.0146, + "step": 14727 + }, + { + "epoch": 12.43918918918919, + "grad_norm": 0.24129340052604675, + "learning_rate": 2.8425333333333334e-05, + "loss": 0.0183, + "step": 14728 + }, + { + "epoch": 12.440033783783784, + "grad_norm": 0.23790889978408813, + "learning_rate": 2.8425e-05, + "loss": 0.0222, + "step": 14729 + }, + { + "epoch": 12.440878378378379, + "grad_norm": 0.3339013159275055, + "learning_rate": 2.842466666666667e-05, + "loss": 0.0319, + "step": 14730 + }, + { + "epoch": 12.441722972972974, + "grad_norm": 0.28654080629348755, + "learning_rate": 2.842433333333333e-05, + "loss": 0.0248, + "step": 14731 + }, + { + "epoch": 12.442567567567568, + "grad_norm": 0.2281852662563324, + "learning_rate": 2.8424e-05, + "loss": 0.0251, + "step": 14732 + }, + { + "epoch": 12.443412162162161, + "grad_norm": 0.361899197101593, + "learning_rate": 2.842366666666667e-05, + "loss": 0.0192, + "step": 14733 + }, + { + "epoch": 12.444256756756756, + "grad_norm": 0.2901793420314789, + "learning_rate": 2.8423333333333333e-05, + "loss": 0.0225, + "step": 14734 + }, + { + "epoch": 12.44510135135135, + "grad_norm": 0.7568072080612183, + "learning_rate": 2.8423000000000002e-05, + "loss": 0.0313, + "step": 14735 + }, + { + "epoch": 12.445945945945946, + "grad_norm": 0.377724289894104, + "learning_rate": 2.8422666666666668e-05, + "loss": 0.0182, + "step": 14736 + }, + { + "epoch": 12.44679054054054, + "grad_norm": 0.3948017656803131, + "learning_rate": 2.8422333333333334e-05, + "loss": 0.0225, + "step": 14737 + }, + { + "epoch": 12.447635135135135, + "grad_norm": 0.25688040256500244, + "learning_rate": 2.8422e-05, + "loss": 0.0224, + "step": 14738 + }, + { + "epoch": 12.44847972972973, + "grad_norm": 0.2986907660961151, + "learning_rate": 2.8421666666666666e-05, + "loss": 0.0222, + "step": 14739 + }, + { + "epoch": 12.449324324324325, + "grad_norm": 0.286438912153244, + "learning_rate": 2.8421333333333335e-05, + "loss": 0.029, + "step": 14740 + }, + { + "epoch": 12.45016891891892, + "grad_norm": 0.2659013271331787, + "learning_rate": 2.8421e-05, + "loss": 0.0148, + "step": 14741 + }, + { + "epoch": 12.451013513513514, + "grad_norm": 0.18541663885116577, + "learning_rate": 2.8420666666666667e-05, + "loss": 0.0136, + "step": 14742 + }, + { + "epoch": 12.451858108108109, + "grad_norm": 0.1755150407552719, + "learning_rate": 2.8420333333333333e-05, + "loss": 0.0149, + "step": 14743 + }, + { + "epoch": 12.452702702702704, + "grad_norm": 1.3189488649368286, + "learning_rate": 2.8420000000000002e-05, + "loss": 0.0369, + "step": 14744 + }, + { + "epoch": 12.453547297297296, + "grad_norm": 0.4305727779865265, + "learning_rate": 2.8419666666666665e-05, + "loss": 0.0146, + "step": 14745 + }, + { + "epoch": 12.454391891891891, + "grad_norm": 0.2800538241863251, + "learning_rate": 2.8419333333333334e-05, + "loss": 0.0192, + "step": 14746 + }, + { + "epoch": 12.455236486486486, + "grad_norm": 0.27857643365859985, + "learning_rate": 2.8419e-05, + "loss": 0.0174, + "step": 14747 + }, + { + "epoch": 12.45608108108108, + "grad_norm": 0.3171418011188507, + "learning_rate": 2.8418666666666666e-05, + "loss": 0.0171, + "step": 14748 + }, + { + "epoch": 12.456925675675675, + "grad_norm": 0.3559357523918152, + "learning_rate": 2.8418333333333335e-05, + "loss": 0.0246, + "step": 14749 + }, + { + "epoch": 12.45777027027027, + "grad_norm": 0.19934989511966705, + "learning_rate": 2.8418e-05, + "loss": 0.0124, + "step": 14750 + }, + { + "epoch": 12.458614864864865, + "grad_norm": 0.4585868716239929, + "learning_rate": 2.8417666666666667e-05, + "loss": 0.0252, + "step": 14751 + }, + { + "epoch": 12.45945945945946, + "grad_norm": 0.594674825668335, + "learning_rate": 2.8417333333333333e-05, + "loss": 0.044, + "step": 14752 + }, + { + "epoch": 12.460304054054054, + "grad_norm": 0.4026544690132141, + "learning_rate": 2.8417000000000003e-05, + "loss": 0.016, + "step": 14753 + }, + { + "epoch": 12.46114864864865, + "grad_norm": 0.21946120262145996, + "learning_rate": 2.8416666666666665e-05, + "loss": 0.018, + "step": 14754 + }, + { + "epoch": 12.461993243243244, + "grad_norm": 0.6224389672279358, + "learning_rate": 2.8416333333333334e-05, + "loss": 0.0356, + "step": 14755 + }, + { + "epoch": 12.462837837837839, + "grad_norm": 0.3664194643497467, + "learning_rate": 2.8416e-05, + "loss": 0.0202, + "step": 14756 + }, + { + "epoch": 12.463682432432432, + "grad_norm": 0.6130070090293884, + "learning_rate": 2.8415666666666666e-05, + "loss": 0.0365, + "step": 14757 + }, + { + "epoch": 12.464527027027026, + "grad_norm": 0.6627283692359924, + "learning_rate": 2.8415333333333336e-05, + "loss": 0.0261, + "step": 14758 + }, + { + "epoch": 12.465371621621621, + "grad_norm": 0.7261884212493896, + "learning_rate": 2.8415e-05, + "loss": 0.1998, + "step": 14759 + }, + { + "epoch": 12.466216216216216, + "grad_norm": 0.5190536975860596, + "learning_rate": 2.8414666666666667e-05, + "loss": 0.1265, + "step": 14760 + }, + { + "epoch": 12.46706081081081, + "grad_norm": 0.38199129700660706, + "learning_rate": 2.8414333333333333e-05, + "loss": 0.0655, + "step": 14761 + }, + { + "epoch": 12.467905405405405, + "grad_norm": 0.28939101099967957, + "learning_rate": 2.8414000000000003e-05, + "loss": 0.0321, + "step": 14762 + }, + { + "epoch": 12.46875, + "grad_norm": 0.28707221150398254, + "learning_rate": 2.8413666666666665e-05, + "loss": 0.0242, + "step": 14763 + }, + { + "epoch": 12.469594594594595, + "grad_norm": 0.36276769638061523, + "learning_rate": 2.8413333333333335e-05, + "loss": 0.0338, + "step": 14764 + }, + { + "epoch": 12.47043918918919, + "grad_norm": 0.46506771445274353, + "learning_rate": 2.8413000000000004e-05, + "loss": 0.0299, + "step": 14765 + }, + { + "epoch": 12.471283783783784, + "grad_norm": 0.3932360112667084, + "learning_rate": 2.8412666666666666e-05, + "loss": 0.0281, + "step": 14766 + }, + { + "epoch": 12.472128378378379, + "grad_norm": 0.2240803986787796, + "learning_rate": 2.8412333333333336e-05, + "loss": 0.0183, + "step": 14767 + }, + { + "epoch": 12.472972972972974, + "grad_norm": 0.2918508052825928, + "learning_rate": 2.8412e-05, + "loss": 0.0284, + "step": 14768 + }, + { + "epoch": 12.473817567567568, + "grad_norm": 0.31489095091819763, + "learning_rate": 2.8411666666666668e-05, + "loss": 0.0297, + "step": 14769 + }, + { + "epoch": 12.474662162162161, + "grad_norm": 0.22896738350391388, + "learning_rate": 2.8411333333333334e-05, + "loss": 0.023, + "step": 14770 + }, + { + "epoch": 12.475506756756756, + "grad_norm": 0.7087687849998474, + "learning_rate": 2.8411e-05, + "loss": 0.03, + "step": 14771 + }, + { + "epoch": 12.47635135135135, + "grad_norm": 0.29104459285736084, + "learning_rate": 2.8410666666666665e-05, + "loss": 0.0239, + "step": 14772 + }, + { + "epoch": 12.477195945945946, + "grad_norm": 0.3236690163612366, + "learning_rate": 2.8410333333333335e-05, + "loss": 0.0164, + "step": 14773 + }, + { + "epoch": 12.47804054054054, + "grad_norm": 0.21997900307178497, + "learning_rate": 2.841e-05, + "loss": 0.0162, + "step": 14774 + }, + { + "epoch": 12.478885135135135, + "grad_norm": 0.23419089615345, + "learning_rate": 2.8409666666666667e-05, + "loss": 0.0305, + "step": 14775 + }, + { + "epoch": 12.47972972972973, + "grad_norm": 0.30796852707862854, + "learning_rate": 2.8409333333333336e-05, + "loss": 0.0282, + "step": 14776 + }, + { + "epoch": 12.480574324324325, + "grad_norm": 0.27260103821754456, + "learning_rate": 2.8409e-05, + "loss": 0.0215, + "step": 14777 + }, + { + "epoch": 12.48141891891892, + "grad_norm": 0.42165136337280273, + "learning_rate": 2.8408666666666668e-05, + "loss": 0.0199, + "step": 14778 + }, + { + "epoch": 12.482263513513514, + "grad_norm": 0.235086590051651, + "learning_rate": 2.8408333333333334e-05, + "loss": 0.0196, + "step": 14779 + }, + { + "epoch": 12.483108108108109, + "grad_norm": 0.17886358499526978, + "learning_rate": 2.8408e-05, + "loss": 0.0133, + "step": 14780 + }, + { + "epoch": 12.483952702702704, + "grad_norm": 0.47925713658332825, + "learning_rate": 2.8407666666666666e-05, + "loss": 0.0155, + "step": 14781 + }, + { + "epoch": 12.484797297297296, + "grad_norm": 0.18244752287864685, + "learning_rate": 2.8407333333333335e-05, + "loss": 0.0201, + "step": 14782 + }, + { + "epoch": 12.485641891891891, + "grad_norm": 0.2525389492511749, + "learning_rate": 2.8407e-05, + "loss": 0.0169, + "step": 14783 + }, + { + "epoch": 12.486486486486486, + "grad_norm": 0.3760823905467987, + "learning_rate": 2.8406666666666667e-05, + "loss": 0.0148, + "step": 14784 + }, + { + "epoch": 12.48733108108108, + "grad_norm": 0.23068156838417053, + "learning_rate": 2.8406333333333336e-05, + "loss": 0.0185, + "step": 14785 + }, + { + "epoch": 12.488175675675675, + "grad_norm": 0.2807624042034149, + "learning_rate": 2.8406e-05, + "loss": 0.0202, + "step": 14786 + }, + { + "epoch": 12.48902027027027, + "grad_norm": 0.3921656608581543, + "learning_rate": 2.8405666666666668e-05, + "loss": 0.0182, + "step": 14787 + }, + { + "epoch": 12.489864864864865, + "grad_norm": 0.2229410707950592, + "learning_rate": 2.8405333333333334e-05, + "loss": 0.0169, + "step": 14788 + }, + { + "epoch": 12.49070945945946, + "grad_norm": 0.23983903229236603, + "learning_rate": 2.8405e-05, + "loss": 0.0213, + "step": 14789 + }, + { + "epoch": 12.491554054054054, + "grad_norm": 0.360553503036499, + "learning_rate": 2.840466666666667e-05, + "loss": 0.0229, + "step": 14790 + }, + { + "epoch": 12.49239864864865, + "grad_norm": 0.253480464220047, + "learning_rate": 2.8404333333333335e-05, + "loss": 0.0153, + "step": 14791 + }, + { + "epoch": 12.493243243243244, + "grad_norm": 0.3606550693511963, + "learning_rate": 2.8404e-05, + "loss": 0.0279, + "step": 14792 + }, + { + "epoch": 12.494087837837839, + "grad_norm": 0.28100237250328064, + "learning_rate": 2.8403666666666667e-05, + "loss": 0.0182, + "step": 14793 + }, + { + "epoch": 12.494932432432432, + "grad_norm": 0.39498332142829895, + "learning_rate": 2.8403333333333336e-05, + "loss": 0.0269, + "step": 14794 + }, + { + "epoch": 12.495777027027026, + "grad_norm": 0.2555437982082367, + "learning_rate": 2.8403e-05, + "loss": 0.0141, + "step": 14795 + }, + { + "epoch": 12.496621621621621, + "grad_norm": 0.3482447564601898, + "learning_rate": 2.8402666666666668e-05, + "loss": 0.0167, + "step": 14796 + }, + { + "epoch": 12.497466216216216, + "grad_norm": 0.17642350494861603, + "learning_rate": 2.8402333333333334e-05, + "loss": 0.01, + "step": 14797 + }, + { + "epoch": 12.49831081081081, + "grad_norm": 0.39590689539909363, + "learning_rate": 2.8402e-05, + "loss": 0.0232, + "step": 14798 + }, + { + "epoch": 12.499155405405405, + "grad_norm": 0.24836289882659912, + "learning_rate": 2.840166666666667e-05, + "loss": 0.026, + "step": 14799 + }, + { + "epoch": 12.5, + "grad_norm": 0.2682209014892578, + "learning_rate": 2.8401333333333332e-05, + "loss": 0.0124, + "step": 14800 + }, + { + "epoch": 12.500844594594595, + "grad_norm": 0.8555505871772766, + "learning_rate": 2.8401e-05, + "loss": 0.0308, + "step": 14801 + }, + { + "epoch": 12.50168918918919, + "grad_norm": 0.32549166679382324, + "learning_rate": 2.8400666666666667e-05, + "loss": 0.0172, + "step": 14802 + }, + { + "epoch": 12.502533783783784, + "grad_norm": 0.47280117869377136, + "learning_rate": 2.8400333333333333e-05, + "loss": 0.0226, + "step": 14803 + }, + { + "epoch": 12.503378378378379, + "grad_norm": 0.29126042127609253, + "learning_rate": 2.84e-05, + "loss": 0.0252, + "step": 14804 + }, + { + "epoch": 12.504222972972974, + "grad_norm": 0.4049931764602661, + "learning_rate": 2.839966666666667e-05, + "loss": 0.0313, + "step": 14805 + }, + { + "epoch": 12.505067567567568, + "grad_norm": 0.35307013988494873, + "learning_rate": 2.8399333333333334e-05, + "loss": 0.0249, + "step": 14806 + }, + { + "epoch": 12.505912162162161, + "grad_norm": 0.4946669638156891, + "learning_rate": 2.8399e-05, + "loss": 0.0423, + "step": 14807 + }, + { + "epoch": 12.506756756756756, + "grad_norm": 0.49387556314468384, + "learning_rate": 2.839866666666667e-05, + "loss": 0.0267, + "step": 14808 + }, + { + "epoch": 12.50760135135135, + "grad_norm": 0.6120842099189758, + "learning_rate": 2.8398333333333332e-05, + "loss": 0.1718, + "step": 14809 + }, + { + "epoch": 12.508445945945946, + "grad_norm": 0.7439628839492798, + "learning_rate": 2.8398e-05, + "loss": 0.1128, + "step": 14810 + }, + { + "epoch": 12.50929054054054, + "grad_norm": 0.45725366473197937, + "learning_rate": 2.8397666666666667e-05, + "loss": 0.0672, + "step": 14811 + }, + { + "epoch": 12.510135135135135, + "grad_norm": 0.24718694388866425, + "learning_rate": 2.8397333333333333e-05, + "loss": 0.0368, + "step": 14812 + }, + { + "epoch": 12.51097972972973, + "grad_norm": 0.28383392095565796, + "learning_rate": 2.8397e-05, + "loss": 0.0209, + "step": 14813 + }, + { + "epoch": 12.511824324324325, + "grad_norm": 1.0750863552093506, + "learning_rate": 2.839666666666667e-05, + "loss": 0.0279, + "step": 14814 + }, + { + "epoch": 12.51266891891892, + "grad_norm": 0.3132067918777466, + "learning_rate": 2.8396333333333335e-05, + "loss": 0.033, + "step": 14815 + }, + { + "epoch": 12.513513513513514, + "grad_norm": 0.345535010099411, + "learning_rate": 2.8396e-05, + "loss": 0.0254, + "step": 14816 + }, + { + "epoch": 12.514358108108109, + "grad_norm": 0.5452395081520081, + "learning_rate": 2.839566666666667e-05, + "loss": 0.0144, + "step": 14817 + }, + { + "epoch": 12.515202702702704, + "grad_norm": 0.4429996907711029, + "learning_rate": 2.8395333333333332e-05, + "loss": 0.0258, + "step": 14818 + }, + { + "epoch": 12.516047297297296, + "grad_norm": 0.1747012585401535, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0109, + "step": 14819 + }, + { + "epoch": 12.516891891891891, + "grad_norm": 0.24082808196544647, + "learning_rate": 2.8394666666666668e-05, + "loss": 0.0267, + "step": 14820 + }, + { + "epoch": 12.517736486486486, + "grad_norm": 0.3286198079586029, + "learning_rate": 2.8394333333333334e-05, + "loss": 0.0356, + "step": 14821 + }, + { + "epoch": 12.51858108108108, + "grad_norm": 0.4878372848033905, + "learning_rate": 2.8394e-05, + "loss": 0.0434, + "step": 14822 + }, + { + "epoch": 12.519425675675675, + "grad_norm": 0.1680971086025238, + "learning_rate": 2.839366666666667e-05, + "loss": 0.0131, + "step": 14823 + }, + { + "epoch": 12.52027027027027, + "grad_norm": 0.43354836106300354, + "learning_rate": 2.8393333333333335e-05, + "loss": 0.045, + "step": 14824 + }, + { + "epoch": 12.521114864864865, + "grad_norm": 0.45058074593544006, + "learning_rate": 2.8393e-05, + "loss": 0.0175, + "step": 14825 + }, + { + "epoch": 12.52195945945946, + "grad_norm": 0.1710488349199295, + "learning_rate": 2.839266666666667e-05, + "loss": 0.0142, + "step": 14826 + }, + { + "epoch": 12.522804054054054, + "grad_norm": 0.35250124335289, + "learning_rate": 2.8392333333333333e-05, + "loss": 0.0261, + "step": 14827 + }, + { + "epoch": 12.52364864864865, + "grad_norm": 0.46954140067100525, + "learning_rate": 2.8392000000000002e-05, + "loss": 0.0284, + "step": 14828 + }, + { + "epoch": 12.524493243243244, + "grad_norm": 0.1877991110086441, + "learning_rate": 2.8391666666666664e-05, + "loss": 0.0145, + "step": 14829 + }, + { + "epoch": 12.525337837837839, + "grad_norm": 0.441468209028244, + "learning_rate": 2.8391333333333334e-05, + "loss": 0.0222, + "step": 14830 + }, + { + "epoch": 12.526182432432432, + "grad_norm": 0.16568130254745483, + "learning_rate": 2.8391e-05, + "loss": 0.0147, + "step": 14831 + }, + { + "epoch": 12.527027027027026, + "grad_norm": 0.21411961317062378, + "learning_rate": 2.8390666666666666e-05, + "loss": 0.0137, + "step": 14832 + }, + { + "epoch": 12.527871621621621, + "grad_norm": 0.514860987663269, + "learning_rate": 2.8390333333333335e-05, + "loss": 0.0192, + "step": 14833 + }, + { + "epoch": 12.528716216216216, + "grad_norm": 0.2836135923862457, + "learning_rate": 2.839e-05, + "loss": 0.0212, + "step": 14834 + }, + { + "epoch": 12.52956081081081, + "grad_norm": 0.26016250252723694, + "learning_rate": 2.8389666666666667e-05, + "loss": 0.0183, + "step": 14835 + }, + { + "epoch": 12.530405405405405, + "grad_norm": 0.2246493250131607, + "learning_rate": 2.8389333333333333e-05, + "loss": 0.0139, + "step": 14836 + }, + { + "epoch": 12.53125, + "grad_norm": 0.6430110931396484, + "learning_rate": 2.8389000000000002e-05, + "loss": 0.0202, + "step": 14837 + }, + { + "epoch": 12.532094594594595, + "grad_norm": 0.175309956073761, + "learning_rate": 2.8388666666666665e-05, + "loss": 0.0156, + "step": 14838 + }, + { + "epoch": 12.53293918918919, + "grad_norm": 0.4723362326622009, + "learning_rate": 2.8388333333333334e-05, + "loss": 0.0373, + "step": 14839 + }, + { + "epoch": 12.533783783783784, + "grad_norm": 0.49053338170051575, + "learning_rate": 2.8388000000000003e-05, + "loss": 0.0236, + "step": 14840 + }, + { + "epoch": 12.534628378378379, + "grad_norm": 0.23931917548179626, + "learning_rate": 2.8387666666666666e-05, + "loss": 0.0094, + "step": 14841 + }, + { + "epoch": 12.535472972972974, + "grad_norm": 0.3824673593044281, + "learning_rate": 2.8387333333333335e-05, + "loss": 0.0312, + "step": 14842 + }, + { + "epoch": 12.536317567567568, + "grad_norm": 0.18712268769741058, + "learning_rate": 2.8387e-05, + "loss": 0.0117, + "step": 14843 + }, + { + "epoch": 12.537162162162161, + "grad_norm": 0.4172927737236023, + "learning_rate": 2.8386666666666667e-05, + "loss": 0.0317, + "step": 14844 + }, + { + "epoch": 12.538006756756756, + "grad_norm": 0.38896605372428894, + "learning_rate": 2.8386333333333333e-05, + "loss": 0.0238, + "step": 14845 + }, + { + "epoch": 12.53885135135135, + "grad_norm": 0.2546064853668213, + "learning_rate": 2.8386000000000002e-05, + "loss": 0.0142, + "step": 14846 + }, + { + "epoch": 12.539695945945946, + "grad_norm": 0.42539623379707336, + "learning_rate": 2.8385666666666665e-05, + "loss": 0.0219, + "step": 14847 + }, + { + "epoch": 12.54054054054054, + "grad_norm": 0.2886580526828766, + "learning_rate": 2.8385333333333334e-05, + "loss": 0.0188, + "step": 14848 + }, + { + "epoch": 12.541385135135135, + "grad_norm": 0.33424487709999084, + "learning_rate": 2.8385000000000004e-05, + "loss": 0.0306, + "step": 14849 + }, + { + "epoch": 12.54222972972973, + "grad_norm": 0.3471667468547821, + "learning_rate": 2.8384666666666666e-05, + "loss": 0.0177, + "step": 14850 + }, + { + "epoch": 12.543074324324325, + "grad_norm": 0.33081212639808655, + "learning_rate": 2.8384333333333335e-05, + "loss": 0.0158, + "step": 14851 + }, + { + "epoch": 12.54391891891892, + "grad_norm": 0.2575814425945282, + "learning_rate": 2.8384e-05, + "loss": 0.017, + "step": 14852 + }, + { + "epoch": 12.544763513513514, + "grad_norm": 0.3071545660495758, + "learning_rate": 2.8383666666666667e-05, + "loss": 0.0286, + "step": 14853 + }, + { + "epoch": 12.545608108108109, + "grad_norm": 0.6037377715110779, + "learning_rate": 2.8383333333333333e-05, + "loss": 0.0229, + "step": 14854 + }, + { + "epoch": 12.546452702702704, + "grad_norm": 0.8882977962493896, + "learning_rate": 2.8383000000000003e-05, + "loss": 0.0303, + "step": 14855 + }, + { + "epoch": 12.547297297297296, + "grad_norm": 0.4282629191875458, + "learning_rate": 2.838266666666667e-05, + "loss": 0.0211, + "step": 14856 + }, + { + "epoch": 12.548141891891891, + "grad_norm": 0.3357529640197754, + "learning_rate": 2.8382333333333334e-05, + "loss": 0.0319, + "step": 14857 + }, + { + "epoch": 12.548986486486486, + "grad_norm": 0.546819806098938, + "learning_rate": 2.8382e-05, + "loss": 0.035, + "step": 14858 + }, + { + "epoch": 12.54983108108108, + "grad_norm": 0.5989054441452026, + "learning_rate": 2.8381666666666666e-05, + "loss": 0.2094, + "step": 14859 + }, + { + "epoch": 12.550675675675675, + "grad_norm": 0.552507758140564, + "learning_rate": 2.8381333333333336e-05, + "loss": 0.1181, + "step": 14860 + }, + { + "epoch": 12.55152027027027, + "grad_norm": 0.3260217308998108, + "learning_rate": 2.8380999999999998e-05, + "loss": 0.0465, + "step": 14861 + }, + { + "epoch": 12.552364864864865, + "grad_norm": 0.3081076741218567, + "learning_rate": 2.8380666666666667e-05, + "loss": 0.0403, + "step": 14862 + }, + { + "epoch": 12.55320945945946, + "grad_norm": 0.3846648633480072, + "learning_rate": 2.8380333333333333e-05, + "loss": 0.0423, + "step": 14863 + }, + { + "epoch": 12.554054054054054, + "grad_norm": 0.2652827501296997, + "learning_rate": 2.838e-05, + "loss": 0.0276, + "step": 14864 + }, + { + "epoch": 12.55489864864865, + "grad_norm": 0.2505081593990326, + "learning_rate": 2.837966666666667e-05, + "loss": 0.0285, + "step": 14865 + }, + { + "epoch": 12.555743243243244, + "grad_norm": 0.16032910346984863, + "learning_rate": 2.8379333333333335e-05, + "loss": 0.018, + "step": 14866 + }, + { + "epoch": 12.556587837837839, + "grad_norm": 0.2223948985338211, + "learning_rate": 2.8379e-05, + "loss": 0.0286, + "step": 14867 + }, + { + "epoch": 12.557432432432432, + "grad_norm": 0.24874070286750793, + "learning_rate": 2.8378666666666666e-05, + "loss": 0.0218, + "step": 14868 + }, + { + "epoch": 12.558277027027026, + "grad_norm": 0.19313457608222961, + "learning_rate": 2.8378333333333336e-05, + "loss": 0.023, + "step": 14869 + }, + { + "epoch": 12.559121621621621, + "grad_norm": 0.34702008962631226, + "learning_rate": 2.8378e-05, + "loss": 0.0269, + "step": 14870 + }, + { + "epoch": 12.559966216216216, + "grad_norm": 0.2318984568119049, + "learning_rate": 2.8377666666666668e-05, + "loss": 0.0179, + "step": 14871 + }, + { + "epoch": 12.56081081081081, + "grad_norm": 0.23792308568954468, + "learning_rate": 2.8377333333333334e-05, + "loss": 0.0231, + "step": 14872 + }, + { + "epoch": 12.561655405405405, + "grad_norm": 0.24733689427375793, + "learning_rate": 2.8377e-05, + "loss": 0.0176, + "step": 14873 + }, + { + "epoch": 12.5625, + "grad_norm": 0.23213377594947815, + "learning_rate": 2.837666666666667e-05, + "loss": 0.0222, + "step": 14874 + }, + { + "epoch": 12.563344594594595, + "grad_norm": 0.35450509190559387, + "learning_rate": 2.8376333333333335e-05, + "loss": 0.0208, + "step": 14875 + }, + { + "epoch": 12.56418918918919, + "grad_norm": 0.31648334860801697, + "learning_rate": 2.8376e-05, + "loss": 0.0183, + "step": 14876 + }, + { + "epoch": 12.565033783783784, + "grad_norm": 0.42108339071273804, + "learning_rate": 2.8375666666666667e-05, + "loss": 0.0261, + "step": 14877 + }, + { + "epoch": 12.565878378378379, + "grad_norm": 0.19600848853588104, + "learning_rate": 2.8375333333333336e-05, + "loss": 0.0151, + "step": 14878 + }, + { + "epoch": 12.566722972972974, + "grad_norm": 0.23623308539390564, + "learning_rate": 2.8375e-05, + "loss": 0.0217, + "step": 14879 + }, + { + "epoch": 12.567567567567568, + "grad_norm": 0.24476848542690277, + "learning_rate": 2.8374666666666668e-05, + "loss": 0.0212, + "step": 14880 + }, + { + "epoch": 12.568412162162161, + "grad_norm": 0.353245347738266, + "learning_rate": 2.8374333333333334e-05, + "loss": 0.0202, + "step": 14881 + }, + { + "epoch": 12.569256756756756, + "grad_norm": 0.6909670829772949, + "learning_rate": 2.8374e-05, + "loss": 0.0163, + "step": 14882 + }, + { + "epoch": 12.57010135135135, + "grad_norm": 0.1992567926645279, + "learning_rate": 2.837366666666667e-05, + "loss": 0.0107, + "step": 14883 + }, + { + "epoch": 12.570945945945946, + "grad_norm": 0.47435516119003296, + "learning_rate": 2.8373333333333335e-05, + "loss": 0.02, + "step": 14884 + }, + { + "epoch": 12.57179054054054, + "grad_norm": 0.5340313911437988, + "learning_rate": 2.8373e-05, + "loss": 0.0349, + "step": 14885 + }, + { + "epoch": 12.572635135135135, + "grad_norm": 0.5812420845031738, + "learning_rate": 2.8372666666666667e-05, + "loss": 0.032, + "step": 14886 + }, + { + "epoch": 12.57347972972973, + "grad_norm": 0.1927167773246765, + "learning_rate": 2.8372333333333336e-05, + "loss": 0.0159, + "step": 14887 + }, + { + "epoch": 12.574324324324325, + "grad_norm": 0.2681322395801544, + "learning_rate": 2.8372e-05, + "loss": 0.0297, + "step": 14888 + }, + { + "epoch": 12.57516891891892, + "grad_norm": 0.24012817442417145, + "learning_rate": 2.8371666666666668e-05, + "loss": 0.0143, + "step": 14889 + }, + { + "epoch": 12.576013513513514, + "grad_norm": 0.3621043860912323, + "learning_rate": 2.8371333333333334e-05, + "loss": 0.0306, + "step": 14890 + }, + { + "epoch": 12.576858108108109, + "grad_norm": 0.2545062303543091, + "learning_rate": 2.8371e-05, + "loss": 0.0196, + "step": 14891 + }, + { + "epoch": 12.577702702702704, + "grad_norm": 0.721332311630249, + "learning_rate": 2.837066666666667e-05, + "loss": 0.0383, + "step": 14892 + }, + { + "epoch": 12.578547297297296, + "grad_norm": 0.280086874961853, + "learning_rate": 2.8370333333333332e-05, + "loss": 0.0199, + "step": 14893 + }, + { + "epoch": 12.579391891891891, + "grad_norm": 0.25728318095207214, + "learning_rate": 2.837e-05, + "loss": 0.0138, + "step": 14894 + }, + { + "epoch": 12.580236486486486, + "grad_norm": 0.5535433888435364, + "learning_rate": 2.8369666666666667e-05, + "loss": 0.0301, + "step": 14895 + }, + { + "epoch": 12.58108108108108, + "grad_norm": 0.2353011965751648, + "learning_rate": 2.8369333333333333e-05, + "loss": 0.0131, + "step": 14896 + }, + { + "epoch": 12.581925675675675, + "grad_norm": 0.47499728202819824, + "learning_rate": 2.8369e-05, + "loss": 0.0159, + "step": 14897 + }, + { + "epoch": 12.58277027027027, + "grad_norm": 0.40595462918281555, + "learning_rate": 2.8368666666666668e-05, + "loss": 0.0326, + "step": 14898 + }, + { + "epoch": 12.583614864864865, + "grad_norm": 0.37604865431785583, + "learning_rate": 2.8368333333333334e-05, + "loss": 0.0243, + "step": 14899 + }, + { + "epoch": 12.58445945945946, + "grad_norm": 0.23820026218891144, + "learning_rate": 2.8368e-05, + "loss": 0.0163, + "step": 14900 + }, + { + "epoch": 12.585304054054054, + "grad_norm": 0.42072632908821106, + "learning_rate": 2.836766666666667e-05, + "loss": 0.0268, + "step": 14901 + }, + { + "epoch": 12.58614864864865, + "grad_norm": 0.6044257283210754, + "learning_rate": 2.8367333333333332e-05, + "loss": 0.0209, + "step": 14902 + }, + { + "epoch": 12.586993243243244, + "grad_norm": 0.36987215280532837, + "learning_rate": 2.8367e-05, + "loss": 0.0314, + "step": 14903 + }, + { + "epoch": 12.587837837837839, + "grad_norm": 0.2722662091255188, + "learning_rate": 2.8366666666666667e-05, + "loss": 0.0143, + "step": 14904 + }, + { + "epoch": 12.588682432432432, + "grad_norm": 0.9526032209396362, + "learning_rate": 2.8366333333333333e-05, + "loss": 0.0324, + "step": 14905 + }, + { + "epoch": 12.589527027027026, + "grad_norm": 0.4612269997596741, + "learning_rate": 2.8366000000000003e-05, + "loss": 0.0261, + "step": 14906 + }, + { + "epoch": 12.590371621621621, + "grad_norm": 0.7255565524101257, + "learning_rate": 2.836566666666667e-05, + "loss": 0.0256, + "step": 14907 + }, + { + "epoch": 12.591216216216216, + "grad_norm": 0.5569741129875183, + "learning_rate": 2.8365333333333334e-05, + "loss": 0.0275, + "step": 14908 + }, + { + "epoch": 12.59206081081081, + "grad_norm": 0.618495762348175, + "learning_rate": 2.8365e-05, + "loss": 0.1881, + "step": 14909 + }, + { + "epoch": 12.592905405405405, + "grad_norm": 0.6409093141555786, + "learning_rate": 2.836466666666667e-05, + "loss": 0.1435, + "step": 14910 + }, + { + "epoch": 12.59375, + "grad_norm": 0.9171792268753052, + "learning_rate": 2.8364333333333332e-05, + "loss": 0.089, + "step": 14911 + }, + { + "epoch": 12.594594594594595, + "grad_norm": 0.28218913078308105, + "learning_rate": 2.8364e-05, + "loss": 0.0442, + "step": 14912 + }, + { + "epoch": 12.59543918918919, + "grad_norm": 0.3779640793800354, + "learning_rate": 2.8363666666666667e-05, + "loss": 0.0198, + "step": 14913 + }, + { + "epoch": 12.596283783783784, + "grad_norm": 0.25518670678138733, + "learning_rate": 2.8363333333333333e-05, + "loss": 0.0183, + "step": 14914 + }, + { + "epoch": 12.597128378378379, + "grad_norm": 0.3031501770019531, + "learning_rate": 2.8363000000000003e-05, + "loss": 0.0203, + "step": 14915 + }, + { + "epoch": 12.597972972972974, + "grad_norm": 0.4125332534313202, + "learning_rate": 2.836266666666667e-05, + "loss": 0.0331, + "step": 14916 + }, + { + "epoch": 12.598817567567568, + "grad_norm": 0.321900337934494, + "learning_rate": 2.8362333333333335e-05, + "loss": 0.02, + "step": 14917 + }, + { + "epoch": 12.599662162162161, + "grad_norm": 0.2541002929210663, + "learning_rate": 2.8362e-05, + "loss": 0.0195, + "step": 14918 + }, + { + "epoch": 12.600506756756756, + "grad_norm": 0.21139279007911682, + "learning_rate": 2.8361666666666666e-05, + "loss": 0.0197, + "step": 14919 + }, + { + "epoch": 12.60135135135135, + "grad_norm": 0.16424104571342468, + "learning_rate": 2.8361333333333332e-05, + "loss": 0.0117, + "step": 14920 + }, + { + "epoch": 12.602195945945946, + "grad_norm": 0.2431010901927948, + "learning_rate": 2.8361000000000002e-05, + "loss": 0.03, + "step": 14921 + }, + { + "epoch": 12.60304054054054, + "grad_norm": 0.3554956018924713, + "learning_rate": 2.8360666666666664e-05, + "loss": 0.0188, + "step": 14922 + }, + { + "epoch": 12.603885135135135, + "grad_norm": 0.24374891817569733, + "learning_rate": 2.8360333333333334e-05, + "loss": 0.02, + "step": 14923 + }, + { + "epoch": 12.60472972972973, + "grad_norm": 0.17567823827266693, + "learning_rate": 2.8360000000000003e-05, + "loss": 0.0117, + "step": 14924 + }, + { + "epoch": 12.605574324324325, + "grad_norm": 0.28996139764785767, + "learning_rate": 2.8359666666666665e-05, + "loss": 0.0252, + "step": 14925 + }, + { + "epoch": 12.60641891891892, + "grad_norm": 0.19123715162277222, + "learning_rate": 2.8359333333333335e-05, + "loss": 0.0169, + "step": 14926 + }, + { + "epoch": 12.607263513513514, + "grad_norm": 0.6859782934188843, + "learning_rate": 2.8359e-05, + "loss": 0.0411, + "step": 14927 + }, + { + "epoch": 12.608108108108109, + "grad_norm": 0.25640004873275757, + "learning_rate": 2.8358666666666667e-05, + "loss": 0.0154, + "step": 14928 + }, + { + "epoch": 12.608952702702704, + "grad_norm": 0.3081602454185486, + "learning_rate": 2.8358333333333333e-05, + "loss": 0.0194, + "step": 14929 + }, + { + "epoch": 12.609797297297296, + "grad_norm": 0.48771053552627563, + "learning_rate": 2.8358000000000002e-05, + "loss": 0.0287, + "step": 14930 + }, + { + "epoch": 12.610641891891891, + "grad_norm": 0.4539593756198883, + "learning_rate": 2.8357666666666668e-05, + "loss": 0.0268, + "step": 14931 + }, + { + "epoch": 12.611486486486486, + "grad_norm": 0.20671813189983368, + "learning_rate": 2.8357333333333334e-05, + "loss": 0.012, + "step": 14932 + }, + { + "epoch": 12.61233108108108, + "grad_norm": 0.36411601305007935, + "learning_rate": 2.8357000000000003e-05, + "loss": 0.0196, + "step": 14933 + }, + { + "epoch": 12.613175675675675, + "grad_norm": 0.20402280986309052, + "learning_rate": 2.8356666666666666e-05, + "loss": 0.0154, + "step": 14934 + }, + { + "epoch": 12.61402027027027, + "grad_norm": 0.30899930000305176, + "learning_rate": 2.8356333333333335e-05, + "loss": 0.0287, + "step": 14935 + }, + { + "epoch": 12.614864864864865, + "grad_norm": 0.2738029360771179, + "learning_rate": 2.8356e-05, + "loss": 0.0197, + "step": 14936 + }, + { + "epoch": 12.61570945945946, + "grad_norm": 0.4460753798484802, + "learning_rate": 2.8355666666666667e-05, + "loss": 0.0245, + "step": 14937 + }, + { + "epoch": 12.616554054054054, + "grad_norm": 0.30188190937042236, + "learning_rate": 2.8355333333333333e-05, + "loss": 0.0117, + "step": 14938 + }, + { + "epoch": 12.61739864864865, + "grad_norm": 0.21965157985687256, + "learning_rate": 2.8355000000000002e-05, + "loss": 0.0181, + "step": 14939 + }, + { + "epoch": 12.618243243243244, + "grad_norm": 0.6369209885597229, + "learning_rate": 2.8354666666666668e-05, + "loss": 0.0375, + "step": 14940 + }, + { + "epoch": 12.619087837837839, + "grad_norm": 0.3079594671726227, + "learning_rate": 2.8354333333333334e-05, + "loss": 0.0228, + "step": 14941 + }, + { + "epoch": 12.619932432432432, + "grad_norm": 0.18642666935920715, + "learning_rate": 2.8354000000000003e-05, + "loss": 0.0135, + "step": 14942 + }, + { + "epoch": 12.620777027027026, + "grad_norm": 0.35967856645584106, + "learning_rate": 2.8353666666666666e-05, + "loss": 0.0199, + "step": 14943 + }, + { + "epoch": 12.621621621621621, + "grad_norm": 0.20729747414588928, + "learning_rate": 2.8353333333333335e-05, + "loss": 0.017, + "step": 14944 + }, + { + "epoch": 12.622466216216216, + "grad_norm": 0.2150869220495224, + "learning_rate": 2.8353e-05, + "loss": 0.0218, + "step": 14945 + }, + { + "epoch": 12.62331081081081, + "grad_norm": 0.29806602001190186, + "learning_rate": 2.8352666666666667e-05, + "loss": 0.018, + "step": 14946 + }, + { + "epoch": 12.624155405405405, + "grad_norm": 0.44390976428985596, + "learning_rate": 2.8352333333333333e-05, + "loss": 0.035, + "step": 14947 + }, + { + "epoch": 12.625, + "grad_norm": 0.3222532272338867, + "learning_rate": 2.8352000000000002e-05, + "loss": 0.0268, + "step": 14948 + }, + { + "epoch": 12.625844594594595, + "grad_norm": 0.17381569743156433, + "learning_rate": 2.8351666666666668e-05, + "loss": 0.019, + "step": 14949 + }, + { + "epoch": 12.62668918918919, + "grad_norm": 0.40264591574668884, + "learning_rate": 2.8351333333333334e-05, + "loss": 0.0233, + "step": 14950 + }, + { + "epoch": 12.627533783783784, + "grad_norm": 0.28498461842536926, + "learning_rate": 2.8351e-05, + "loss": 0.0252, + "step": 14951 + }, + { + "epoch": 12.628378378378379, + "grad_norm": 0.28330472111701965, + "learning_rate": 2.8350666666666666e-05, + "loss": 0.0167, + "step": 14952 + }, + { + "epoch": 12.629222972972974, + "grad_norm": 0.2986074984073639, + "learning_rate": 2.8350333333333335e-05, + "loss": 0.0199, + "step": 14953 + }, + { + "epoch": 12.630067567567568, + "grad_norm": 0.5250025391578674, + "learning_rate": 2.8349999999999998e-05, + "loss": 0.0333, + "step": 14954 + }, + { + "epoch": 12.630912162162161, + "grad_norm": 0.2814478576183319, + "learning_rate": 2.8349666666666667e-05, + "loss": 0.0226, + "step": 14955 + }, + { + "epoch": 12.631756756756756, + "grad_norm": 0.7275882363319397, + "learning_rate": 2.8349333333333337e-05, + "loss": 0.0227, + "step": 14956 + }, + { + "epoch": 12.63260135135135, + "grad_norm": 0.4557420611381531, + "learning_rate": 2.8349e-05, + "loss": 0.0263, + "step": 14957 + }, + { + "epoch": 12.633445945945946, + "grad_norm": 0.42065441608428955, + "learning_rate": 2.834866666666667e-05, + "loss": 0.0692, + "step": 14958 + }, + { + "epoch": 12.63429054054054, + "grad_norm": 0.6600936651229858, + "learning_rate": 2.8348333333333334e-05, + "loss": 0.21, + "step": 14959 + }, + { + "epoch": 12.635135135135135, + "grad_norm": 0.9090738296508789, + "learning_rate": 2.8348e-05, + "loss": 0.1183, + "step": 14960 + }, + { + "epoch": 12.63597972972973, + "grad_norm": 0.5784822106361389, + "learning_rate": 2.8347666666666666e-05, + "loss": 0.0638, + "step": 14961 + }, + { + "epoch": 12.636824324324325, + "grad_norm": 0.4635057747364044, + "learning_rate": 2.8347333333333336e-05, + "loss": 0.0377, + "step": 14962 + }, + { + "epoch": 12.63766891891892, + "grad_norm": 0.2260054498910904, + "learning_rate": 2.8346999999999998e-05, + "loss": 0.0218, + "step": 14963 + }, + { + "epoch": 12.638513513513514, + "grad_norm": 0.2920733690261841, + "learning_rate": 2.8346666666666667e-05, + "loss": 0.029, + "step": 14964 + }, + { + "epoch": 12.639358108108109, + "grad_norm": 0.2767215967178345, + "learning_rate": 2.8346333333333337e-05, + "loss": 0.0262, + "step": 14965 + }, + { + "epoch": 12.640202702702704, + "grad_norm": 0.24413549900054932, + "learning_rate": 2.8346e-05, + "loss": 0.0176, + "step": 14966 + }, + { + "epoch": 12.641047297297296, + "grad_norm": 0.4625939130783081, + "learning_rate": 2.834566666666667e-05, + "loss": 0.0327, + "step": 14967 + }, + { + "epoch": 12.641891891891891, + "grad_norm": 0.26100990176200867, + "learning_rate": 2.8345333333333335e-05, + "loss": 0.0281, + "step": 14968 + }, + { + "epoch": 12.642736486486486, + "grad_norm": 0.17221452295780182, + "learning_rate": 2.8345e-05, + "loss": 0.0148, + "step": 14969 + }, + { + "epoch": 12.64358108108108, + "grad_norm": 0.4187600314617157, + "learning_rate": 2.8344666666666666e-05, + "loss": 0.0245, + "step": 14970 + }, + { + "epoch": 12.644425675675675, + "grad_norm": 0.3938424587249756, + "learning_rate": 2.8344333333333336e-05, + "loss": 0.0381, + "step": 14971 + }, + { + "epoch": 12.64527027027027, + "grad_norm": 0.3080059587955475, + "learning_rate": 2.8344e-05, + "loss": 0.0188, + "step": 14972 + }, + { + "epoch": 12.646114864864865, + "grad_norm": 0.5489099025726318, + "learning_rate": 2.8343666666666668e-05, + "loss": 0.0355, + "step": 14973 + }, + { + "epoch": 12.64695945945946, + "grad_norm": 0.32042983174324036, + "learning_rate": 2.8343333333333337e-05, + "loss": 0.027, + "step": 14974 + }, + { + "epoch": 12.647804054054054, + "grad_norm": 0.2831363081932068, + "learning_rate": 2.8343e-05, + "loss": 0.0401, + "step": 14975 + }, + { + "epoch": 12.64864864864865, + "grad_norm": 0.20123155415058136, + "learning_rate": 2.834266666666667e-05, + "loss": 0.0131, + "step": 14976 + }, + { + "epoch": 12.649493243243244, + "grad_norm": 0.17430122196674347, + "learning_rate": 2.8342333333333335e-05, + "loss": 0.0135, + "step": 14977 + }, + { + "epoch": 12.650337837837839, + "grad_norm": 0.31584861874580383, + "learning_rate": 2.8342e-05, + "loss": 0.0278, + "step": 14978 + }, + { + "epoch": 12.651182432432432, + "grad_norm": 0.17960679531097412, + "learning_rate": 2.8341666666666667e-05, + "loss": 0.0137, + "step": 14979 + }, + { + "epoch": 12.652027027027026, + "grad_norm": 0.3478343188762665, + "learning_rate": 2.8341333333333333e-05, + "loss": 0.0256, + "step": 14980 + }, + { + "epoch": 12.652871621621621, + "grad_norm": 0.47343629598617554, + "learning_rate": 2.8341000000000002e-05, + "loss": 0.0231, + "step": 14981 + }, + { + "epoch": 12.653716216216216, + "grad_norm": 0.4849473237991333, + "learning_rate": 2.8340666666666668e-05, + "loss": 0.0153, + "step": 14982 + }, + { + "epoch": 12.65456081081081, + "grad_norm": 0.11988754570484161, + "learning_rate": 2.8340333333333334e-05, + "loss": 0.0134, + "step": 14983 + }, + { + "epoch": 12.655405405405405, + "grad_norm": 0.2314288467168808, + "learning_rate": 2.834e-05, + "loss": 0.0166, + "step": 14984 + }, + { + "epoch": 12.65625, + "grad_norm": 0.22112786769866943, + "learning_rate": 2.833966666666667e-05, + "loss": 0.0219, + "step": 14985 + }, + { + "epoch": 12.657094594594595, + "grad_norm": 0.2503642737865448, + "learning_rate": 2.833933333333333e-05, + "loss": 0.0225, + "step": 14986 + }, + { + "epoch": 12.65793918918919, + "grad_norm": 0.36024054884910583, + "learning_rate": 2.8339e-05, + "loss": 0.0201, + "step": 14987 + }, + { + "epoch": 12.658783783783784, + "grad_norm": 0.16061517596244812, + "learning_rate": 2.8338666666666667e-05, + "loss": 0.0129, + "step": 14988 + }, + { + "epoch": 12.659628378378379, + "grad_norm": 0.3555668592453003, + "learning_rate": 2.8338333333333333e-05, + "loss": 0.0281, + "step": 14989 + }, + { + "epoch": 12.660472972972974, + "grad_norm": 0.4124050736427307, + "learning_rate": 2.8338000000000002e-05, + "loss": 0.0214, + "step": 14990 + }, + { + "epoch": 12.661317567567568, + "grad_norm": 0.34966132044792175, + "learning_rate": 2.8337666666666668e-05, + "loss": 0.0189, + "step": 14991 + }, + { + "epoch": 12.662162162162161, + "grad_norm": 0.43139559030532837, + "learning_rate": 2.8337333333333334e-05, + "loss": 0.0262, + "step": 14992 + }, + { + "epoch": 12.663006756756756, + "grad_norm": 0.2916378974914551, + "learning_rate": 2.8337e-05, + "loss": 0.0144, + "step": 14993 + }, + { + "epoch": 12.66385135135135, + "grad_norm": 0.39554038643836975, + "learning_rate": 2.833666666666667e-05, + "loss": 0.0126, + "step": 14994 + }, + { + "epoch": 12.664695945945946, + "grad_norm": 0.4735702872276306, + "learning_rate": 2.8336333333333332e-05, + "loss": 0.033, + "step": 14995 + }, + { + "epoch": 12.66554054054054, + "grad_norm": 0.3246217370033264, + "learning_rate": 2.8336e-05, + "loss": 0.0199, + "step": 14996 + }, + { + "epoch": 12.666385135135135, + "grad_norm": 0.2833232283592224, + "learning_rate": 2.8335666666666667e-05, + "loss": 0.0279, + "step": 14997 + }, + { + "epoch": 12.66722972972973, + "grad_norm": 0.32848745584487915, + "learning_rate": 2.8335333333333333e-05, + "loss": 0.0298, + "step": 14998 + }, + { + "epoch": 12.668074324324325, + "grad_norm": 0.62159663438797, + "learning_rate": 2.8335000000000002e-05, + "loss": 0.0176, + "step": 14999 + }, + { + "epoch": 12.66891891891892, + "grad_norm": 0.3328787386417389, + "learning_rate": 2.8334666666666668e-05, + "loss": 0.0201, + "step": 15000 + }, + { + "epoch": 12.66891891891892, + "eval_cer": 0.09220142960404787, + "eval_loss": 0.2574966847896576, + "eval_runtime": 8.4067, + "eval_samples_per_second": 12.014, + "eval_steps_per_second": 0.476, + "eval_wer": 0.30070546737213405, + "step": 15000 + }, + { + "epoch": 12.669763513513514, + "grad_norm": 0.6326814293861389, + "learning_rate": 2.8334333333333334e-05, + "loss": 0.0341, + "step": 15001 + }, + { + "epoch": 12.670608108108109, + "grad_norm": 1.0772217512130737, + "learning_rate": 2.8334e-05, + "loss": 0.0255, + "step": 15002 + }, + { + "epoch": 12.671452702702704, + "grad_norm": 0.35435351729393005, + "learning_rate": 2.833366666666667e-05, + "loss": 0.0186, + "step": 15003 + }, + { + "epoch": 12.672297297297296, + "grad_norm": 0.41508427262306213, + "learning_rate": 2.8333333333333332e-05, + "loss": 0.0219, + "step": 15004 + }, + { + "epoch": 12.673141891891891, + "grad_norm": 0.45756468176841736, + "learning_rate": 2.8333e-05, + "loss": 0.0362, + "step": 15005 + }, + { + "epoch": 12.673986486486486, + "grad_norm": 0.37681931257247925, + "learning_rate": 2.833266666666667e-05, + "loss": 0.0243, + "step": 15006 + }, + { + "epoch": 12.67483108108108, + "grad_norm": 0.8109964728355408, + "learning_rate": 2.8332333333333333e-05, + "loss": 0.0474, + "step": 15007 + }, + { + "epoch": 12.675675675675675, + "grad_norm": 0.4983816146850586, + "learning_rate": 2.8332000000000002e-05, + "loss": 0.0291, + "step": 15008 + }, + { + "epoch": 12.67652027027027, + "grad_norm": 0.5638625025749207, + "learning_rate": 2.833166666666667e-05, + "loss": 0.1793, + "step": 15009 + }, + { + "epoch": 12.677364864864865, + "grad_norm": 0.5516330003738403, + "learning_rate": 2.8331333333333334e-05, + "loss": 0.1476, + "step": 15010 + }, + { + "epoch": 12.67820945945946, + "grad_norm": 0.27847370505332947, + "learning_rate": 2.8331e-05, + "loss": 0.0547, + "step": 15011 + }, + { + "epoch": 12.679054054054054, + "grad_norm": 0.3800394535064697, + "learning_rate": 2.8330666666666666e-05, + "loss": 0.0476, + "step": 15012 + }, + { + "epoch": 12.67989864864865, + "grad_norm": 0.46501055359840393, + "learning_rate": 2.8330333333333332e-05, + "loss": 0.0473, + "step": 15013 + }, + { + "epoch": 12.680743243243244, + "grad_norm": 0.2576601505279541, + "learning_rate": 2.833e-05, + "loss": 0.039, + "step": 15014 + }, + { + "epoch": 12.681587837837839, + "grad_norm": 0.3290341794490814, + "learning_rate": 2.8329666666666667e-05, + "loss": 0.0235, + "step": 15015 + }, + { + "epoch": 12.682432432432432, + "grad_norm": 0.1720137745141983, + "learning_rate": 2.8329333333333333e-05, + "loss": 0.0209, + "step": 15016 + }, + { + "epoch": 12.683277027027026, + "grad_norm": 0.5094953179359436, + "learning_rate": 2.8329000000000003e-05, + "loss": 0.0223, + "step": 15017 + }, + { + "epoch": 12.684121621621621, + "grad_norm": 0.29774823784828186, + "learning_rate": 2.8328666666666665e-05, + "loss": 0.0247, + "step": 15018 + }, + { + "epoch": 12.684966216216216, + "grad_norm": 0.21568231284618378, + "learning_rate": 2.8328333333333335e-05, + "loss": 0.0258, + "step": 15019 + }, + { + "epoch": 12.68581081081081, + "grad_norm": 0.34914228320121765, + "learning_rate": 2.8328e-05, + "loss": 0.0166, + "step": 15020 + }, + { + "epoch": 12.686655405405405, + "grad_norm": 0.3828310966491699, + "learning_rate": 2.8327666666666666e-05, + "loss": 0.023, + "step": 15021 + }, + { + "epoch": 12.6875, + "grad_norm": 0.21427297592163086, + "learning_rate": 2.8327333333333332e-05, + "loss": 0.0248, + "step": 15022 + }, + { + "epoch": 12.688344594594595, + "grad_norm": 0.47199907898902893, + "learning_rate": 2.8327000000000002e-05, + "loss": 0.027, + "step": 15023 + }, + { + "epoch": 12.68918918918919, + "grad_norm": 0.43606629967689514, + "learning_rate": 2.8326666666666668e-05, + "loss": 0.0226, + "step": 15024 + }, + { + "epoch": 12.690033783783784, + "grad_norm": 0.2754044830799103, + "learning_rate": 2.8326333333333334e-05, + "loss": 0.0216, + "step": 15025 + }, + { + "epoch": 12.690878378378379, + "grad_norm": 0.28095191717147827, + "learning_rate": 2.8326000000000003e-05, + "loss": 0.0195, + "step": 15026 + }, + { + "epoch": 12.691722972972974, + "grad_norm": 0.48374101519584656, + "learning_rate": 2.8325666666666665e-05, + "loss": 0.0137, + "step": 15027 + }, + { + "epoch": 12.692567567567568, + "grad_norm": 0.36700624227523804, + "learning_rate": 2.8325333333333335e-05, + "loss": 0.0244, + "step": 15028 + }, + { + "epoch": 12.693412162162161, + "grad_norm": 0.25021669268608093, + "learning_rate": 2.8325e-05, + "loss": 0.0177, + "step": 15029 + }, + { + "epoch": 12.694256756756756, + "grad_norm": 0.251939058303833, + "learning_rate": 2.8324666666666667e-05, + "loss": 0.0246, + "step": 15030 + }, + { + "epoch": 12.69510135135135, + "grad_norm": 0.509070098400116, + "learning_rate": 2.8324333333333336e-05, + "loss": 0.0279, + "step": 15031 + }, + { + "epoch": 12.695945945945946, + "grad_norm": 0.39264002442359924, + "learning_rate": 2.8324000000000002e-05, + "loss": 0.0221, + "step": 15032 + }, + { + "epoch": 12.69679054054054, + "grad_norm": 0.2906261384487152, + "learning_rate": 2.8323666666666668e-05, + "loss": 0.0175, + "step": 15033 + }, + { + "epoch": 12.697635135135135, + "grad_norm": 0.4989927411079407, + "learning_rate": 2.8323333333333334e-05, + "loss": 0.027, + "step": 15034 + }, + { + "epoch": 12.69847972972973, + "grad_norm": 0.2530323266983032, + "learning_rate": 2.8323000000000003e-05, + "loss": 0.0153, + "step": 15035 + }, + { + "epoch": 12.699324324324325, + "grad_norm": 0.38573014736175537, + "learning_rate": 2.8322666666666666e-05, + "loss": 0.0199, + "step": 15036 + }, + { + "epoch": 12.70016891891892, + "grad_norm": 0.33361679315567017, + "learning_rate": 2.8322333333333335e-05, + "loss": 0.0321, + "step": 15037 + }, + { + "epoch": 12.701013513513514, + "grad_norm": 0.7078641057014465, + "learning_rate": 2.8322e-05, + "loss": 0.0214, + "step": 15038 + }, + { + "epoch": 12.701858108108109, + "grad_norm": 0.26415565609931946, + "learning_rate": 2.8321666666666667e-05, + "loss": 0.0165, + "step": 15039 + }, + { + "epoch": 12.702702702702704, + "grad_norm": 0.5878620147705078, + "learning_rate": 2.8321333333333336e-05, + "loss": 0.0197, + "step": 15040 + }, + { + "epoch": 12.703547297297296, + "grad_norm": 0.2775714099407196, + "learning_rate": 2.8321e-05, + "loss": 0.0203, + "step": 15041 + }, + { + "epoch": 12.704391891891891, + "grad_norm": 0.36838498711586, + "learning_rate": 2.8320666666666668e-05, + "loss": 0.0223, + "step": 15042 + }, + { + "epoch": 12.705236486486486, + "grad_norm": 0.3892382085323334, + "learning_rate": 2.8320333333333334e-05, + "loss": 0.0307, + "step": 15043 + }, + { + "epoch": 12.70608108108108, + "grad_norm": 0.1988212913274765, + "learning_rate": 2.832e-05, + "loss": 0.0192, + "step": 15044 + }, + { + "epoch": 12.706925675675675, + "grad_norm": 0.4764004647731781, + "learning_rate": 2.8319666666666666e-05, + "loss": 0.0214, + "step": 15045 + }, + { + "epoch": 12.70777027027027, + "grad_norm": 0.1469837725162506, + "learning_rate": 2.8319333333333335e-05, + "loss": 0.0112, + "step": 15046 + }, + { + "epoch": 12.708614864864865, + "grad_norm": 0.3168465793132782, + "learning_rate": 2.8318999999999998e-05, + "loss": 0.0248, + "step": 15047 + }, + { + "epoch": 12.70945945945946, + "grad_norm": 0.3637043535709381, + "learning_rate": 2.8318666666666667e-05, + "loss": 0.0265, + "step": 15048 + }, + { + "epoch": 12.710304054054054, + "grad_norm": 0.510797381401062, + "learning_rate": 2.8318333333333336e-05, + "loss": 0.0307, + "step": 15049 + }, + { + "epoch": 12.71114864864865, + "grad_norm": 0.30376744270324707, + "learning_rate": 2.8318e-05, + "loss": 0.0229, + "step": 15050 + }, + { + "epoch": 12.711993243243244, + "grad_norm": 0.2651015818119049, + "learning_rate": 2.8317666666666668e-05, + "loss": 0.0114, + "step": 15051 + }, + { + "epoch": 12.712837837837839, + "grad_norm": 0.4465654194355011, + "learning_rate": 2.8317333333333334e-05, + "loss": 0.0238, + "step": 15052 + }, + { + "epoch": 12.713682432432432, + "grad_norm": 0.42797982692718506, + "learning_rate": 2.8317e-05, + "loss": 0.0267, + "step": 15053 + }, + { + "epoch": 12.714527027027026, + "grad_norm": 0.5875527262687683, + "learning_rate": 2.8316666666666666e-05, + "loss": 0.0344, + "step": 15054 + }, + { + "epoch": 12.715371621621621, + "grad_norm": 0.47251635789871216, + "learning_rate": 2.8316333333333335e-05, + "loss": 0.0226, + "step": 15055 + }, + { + "epoch": 12.716216216216216, + "grad_norm": 0.6565033793449402, + "learning_rate": 2.8316e-05, + "loss": 0.0266, + "step": 15056 + }, + { + "epoch": 12.71706081081081, + "grad_norm": 0.22564230859279633, + "learning_rate": 2.8315666666666667e-05, + "loss": 0.021, + "step": 15057 + }, + { + "epoch": 12.717905405405405, + "grad_norm": 0.7443779110908508, + "learning_rate": 2.8315333333333337e-05, + "loss": 0.028, + "step": 15058 + }, + { + "epoch": 12.71875, + "grad_norm": 0.772085428237915, + "learning_rate": 2.8315e-05, + "loss": 0.2595, + "step": 15059 + }, + { + "epoch": 12.719594594594595, + "grad_norm": 0.5060568451881409, + "learning_rate": 2.831466666666667e-05, + "loss": 0.1146, + "step": 15060 + }, + { + "epoch": 12.72043918918919, + "grad_norm": 0.4227498471736908, + "learning_rate": 2.8314333333333334e-05, + "loss": 0.0593, + "step": 15061 + }, + { + "epoch": 12.721283783783784, + "grad_norm": 0.3143518269062042, + "learning_rate": 2.8314e-05, + "loss": 0.0341, + "step": 15062 + }, + { + "epoch": 12.722128378378379, + "grad_norm": 0.2858039438724518, + "learning_rate": 2.8313666666666666e-05, + "loss": 0.0364, + "step": 15063 + }, + { + "epoch": 12.722972972972974, + "grad_norm": 0.6154340505599976, + "learning_rate": 2.8313333333333336e-05, + "loss": 0.0333, + "step": 15064 + }, + { + "epoch": 12.723817567567568, + "grad_norm": 0.21845683455467224, + "learning_rate": 2.8313e-05, + "loss": 0.0202, + "step": 15065 + }, + { + "epoch": 12.724662162162161, + "grad_norm": 0.37755221128463745, + "learning_rate": 2.8312666666666667e-05, + "loss": 0.0281, + "step": 15066 + }, + { + "epoch": 12.725506756756756, + "grad_norm": 0.21713228523731232, + "learning_rate": 2.8312333333333337e-05, + "loss": 0.0208, + "step": 15067 + }, + { + "epoch": 12.72635135135135, + "grad_norm": 0.4004918336868286, + "learning_rate": 2.8312e-05, + "loss": 0.0453, + "step": 15068 + }, + { + "epoch": 12.727195945945946, + "grad_norm": 0.2983355224132538, + "learning_rate": 2.831166666666667e-05, + "loss": 0.0227, + "step": 15069 + }, + { + "epoch": 12.72804054054054, + "grad_norm": 0.23310789465904236, + "learning_rate": 2.831133333333333e-05, + "loss": 0.0189, + "step": 15070 + }, + { + "epoch": 12.728885135135135, + "grad_norm": 0.26457417011260986, + "learning_rate": 2.8311e-05, + "loss": 0.028, + "step": 15071 + }, + { + "epoch": 12.72972972972973, + "grad_norm": 0.2424260377883911, + "learning_rate": 2.8310666666666666e-05, + "loss": 0.0265, + "step": 15072 + }, + { + "epoch": 12.730574324324325, + "grad_norm": 0.17532537877559662, + "learning_rate": 2.8310333333333332e-05, + "loss": 0.015, + "step": 15073 + }, + { + "epoch": 12.73141891891892, + "grad_norm": 0.236322283744812, + "learning_rate": 2.8310000000000002e-05, + "loss": 0.0165, + "step": 15074 + }, + { + "epoch": 12.732263513513514, + "grad_norm": 0.2404046505689621, + "learning_rate": 2.8309666666666668e-05, + "loss": 0.018, + "step": 15075 + }, + { + "epoch": 12.733108108108109, + "grad_norm": 0.3324275314807892, + "learning_rate": 2.8309333333333334e-05, + "loss": 0.0236, + "step": 15076 + }, + { + "epoch": 12.733952702702704, + "grad_norm": 0.29665303230285645, + "learning_rate": 2.8309e-05, + "loss": 0.0207, + "step": 15077 + }, + { + "epoch": 12.734797297297296, + "grad_norm": 0.33040085434913635, + "learning_rate": 2.830866666666667e-05, + "loss": 0.0219, + "step": 15078 + }, + { + "epoch": 12.735641891891891, + "grad_norm": 0.35195669531822205, + "learning_rate": 2.830833333333333e-05, + "loss": 0.029, + "step": 15079 + }, + { + "epoch": 12.736486486486486, + "grad_norm": 0.33745744824409485, + "learning_rate": 2.8308e-05, + "loss": 0.0399, + "step": 15080 + }, + { + "epoch": 12.73733108108108, + "grad_norm": 0.17295444011688232, + "learning_rate": 2.830766666666667e-05, + "loss": 0.0158, + "step": 15081 + }, + { + "epoch": 12.738175675675675, + "grad_norm": 0.3703295886516571, + "learning_rate": 2.8307333333333333e-05, + "loss": 0.0201, + "step": 15082 + }, + { + "epoch": 12.73902027027027, + "grad_norm": 0.39759448170661926, + "learning_rate": 2.8307000000000002e-05, + "loss": 0.0197, + "step": 15083 + }, + { + "epoch": 12.739864864864865, + "grad_norm": 0.32165849208831787, + "learning_rate": 2.8306666666666668e-05, + "loss": 0.0214, + "step": 15084 + }, + { + "epoch": 12.74070945945946, + "grad_norm": 0.30374184250831604, + "learning_rate": 2.8306333333333334e-05, + "loss": 0.0215, + "step": 15085 + }, + { + "epoch": 12.741554054054054, + "grad_norm": 0.3173753619194031, + "learning_rate": 2.8306e-05, + "loss": 0.0143, + "step": 15086 + }, + { + "epoch": 12.74239864864865, + "grad_norm": 0.15441378951072693, + "learning_rate": 2.830566666666667e-05, + "loss": 0.0132, + "step": 15087 + }, + { + "epoch": 12.743243243243244, + "grad_norm": 0.4593358635902405, + "learning_rate": 2.830533333333333e-05, + "loss": 0.0206, + "step": 15088 + }, + { + "epoch": 12.744087837837839, + "grad_norm": 0.32563039660453796, + "learning_rate": 2.8305e-05, + "loss": 0.0187, + "step": 15089 + }, + { + "epoch": 12.744932432432432, + "grad_norm": 0.23580902814865112, + "learning_rate": 2.830466666666667e-05, + "loss": 0.0172, + "step": 15090 + }, + { + "epoch": 12.745777027027026, + "grad_norm": 0.5696074962615967, + "learning_rate": 2.8304333333333333e-05, + "loss": 0.0279, + "step": 15091 + }, + { + "epoch": 12.746621621621621, + "grad_norm": 0.3511073887348175, + "learning_rate": 2.8304000000000002e-05, + "loss": 0.0222, + "step": 15092 + }, + { + "epoch": 12.747466216216216, + "grad_norm": 0.23692025244235992, + "learning_rate": 2.8303666666666668e-05, + "loss": 0.0107, + "step": 15093 + }, + { + "epoch": 12.74831081081081, + "grad_norm": 0.3187984228134155, + "learning_rate": 2.8303333333333334e-05, + "loss": 0.0196, + "step": 15094 + }, + { + "epoch": 12.749155405405405, + "grad_norm": 0.424081027507782, + "learning_rate": 2.8303e-05, + "loss": 0.0208, + "step": 15095 + }, + { + "epoch": 12.75, + "grad_norm": 0.34311747550964355, + "learning_rate": 2.830266666666667e-05, + "loss": 0.0308, + "step": 15096 + }, + { + "epoch": 12.750844594594595, + "grad_norm": 0.6613331437110901, + "learning_rate": 2.8302333333333332e-05, + "loss": 0.0266, + "step": 15097 + }, + { + "epoch": 12.75168918918919, + "grad_norm": 0.2181309014558792, + "learning_rate": 2.8302e-05, + "loss": 0.0169, + "step": 15098 + }, + { + "epoch": 12.752533783783784, + "grad_norm": 0.4273007810115814, + "learning_rate": 2.830166666666667e-05, + "loss": 0.0237, + "step": 15099 + }, + { + "epoch": 12.753378378378379, + "grad_norm": 0.3754539489746094, + "learning_rate": 2.8301333333333333e-05, + "loss": 0.0178, + "step": 15100 + }, + { + "epoch": 12.754222972972974, + "grad_norm": 0.7188889980316162, + "learning_rate": 2.8301000000000002e-05, + "loss": 0.0183, + "step": 15101 + }, + { + "epoch": 12.755067567567568, + "grad_norm": 0.3213954269886017, + "learning_rate": 2.8300666666666665e-05, + "loss": 0.0227, + "step": 15102 + }, + { + "epoch": 12.755912162162161, + "grad_norm": 0.4332452118396759, + "learning_rate": 2.8300333333333334e-05, + "loss": 0.0284, + "step": 15103 + }, + { + "epoch": 12.756756756756756, + "grad_norm": 0.3329532742500305, + "learning_rate": 2.83e-05, + "loss": 0.0239, + "step": 15104 + }, + { + "epoch": 12.75760135135135, + "grad_norm": 0.3904661536216736, + "learning_rate": 2.8299666666666666e-05, + "loss": 0.0206, + "step": 15105 + }, + { + "epoch": 12.758445945945946, + "grad_norm": 0.3326680064201355, + "learning_rate": 2.8299333333333335e-05, + "loss": 0.0277, + "step": 15106 + }, + { + "epoch": 12.75929054054054, + "grad_norm": 0.8711507320404053, + "learning_rate": 2.8299e-05, + "loss": 0.0492, + "step": 15107 + }, + { + "epoch": 12.760135135135135, + "grad_norm": 0.9833112955093384, + "learning_rate": 2.8298666666666667e-05, + "loss": 0.0667, + "step": 15108 + }, + { + "epoch": 12.76097972972973, + "grad_norm": 0.6823424100875854, + "learning_rate": 2.8298333333333333e-05, + "loss": 0.1955, + "step": 15109 + }, + { + "epoch": 12.761824324324325, + "grad_norm": 0.7076215147972107, + "learning_rate": 2.8298000000000002e-05, + "loss": 0.197, + "step": 15110 + }, + { + "epoch": 12.76266891891892, + "grad_norm": 0.34051987528800964, + "learning_rate": 2.8297666666666665e-05, + "loss": 0.054, + "step": 15111 + }, + { + "epoch": 12.763513513513514, + "grad_norm": 0.2051030993461609, + "learning_rate": 2.8297333333333334e-05, + "loss": 0.0276, + "step": 15112 + }, + { + "epoch": 12.764358108108109, + "grad_norm": 0.4169202148914337, + "learning_rate": 2.8297e-05, + "loss": 0.0448, + "step": 15113 + }, + { + "epoch": 12.765202702702704, + "grad_norm": 0.27533915638923645, + "learning_rate": 2.8296666666666666e-05, + "loss": 0.0258, + "step": 15114 + }, + { + "epoch": 12.766047297297296, + "grad_norm": 0.3769071698188782, + "learning_rate": 2.8296333333333336e-05, + "loss": 0.0198, + "step": 15115 + }, + { + "epoch": 12.766891891891891, + "grad_norm": 0.2748459577560425, + "learning_rate": 2.8296e-05, + "loss": 0.0296, + "step": 15116 + }, + { + "epoch": 12.767736486486486, + "grad_norm": 0.16949410736560822, + "learning_rate": 2.8295666666666667e-05, + "loss": 0.0147, + "step": 15117 + }, + { + "epoch": 12.76858108108108, + "grad_norm": 0.21878044307231903, + "learning_rate": 2.8295333333333333e-05, + "loss": 0.0179, + "step": 15118 + }, + { + "epoch": 12.769425675675675, + "grad_norm": 0.21285048127174377, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.02, + "step": 15119 + }, + { + "epoch": 12.77027027027027, + "grad_norm": 0.27930113673210144, + "learning_rate": 2.8294666666666665e-05, + "loss": 0.0206, + "step": 15120 + }, + { + "epoch": 12.771114864864865, + "grad_norm": 0.29860401153564453, + "learning_rate": 2.8294333333333335e-05, + "loss": 0.023, + "step": 15121 + }, + { + "epoch": 12.77195945945946, + "grad_norm": 0.3095928430557251, + "learning_rate": 2.8294e-05, + "loss": 0.023, + "step": 15122 + }, + { + "epoch": 12.772804054054054, + "grad_norm": 0.40246936678886414, + "learning_rate": 2.8293666666666666e-05, + "loss": 0.0235, + "step": 15123 + }, + { + "epoch": 12.77364864864865, + "grad_norm": 0.19923821091651917, + "learning_rate": 2.8293333333333336e-05, + "loss": 0.0142, + "step": 15124 + }, + { + "epoch": 12.774493243243244, + "grad_norm": 0.36072856187820435, + "learning_rate": 2.8293e-05, + "loss": 0.0235, + "step": 15125 + }, + { + "epoch": 12.775337837837839, + "grad_norm": 0.3156173825263977, + "learning_rate": 2.8292666666666668e-05, + "loss": 0.0191, + "step": 15126 + }, + { + "epoch": 12.776182432432432, + "grad_norm": 0.6561729907989502, + "learning_rate": 2.8292333333333334e-05, + "loss": 0.0468, + "step": 15127 + }, + { + "epoch": 12.777027027027026, + "grad_norm": 0.47365209460258484, + "learning_rate": 2.8292000000000003e-05, + "loss": 0.022, + "step": 15128 + }, + { + "epoch": 12.777871621621621, + "grad_norm": 0.3336365222930908, + "learning_rate": 2.8291666666666665e-05, + "loss": 0.0198, + "step": 15129 + }, + { + "epoch": 12.778716216216216, + "grad_norm": 1.022221565246582, + "learning_rate": 2.8291333333333335e-05, + "loss": 0.0332, + "step": 15130 + }, + { + "epoch": 12.77956081081081, + "grad_norm": 0.493819922208786, + "learning_rate": 2.8291e-05, + "loss": 0.0206, + "step": 15131 + }, + { + "epoch": 12.780405405405405, + "grad_norm": 0.22920246422290802, + "learning_rate": 2.8290666666666667e-05, + "loss": 0.0159, + "step": 15132 + }, + { + "epoch": 12.78125, + "grad_norm": 0.5415123701095581, + "learning_rate": 2.8290333333333336e-05, + "loss": 0.0214, + "step": 15133 + }, + { + "epoch": 12.782094594594595, + "grad_norm": 0.29689544439315796, + "learning_rate": 2.829e-05, + "loss": 0.0171, + "step": 15134 + }, + { + "epoch": 12.78293918918919, + "grad_norm": 0.43709883093833923, + "learning_rate": 2.8289666666666668e-05, + "loss": 0.0189, + "step": 15135 + }, + { + "epoch": 12.783783783783784, + "grad_norm": 0.25490593910217285, + "learning_rate": 2.8289333333333334e-05, + "loss": 0.0116, + "step": 15136 + }, + { + "epoch": 12.784628378378379, + "grad_norm": 0.2566286623477936, + "learning_rate": 2.8289e-05, + "loss": 0.0149, + "step": 15137 + }, + { + "epoch": 12.785472972972974, + "grad_norm": 0.37905651330947876, + "learning_rate": 2.8288666666666666e-05, + "loss": 0.0178, + "step": 15138 + }, + { + "epoch": 12.786317567567568, + "grad_norm": 0.4141225814819336, + "learning_rate": 2.8288333333333335e-05, + "loss": 0.021, + "step": 15139 + }, + { + "epoch": 12.787162162162161, + "grad_norm": 0.4986344575881958, + "learning_rate": 2.8288e-05, + "loss": 0.0224, + "step": 15140 + }, + { + "epoch": 12.788006756756756, + "grad_norm": 0.1338685005903244, + "learning_rate": 2.8287666666666667e-05, + "loss": 0.0129, + "step": 15141 + }, + { + "epoch": 12.78885135135135, + "grad_norm": 0.3928334712982178, + "learning_rate": 2.8287333333333336e-05, + "loss": 0.0204, + "step": 15142 + }, + { + "epoch": 12.789695945945946, + "grad_norm": 0.42170071601867676, + "learning_rate": 2.8287e-05, + "loss": 0.0303, + "step": 15143 + }, + { + "epoch": 12.79054054054054, + "grad_norm": 0.3186054229736328, + "learning_rate": 2.8286666666666668e-05, + "loss": 0.0224, + "step": 15144 + }, + { + "epoch": 12.791385135135135, + "grad_norm": 0.3368656635284424, + "learning_rate": 2.8286333333333334e-05, + "loss": 0.0174, + "step": 15145 + }, + { + "epoch": 12.79222972972973, + "grad_norm": 0.2936509847640991, + "learning_rate": 2.8286e-05, + "loss": 0.0285, + "step": 15146 + }, + { + "epoch": 12.793074324324325, + "grad_norm": 0.33600059151649475, + "learning_rate": 2.8285666666666666e-05, + "loss": 0.017, + "step": 15147 + }, + { + "epoch": 12.79391891891892, + "grad_norm": 0.4965232312679291, + "learning_rate": 2.8285333333333335e-05, + "loss": 0.0141, + "step": 15148 + }, + { + "epoch": 12.794763513513514, + "grad_norm": 0.23200002312660217, + "learning_rate": 2.8285e-05, + "loss": 0.0179, + "step": 15149 + }, + { + "epoch": 12.795608108108109, + "grad_norm": 0.36552825570106506, + "learning_rate": 2.8284666666666667e-05, + "loss": 0.0226, + "step": 15150 + }, + { + "epoch": 12.796452702702704, + "grad_norm": 0.3257399797439575, + "learning_rate": 2.8284333333333336e-05, + "loss": 0.0251, + "step": 15151 + }, + { + "epoch": 12.797297297297296, + "grad_norm": 0.3518884778022766, + "learning_rate": 2.8284e-05, + "loss": 0.0178, + "step": 15152 + }, + { + "epoch": 12.798141891891891, + "grad_norm": 0.28954094648361206, + "learning_rate": 2.8283666666666668e-05, + "loss": 0.0179, + "step": 15153 + }, + { + "epoch": 12.798986486486486, + "grad_norm": 0.4084048569202423, + "learning_rate": 2.8283333333333334e-05, + "loss": 0.0227, + "step": 15154 + }, + { + "epoch": 12.79983108108108, + "grad_norm": 0.402720183134079, + "learning_rate": 2.8283e-05, + "loss": 0.0207, + "step": 15155 + }, + { + "epoch": 12.800675675675675, + "grad_norm": 0.3020060956478119, + "learning_rate": 2.828266666666667e-05, + "loss": 0.0223, + "step": 15156 + }, + { + "epoch": 12.80152027027027, + "grad_norm": 0.2900524437427521, + "learning_rate": 2.8282333333333335e-05, + "loss": 0.0193, + "step": 15157 + }, + { + "epoch": 12.802364864864865, + "grad_norm": 0.6602749824523926, + "learning_rate": 2.8282e-05, + "loss": 0.0307, + "step": 15158 + }, + { + "epoch": 12.80320945945946, + "grad_norm": 0.71327805519104, + "learning_rate": 2.8281666666666667e-05, + "loss": 0.2007, + "step": 15159 + }, + { + "epoch": 12.804054054054054, + "grad_norm": 0.6405662298202515, + "learning_rate": 2.8281333333333337e-05, + "loss": 0.129, + "step": 15160 + }, + { + "epoch": 12.80489864864865, + "grad_norm": 0.7471494674682617, + "learning_rate": 2.8281e-05, + "loss": 0.0579, + "step": 15161 + }, + { + "epoch": 12.805743243243244, + "grad_norm": 0.4732683002948761, + "learning_rate": 2.828066666666667e-05, + "loss": 0.052, + "step": 15162 + }, + { + "epoch": 12.806587837837839, + "grad_norm": 0.3792153000831604, + "learning_rate": 2.828033333333333e-05, + "loss": 0.0243, + "step": 15163 + }, + { + "epoch": 12.807432432432432, + "grad_norm": 0.4876505434513092, + "learning_rate": 2.828e-05, + "loss": 0.0328, + "step": 15164 + }, + { + "epoch": 12.808277027027026, + "grad_norm": 0.39124709367752075, + "learning_rate": 2.827966666666667e-05, + "loss": 0.0299, + "step": 15165 + }, + { + "epoch": 12.809121621621621, + "grad_norm": 0.3517492711544037, + "learning_rate": 2.8279333333333332e-05, + "loss": 0.0277, + "step": 15166 + }, + { + "epoch": 12.809966216216216, + "grad_norm": 0.2290593534708023, + "learning_rate": 2.8279e-05, + "loss": 0.0273, + "step": 15167 + }, + { + "epoch": 12.81081081081081, + "grad_norm": 0.23892809450626373, + "learning_rate": 2.8278666666666667e-05, + "loss": 0.0235, + "step": 15168 + }, + { + "epoch": 12.811655405405405, + "grad_norm": 0.475658118724823, + "learning_rate": 2.8278333333333333e-05, + "loss": 0.023, + "step": 15169 + }, + { + "epoch": 12.8125, + "grad_norm": 0.6040613651275635, + "learning_rate": 2.8278e-05, + "loss": 0.0241, + "step": 15170 + }, + { + "epoch": 12.813344594594595, + "grad_norm": 0.2755105197429657, + "learning_rate": 2.827766666666667e-05, + "loss": 0.0319, + "step": 15171 + }, + { + "epoch": 12.81418918918919, + "grad_norm": 0.266318142414093, + "learning_rate": 2.827733333333333e-05, + "loss": 0.0203, + "step": 15172 + }, + { + "epoch": 12.815033783783784, + "grad_norm": 0.2229698747396469, + "learning_rate": 2.8277e-05, + "loss": 0.0151, + "step": 15173 + }, + { + "epoch": 12.815878378378379, + "grad_norm": 0.2099798321723938, + "learning_rate": 2.827666666666667e-05, + "loss": 0.0198, + "step": 15174 + }, + { + "epoch": 12.816722972972974, + "grad_norm": 0.36430391669273376, + "learning_rate": 2.8276333333333332e-05, + "loss": 0.0289, + "step": 15175 + }, + { + "epoch": 12.817567567567568, + "grad_norm": 0.2079252451658249, + "learning_rate": 2.8276e-05, + "loss": 0.0136, + "step": 15176 + }, + { + "epoch": 12.818412162162161, + "grad_norm": 0.26105257868766785, + "learning_rate": 2.8275666666666668e-05, + "loss": 0.024, + "step": 15177 + }, + { + "epoch": 12.819256756756756, + "grad_norm": 0.21133974194526672, + "learning_rate": 2.8275333333333334e-05, + "loss": 0.0334, + "step": 15178 + }, + { + "epoch": 12.82010135135135, + "grad_norm": 0.23534099757671356, + "learning_rate": 2.8275e-05, + "loss": 0.0207, + "step": 15179 + }, + { + "epoch": 12.820945945945946, + "grad_norm": 0.27990788221359253, + "learning_rate": 2.827466666666667e-05, + "loss": 0.0264, + "step": 15180 + }, + { + "epoch": 12.82179054054054, + "grad_norm": 0.18689344823360443, + "learning_rate": 2.8274333333333335e-05, + "loss": 0.0217, + "step": 15181 + }, + { + "epoch": 12.822635135135135, + "grad_norm": 0.1412864327430725, + "learning_rate": 2.8274e-05, + "loss": 0.0114, + "step": 15182 + }, + { + "epoch": 12.82347972972973, + "grad_norm": 0.29447343945503235, + "learning_rate": 2.827366666666667e-05, + "loss": 0.0456, + "step": 15183 + }, + { + "epoch": 12.824324324324325, + "grad_norm": 0.29761165380477905, + "learning_rate": 2.8273333333333333e-05, + "loss": 0.0193, + "step": 15184 + }, + { + "epoch": 12.82516891891892, + "grad_norm": 0.2731521725654602, + "learning_rate": 2.8273000000000002e-05, + "loss": 0.026, + "step": 15185 + }, + { + "epoch": 12.826013513513514, + "grad_norm": 0.2574947476387024, + "learning_rate": 2.8272666666666668e-05, + "loss": 0.0197, + "step": 15186 + }, + { + "epoch": 12.826858108108109, + "grad_norm": 0.32427629828453064, + "learning_rate": 2.8272333333333334e-05, + "loss": 0.0225, + "step": 15187 + }, + { + "epoch": 12.827702702702704, + "grad_norm": 0.2770855724811554, + "learning_rate": 2.8272e-05, + "loss": 0.0289, + "step": 15188 + }, + { + "epoch": 12.828547297297296, + "grad_norm": 0.29412034153938293, + "learning_rate": 2.827166666666667e-05, + "loss": 0.016, + "step": 15189 + }, + { + "epoch": 12.829391891891891, + "grad_norm": 0.2845584452152252, + "learning_rate": 2.8271333333333335e-05, + "loss": 0.0175, + "step": 15190 + }, + { + "epoch": 12.830236486486486, + "grad_norm": 0.3215128481388092, + "learning_rate": 2.8271e-05, + "loss": 0.0149, + "step": 15191 + }, + { + "epoch": 12.83108108108108, + "grad_norm": 0.3648020923137665, + "learning_rate": 2.8270666666666667e-05, + "loss": 0.0284, + "step": 15192 + }, + { + "epoch": 12.831925675675675, + "grad_norm": 0.43006762862205505, + "learning_rate": 2.8270333333333333e-05, + "loss": 0.0242, + "step": 15193 + }, + { + "epoch": 12.83277027027027, + "grad_norm": 0.27138352394104004, + "learning_rate": 2.8270000000000002e-05, + "loss": 0.0128, + "step": 15194 + }, + { + "epoch": 12.833614864864865, + "grad_norm": 0.2372065782546997, + "learning_rate": 2.8269666666666665e-05, + "loss": 0.0173, + "step": 15195 + }, + { + "epoch": 12.83445945945946, + "grad_norm": 0.4809146523475647, + "learning_rate": 2.8269333333333334e-05, + "loss": 0.0135, + "step": 15196 + }, + { + "epoch": 12.835304054054054, + "grad_norm": 0.3738202750682831, + "learning_rate": 2.8269e-05, + "loss": 0.0278, + "step": 15197 + }, + { + "epoch": 12.83614864864865, + "grad_norm": 0.39466941356658936, + "learning_rate": 2.8268666666666666e-05, + "loss": 0.0344, + "step": 15198 + }, + { + "epoch": 12.836993243243244, + "grad_norm": 0.43987250328063965, + "learning_rate": 2.8268333333333335e-05, + "loss": 0.0242, + "step": 15199 + }, + { + "epoch": 12.837837837837839, + "grad_norm": 0.28748244047164917, + "learning_rate": 2.8268e-05, + "loss": 0.0212, + "step": 15200 + }, + { + "epoch": 12.838682432432432, + "grad_norm": 0.2385370433330536, + "learning_rate": 2.8267666666666667e-05, + "loss": 0.0179, + "step": 15201 + }, + { + "epoch": 12.839527027027026, + "grad_norm": 0.41385018825531006, + "learning_rate": 2.8267333333333333e-05, + "loss": 0.0644, + "step": 15202 + }, + { + "epoch": 12.840371621621621, + "grad_norm": 0.18079163134098053, + "learning_rate": 2.8267000000000002e-05, + "loss": 0.0166, + "step": 15203 + }, + { + "epoch": 12.841216216216216, + "grad_norm": 0.30817121267318726, + "learning_rate": 2.8266666666666665e-05, + "loss": 0.0225, + "step": 15204 + }, + { + "epoch": 12.84206081081081, + "grad_norm": 0.418559730052948, + "learning_rate": 2.8266333333333334e-05, + "loss": 0.0372, + "step": 15205 + }, + { + "epoch": 12.842905405405405, + "grad_norm": 0.4709526598453522, + "learning_rate": 2.8266000000000003e-05, + "loss": 0.0241, + "step": 15206 + }, + { + "epoch": 12.84375, + "grad_norm": 0.2020387500524521, + "learning_rate": 2.8265666666666666e-05, + "loss": 0.0191, + "step": 15207 + }, + { + "epoch": 12.844594594594595, + "grad_norm": 0.7544375658035278, + "learning_rate": 2.8265333333333335e-05, + "loss": 0.0386, + "step": 15208 + }, + { + "epoch": 12.84543918918919, + "grad_norm": 0.6373528838157654, + "learning_rate": 2.8265e-05, + "loss": 0.2007, + "step": 15209 + }, + { + "epoch": 12.846283783783784, + "grad_norm": 0.3916146755218506, + "learning_rate": 2.8264666666666667e-05, + "loss": 0.1207, + "step": 15210 + }, + { + "epoch": 12.847128378378379, + "grad_norm": 0.30850282311439514, + "learning_rate": 2.8264333333333333e-05, + "loss": 0.0693, + "step": 15211 + }, + { + "epoch": 12.847972972972974, + "grad_norm": 0.3040619492530823, + "learning_rate": 2.8264000000000002e-05, + "loss": 0.0407, + "step": 15212 + }, + { + "epoch": 12.848817567567568, + "grad_norm": 0.2923494279384613, + "learning_rate": 2.8263666666666665e-05, + "loss": 0.0266, + "step": 15213 + }, + { + "epoch": 12.849662162162161, + "grad_norm": 0.2819957137107849, + "learning_rate": 2.8263333333333334e-05, + "loss": 0.027, + "step": 15214 + }, + { + "epoch": 12.850506756756756, + "grad_norm": 0.21674203872680664, + "learning_rate": 2.8263000000000004e-05, + "loss": 0.0373, + "step": 15215 + }, + { + "epoch": 12.85135135135135, + "grad_norm": 0.28111740946769714, + "learning_rate": 2.8262666666666666e-05, + "loss": 0.026, + "step": 15216 + }, + { + "epoch": 12.852195945945946, + "grad_norm": 0.21760018169879913, + "learning_rate": 2.8262333333333336e-05, + "loss": 0.0197, + "step": 15217 + }, + { + "epoch": 12.85304054054054, + "grad_norm": 0.21965532004833221, + "learning_rate": 2.8262e-05, + "loss": 0.0195, + "step": 15218 + }, + { + "epoch": 12.853885135135135, + "grad_norm": 0.7356077432632446, + "learning_rate": 2.8261666666666667e-05, + "loss": 0.0379, + "step": 15219 + }, + { + "epoch": 12.85472972972973, + "grad_norm": 0.24454060196876526, + "learning_rate": 2.8261333333333333e-05, + "loss": 0.0181, + "step": 15220 + }, + { + "epoch": 12.855574324324325, + "grad_norm": 0.33258718252182007, + "learning_rate": 2.8261e-05, + "loss": 0.0365, + "step": 15221 + }, + { + "epoch": 12.85641891891892, + "grad_norm": 0.27849018573760986, + "learning_rate": 2.8260666666666665e-05, + "loss": 0.0179, + "step": 15222 + }, + { + "epoch": 12.857263513513514, + "grad_norm": 0.14424030482769012, + "learning_rate": 2.8260333333333335e-05, + "loss": 0.0118, + "step": 15223 + }, + { + "epoch": 12.858108108108109, + "grad_norm": 0.1721939742565155, + "learning_rate": 2.826e-05, + "loss": 0.014, + "step": 15224 + }, + { + "epoch": 12.858952702702704, + "grad_norm": 0.3949067294597626, + "learning_rate": 2.8259666666666666e-05, + "loss": 0.0228, + "step": 15225 + }, + { + "epoch": 12.859797297297296, + "grad_norm": 0.20911461114883423, + "learning_rate": 2.8259333333333336e-05, + "loss": 0.0151, + "step": 15226 + }, + { + "epoch": 12.860641891891891, + "grad_norm": 0.3321620523929596, + "learning_rate": 2.8258999999999998e-05, + "loss": 0.0214, + "step": 15227 + }, + { + "epoch": 12.861486486486486, + "grad_norm": 0.27956002950668335, + "learning_rate": 2.8258666666666668e-05, + "loss": 0.019, + "step": 15228 + }, + { + "epoch": 12.86233108108108, + "grad_norm": 0.26359400153160095, + "learning_rate": 2.8258333333333334e-05, + "loss": 0.0153, + "step": 15229 + }, + { + "epoch": 12.863175675675675, + "grad_norm": 0.25361981987953186, + "learning_rate": 2.8258e-05, + "loss": 0.0218, + "step": 15230 + }, + { + "epoch": 12.86402027027027, + "grad_norm": 0.2861981689929962, + "learning_rate": 2.825766666666667e-05, + "loss": 0.016, + "step": 15231 + }, + { + "epoch": 12.864864864864865, + "grad_norm": 0.5017393231391907, + "learning_rate": 2.8257333333333335e-05, + "loss": 0.0232, + "step": 15232 + }, + { + "epoch": 12.86570945945946, + "grad_norm": 0.14477689564228058, + "learning_rate": 2.8257e-05, + "loss": 0.0097, + "step": 15233 + }, + { + "epoch": 12.866554054054054, + "grad_norm": 0.39715254306793213, + "learning_rate": 2.8256666666666667e-05, + "loss": 0.0339, + "step": 15234 + }, + { + "epoch": 12.86739864864865, + "grad_norm": 0.2665027678012848, + "learning_rate": 2.8256333333333336e-05, + "loss": 0.0207, + "step": 15235 + }, + { + "epoch": 12.868243243243244, + "grad_norm": 0.3251749277114868, + "learning_rate": 2.8256e-05, + "loss": 0.0202, + "step": 15236 + }, + { + "epoch": 12.869087837837839, + "grad_norm": 0.5246041417121887, + "learning_rate": 2.8255666666666668e-05, + "loss": 0.018, + "step": 15237 + }, + { + "epoch": 12.869932432432432, + "grad_norm": 0.4488825798034668, + "learning_rate": 2.8255333333333334e-05, + "loss": 0.0227, + "step": 15238 + }, + { + "epoch": 12.870777027027026, + "grad_norm": 0.4022410809993744, + "learning_rate": 2.8255e-05, + "loss": 0.0179, + "step": 15239 + }, + { + "epoch": 12.871621621621621, + "grad_norm": 0.30367159843444824, + "learning_rate": 2.825466666666667e-05, + "loss": 0.0149, + "step": 15240 + }, + { + "epoch": 12.872466216216216, + "grad_norm": 0.2896629571914673, + "learning_rate": 2.8254333333333335e-05, + "loss": 0.0224, + "step": 15241 + }, + { + "epoch": 12.87331081081081, + "grad_norm": 0.23483198881149292, + "learning_rate": 2.8254e-05, + "loss": 0.0244, + "step": 15242 + }, + { + "epoch": 12.874155405405405, + "grad_norm": 0.3433634340763092, + "learning_rate": 2.8253666666666667e-05, + "loss": 0.0193, + "step": 15243 + }, + { + "epoch": 12.875, + "grad_norm": 0.2557063698768616, + "learning_rate": 2.8253333333333336e-05, + "loss": 0.0294, + "step": 15244 + }, + { + "epoch": 12.875844594594595, + "grad_norm": 0.2176133543252945, + "learning_rate": 2.8253e-05, + "loss": 0.0242, + "step": 15245 + }, + { + "epoch": 12.87668918918919, + "grad_norm": 0.3486867845058441, + "learning_rate": 2.8252666666666668e-05, + "loss": 0.012, + "step": 15246 + }, + { + "epoch": 12.877533783783784, + "grad_norm": 0.4129074215888977, + "learning_rate": 2.8252333333333334e-05, + "loss": 0.0206, + "step": 15247 + }, + { + "epoch": 12.878378378378379, + "grad_norm": 0.6719138622283936, + "learning_rate": 2.8252e-05, + "loss": 0.0174, + "step": 15248 + }, + { + "epoch": 12.879222972972974, + "grad_norm": 0.9030787348747253, + "learning_rate": 2.825166666666667e-05, + "loss": 0.0335, + "step": 15249 + }, + { + "epoch": 12.880067567567568, + "grad_norm": 0.21728575229644775, + "learning_rate": 2.8251333333333335e-05, + "loss": 0.0175, + "step": 15250 + }, + { + "epoch": 12.880912162162161, + "grad_norm": 0.289566308259964, + "learning_rate": 2.8251e-05, + "loss": 0.0185, + "step": 15251 + }, + { + "epoch": 12.881756756756756, + "grad_norm": 0.36058688163757324, + "learning_rate": 2.8250666666666667e-05, + "loss": 0.0241, + "step": 15252 + }, + { + "epoch": 12.88260135135135, + "grad_norm": 0.6470731496810913, + "learning_rate": 2.8250333333333333e-05, + "loss": 0.0131, + "step": 15253 + }, + { + "epoch": 12.883445945945946, + "grad_norm": 0.3003403842449188, + "learning_rate": 2.825e-05, + "loss": 0.0196, + "step": 15254 + }, + { + "epoch": 12.88429054054054, + "grad_norm": 0.4157435894012451, + "learning_rate": 2.8249666666666668e-05, + "loss": 0.0175, + "step": 15255 + }, + { + "epoch": 12.885135135135135, + "grad_norm": 0.7594852447509766, + "learning_rate": 2.8249333333333334e-05, + "loss": 0.0257, + "step": 15256 + }, + { + "epoch": 12.88597972972973, + "grad_norm": 0.2907940149307251, + "learning_rate": 2.8249e-05, + "loss": 0.019, + "step": 15257 + }, + { + "epoch": 12.886824324324325, + "grad_norm": 0.390583336353302, + "learning_rate": 2.824866666666667e-05, + "loss": 0.029, + "step": 15258 + }, + { + "epoch": 12.88766891891892, + "grad_norm": 0.7423033118247986, + "learning_rate": 2.8248333333333332e-05, + "loss": 0.2025, + "step": 15259 + }, + { + "epoch": 12.888513513513514, + "grad_norm": 0.6797662377357483, + "learning_rate": 2.8248e-05, + "loss": 0.1358, + "step": 15260 + }, + { + "epoch": 12.889358108108109, + "grad_norm": 0.34269797801971436, + "learning_rate": 2.8247666666666667e-05, + "loss": 0.0623, + "step": 15261 + }, + { + "epoch": 12.890202702702704, + "grad_norm": 0.3679341971874237, + "learning_rate": 2.8247333333333333e-05, + "loss": 0.0345, + "step": 15262 + }, + { + "epoch": 12.891047297297296, + "grad_norm": 0.2525983452796936, + "learning_rate": 2.8247e-05, + "loss": 0.0471, + "step": 15263 + }, + { + "epoch": 12.891891891891891, + "grad_norm": 0.6370448470115662, + "learning_rate": 2.824666666666667e-05, + "loss": 0.0358, + "step": 15264 + }, + { + "epoch": 12.892736486486486, + "grad_norm": 0.23172907531261444, + "learning_rate": 2.8246333333333334e-05, + "loss": 0.0098, + "step": 15265 + }, + { + "epoch": 12.89358108108108, + "grad_norm": 0.6785866022109985, + "learning_rate": 2.8246e-05, + "loss": 0.0345, + "step": 15266 + }, + { + "epoch": 12.894425675675675, + "grad_norm": 0.34686386585235596, + "learning_rate": 2.824566666666667e-05, + "loss": 0.0247, + "step": 15267 + }, + { + "epoch": 12.89527027027027, + "grad_norm": 0.48811039328575134, + "learning_rate": 2.8245333333333332e-05, + "loss": 0.0258, + "step": 15268 + }, + { + "epoch": 12.896114864864865, + "grad_norm": 0.2579672932624817, + "learning_rate": 2.8245e-05, + "loss": 0.0224, + "step": 15269 + }, + { + "epoch": 12.89695945945946, + "grad_norm": 0.1976732313632965, + "learning_rate": 2.8244666666666667e-05, + "loss": 0.0185, + "step": 15270 + }, + { + "epoch": 12.897804054054054, + "grad_norm": 0.5002905130386353, + "learning_rate": 2.8244333333333333e-05, + "loss": 0.025, + "step": 15271 + }, + { + "epoch": 12.89864864864865, + "grad_norm": 0.6356616616249084, + "learning_rate": 2.8244e-05, + "loss": 0.0249, + "step": 15272 + }, + { + "epoch": 12.899493243243244, + "grad_norm": 0.3569738566875458, + "learning_rate": 2.824366666666667e-05, + "loss": 0.0325, + "step": 15273 + }, + { + "epoch": 12.900337837837839, + "grad_norm": 0.31565994024276733, + "learning_rate": 2.8243333333333335e-05, + "loss": 0.0195, + "step": 15274 + }, + { + "epoch": 12.901182432432432, + "grad_norm": 0.20841079950332642, + "learning_rate": 2.8243e-05, + "loss": 0.0132, + "step": 15275 + }, + { + "epoch": 12.902027027027026, + "grad_norm": 0.40390297770500183, + "learning_rate": 2.824266666666667e-05, + "loss": 0.0208, + "step": 15276 + }, + { + "epoch": 12.902871621621621, + "grad_norm": 0.26738420128822327, + "learning_rate": 2.8242333333333332e-05, + "loss": 0.0191, + "step": 15277 + }, + { + "epoch": 12.903716216216216, + "grad_norm": 0.45458585023880005, + "learning_rate": 2.8242e-05, + "loss": 0.0204, + "step": 15278 + }, + { + "epoch": 12.90456081081081, + "grad_norm": 0.3574301600456238, + "learning_rate": 2.8241666666666668e-05, + "loss": 0.0238, + "step": 15279 + }, + { + "epoch": 12.905405405405405, + "grad_norm": 0.16457128524780273, + "learning_rate": 2.8241333333333334e-05, + "loss": 0.018, + "step": 15280 + }, + { + "epoch": 12.90625, + "grad_norm": 0.24462874233722687, + "learning_rate": 2.8241000000000003e-05, + "loss": 0.0147, + "step": 15281 + }, + { + "epoch": 12.907094594594595, + "grad_norm": 0.22109410166740417, + "learning_rate": 2.8240666666666665e-05, + "loss": 0.0235, + "step": 15282 + }, + { + "epoch": 12.90793918918919, + "grad_norm": 0.2855297029018402, + "learning_rate": 2.8240333333333335e-05, + "loss": 0.02, + "step": 15283 + }, + { + "epoch": 12.908783783783784, + "grad_norm": 0.4239870607852936, + "learning_rate": 2.824e-05, + "loss": 0.0251, + "step": 15284 + }, + { + "epoch": 12.909628378378379, + "grad_norm": 0.22729340195655823, + "learning_rate": 2.8239666666666667e-05, + "loss": 0.0239, + "step": 15285 + }, + { + "epoch": 12.910472972972974, + "grad_norm": 0.20435042679309845, + "learning_rate": 2.8239333333333333e-05, + "loss": 0.0165, + "step": 15286 + }, + { + "epoch": 12.911317567567568, + "grad_norm": 0.30683764815330505, + "learning_rate": 2.8239000000000002e-05, + "loss": 0.0247, + "step": 15287 + }, + { + "epoch": 12.912162162162161, + "grad_norm": 0.29762762784957886, + "learning_rate": 2.8238666666666664e-05, + "loss": 0.0245, + "step": 15288 + }, + { + "epoch": 12.913006756756756, + "grad_norm": 0.27713742852211, + "learning_rate": 2.8238333333333334e-05, + "loss": 0.0228, + "step": 15289 + }, + { + "epoch": 12.91385135135135, + "grad_norm": 0.3218226730823517, + "learning_rate": 2.8238000000000003e-05, + "loss": 0.0206, + "step": 15290 + }, + { + "epoch": 12.914695945945946, + "grad_norm": 0.40129172801971436, + "learning_rate": 2.8237666666666666e-05, + "loss": 0.0259, + "step": 15291 + }, + { + "epoch": 12.91554054054054, + "grad_norm": 0.24943053722381592, + "learning_rate": 2.8237333333333335e-05, + "loss": 0.0248, + "step": 15292 + }, + { + "epoch": 12.916385135135135, + "grad_norm": 0.23591411113739014, + "learning_rate": 2.8237e-05, + "loss": 0.0291, + "step": 15293 + }, + { + "epoch": 12.91722972972973, + "grad_norm": 0.22600892186164856, + "learning_rate": 2.8236666666666667e-05, + "loss": 0.0182, + "step": 15294 + }, + { + "epoch": 12.918074324324325, + "grad_norm": 0.47531771659851074, + "learning_rate": 2.8236333333333333e-05, + "loss": 0.0239, + "step": 15295 + }, + { + "epoch": 12.91891891891892, + "grad_norm": 0.13626661896705627, + "learning_rate": 2.8236000000000002e-05, + "loss": 0.0129, + "step": 15296 + }, + { + "epoch": 12.919763513513514, + "grad_norm": 0.4921260476112366, + "learning_rate": 2.8235666666666665e-05, + "loss": 0.0268, + "step": 15297 + }, + { + "epoch": 12.920608108108109, + "grad_norm": 0.35812634229660034, + "learning_rate": 2.8235333333333334e-05, + "loss": 0.023, + "step": 15298 + }, + { + "epoch": 12.921452702702704, + "grad_norm": 0.4342098534107208, + "learning_rate": 2.8235000000000003e-05, + "loss": 0.0173, + "step": 15299 + }, + { + "epoch": 12.922297297297296, + "grad_norm": 0.3485162854194641, + "learning_rate": 2.8234666666666666e-05, + "loss": 0.0245, + "step": 15300 + }, + { + "epoch": 12.923141891891891, + "grad_norm": 0.4823046624660492, + "learning_rate": 2.8234333333333335e-05, + "loss": 0.0175, + "step": 15301 + }, + { + "epoch": 12.923986486486486, + "grad_norm": 0.2784380316734314, + "learning_rate": 2.8234e-05, + "loss": 0.0207, + "step": 15302 + }, + { + "epoch": 12.92483108108108, + "grad_norm": 0.47073808312416077, + "learning_rate": 2.8233666666666667e-05, + "loss": 0.0252, + "step": 15303 + }, + { + "epoch": 12.925675675675675, + "grad_norm": 0.5794417858123779, + "learning_rate": 2.8233333333333333e-05, + "loss": 0.0233, + "step": 15304 + }, + { + "epoch": 12.92652027027027, + "grad_norm": 0.9827991724014282, + "learning_rate": 2.8233000000000002e-05, + "loss": 0.0533, + "step": 15305 + }, + { + "epoch": 12.927364864864865, + "grad_norm": 0.5814405679702759, + "learning_rate": 2.8232666666666668e-05, + "loss": 0.0215, + "step": 15306 + }, + { + "epoch": 12.92820945945946, + "grad_norm": 0.5886610150337219, + "learning_rate": 2.8232333333333334e-05, + "loss": 0.0311, + "step": 15307 + }, + { + "epoch": 12.929054054054054, + "grad_norm": 0.530943751335144, + "learning_rate": 2.8232000000000003e-05, + "loss": 0.0395, + "step": 15308 + }, + { + "epoch": 12.92989864864865, + "grad_norm": 0.6637847423553467, + "learning_rate": 2.8231666666666666e-05, + "loss": 0.1988, + "step": 15309 + }, + { + "epoch": 12.930743243243244, + "grad_norm": 0.4880312383174896, + "learning_rate": 2.8231333333333335e-05, + "loss": 0.1137, + "step": 15310 + }, + { + "epoch": 12.931587837837839, + "grad_norm": 0.3701972961425781, + "learning_rate": 2.8231e-05, + "loss": 0.0573, + "step": 15311 + }, + { + "epoch": 12.932432432432432, + "grad_norm": 0.3186551034450531, + "learning_rate": 2.8230666666666667e-05, + "loss": 0.0384, + "step": 15312 + }, + { + "epoch": 12.933277027027026, + "grad_norm": 0.20006801187992096, + "learning_rate": 2.8230333333333333e-05, + "loss": 0.0204, + "step": 15313 + }, + { + "epoch": 12.934121621621621, + "grad_norm": 0.266824334859848, + "learning_rate": 2.823e-05, + "loss": 0.0179, + "step": 15314 + }, + { + "epoch": 12.934966216216216, + "grad_norm": 0.5018215179443359, + "learning_rate": 2.822966666666667e-05, + "loss": 0.0333, + "step": 15315 + }, + { + "epoch": 12.93581081081081, + "grad_norm": 0.2053609937429428, + "learning_rate": 2.8229333333333334e-05, + "loss": 0.0098, + "step": 15316 + }, + { + "epoch": 12.936655405405405, + "grad_norm": 0.37949708104133606, + "learning_rate": 2.8229e-05, + "loss": 0.0436, + "step": 15317 + }, + { + "epoch": 12.9375, + "grad_norm": 0.19240982830524445, + "learning_rate": 2.8228666666666666e-05, + "loss": 0.023, + "step": 15318 + }, + { + "epoch": 12.938344594594595, + "grad_norm": 0.4039762020111084, + "learning_rate": 2.8228333333333336e-05, + "loss": 0.0267, + "step": 15319 + }, + { + "epoch": 12.93918918918919, + "grad_norm": 0.3271941542625427, + "learning_rate": 2.8227999999999998e-05, + "loss": 0.0256, + "step": 15320 + }, + { + "epoch": 12.940033783783784, + "grad_norm": 0.2653622031211853, + "learning_rate": 2.8227666666666667e-05, + "loss": 0.0171, + "step": 15321 + }, + { + "epoch": 12.940878378378379, + "grad_norm": 0.2101365476846695, + "learning_rate": 2.8227333333333333e-05, + "loss": 0.0237, + "step": 15322 + }, + { + "epoch": 12.941722972972974, + "grad_norm": 0.3864898681640625, + "learning_rate": 2.8227e-05, + "loss": 0.0189, + "step": 15323 + }, + { + "epoch": 12.942567567567568, + "grad_norm": 0.2672373056411743, + "learning_rate": 2.822666666666667e-05, + "loss": 0.0207, + "step": 15324 + }, + { + "epoch": 12.943412162162161, + "grad_norm": 0.2505623698234558, + "learning_rate": 2.8226333333333335e-05, + "loss": 0.0267, + "step": 15325 + }, + { + "epoch": 12.944256756756756, + "grad_norm": 0.24183696508407593, + "learning_rate": 2.8226e-05, + "loss": 0.0228, + "step": 15326 + }, + { + "epoch": 12.94510135135135, + "grad_norm": 0.163090780377388, + "learning_rate": 2.8225666666666666e-05, + "loss": 0.0164, + "step": 15327 + }, + { + "epoch": 12.945945945945946, + "grad_norm": 0.4545929431915283, + "learning_rate": 2.8225333333333336e-05, + "loss": 0.0148, + "step": 15328 + }, + { + "epoch": 12.94679054054054, + "grad_norm": 0.22719541192054749, + "learning_rate": 2.8224999999999998e-05, + "loss": 0.023, + "step": 15329 + }, + { + "epoch": 12.947635135135135, + "grad_norm": 0.23941776156425476, + "learning_rate": 2.8224666666666668e-05, + "loss": 0.0167, + "step": 15330 + }, + { + "epoch": 12.94847972972973, + "grad_norm": 0.2921576201915741, + "learning_rate": 2.8224333333333337e-05, + "loss": 0.0174, + "step": 15331 + }, + { + "epoch": 12.949324324324325, + "grad_norm": 0.3153209686279297, + "learning_rate": 2.8224e-05, + "loss": 0.0193, + "step": 15332 + }, + { + "epoch": 12.95016891891892, + "grad_norm": 0.4338565468788147, + "learning_rate": 2.822366666666667e-05, + "loss": 0.0335, + "step": 15333 + }, + { + "epoch": 12.951013513513514, + "grad_norm": 0.23393532633781433, + "learning_rate": 2.8223333333333335e-05, + "loss": 0.0167, + "step": 15334 + }, + { + "epoch": 12.951858108108109, + "grad_norm": 0.16458866000175476, + "learning_rate": 2.8223e-05, + "loss": 0.0135, + "step": 15335 + }, + { + "epoch": 12.952702702702704, + "grad_norm": 0.24476459622383118, + "learning_rate": 2.8222666666666667e-05, + "loss": 0.0217, + "step": 15336 + }, + { + "epoch": 12.953547297297296, + "grad_norm": 0.23959727585315704, + "learning_rate": 2.8222333333333336e-05, + "loss": 0.0141, + "step": 15337 + }, + { + "epoch": 12.954391891891891, + "grad_norm": 0.2818269431591034, + "learning_rate": 2.8222e-05, + "loss": 0.0128, + "step": 15338 + }, + { + "epoch": 12.955236486486486, + "grad_norm": 0.38421863317489624, + "learning_rate": 2.8221666666666668e-05, + "loss": 0.0194, + "step": 15339 + }, + { + "epoch": 12.95608108108108, + "grad_norm": 0.32317644357681274, + "learning_rate": 2.8221333333333337e-05, + "loss": 0.0219, + "step": 15340 + }, + { + "epoch": 12.956925675675675, + "grad_norm": 0.2415333092212677, + "learning_rate": 2.8221e-05, + "loss": 0.0151, + "step": 15341 + }, + { + "epoch": 12.95777027027027, + "grad_norm": 0.43797579407691956, + "learning_rate": 2.822066666666667e-05, + "loss": 0.0127, + "step": 15342 + }, + { + "epoch": 12.958614864864865, + "grad_norm": 0.36749783158302307, + "learning_rate": 2.822033333333333e-05, + "loss": 0.0327, + "step": 15343 + }, + { + "epoch": 12.95945945945946, + "grad_norm": 0.5155549645423889, + "learning_rate": 2.822e-05, + "loss": 0.0212, + "step": 15344 + }, + { + "epoch": 12.960304054054054, + "grad_norm": 0.3881407380104065, + "learning_rate": 2.8219666666666667e-05, + "loss": 0.0185, + "step": 15345 + }, + { + "epoch": 12.96114864864865, + "grad_norm": 0.3340715169906616, + "learning_rate": 2.8219333333333333e-05, + "loss": 0.0248, + "step": 15346 + }, + { + "epoch": 12.961993243243244, + "grad_norm": 0.30920276045799255, + "learning_rate": 2.8219e-05, + "loss": 0.0231, + "step": 15347 + }, + { + "epoch": 12.962837837837839, + "grad_norm": 0.5611916780471802, + "learning_rate": 2.8218666666666668e-05, + "loss": 0.0356, + "step": 15348 + }, + { + "epoch": 12.963682432432432, + "grad_norm": 0.29951930046081543, + "learning_rate": 2.8218333333333334e-05, + "loss": 0.0137, + "step": 15349 + }, + { + "epoch": 12.964527027027026, + "grad_norm": 0.27920329570770264, + "learning_rate": 2.8218e-05, + "loss": 0.0092, + "step": 15350 + }, + { + "epoch": 12.965371621621621, + "grad_norm": 1.2507508993148804, + "learning_rate": 2.821766666666667e-05, + "loss": 0.0369, + "step": 15351 + }, + { + "epoch": 12.966216216216216, + "grad_norm": 0.4206985533237457, + "learning_rate": 2.8217333333333332e-05, + "loss": 0.0246, + "step": 15352 + }, + { + "epoch": 12.96706081081081, + "grad_norm": 0.3598859906196594, + "learning_rate": 2.8217e-05, + "loss": 0.0242, + "step": 15353 + }, + { + "epoch": 12.967905405405405, + "grad_norm": 0.36776649951934814, + "learning_rate": 2.8216666666666667e-05, + "loss": 0.0215, + "step": 15354 + }, + { + "epoch": 12.96875, + "grad_norm": 0.2548978924751282, + "learning_rate": 2.8216333333333333e-05, + "loss": 0.0191, + "step": 15355 + }, + { + "epoch": 12.969594594594595, + "grad_norm": 0.49632754921913147, + "learning_rate": 2.8216000000000002e-05, + "loss": 0.0298, + "step": 15356 + }, + { + "epoch": 12.97043918918919, + "grad_norm": 0.7460479736328125, + "learning_rate": 2.8215666666666668e-05, + "loss": 0.0374, + "step": 15357 + }, + { + "epoch": 12.971283783783784, + "grad_norm": 0.8948992490768433, + "learning_rate": 2.8215333333333334e-05, + "loss": 0.0463, + "step": 15358 + }, + { + "epoch": 12.972128378378379, + "grad_norm": 0.6195207238197327, + "learning_rate": 2.8215e-05, + "loss": 0.1468, + "step": 15359 + }, + { + "epoch": 12.972972972972974, + "grad_norm": 0.45436447858810425, + "learning_rate": 2.821466666666667e-05, + "loss": 0.0917, + "step": 15360 + }, + { + "epoch": 12.973817567567568, + "grad_norm": 0.4361341893672943, + "learning_rate": 2.8214333333333332e-05, + "loss": 0.0607, + "step": 15361 + }, + { + "epoch": 12.974662162162161, + "grad_norm": 0.3672994375228882, + "learning_rate": 2.8214e-05, + "loss": 0.0361, + "step": 15362 + }, + { + "epoch": 12.975506756756756, + "grad_norm": 0.2762325406074524, + "learning_rate": 2.8213666666666667e-05, + "loss": 0.0253, + "step": 15363 + }, + { + "epoch": 12.97635135135135, + "grad_norm": 0.8114959597587585, + "learning_rate": 2.8213333333333333e-05, + "loss": 0.0187, + "step": 15364 + }, + { + "epoch": 12.977195945945946, + "grad_norm": 0.2597557604312897, + "learning_rate": 2.8213000000000002e-05, + "loss": 0.0249, + "step": 15365 + }, + { + "epoch": 12.97804054054054, + "grad_norm": 0.21889574825763702, + "learning_rate": 2.821266666666667e-05, + "loss": 0.0202, + "step": 15366 + }, + { + "epoch": 12.978885135135135, + "grad_norm": 0.38726577162742615, + "learning_rate": 2.8212333333333334e-05, + "loss": 0.0304, + "step": 15367 + }, + { + "epoch": 12.97972972972973, + "grad_norm": 0.1888512521982193, + "learning_rate": 2.8212e-05, + "loss": 0.0153, + "step": 15368 + }, + { + "epoch": 12.980574324324325, + "grad_norm": 0.2297268956899643, + "learning_rate": 2.821166666666667e-05, + "loss": 0.022, + "step": 15369 + }, + { + "epoch": 12.98141891891892, + "grad_norm": 0.24381379783153534, + "learning_rate": 2.8211333333333332e-05, + "loss": 0.023, + "step": 15370 + }, + { + "epoch": 12.982263513513514, + "grad_norm": 0.1579710990190506, + "learning_rate": 2.8211e-05, + "loss": 0.0124, + "step": 15371 + }, + { + "epoch": 12.983108108108109, + "grad_norm": 0.46032652258872986, + "learning_rate": 2.8210666666666667e-05, + "loss": 0.0273, + "step": 15372 + }, + { + "epoch": 12.983952702702704, + "grad_norm": 0.3234539031982422, + "learning_rate": 2.8210333333333333e-05, + "loss": 0.0299, + "step": 15373 + }, + { + "epoch": 12.984797297297296, + "grad_norm": 0.27569830417633057, + "learning_rate": 2.8210000000000003e-05, + "loss": 0.0161, + "step": 15374 + }, + { + "epoch": 12.985641891891891, + "grad_norm": 0.41363435983657837, + "learning_rate": 2.8209666666666665e-05, + "loss": 0.0234, + "step": 15375 + }, + { + "epoch": 12.986486486486486, + "grad_norm": 0.2886505722999573, + "learning_rate": 2.8209333333333335e-05, + "loss": 0.0214, + "step": 15376 + }, + { + "epoch": 12.98733108108108, + "grad_norm": 0.45584020018577576, + "learning_rate": 2.8209e-05, + "loss": 0.0236, + "step": 15377 + }, + { + "epoch": 12.988175675675675, + "grad_norm": 0.3869459927082062, + "learning_rate": 2.8208666666666666e-05, + "loss": 0.0184, + "step": 15378 + }, + { + "epoch": 12.98902027027027, + "grad_norm": 0.40003517270088196, + "learning_rate": 2.8208333333333332e-05, + "loss": 0.0319, + "step": 15379 + }, + { + "epoch": 12.989864864864865, + "grad_norm": 0.1195167824625969, + "learning_rate": 2.8208e-05, + "loss": 0.0115, + "step": 15380 + }, + { + "epoch": 12.99070945945946, + "grad_norm": 0.4991036057472229, + "learning_rate": 2.8207666666666668e-05, + "loss": 0.0205, + "step": 15381 + }, + { + "epoch": 12.991554054054054, + "grad_norm": 0.21327470242977142, + "learning_rate": 2.8207333333333334e-05, + "loss": 0.0213, + "step": 15382 + }, + { + "epoch": 12.99239864864865, + "grad_norm": 0.2887636125087738, + "learning_rate": 2.8207000000000003e-05, + "loss": 0.0172, + "step": 15383 + }, + { + "epoch": 12.993243243243244, + "grad_norm": 0.30262792110443115, + "learning_rate": 2.8206666666666665e-05, + "loss": 0.0206, + "step": 15384 + }, + { + "epoch": 12.994087837837839, + "grad_norm": 0.3175365626811981, + "learning_rate": 2.8206333333333335e-05, + "loss": 0.0122, + "step": 15385 + }, + { + "epoch": 12.994932432432432, + "grad_norm": 0.22485525906085968, + "learning_rate": 2.8206e-05, + "loss": 0.0146, + "step": 15386 + }, + { + "epoch": 12.995777027027026, + "grad_norm": 0.19331273436546326, + "learning_rate": 2.8205666666666667e-05, + "loss": 0.0132, + "step": 15387 + }, + { + "epoch": 12.996621621621621, + "grad_norm": 0.2287691980600357, + "learning_rate": 2.8205333333333333e-05, + "loss": 0.0226, + "step": 15388 + }, + { + "epoch": 12.997466216216216, + "grad_norm": 0.4255349040031433, + "learning_rate": 2.8205000000000002e-05, + "loss": 0.0221, + "step": 15389 + }, + { + "epoch": 12.99831081081081, + "grad_norm": 0.5133192539215088, + "learning_rate": 2.8204666666666668e-05, + "loss": 0.0284, + "step": 15390 + }, + { + "epoch": 12.999155405405405, + "grad_norm": 0.39143532514572144, + "learning_rate": 2.8204333333333334e-05, + "loss": 0.0234, + "step": 15391 + }, + { + "epoch": 13.0, + "grad_norm": 1.205501914024353, + "learning_rate": 2.8204000000000003e-05, + "loss": 0.0281, + "step": 15392 + }, + { + "epoch": 13.000844594594595, + "grad_norm": 0.6976914405822754, + "learning_rate": 2.8203666666666666e-05, + "loss": 0.2086, + "step": 15393 + }, + { + "epoch": 13.00168918918919, + "grad_norm": 0.5168591737747192, + "learning_rate": 2.8203333333333335e-05, + "loss": 0.1393, + "step": 15394 + }, + { + "epoch": 13.002533783783784, + "grad_norm": 0.21927127242088318, + "learning_rate": 2.8203e-05, + "loss": 0.0418, + "step": 15395 + }, + { + "epoch": 13.003378378378379, + "grad_norm": 0.2523837983608246, + "learning_rate": 2.8202666666666667e-05, + "loss": 0.0309, + "step": 15396 + }, + { + "epoch": 13.004222972972974, + "grad_norm": 0.2431890219449997, + "learning_rate": 2.8202333333333333e-05, + "loss": 0.0174, + "step": 15397 + }, + { + "epoch": 13.005067567567568, + "grad_norm": 0.19421662390232086, + "learning_rate": 2.8202000000000002e-05, + "loss": 0.0213, + "step": 15398 + }, + { + "epoch": 13.005912162162161, + "grad_norm": 0.37980392575263977, + "learning_rate": 2.8201666666666668e-05, + "loss": 0.0398, + "step": 15399 + }, + { + "epoch": 13.006756756756756, + "grad_norm": 0.4174080193042755, + "learning_rate": 2.8201333333333334e-05, + "loss": 0.0228, + "step": 15400 + }, + { + "epoch": 13.00760135135135, + "grad_norm": 0.4241948425769806, + "learning_rate": 2.8201000000000003e-05, + "loss": 0.023, + "step": 15401 + }, + { + "epoch": 13.008445945945946, + "grad_norm": 0.25243332982063293, + "learning_rate": 2.8200666666666666e-05, + "loss": 0.0205, + "step": 15402 + }, + { + "epoch": 13.00929054054054, + "grad_norm": 0.3739125728607178, + "learning_rate": 2.8200333333333335e-05, + "loss": 0.0368, + "step": 15403 + }, + { + "epoch": 13.010135135135135, + "grad_norm": 0.23031096160411835, + "learning_rate": 2.8199999999999998e-05, + "loss": 0.0288, + "step": 15404 + }, + { + "epoch": 13.01097972972973, + "grad_norm": 0.24754607677459717, + "learning_rate": 2.8199666666666667e-05, + "loss": 0.0272, + "step": 15405 + }, + { + "epoch": 13.011824324324325, + "grad_norm": 0.27696743607521057, + "learning_rate": 2.8199333333333336e-05, + "loss": 0.0162, + "step": 15406 + }, + { + "epoch": 13.01266891891892, + "grad_norm": 0.3065585792064667, + "learning_rate": 2.8199e-05, + "loss": 0.035, + "step": 15407 + }, + { + "epoch": 13.013513513513514, + "grad_norm": 0.1506434977054596, + "learning_rate": 2.8198666666666668e-05, + "loss": 0.0169, + "step": 15408 + }, + { + "epoch": 13.014358108108109, + "grad_norm": 0.17348265647888184, + "learning_rate": 2.8198333333333334e-05, + "loss": 0.0194, + "step": 15409 + }, + { + "epoch": 13.015202702702704, + "grad_norm": 0.1675160825252533, + "learning_rate": 2.8198e-05, + "loss": 0.0146, + "step": 15410 + }, + { + "epoch": 13.016047297297296, + "grad_norm": 0.20959049463272095, + "learning_rate": 2.8197666666666666e-05, + "loss": 0.0215, + "step": 15411 + }, + { + "epoch": 13.016891891891891, + "grad_norm": 0.40830525755882263, + "learning_rate": 2.8197333333333335e-05, + "loss": 0.032, + "step": 15412 + }, + { + "epoch": 13.017736486486486, + "grad_norm": 0.20706205070018768, + "learning_rate": 2.8196999999999998e-05, + "loss": 0.0186, + "step": 15413 + }, + { + "epoch": 13.01858108108108, + "grad_norm": 0.2346397340297699, + "learning_rate": 2.8196666666666667e-05, + "loss": 0.014, + "step": 15414 + }, + { + "epoch": 13.019425675675675, + "grad_norm": 0.1732100397348404, + "learning_rate": 2.8196333333333337e-05, + "loss": 0.0146, + "step": 15415 + }, + { + "epoch": 13.02027027027027, + "grad_norm": 0.19754362106323242, + "learning_rate": 2.8196e-05, + "loss": 0.023, + "step": 15416 + }, + { + "epoch": 13.021114864864865, + "grad_norm": 0.3229426443576813, + "learning_rate": 2.819566666666667e-05, + "loss": 0.02, + "step": 15417 + }, + { + "epoch": 13.02195945945946, + "grad_norm": 0.2809557318687439, + "learning_rate": 2.8195333333333334e-05, + "loss": 0.0188, + "step": 15418 + }, + { + "epoch": 13.022804054054054, + "grad_norm": 0.22370906174182892, + "learning_rate": 2.8195e-05, + "loss": 0.0216, + "step": 15419 + }, + { + "epoch": 13.02364864864865, + "grad_norm": 0.23775403201580048, + "learning_rate": 2.8194666666666666e-05, + "loss": 0.0125, + "step": 15420 + }, + { + "epoch": 13.024493243243244, + "grad_norm": 0.22905439138412476, + "learning_rate": 2.8194333333333336e-05, + "loss": 0.0177, + "step": 15421 + }, + { + "epoch": 13.025337837837839, + "grad_norm": 0.26966774463653564, + "learning_rate": 2.8194e-05, + "loss": 0.0132, + "step": 15422 + }, + { + "epoch": 13.026182432432432, + "grad_norm": 0.2386893332004547, + "learning_rate": 2.8193666666666667e-05, + "loss": 0.0214, + "step": 15423 + }, + { + "epoch": 13.027027027027026, + "grad_norm": 0.38013720512390137, + "learning_rate": 2.8193333333333337e-05, + "loss": 0.0165, + "step": 15424 + }, + { + "epoch": 13.027871621621621, + "grad_norm": 0.2390633076429367, + "learning_rate": 2.8193e-05, + "loss": 0.0135, + "step": 15425 + }, + { + "epoch": 13.028716216216216, + "grad_norm": 0.2213977724313736, + "learning_rate": 2.819266666666667e-05, + "loss": 0.0118, + "step": 15426 + }, + { + "epoch": 13.02956081081081, + "grad_norm": 0.19981034100055695, + "learning_rate": 2.8192333333333335e-05, + "loss": 0.0175, + "step": 15427 + }, + { + "epoch": 13.030405405405405, + "grad_norm": 0.35168877243995667, + "learning_rate": 2.8192e-05, + "loss": 0.0255, + "step": 15428 + }, + { + "epoch": 13.03125, + "grad_norm": 0.19682346284389496, + "learning_rate": 2.8191666666666666e-05, + "loss": 0.0168, + "step": 15429 + }, + { + "epoch": 13.032094594594595, + "grad_norm": 0.6871364712715149, + "learning_rate": 2.8191333333333336e-05, + "loss": 0.0288, + "step": 15430 + }, + { + "epoch": 13.03293918918919, + "grad_norm": 0.3230350613594055, + "learning_rate": 2.8191e-05, + "loss": 0.0228, + "step": 15431 + }, + { + "epoch": 13.033783783783784, + "grad_norm": 0.21725794672966003, + "learning_rate": 2.8190666666666668e-05, + "loss": 0.0119, + "step": 15432 + }, + { + "epoch": 13.034628378378379, + "grad_norm": 0.312778502702713, + "learning_rate": 2.8190333333333334e-05, + "loss": 0.0153, + "step": 15433 + }, + { + "epoch": 13.035472972972974, + "grad_norm": 0.29565513134002686, + "learning_rate": 2.819e-05, + "loss": 0.018, + "step": 15434 + }, + { + "epoch": 13.036317567567568, + "grad_norm": 0.1576070487499237, + "learning_rate": 2.818966666666667e-05, + "loss": 0.0086, + "step": 15435 + }, + { + "epoch": 13.037162162162161, + "grad_norm": 0.29053881764411926, + "learning_rate": 2.818933333333333e-05, + "loss": 0.0138, + "step": 15436 + }, + { + "epoch": 13.038006756756756, + "grad_norm": 0.4271653890609741, + "learning_rate": 2.8189e-05, + "loss": 0.0204, + "step": 15437 + }, + { + "epoch": 13.03885135135135, + "grad_norm": 0.323064386844635, + "learning_rate": 2.8188666666666667e-05, + "loss": 0.0192, + "step": 15438 + }, + { + "epoch": 13.039695945945946, + "grad_norm": 0.3982172906398773, + "learning_rate": 2.8188333333333333e-05, + "loss": 0.0276, + "step": 15439 + }, + { + "epoch": 13.04054054054054, + "grad_norm": 0.952016294002533, + "learning_rate": 2.8188000000000002e-05, + "loss": 0.0266, + "step": 15440 + }, + { + "epoch": 13.041385135135135, + "grad_norm": 0.4734238088130951, + "learning_rate": 2.8187666666666668e-05, + "loss": 0.0306, + "step": 15441 + }, + { + "epoch": 13.04222972972973, + "grad_norm": 0.7163010239601135, + "learning_rate": 2.8187333333333334e-05, + "loss": 0.0331, + "step": 15442 + }, + { + "epoch": 13.043074324324325, + "grad_norm": 0.611134946346283, + "learning_rate": 2.8187e-05, + "loss": 0.2052, + "step": 15443 + }, + { + "epoch": 13.04391891891892, + "grad_norm": 0.4360331594944, + "learning_rate": 2.818666666666667e-05, + "loss": 0.093, + "step": 15444 + }, + { + "epoch": 13.044763513513514, + "grad_norm": 0.312637060880661, + "learning_rate": 2.818633333333333e-05, + "loss": 0.0625, + "step": 15445 + }, + { + "epoch": 13.045608108108109, + "grad_norm": 0.19269613921642303, + "learning_rate": 2.8186e-05, + "loss": 0.0154, + "step": 15446 + }, + { + "epoch": 13.046452702702704, + "grad_norm": 0.2998218238353729, + "learning_rate": 2.8185666666666667e-05, + "loss": 0.0288, + "step": 15447 + }, + { + "epoch": 13.047297297297296, + "grad_norm": 0.250250905752182, + "learning_rate": 2.8185333333333333e-05, + "loss": 0.0292, + "step": 15448 + }, + { + "epoch": 13.048141891891891, + "grad_norm": 0.6802833676338196, + "learning_rate": 2.8185000000000002e-05, + "loss": 0.0188, + "step": 15449 + }, + { + "epoch": 13.048986486486486, + "grad_norm": 0.1590253859758377, + "learning_rate": 2.8184666666666668e-05, + "loss": 0.0253, + "step": 15450 + }, + { + "epoch": 13.04983108108108, + "grad_norm": 0.7828405499458313, + "learning_rate": 2.8184333333333334e-05, + "loss": 0.0243, + "step": 15451 + }, + { + "epoch": 13.050675675675675, + "grad_norm": 0.24242068827152252, + "learning_rate": 2.8184e-05, + "loss": 0.0216, + "step": 15452 + }, + { + "epoch": 13.05152027027027, + "grad_norm": 0.1690211147069931, + "learning_rate": 2.818366666666667e-05, + "loss": 0.016, + "step": 15453 + }, + { + "epoch": 13.052364864864865, + "grad_norm": 0.22217336297035217, + "learning_rate": 2.8183333333333332e-05, + "loss": 0.0148, + "step": 15454 + }, + { + "epoch": 13.05320945945946, + "grad_norm": 0.2884264886379242, + "learning_rate": 2.8183e-05, + "loss": 0.0142, + "step": 15455 + }, + { + "epoch": 13.054054054054054, + "grad_norm": 0.27748140692710876, + "learning_rate": 2.818266666666667e-05, + "loss": 0.018, + "step": 15456 + }, + { + "epoch": 13.05489864864865, + "grad_norm": 0.3390190601348877, + "learning_rate": 2.8182333333333333e-05, + "loss": 0.0176, + "step": 15457 + }, + { + "epoch": 13.055743243243244, + "grad_norm": 0.23349325358867645, + "learning_rate": 2.8182000000000002e-05, + "loss": 0.0136, + "step": 15458 + }, + { + "epoch": 13.056587837837839, + "grad_norm": 0.20865455269813538, + "learning_rate": 2.8181666666666668e-05, + "loss": 0.0267, + "step": 15459 + }, + { + "epoch": 13.057432432432432, + "grad_norm": 0.18822599947452545, + "learning_rate": 2.8181333333333334e-05, + "loss": 0.0161, + "step": 15460 + }, + { + "epoch": 13.058277027027026, + "grad_norm": 0.24974612891674042, + "learning_rate": 2.8181e-05, + "loss": 0.0174, + "step": 15461 + }, + { + "epoch": 13.059121621621621, + "grad_norm": 0.27954810857772827, + "learning_rate": 2.818066666666667e-05, + "loss": 0.0235, + "step": 15462 + }, + { + "epoch": 13.059966216216216, + "grad_norm": 0.385824978351593, + "learning_rate": 2.8180333333333332e-05, + "loss": 0.0238, + "step": 15463 + }, + { + "epoch": 13.06081081081081, + "grad_norm": 0.15470640361309052, + "learning_rate": 2.818e-05, + "loss": 0.0148, + "step": 15464 + }, + { + "epoch": 13.061655405405405, + "grad_norm": 0.2737314999103546, + "learning_rate": 2.8179666666666667e-05, + "loss": 0.0227, + "step": 15465 + }, + { + "epoch": 13.0625, + "grad_norm": 0.2790479063987732, + "learning_rate": 2.8179333333333333e-05, + "loss": 0.017, + "step": 15466 + }, + { + "epoch": 13.063344594594595, + "grad_norm": 0.14522835612297058, + "learning_rate": 2.8179000000000002e-05, + "loss": 0.0112, + "step": 15467 + }, + { + "epoch": 13.06418918918919, + "grad_norm": 0.23694531619548798, + "learning_rate": 2.8178666666666665e-05, + "loss": 0.0118, + "step": 15468 + }, + { + "epoch": 13.065033783783784, + "grad_norm": 0.1978273093700409, + "learning_rate": 2.8178333333333334e-05, + "loss": 0.019, + "step": 15469 + }, + { + "epoch": 13.065878378378379, + "grad_norm": 0.22108249366283417, + "learning_rate": 2.8178e-05, + "loss": 0.0155, + "step": 15470 + }, + { + "epoch": 13.066722972972974, + "grad_norm": 0.28228604793548584, + "learning_rate": 2.8177666666666666e-05, + "loss": 0.0153, + "step": 15471 + }, + { + "epoch": 13.067567567567568, + "grad_norm": 0.3597482442855835, + "learning_rate": 2.8177333333333336e-05, + "loss": 0.0237, + "step": 15472 + }, + { + "epoch": 13.068412162162161, + "grad_norm": 0.2905884385108948, + "learning_rate": 2.8177e-05, + "loss": 0.0174, + "step": 15473 + }, + { + "epoch": 13.069256756756756, + "grad_norm": 0.3563370406627655, + "learning_rate": 2.8176666666666667e-05, + "loss": 0.0186, + "step": 15474 + }, + { + "epoch": 13.07010135135135, + "grad_norm": 0.21791431307792664, + "learning_rate": 2.8176333333333333e-05, + "loss": 0.013, + "step": 15475 + }, + { + "epoch": 13.070945945945946, + "grad_norm": 0.252359002828598, + "learning_rate": 2.8176000000000003e-05, + "loss": 0.0134, + "step": 15476 + }, + { + "epoch": 13.07179054054054, + "grad_norm": 0.13663597404956818, + "learning_rate": 2.8175666666666665e-05, + "loss": 0.0094, + "step": 15477 + }, + { + "epoch": 13.072635135135135, + "grad_norm": 0.4558465778827667, + "learning_rate": 2.8175333333333335e-05, + "loss": 0.0179, + "step": 15478 + }, + { + "epoch": 13.07347972972973, + "grad_norm": 0.2654597759246826, + "learning_rate": 2.8175e-05, + "loss": 0.0175, + "step": 15479 + }, + { + "epoch": 13.074324324324325, + "grad_norm": 0.25081029534339905, + "learning_rate": 2.8174666666666666e-05, + "loss": 0.0135, + "step": 15480 + }, + { + "epoch": 13.07516891891892, + "grad_norm": 0.2059129774570465, + "learning_rate": 2.8174333333333336e-05, + "loss": 0.0109, + "step": 15481 + }, + { + "epoch": 13.076013513513514, + "grad_norm": 0.20572488009929657, + "learning_rate": 2.8174e-05, + "loss": 0.0141, + "step": 15482 + }, + { + "epoch": 13.076858108108109, + "grad_norm": 0.49564725160598755, + "learning_rate": 2.8173666666666668e-05, + "loss": 0.0237, + "step": 15483 + }, + { + "epoch": 13.077702702702704, + "grad_norm": 0.3182987868785858, + "learning_rate": 2.8173333333333334e-05, + "loss": 0.0183, + "step": 15484 + }, + { + "epoch": 13.078547297297296, + "grad_norm": 0.33598703145980835, + "learning_rate": 2.8173000000000003e-05, + "loss": 0.0179, + "step": 15485 + }, + { + "epoch": 13.079391891891891, + "grad_norm": 0.48789772391319275, + "learning_rate": 2.8172666666666665e-05, + "loss": 0.014, + "step": 15486 + }, + { + "epoch": 13.080236486486486, + "grad_norm": 0.299970805644989, + "learning_rate": 2.8172333333333335e-05, + "loss": 0.0136, + "step": 15487 + }, + { + "epoch": 13.08108108108108, + "grad_norm": 0.5373780727386475, + "learning_rate": 2.8172e-05, + "loss": 0.0209, + "step": 15488 + }, + { + "epoch": 13.081925675675675, + "grad_norm": 0.3580116629600525, + "learning_rate": 2.8171666666666667e-05, + "loss": 0.0301, + "step": 15489 + }, + { + "epoch": 13.08277027027027, + "grad_norm": 0.8054562211036682, + "learning_rate": 2.8171333333333336e-05, + "loss": 0.0187, + "step": 15490 + }, + { + "epoch": 13.083614864864865, + "grad_norm": 0.5218288898468018, + "learning_rate": 2.8171000000000002e-05, + "loss": 0.0322, + "step": 15491 + }, + { + "epoch": 13.08445945945946, + "grad_norm": 0.4692350924015045, + "learning_rate": 2.8170666666666668e-05, + "loss": 0.0326, + "step": 15492 + }, + { + "epoch": 13.085304054054054, + "grad_norm": 0.4497619569301605, + "learning_rate": 2.8170333333333334e-05, + "loss": 0.1744, + "step": 15493 + }, + { + "epoch": 13.08614864864865, + "grad_norm": 0.38046160340309143, + "learning_rate": 2.817e-05, + "loss": 0.0836, + "step": 15494 + }, + { + "epoch": 13.086993243243244, + "grad_norm": 0.788853108882904, + "learning_rate": 2.8169666666666666e-05, + "loss": 0.0689, + "step": 15495 + }, + { + "epoch": 13.087837837837839, + "grad_norm": 0.3116481900215149, + "learning_rate": 2.8169333333333335e-05, + "loss": 0.046, + "step": 15496 + }, + { + "epoch": 13.088682432432432, + "grad_norm": 0.17364145815372467, + "learning_rate": 2.8169e-05, + "loss": 0.0128, + "step": 15497 + }, + { + "epoch": 13.089527027027026, + "grad_norm": 0.5461792945861816, + "learning_rate": 2.8168666666666667e-05, + "loss": 0.0228, + "step": 15498 + }, + { + "epoch": 13.090371621621621, + "grad_norm": 0.13051196932792664, + "learning_rate": 2.8168333333333336e-05, + "loss": 0.0126, + "step": 15499 + }, + { + "epoch": 13.091216216216216, + "grad_norm": 0.22161102294921875, + "learning_rate": 2.8168e-05, + "loss": 0.0187, + "step": 15500 + }, + { + "epoch": 13.09206081081081, + "grad_norm": 0.24127908051013947, + "learning_rate": 2.8167666666666668e-05, + "loss": 0.0221, + "step": 15501 + }, + { + "epoch": 13.092905405405405, + "grad_norm": 0.19637352228164673, + "learning_rate": 2.8167333333333334e-05, + "loss": 0.0241, + "step": 15502 + }, + { + "epoch": 13.09375, + "grad_norm": 0.19369326531887054, + "learning_rate": 2.8167e-05, + "loss": 0.0075, + "step": 15503 + }, + { + "epoch": 13.094594594594595, + "grad_norm": 0.45971745252609253, + "learning_rate": 2.8166666666666666e-05, + "loss": 0.0124, + "step": 15504 + }, + { + "epoch": 13.09543918918919, + "grad_norm": 0.5192379355430603, + "learning_rate": 2.8166333333333335e-05, + "loss": 0.0241, + "step": 15505 + }, + { + "epoch": 13.096283783783784, + "grad_norm": 0.2407829463481903, + "learning_rate": 2.8166e-05, + "loss": 0.0164, + "step": 15506 + }, + { + "epoch": 13.097128378378379, + "grad_norm": 0.560684859752655, + "learning_rate": 2.8165666666666667e-05, + "loss": 0.0136, + "step": 15507 + }, + { + "epoch": 13.097972972972974, + "grad_norm": 0.2194863110780716, + "learning_rate": 2.8165333333333336e-05, + "loss": 0.0132, + "step": 15508 + }, + { + "epoch": 13.098817567567568, + "grad_norm": 0.268200546503067, + "learning_rate": 2.8165e-05, + "loss": 0.0102, + "step": 15509 + }, + { + "epoch": 13.099662162162161, + "grad_norm": 0.40073999762535095, + "learning_rate": 2.8164666666666668e-05, + "loss": 0.0201, + "step": 15510 + }, + { + "epoch": 13.100506756756756, + "grad_norm": 0.3846701979637146, + "learning_rate": 2.8164333333333334e-05, + "loss": 0.0114, + "step": 15511 + }, + { + "epoch": 13.10135135135135, + "grad_norm": 0.3350304961204529, + "learning_rate": 2.8164e-05, + "loss": 0.0125, + "step": 15512 + }, + { + "epoch": 13.102195945945946, + "grad_norm": 0.25888219475746155, + "learning_rate": 2.8163666666666666e-05, + "loss": 0.0185, + "step": 15513 + }, + { + "epoch": 13.10304054054054, + "grad_norm": 0.2572025954723358, + "learning_rate": 2.8163333333333335e-05, + "loss": 0.0156, + "step": 15514 + }, + { + "epoch": 13.103885135135135, + "grad_norm": 0.32424139976501465, + "learning_rate": 2.8163e-05, + "loss": 0.0157, + "step": 15515 + }, + { + "epoch": 13.10472972972973, + "grad_norm": 0.445551335811615, + "learning_rate": 2.8162666666666667e-05, + "loss": 0.0276, + "step": 15516 + }, + { + "epoch": 13.105574324324325, + "grad_norm": 0.27080801129341125, + "learning_rate": 2.8162333333333336e-05, + "loss": 0.0233, + "step": 15517 + }, + { + "epoch": 13.10641891891892, + "grad_norm": 0.23056654632091522, + "learning_rate": 2.8162e-05, + "loss": 0.0168, + "step": 15518 + }, + { + "epoch": 13.107263513513514, + "grad_norm": 0.3010304570198059, + "learning_rate": 2.816166666666667e-05, + "loss": 0.017, + "step": 15519 + }, + { + "epoch": 13.108108108108109, + "grad_norm": 0.2439066618680954, + "learning_rate": 2.8161333333333334e-05, + "loss": 0.0134, + "step": 15520 + }, + { + "epoch": 13.108952702702704, + "grad_norm": 0.24250349402427673, + "learning_rate": 2.8161e-05, + "loss": 0.0146, + "step": 15521 + }, + { + "epoch": 13.109797297297296, + "grad_norm": 0.2099681794643402, + "learning_rate": 2.816066666666667e-05, + "loss": 0.0122, + "step": 15522 + }, + { + "epoch": 13.110641891891891, + "grad_norm": 0.5100905299186707, + "learning_rate": 2.8160333333333336e-05, + "loss": 0.0152, + "step": 15523 + }, + { + "epoch": 13.111486486486486, + "grad_norm": 0.2591760754585266, + "learning_rate": 2.816e-05, + "loss": 0.0208, + "step": 15524 + }, + { + "epoch": 13.11233108108108, + "grad_norm": 0.4530916213989258, + "learning_rate": 2.8159666666666667e-05, + "loss": 0.0176, + "step": 15525 + }, + { + "epoch": 13.113175675675675, + "grad_norm": 0.38984793424606323, + "learning_rate": 2.8159333333333333e-05, + "loss": 0.0173, + "step": 15526 + }, + { + "epoch": 13.11402027027027, + "grad_norm": 0.19288401305675507, + "learning_rate": 2.8159e-05, + "loss": 0.0137, + "step": 15527 + }, + { + "epoch": 13.114864864864865, + "grad_norm": 0.2646792232990265, + "learning_rate": 2.815866666666667e-05, + "loss": 0.0181, + "step": 15528 + }, + { + "epoch": 13.11570945945946, + "grad_norm": 0.2772451937198639, + "learning_rate": 2.815833333333333e-05, + "loss": 0.0174, + "step": 15529 + }, + { + "epoch": 13.116554054054054, + "grad_norm": 0.17706692218780518, + "learning_rate": 2.8158e-05, + "loss": 0.0091, + "step": 15530 + }, + { + "epoch": 13.11739864864865, + "grad_norm": 0.259848028421402, + "learning_rate": 2.815766666666667e-05, + "loss": 0.0212, + "step": 15531 + }, + { + "epoch": 13.118243243243244, + "grad_norm": 0.18079239130020142, + "learning_rate": 2.8157333333333332e-05, + "loss": 0.0116, + "step": 15532 + }, + { + "epoch": 13.119087837837839, + "grad_norm": 0.21625962853431702, + "learning_rate": 2.8157e-05, + "loss": 0.0154, + "step": 15533 + }, + { + "epoch": 13.119932432432432, + "grad_norm": 0.2751086950302124, + "learning_rate": 2.8156666666666668e-05, + "loss": 0.0105, + "step": 15534 + }, + { + "epoch": 13.120777027027026, + "grad_norm": 0.3072381019592285, + "learning_rate": 2.8156333333333334e-05, + "loss": 0.0221, + "step": 15535 + }, + { + "epoch": 13.121621621621621, + "grad_norm": 0.3995543420314789, + "learning_rate": 2.8156e-05, + "loss": 0.0216, + "step": 15536 + }, + { + "epoch": 13.122466216216216, + "grad_norm": 0.27310603857040405, + "learning_rate": 2.815566666666667e-05, + "loss": 0.0109, + "step": 15537 + }, + { + "epoch": 13.12331081081081, + "grad_norm": 0.5866914987564087, + "learning_rate": 2.815533333333333e-05, + "loss": 0.0202, + "step": 15538 + }, + { + "epoch": 13.124155405405405, + "grad_norm": 0.18171624839305878, + "learning_rate": 2.8155e-05, + "loss": 0.0121, + "step": 15539 + }, + { + "epoch": 13.125, + "grad_norm": 0.22477388381958008, + "learning_rate": 2.815466666666667e-05, + "loss": 0.0218, + "step": 15540 + }, + { + "epoch": 13.125844594594595, + "grad_norm": 0.4635719954967499, + "learning_rate": 2.8154333333333333e-05, + "loss": 0.0168, + "step": 15541 + }, + { + "epoch": 13.12668918918919, + "grad_norm": 0.7711209058761597, + "learning_rate": 2.8154000000000002e-05, + "loss": 0.0275, + "step": 15542 + }, + { + "epoch": 13.127533783783784, + "grad_norm": 0.6594995856285095, + "learning_rate": 2.8153666666666668e-05, + "loss": 0.1975, + "step": 15543 + }, + { + "epoch": 13.128378378378379, + "grad_norm": 0.9595744013786316, + "learning_rate": 2.8153333333333334e-05, + "loss": 0.0897, + "step": 15544 + }, + { + "epoch": 13.129222972972974, + "grad_norm": 0.20324605703353882, + "learning_rate": 2.8153e-05, + "loss": 0.034, + "step": 15545 + }, + { + "epoch": 13.130067567567568, + "grad_norm": 0.2905711829662323, + "learning_rate": 2.815266666666667e-05, + "loss": 0.0334, + "step": 15546 + }, + { + "epoch": 13.130912162162161, + "grad_norm": 0.3644592761993408, + "learning_rate": 2.8152333333333335e-05, + "loss": 0.0349, + "step": 15547 + }, + { + "epoch": 13.131756756756756, + "grad_norm": 0.4478592276573181, + "learning_rate": 2.8152e-05, + "loss": 0.037, + "step": 15548 + }, + { + "epoch": 13.13260135135135, + "grad_norm": 0.2602452039718628, + "learning_rate": 2.815166666666667e-05, + "loss": 0.0258, + "step": 15549 + }, + { + "epoch": 13.133445945945946, + "grad_norm": 0.24951355159282684, + "learning_rate": 2.8151333333333333e-05, + "loss": 0.0221, + "step": 15550 + }, + { + "epoch": 13.13429054054054, + "grad_norm": 0.27963387966156006, + "learning_rate": 2.8151000000000002e-05, + "loss": 0.0173, + "step": 15551 + }, + { + "epoch": 13.135135135135135, + "grad_norm": 0.5450161099433899, + "learning_rate": 2.8150666666666668e-05, + "loss": 0.0227, + "step": 15552 + }, + { + "epoch": 13.13597972972973, + "grad_norm": 0.1605963110923767, + "learning_rate": 2.8150333333333334e-05, + "loss": 0.0131, + "step": 15553 + }, + { + "epoch": 13.136824324324325, + "grad_norm": 0.17750340700149536, + "learning_rate": 2.815e-05, + "loss": 0.0117, + "step": 15554 + }, + { + "epoch": 13.13766891891892, + "grad_norm": 0.2250278741121292, + "learning_rate": 2.8149666666666666e-05, + "loss": 0.0103, + "step": 15555 + }, + { + "epoch": 13.138513513513514, + "grad_norm": 0.49034348130226135, + "learning_rate": 2.8149333333333335e-05, + "loss": 0.0264, + "step": 15556 + }, + { + "epoch": 13.139358108108109, + "grad_norm": 0.22631964087486267, + "learning_rate": 2.8149e-05, + "loss": 0.019, + "step": 15557 + }, + { + "epoch": 13.140202702702704, + "grad_norm": 0.42268362641334534, + "learning_rate": 2.8148666666666667e-05, + "loss": 0.0257, + "step": 15558 + }, + { + "epoch": 13.141047297297296, + "grad_norm": 0.32149356603622437, + "learning_rate": 2.8148333333333333e-05, + "loss": 0.0247, + "step": 15559 + }, + { + "epoch": 13.141891891891891, + "grad_norm": 0.3400772213935852, + "learning_rate": 2.8148000000000002e-05, + "loss": 0.0233, + "step": 15560 + }, + { + "epoch": 13.142736486486486, + "grad_norm": 0.6486133933067322, + "learning_rate": 2.8147666666666665e-05, + "loss": 0.0154, + "step": 15561 + }, + { + "epoch": 13.14358108108108, + "grad_norm": 1.0135257244110107, + "learning_rate": 2.8147333333333334e-05, + "loss": 0.0233, + "step": 15562 + }, + { + "epoch": 13.144425675675675, + "grad_norm": 0.26325374841690063, + "learning_rate": 2.8147e-05, + "loss": 0.0124, + "step": 15563 + }, + { + "epoch": 13.14527027027027, + "grad_norm": 0.22659212350845337, + "learning_rate": 2.8146666666666666e-05, + "loss": 0.0208, + "step": 15564 + }, + { + "epoch": 13.146114864864865, + "grad_norm": 0.4297511577606201, + "learning_rate": 2.8146333333333335e-05, + "loss": 0.0152, + "step": 15565 + }, + { + "epoch": 13.14695945945946, + "grad_norm": 0.29564806818962097, + "learning_rate": 2.8146e-05, + "loss": 0.0146, + "step": 15566 + }, + { + "epoch": 13.147804054054054, + "grad_norm": 0.2539292573928833, + "learning_rate": 2.8145666666666667e-05, + "loss": 0.0145, + "step": 15567 + }, + { + "epoch": 13.14864864864865, + "grad_norm": 0.41298434138298035, + "learning_rate": 2.8145333333333333e-05, + "loss": 0.027, + "step": 15568 + }, + { + "epoch": 13.149493243243244, + "grad_norm": 0.2530510723590851, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0185, + "step": 15569 + }, + { + "epoch": 13.150337837837839, + "grad_norm": 0.19325169920921326, + "learning_rate": 2.8144666666666665e-05, + "loss": 0.0132, + "step": 15570 + }, + { + "epoch": 13.151182432432432, + "grad_norm": 0.2689879238605499, + "learning_rate": 2.8144333333333334e-05, + "loss": 0.0139, + "step": 15571 + }, + { + "epoch": 13.152027027027026, + "grad_norm": 0.3643855154514313, + "learning_rate": 2.8144000000000004e-05, + "loss": 0.027, + "step": 15572 + }, + { + "epoch": 13.152871621621621, + "grad_norm": 0.578828752040863, + "learning_rate": 2.8143666666666666e-05, + "loss": 0.0429, + "step": 15573 + }, + { + "epoch": 13.153716216216216, + "grad_norm": 0.37823793292045593, + "learning_rate": 2.8143333333333335e-05, + "loss": 0.0185, + "step": 15574 + }, + { + "epoch": 13.15456081081081, + "grad_norm": 0.28256961703300476, + "learning_rate": 2.8143e-05, + "loss": 0.0124, + "step": 15575 + }, + { + "epoch": 13.155405405405405, + "grad_norm": 0.21592898666858673, + "learning_rate": 2.8142666666666667e-05, + "loss": 0.0265, + "step": 15576 + }, + { + "epoch": 13.15625, + "grad_norm": 0.32616519927978516, + "learning_rate": 2.8142333333333333e-05, + "loss": 0.0211, + "step": 15577 + }, + { + "epoch": 13.157094594594595, + "grad_norm": 0.24334919452667236, + "learning_rate": 2.8142000000000003e-05, + "loss": 0.0176, + "step": 15578 + }, + { + "epoch": 13.15793918918919, + "grad_norm": 0.6225990653038025, + "learning_rate": 2.8141666666666665e-05, + "loss": 0.0097, + "step": 15579 + }, + { + "epoch": 13.158783783783784, + "grad_norm": 0.19872434437274933, + "learning_rate": 2.8141333333333334e-05, + "loss": 0.0093, + "step": 15580 + }, + { + "epoch": 13.159628378378379, + "grad_norm": 0.6967335343360901, + "learning_rate": 2.8141000000000004e-05, + "loss": 0.0245, + "step": 15581 + }, + { + "epoch": 13.160472972972974, + "grad_norm": 0.35573074221611023, + "learning_rate": 2.8140666666666666e-05, + "loss": 0.0243, + "step": 15582 + }, + { + "epoch": 13.161317567567568, + "grad_norm": 0.35752493143081665, + "learning_rate": 2.8140333333333336e-05, + "loss": 0.0133, + "step": 15583 + }, + { + "epoch": 13.162162162162161, + "grad_norm": 0.4822300374507904, + "learning_rate": 2.8139999999999998e-05, + "loss": 0.0192, + "step": 15584 + }, + { + "epoch": 13.163006756756756, + "grad_norm": 0.40969735383987427, + "learning_rate": 2.8139666666666668e-05, + "loss": 0.018, + "step": 15585 + }, + { + "epoch": 13.16385135135135, + "grad_norm": 0.2160245031118393, + "learning_rate": 2.8139333333333333e-05, + "loss": 0.0118, + "step": 15586 + }, + { + "epoch": 13.164695945945946, + "grad_norm": 0.522192120552063, + "learning_rate": 2.8139e-05, + "loss": 0.0132, + "step": 15587 + }, + { + "epoch": 13.16554054054054, + "grad_norm": 0.42155829071998596, + "learning_rate": 2.8138666666666665e-05, + "loss": 0.0173, + "step": 15588 + }, + { + "epoch": 13.166385135135135, + "grad_norm": 0.512875497341156, + "learning_rate": 2.8138333333333335e-05, + "loss": 0.0295, + "step": 15589 + }, + { + "epoch": 13.16722972972973, + "grad_norm": 0.7221590280532837, + "learning_rate": 2.8138e-05, + "loss": 0.0322, + "step": 15590 + }, + { + "epoch": 13.168074324324325, + "grad_norm": 0.43677249550819397, + "learning_rate": 2.8137666666666667e-05, + "loss": 0.0135, + "step": 15591 + }, + { + "epoch": 13.16891891891892, + "grad_norm": 1.2110378742218018, + "learning_rate": 2.8137333333333336e-05, + "loss": 0.0287, + "step": 15592 + }, + { + "epoch": 13.169763513513514, + "grad_norm": 0.522948145866394, + "learning_rate": 2.8137e-05, + "loss": 0.2211, + "step": 15593 + }, + { + "epoch": 13.170608108108109, + "grad_norm": 0.5376609563827515, + "learning_rate": 2.8136666666666668e-05, + "loss": 0.1176, + "step": 15594 + }, + { + "epoch": 13.171452702702704, + "grad_norm": 0.9254939556121826, + "learning_rate": 2.8136333333333334e-05, + "loss": 0.071, + "step": 15595 + }, + { + "epoch": 13.172297297297296, + "grad_norm": 0.3304113447666168, + "learning_rate": 2.8136e-05, + "loss": 0.0477, + "step": 15596 + }, + { + "epoch": 13.173141891891891, + "grad_norm": 0.384090393781662, + "learning_rate": 2.813566666666667e-05, + "loss": 0.034, + "step": 15597 + }, + { + "epoch": 13.173986486486486, + "grad_norm": 0.404187947511673, + "learning_rate": 2.8135333333333335e-05, + "loss": 0.036, + "step": 15598 + }, + { + "epoch": 13.17483108108108, + "grad_norm": 0.28797295689582825, + "learning_rate": 2.8135e-05, + "loss": 0.0276, + "step": 15599 + }, + { + "epoch": 13.175675675675675, + "grad_norm": 0.2739394009113312, + "learning_rate": 2.8134666666666667e-05, + "loss": 0.0135, + "step": 15600 + }, + { + "epoch": 13.17652027027027, + "grad_norm": 0.40703949332237244, + "learning_rate": 2.8134333333333336e-05, + "loss": 0.0218, + "step": 15601 + }, + { + "epoch": 13.177364864864865, + "grad_norm": 0.3338856101036072, + "learning_rate": 2.8134e-05, + "loss": 0.025, + "step": 15602 + }, + { + "epoch": 13.17820945945946, + "grad_norm": 0.25709569454193115, + "learning_rate": 2.8133666666666668e-05, + "loss": 0.0261, + "step": 15603 + }, + { + "epoch": 13.179054054054054, + "grad_norm": 0.32527512311935425, + "learning_rate": 2.8133333333333334e-05, + "loss": 0.0345, + "step": 15604 + }, + { + "epoch": 13.17989864864865, + "grad_norm": 0.24307002127170563, + "learning_rate": 2.8133e-05, + "loss": 0.0223, + "step": 15605 + }, + { + "epoch": 13.180743243243244, + "grad_norm": 0.2219260036945343, + "learning_rate": 2.813266666666667e-05, + "loss": 0.0135, + "step": 15606 + }, + { + "epoch": 13.181587837837839, + "grad_norm": 0.1974135935306549, + "learning_rate": 2.8132333333333335e-05, + "loss": 0.0224, + "step": 15607 + }, + { + "epoch": 13.182432432432432, + "grad_norm": 0.35760319232940674, + "learning_rate": 2.8132e-05, + "loss": 0.0305, + "step": 15608 + }, + { + "epoch": 13.183277027027026, + "grad_norm": 0.31328514218330383, + "learning_rate": 2.8131666666666667e-05, + "loss": 0.019, + "step": 15609 + }, + { + "epoch": 13.184121621621621, + "grad_norm": 0.34353065490722656, + "learning_rate": 2.8131333333333336e-05, + "loss": 0.0225, + "step": 15610 + }, + { + "epoch": 13.184966216216216, + "grad_norm": 0.49174195528030396, + "learning_rate": 2.8131e-05, + "loss": 0.0159, + "step": 15611 + }, + { + "epoch": 13.18581081081081, + "grad_norm": 0.31881317496299744, + "learning_rate": 2.8130666666666668e-05, + "loss": 0.0182, + "step": 15612 + }, + { + "epoch": 13.186655405405405, + "grad_norm": 0.8637452125549316, + "learning_rate": 2.8130333333333334e-05, + "loss": 0.017, + "step": 15613 + }, + { + "epoch": 13.1875, + "grad_norm": 0.38438037037849426, + "learning_rate": 2.813e-05, + "loss": 0.012, + "step": 15614 + }, + { + "epoch": 13.188344594594595, + "grad_norm": 0.24541445076465607, + "learning_rate": 2.812966666666667e-05, + "loss": 0.0202, + "step": 15615 + }, + { + "epoch": 13.18918918918919, + "grad_norm": 0.2500370740890503, + "learning_rate": 2.8129333333333332e-05, + "loss": 0.0303, + "step": 15616 + }, + { + "epoch": 13.190033783783784, + "grad_norm": 0.36300814151763916, + "learning_rate": 2.8129e-05, + "loss": 0.0251, + "step": 15617 + }, + { + "epoch": 13.190878378378379, + "grad_norm": 0.28011855483055115, + "learning_rate": 2.8128666666666667e-05, + "loss": 0.0203, + "step": 15618 + }, + { + "epoch": 13.191722972972974, + "grad_norm": 0.23478814959526062, + "learning_rate": 2.8128333333333333e-05, + "loss": 0.0145, + "step": 15619 + }, + { + "epoch": 13.192567567567568, + "grad_norm": 0.6298639178276062, + "learning_rate": 2.8128e-05, + "loss": 0.0242, + "step": 15620 + }, + { + "epoch": 13.193412162162161, + "grad_norm": 0.23019447922706604, + "learning_rate": 2.812766666666667e-05, + "loss": 0.0182, + "step": 15621 + }, + { + "epoch": 13.194256756756756, + "grad_norm": 0.3013075292110443, + "learning_rate": 2.8127333333333334e-05, + "loss": 0.0212, + "step": 15622 + }, + { + "epoch": 13.19510135135135, + "grad_norm": 0.36425694823265076, + "learning_rate": 2.8127e-05, + "loss": 0.0202, + "step": 15623 + }, + { + "epoch": 13.195945945945946, + "grad_norm": 0.5416908264160156, + "learning_rate": 2.812666666666667e-05, + "loss": 0.0253, + "step": 15624 + }, + { + "epoch": 13.19679054054054, + "grad_norm": 0.1419326364994049, + "learning_rate": 2.8126333333333332e-05, + "loss": 0.0105, + "step": 15625 + }, + { + "epoch": 13.197635135135135, + "grad_norm": 0.2327638417482376, + "learning_rate": 2.8126e-05, + "loss": 0.0203, + "step": 15626 + }, + { + "epoch": 13.19847972972973, + "grad_norm": 0.486268550157547, + "learning_rate": 2.8125666666666667e-05, + "loss": 0.0209, + "step": 15627 + }, + { + "epoch": 13.199324324324325, + "grad_norm": 0.8901405334472656, + "learning_rate": 2.8125333333333333e-05, + "loss": 0.0181, + "step": 15628 + }, + { + "epoch": 13.20016891891892, + "grad_norm": 0.36936402320861816, + "learning_rate": 2.8125e-05, + "loss": 0.0278, + "step": 15629 + }, + { + "epoch": 13.201013513513514, + "grad_norm": 0.33903324604034424, + "learning_rate": 2.812466666666667e-05, + "loss": 0.0248, + "step": 15630 + }, + { + "epoch": 13.201858108108109, + "grad_norm": 0.3788880705833435, + "learning_rate": 2.8124333333333334e-05, + "loss": 0.0197, + "step": 15631 + }, + { + "epoch": 13.202702702702704, + "grad_norm": 0.2500455677509308, + "learning_rate": 2.8124e-05, + "loss": 0.0162, + "step": 15632 + }, + { + "epoch": 13.203547297297296, + "grad_norm": 0.6826362013816833, + "learning_rate": 2.812366666666667e-05, + "loss": 0.0226, + "step": 15633 + }, + { + "epoch": 13.204391891891891, + "grad_norm": 0.22736135125160217, + "learning_rate": 2.8123333333333332e-05, + "loss": 0.0156, + "step": 15634 + }, + { + "epoch": 13.205236486486486, + "grad_norm": 0.24613763391971588, + "learning_rate": 2.8123e-05, + "loss": 0.0161, + "step": 15635 + }, + { + "epoch": 13.20608108108108, + "grad_norm": 0.6949020624160767, + "learning_rate": 2.8122666666666668e-05, + "loss": 0.0353, + "step": 15636 + }, + { + "epoch": 13.206925675675675, + "grad_norm": 0.8968488574028015, + "learning_rate": 2.8122333333333333e-05, + "loss": 0.0297, + "step": 15637 + }, + { + "epoch": 13.20777027027027, + "grad_norm": 0.3564552068710327, + "learning_rate": 2.8122e-05, + "loss": 0.0238, + "step": 15638 + }, + { + "epoch": 13.208614864864865, + "grad_norm": 1.1644171476364136, + "learning_rate": 2.812166666666667e-05, + "loss": 0.0531, + "step": 15639 + }, + { + "epoch": 13.20945945945946, + "grad_norm": 0.40911588072776794, + "learning_rate": 2.8121333333333335e-05, + "loss": 0.0231, + "step": 15640 + }, + { + "epoch": 13.210304054054054, + "grad_norm": 0.44320884346961975, + "learning_rate": 2.8121e-05, + "loss": 0.0301, + "step": 15641 + }, + { + "epoch": 13.21114864864865, + "grad_norm": 0.5167083740234375, + "learning_rate": 2.812066666666667e-05, + "loss": 0.0201, + "step": 15642 + }, + { + "epoch": 13.211993243243244, + "grad_norm": 0.5136405825614929, + "learning_rate": 2.8120333333333332e-05, + "loss": 0.1826, + "step": 15643 + }, + { + "epoch": 13.212837837837839, + "grad_norm": 0.5813616514205933, + "learning_rate": 2.8120000000000002e-05, + "loss": 0.1243, + "step": 15644 + }, + { + "epoch": 13.213682432432432, + "grad_norm": 0.8979243636131287, + "learning_rate": 2.8119666666666664e-05, + "loss": 0.0811, + "step": 15645 + }, + { + "epoch": 13.214527027027026, + "grad_norm": 0.26327258348464966, + "learning_rate": 2.8119333333333334e-05, + "loss": 0.0294, + "step": 15646 + }, + { + "epoch": 13.215371621621621, + "grad_norm": 0.27506715059280396, + "learning_rate": 2.8119000000000003e-05, + "loss": 0.027, + "step": 15647 + }, + { + "epoch": 13.216216216216216, + "grad_norm": 0.32339781522750854, + "learning_rate": 2.8118666666666666e-05, + "loss": 0.0226, + "step": 15648 + }, + { + "epoch": 13.21706081081081, + "grad_norm": 0.3677653670310974, + "learning_rate": 2.8118333333333335e-05, + "loss": 0.0315, + "step": 15649 + }, + { + "epoch": 13.217905405405405, + "grad_norm": 0.16816146671772003, + "learning_rate": 2.8118e-05, + "loss": 0.0125, + "step": 15650 + }, + { + "epoch": 13.21875, + "grad_norm": 0.4118160307407379, + "learning_rate": 2.8117666666666667e-05, + "loss": 0.0486, + "step": 15651 + }, + { + "epoch": 13.219594594594595, + "grad_norm": 0.2981637418270111, + "learning_rate": 2.8117333333333333e-05, + "loss": 0.0317, + "step": 15652 + }, + { + "epoch": 13.22043918918919, + "grad_norm": 0.2160474807024002, + "learning_rate": 2.8117000000000002e-05, + "loss": 0.0128, + "step": 15653 + }, + { + "epoch": 13.221283783783784, + "grad_norm": 0.3572324514389038, + "learning_rate": 2.8116666666666665e-05, + "loss": 0.0243, + "step": 15654 + }, + { + "epoch": 13.222128378378379, + "grad_norm": 0.575950562953949, + "learning_rate": 2.8116333333333334e-05, + "loss": 0.0364, + "step": 15655 + }, + { + "epoch": 13.222972972972974, + "grad_norm": 0.3436470627784729, + "learning_rate": 2.8116000000000003e-05, + "loss": 0.0241, + "step": 15656 + }, + { + "epoch": 13.223817567567568, + "grad_norm": 0.2713339626789093, + "learning_rate": 2.8115666666666666e-05, + "loss": 0.0203, + "step": 15657 + }, + { + "epoch": 13.224662162162161, + "grad_norm": 0.5733011960983276, + "learning_rate": 2.8115333333333335e-05, + "loss": 0.0226, + "step": 15658 + }, + { + "epoch": 13.225506756756756, + "grad_norm": 0.3078661561012268, + "learning_rate": 2.8115e-05, + "loss": 0.0241, + "step": 15659 + }, + { + "epoch": 13.22635135135135, + "grad_norm": 0.3757798969745636, + "learning_rate": 2.8114666666666667e-05, + "loss": 0.0355, + "step": 15660 + }, + { + "epoch": 13.227195945945946, + "grad_norm": 0.19712509214878082, + "learning_rate": 2.8114333333333333e-05, + "loss": 0.0105, + "step": 15661 + }, + { + "epoch": 13.22804054054054, + "grad_norm": 0.2832326292991638, + "learning_rate": 2.8114000000000002e-05, + "loss": 0.0274, + "step": 15662 + }, + { + "epoch": 13.228885135135135, + "grad_norm": 0.14067675173282623, + "learning_rate": 2.8113666666666665e-05, + "loss": 0.0122, + "step": 15663 + }, + { + "epoch": 13.22972972972973, + "grad_norm": 0.2860272228717804, + "learning_rate": 2.8113333333333334e-05, + "loss": 0.0235, + "step": 15664 + }, + { + "epoch": 13.230574324324325, + "grad_norm": 0.26523271203041077, + "learning_rate": 2.8113000000000003e-05, + "loss": 0.0183, + "step": 15665 + }, + { + "epoch": 13.23141891891892, + "grad_norm": 0.19746506214141846, + "learning_rate": 2.8112666666666666e-05, + "loss": 0.022, + "step": 15666 + }, + { + "epoch": 13.232263513513514, + "grad_norm": 0.5444831252098083, + "learning_rate": 2.8112333333333335e-05, + "loss": 0.0324, + "step": 15667 + }, + { + "epoch": 13.233108108108109, + "grad_norm": 0.3470064699649811, + "learning_rate": 2.8112e-05, + "loss": 0.0166, + "step": 15668 + }, + { + "epoch": 13.233952702702704, + "grad_norm": 0.17795640230178833, + "learning_rate": 2.8111666666666667e-05, + "loss": 0.0156, + "step": 15669 + }, + { + "epoch": 13.234797297297296, + "grad_norm": 0.24038852751255035, + "learning_rate": 2.8111333333333333e-05, + "loss": 0.0163, + "step": 15670 + }, + { + "epoch": 13.235641891891891, + "grad_norm": 0.18348939716815948, + "learning_rate": 2.8111000000000002e-05, + "loss": 0.0068, + "step": 15671 + }, + { + "epoch": 13.236486486486486, + "grad_norm": 0.40428316593170166, + "learning_rate": 2.811066666666667e-05, + "loss": 0.0385, + "step": 15672 + }, + { + "epoch": 13.23733108108108, + "grad_norm": 0.36096784472465515, + "learning_rate": 2.8110333333333334e-05, + "loss": 0.0168, + "step": 15673 + }, + { + "epoch": 13.238175675675675, + "grad_norm": 0.3632524907588959, + "learning_rate": 2.8110000000000004e-05, + "loss": 0.0199, + "step": 15674 + }, + { + "epoch": 13.23902027027027, + "grad_norm": 0.16194535791873932, + "learning_rate": 2.8109666666666666e-05, + "loss": 0.0135, + "step": 15675 + }, + { + "epoch": 13.239864864864865, + "grad_norm": 0.2820940315723419, + "learning_rate": 2.8109333333333335e-05, + "loss": 0.0133, + "step": 15676 + }, + { + "epoch": 13.24070945945946, + "grad_norm": 0.169198676943779, + "learning_rate": 2.8108999999999998e-05, + "loss": 0.0084, + "step": 15677 + }, + { + "epoch": 13.241554054054054, + "grad_norm": 0.3834323287010193, + "learning_rate": 2.8108666666666667e-05, + "loss": 0.03, + "step": 15678 + }, + { + "epoch": 13.24239864864865, + "grad_norm": 0.21321485936641693, + "learning_rate": 2.8108333333333333e-05, + "loss": 0.0152, + "step": 15679 + }, + { + "epoch": 13.243243243243244, + "grad_norm": 0.24257628619670868, + "learning_rate": 2.8108e-05, + "loss": 0.0231, + "step": 15680 + }, + { + "epoch": 13.244087837837839, + "grad_norm": 0.4048645794391632, + "learning_rate": 2.810766666666667e-05, + "loss": 0.0123, + "step": 15681 + }, + { + "epoch": 13.244932432432432, + "grad_norm": 0.515625, + "learning_rate": 2.8107333333333334e-05, + "loss": 0.0215, + "step": 15682 + }, + { + "epoch": 13.245777027027026, + "grad_norm": 0.4084699749946594, + "learning_rate": 2.8107e-05, + "loss": 0.0104, + "step": 15683 + }, + { + "epoch": 13.246621621621621, + "grad_norm": 0.61988365650177, + "learning_rate": 2.8106666666666666e-05, + "loss": 0.0226, + "step": 15684 + }, + { + "epoch": 13.247466216216216, + "grad_norm": 0.3020993769168854, + "learning_rate": 2.8106333333333336e-05, + "loss": 0.0263, + "step": 15685 + }, + { + "epoch": 13.24831081081081, + "grad_norm": 0.2127891629934311, + "learning_rate": 2.8105999999999998e-05, + "loss": 0.0097, + "step": 15686 + }, + { + "epoch": 13.249155405405405, + "grad_norm": 0.27286866307258606, + "learning_rate": 2.8105666666666668e-05, + "loss": 0.0205, + "step": 15687 + }, + { + "epoch": 13.25, + "grad_norm": 0.17843709886074066, + "learning_rate": 2.8105333333333333e-05, + "loss": 0.014, + "step": 15688 + }, + { + "epoch": 13.250844594594595, + "grad_norm": 0.21818719804286957, + "learning_rate": 2.8105e-05, + "loss": 0.0231, + "step": 15689 + }, + { + "epoch": 13.25168918918919, + "grad_norm": 0.4368768334388733, + "learning_rate": 2.810466666666667e-05, + "loss": 0.0212, + "step": 15690 + }, + { + "epoch": 13.252533783783784, + "grad_norm": 0.46360135078430176, + "learning_rate": 2.8104333333333335e-05, + "loss": 0.0281, + "step": 15691 + }, + { + "epoch": 13.253378378378379, + "grad_norm": 1.0046542882919312, + "learning_rate": 2.8104e-05, + "loss": 0.0265, + "step": 15692 + }, + { + "epoch": 13.254222972972974, + "grad_norm": 0.7610665559768677, + "learning_rate": 2.8103666666666667e-05, + "loss": 0.1836, + "step": 15693 + }, + { + "epoch": 13.255067567567568, + "grad_norm": 0.3691574037075043, + "learning_rate": 2.8103333333333336e-05, + "loss": 0.1114, + "step": 15694 + }, + { + "epoch": 13.255912162162161, + "grad_norm": 0.283875435590744, + "learning_rate": 2.8103e-05, + "loss": 0.0436, + "step": 15695 + }, + { + "epoch": 13.256756756756756, + "grad_norm": 0.30248066782951355, + "learning_rate": 2.8102666666666668e-05, + "loss": 0.034, + "step": 15696 + }, + { + "epoch": 13.25760135135135, + "grad_norm": 0.4668428897857666, + "learning_rate": 2.8102333333333337e-05, + "loss": 0.0541, + "step": 15697 + }, + { + "epoch": 13.258445945945946, + "grad_norm": 0.27735400199890137, + "learning_rate": 2.8102e-05, + "loss": 0.0216, + "step": 15698 + }, + { + "epoch": 13.25929054054054, + "grad_norm": 0.3925253748893738, + "learning_rate": 2.810166666666667e-05, + "loss": 0.021, + "step": 15699 + }, + { + "epoch": 13.260135135135135, + "grad_norm": 0.2955304980278015, + "learning_rate": 2.8101333333333335e-05, + "loss": 0.0242, + "step": 15700 + }, + { + "epoch": 13.26097972972973, + "grad_norm": 0.6154532432556152, + "learning_rate": 2.8101e-05, + "loss": 0.0219, + "step": 15701 + }, + { + "epoch": 13.261824324324325, + "grad_norm": 0.29190266132354736, + "learning_rate": 2.8100666666666667e-05, + "loss": 0.0201, + "step": 15702 + }, + { + "epoch": 13.26266891891892, + "grad_norm": 0.3205801844596863, + "learning_rate": 2.8100333333333336e-05, + "loss": 0.0241, + "step": 15703 + }, + { + "epoch": 13.263513513513514, + "grad_norm": 0.29748666286468506, + "learning_rate": 2.81e-05, + "loss": 0.012, + "step": 15704 + }, + { + "epoch": 13.264358108108109, + "grad_norm": 0.3068901300430298, + "learning_rate": 2.8099666666666668e-05, + "loss": 0.0178, + "step": 15705 + }, + { + "epoch": 13.265202702702704, + "grad_norm": 0.20116625726222992, + "learning_rate": 2.8099333333333334e-05, + "loss": 0.0147, + "step": 15706 + }, + { + "epoch": 13.266047297297296, + "grad_norm": 0.29002997279167175, + "learning_rate": 2.8099e-05, + "loss": 0.0264, + "step": 15707 + }, + { + "epoch": 13.266891891891891, + "grad_norm": 0.33208170533180237, + "learning_rate": 2.809866666666667e-05, + "loss": 0.0219, + "step": 15708 + }, + { + "epoch": 13.267736486486486, + "grad_norm": 0.5819012522697449, + "learning_rate": 2.809833333333333e-05, + "loss": 0.0369, + "step": 15709 + }, + { + "epoch": 13.26858108108108, + "grad_norm": 0.3663017451763153, + "learning_rate": 2.8098e-05, + "loss": 0.0184, + "step": 15710 + }, + { + "epoch": 13.269425675675675, + "grad_norm": 0.2526068389415741, + "learning_rate": 2.8097666666666667e-05, + "loss": 0.0181, + "step": 15711 + }, + { + "epoch": 13.27027027027027, + "grad_norm": 0.3066852390766144, + "learning_rate": 2.8097333333333333e-05, + "loss": 0.0193, + "step": 15712 + }, + { + "epoch": 13.271114864864865, + "grad_norm": 0.6115442514419556, + "learning_rate": 2.8097e-05, + "loss": 0.0261, + "step": 15713 + }, + { + "epoch": 13.27195945945946, + "grad_norm": 0.31536176800727844, + "learning_rate": 2.8096666666666668e-05, + "loss": 0.0167, + "step": 15714 + }, + { + "epoch": 13.272804054054054, + "grad_norm": 0.3682655692100525, + "learning_rate": 2.8096333333333334e-05, + "loss": 0.0226, + "step": 15715 + }, + { + "epoch": 13.27364864864865, + "grad_norm": 0.3536514937877655, + "learning_rate": 2.8096e-05, + "loss": 0.0589, + "step": 15716 + }, + { + "epoch": 13.274493243243244, + "grad_norm": 0.37614762783050537, + "learning_rate": 2.809566666666667e-05, + "loss": 0.0228, + "step": 15717 + }, + { + "epoch": 13.275337837837839, + "grad_norm": 0.1807490736246109, + "learning_rate": 2.8095333333333332e-05, + "loss": 0.0183, + "step": 15718 + }, + { + "epoch": 13.276182432432432, + "grad_norm": 0.1787104308605194, + "learning_rate": 2.8095e-05, + "loss": 0.0185, + "step": 15719 + }, + { + "epoch": 13.277027027027026, + "grad_norm": 0.21833881735801697, + "learning_rate": 2.8094666666666667e-05, + "loss": 0.0122, + "step": 15720 + }, + { + "epoch": 13.277871621621621, + "grad_norm": 0.2550562918186188, + "learning_rate": 2.8094333333333333e-05, + "loss": 0.0127, + "step": 15721 + }, + { + "epoch": 13.278716216216216, + "grad_norm": 0.4407607614994049, + "learning_rate": 2.8094000000000002e-05, + "loss": 0.0211, + "step": 15722 + }, + { + "epoch": 13.27956081081081, + "grad_norm": 0.2366940975189209, + "learning_rate": 2.809366666666667e-05, + "loss": 0.0161, + "step": 15723 + }, + { + "epoch": 13.280405405405405, + "grad_norm": 0.33549782633781433, + "learning_rate": 2.8093333333333334e-05, + "loss": 0.0171, + "step": 15724 + }, + { + "epoch": 13.28125, + "grad_norm": 0.4432777166366577, + "learning_rate": 2.8093e-05, + "loss": 0.0148, + "step": 15725 + }, + { + "epoch": 13.282094594594595, + "grad_norm": 0.224834144115448, + "learning_rate": 2.809266666666667e-05, + "loss": 0.0164, + "step": 15726 + }, + { + "epoch": 13.28293918918919, + "grad_norm": 0.4876609146595001, + "learning_rate": 2.8092333333333332e-05, + "loss": 0.0216, + "step": 15727 + }, + { + "epoch": 13.283783783783784, + "grad_norm": 0.19154037535190582, + "learning_rate": 2.8092e-05, + "loss": 0.0191, + "step": 15728 + }, + { + "epoch": 13.284628378378379, + "grad_norm": 0.42850059270858765, + "learning_rate": 2.8091666666666667e-05, + "loss": 0.0126, + "step": 15729 + }, + { + "epoch": 13.285472972972974, + "grad_norm": 0.18324808776378632, + "learning_rate": 2.8091333333333333e-05, + "loss": 0.0157, + "step": 15730 + }, + { + "epoch": 13.286317567567568, + "grad_norm": 0.6091440916061401, + "learning_rate": 2.8091000000000003e-05, + "loss": 0.0217, + "step": 15731 + }, + { + "epoch": 13.287162162162161, + "grad_norm": 0.23198416829109192, + "learning_rate": 2.809066666666667e-05, + "loss": 0.0146, + "step": 15732 + }, + { + "epoch": 13.288006756756756, + "grad_norm": 1.1533719301223755, + "learning_rate": 2.8090333333333334e-05, + "loss": 0.0456, + "step": 15733 + }, + { + "epoch": 13.28885135135135, + "grad_norm": 0.24294205009937286, + "learning_rate": 2.809e-05, + "loss": 0.0098, + "step": 15734 + }, + { + "epoch": 13.289695945945946, + "grad_norm": 0.31196293234825134, + "learning_rate": 2.8089666666666666e-05, + "loss": 0.0164, + "step": 15735 + }, + { + "epoch": 13.29054054054054, + "grad_norm": 0.33212751150131226, + "learning_rate": 2.8089333333333332e-05, + "loss": 0.0213, + "step": 15736 + }, + { + "epoch": 13.291385135135135, + "grad_norm": 0.2021138072013855, + "learning_rate": 2.8089e-05, + "loss": 0.0125, + "step": 15737 + }, + { + "epoch": 13.29222972972973, + "grad_norm": 0.3796827793121338, + "learning_rate": 2.8088666666666664e-05, + "loss": 0.0268, + "step": 15738 + }, + { + "epoch": 13.293074324324325, + "grad_norm": 0.22685706615447998, + "learning_rate": 2.8088333333333333e-05, + "loss": 0.0132, + "step": 15739 + }, + { + "epoch": 13.29391891891892, + "grad_norm": 0.4598744511604309, + "learning_rate": 2.8088000000000003e-05, + "loss": 0.0239, + "step": 15740 + }, + { + "epoch": 13.294763513513514, + "grad_norm": 0.6139001846313477, + "learning_rate": 2.8087666666666665e-05, + "loss": 0.0347, + "step": 15741 + }, + { + "epoch": 13.295608108108109, + "grad_norm": 0.5347630977630615, + "learning_rate": 2.8087333333333335e-05, + "loss": 0.023, + "step": 15742 + }, + { + "epoch": 13.296452702702704, + "grad_norm": 0.6450557708740234, + "learning_rate": 2.8087e-05, + "loss": 0.1688, + "step": 15743 + }, + { + "epoch": 13.297297297297296, + "grad_norm": 0.714684784412384, + "learning_rate": 2.8086666666666667e-05, + "loss": 0.085, + "step": 15744 + }, + { + "epoch": 13.298141891891891, + "grad_norm": 0.3291566073894501, + "learning_rate": 2.8086333333333332e-05, + "loss": 0.0355, + "step": 15745 + }, + { + "epoch": 13.298986486486486, + "grad_norm": 0.38855063915252686, + "learning_rate": 2.8086000000000002e-05, + "loss": 0.0328, + "step": 15746 + }, + { + "epoch": 13.29983108108108, + "grad_norm": 0.6569535732269287, + "learning_rate": 2.8085666666666668e-05, + "loss": 0.0371, + "step": 15747 + }, + { + "epoch": 13.300675675675675, + "grad_norm": 0.25141286849975586, + "learning_rate": 2.8085333333333334e-05, + "loss": 0.0166, + "step": 15748 + }, + { + "epoch": 13.30152027027027, + "grad_norm": 0.24603328108787537, + "learning_rate": 2.8085000000000003e-05, + "loss": 0.0186, + "step": 15749 + }, + { + "epoch": 13.302364864864865, + "grad_norm": 0.27867481112480164, + "learning_rate": 2.8084666666666666e-05, + "loss": 0.0159, + "step": 15750 + }, + { + "epoch": 13.30320945945946, + "grad_norm": 0.736771285533905, + "learning_rate": 2.8084333333333335e-05, + "loss": 0.0177, + "step": 15751 + }, + { + "epoch": 13.304054054054054, + "grad_norm": 0.4161054491996765, + "learning_rate": 2.8084e-05, + "loss": 0.0207, + "step": 15752 + }, + { + "epoch": 13.30489864864865, + "grad_norm": 0.2811409831047058, + "learning_rate": 2.8083666666666667e-05, + "loss": 0.0196, + "step": 15753 + }, + { + "epoch": 13.305743243243244, + "grad_norm": 0.2954884171485901, + "learning_rate": 2.8083333333333333e-05, + "loss": 0.0301, + "step": 15754 + }, + { + "epoch": 13.306587837837839, + "grad_norm": 0.19117748737335205, + "learning_rate": 2.8083000000000002e-05, + "loss": 0.0192, + "step": 15755 + }, + { + "epoch": 13.307432432432432, + "grad_norm": 0.16135680675506592, + "learning_rate": 2.8082666666666668e-05, + "loss": 0.0106, + "step": 15756 + }, + { + "epoch": 13.308277027027026, + "grad_norm": 0.24245710670948029, + "learning_rate": 2.8082333333333334e-05, + "loss": 0.0129, + "step": 15757 + }, + { + "epoch": 13.309121621621621, + "grad_norm": 0.3490409553050995, + "learning_rate": 2.8082000000000003e-05, + "loss": 0.0173, + "step": 15758 + }, + { + "epoch": 13.309966216216216, + "grad_norm": 0.25287926197052, + "learning_rate": 2.8081666666666666e-05, + "loss": 0.0216, + "step": 15759 + }, + { + "epoch": 13.31081081081081, + "grad_norm": 0.4848189651966095, + "learning_rate": 2.8081333333333335e-05, + "loss": 0.0238, + "step": 15760 + }, + { + "epoch": 13.311655405405405, + "grad_norm": 0.25321996212005615, + "learning_rate": 2.8081e-05, + "loss": 0.0177, + "step": 15761 + }, + { + "epoch": 13.3125, + "grad_norm": 0.22356770932674408, + "learning_rate": 2.8080666666666667e-05, + "loss": 0.0222, + "step": 15762 + }, + { + "epoch": 13.313344594594595, + "grad_norm": 0.7963550090789795, + "learning_rate": 2.8080333333333333e-05, + "loss": 0.0215, + "step": 15763 + }, + { + "epoch": 13.31418918918919, + "grad_norm": 0.22669649124145508, + "learning_rate": 2.8080000000000002e-05, + "loss": 0.0143, + "step": 15764 + }, + { + "epoch": 13.315033783783784, + "grad_norm": 0.2773763835430145, + "learning_rate": 2.8079666666666668e-05, + "loss": 0.0225, + "step": 15765 + }, + { + "epoch": 13.315878378378379, + "grad_norm": 0.251603901386261, + "learning_rate": 2.8079333333333334e-05, + "loss": 0.0156, + "step": 15766 + }, + { + "epoch": 13.316722972972974, + "grad_norm": 0.14407871663570404, + "learning_rate": 2.8079e-05, + "loss": 0.0086, + "step": 15767 + }, + { + "epoch": 13.317567567567568, + "grad_norm": 0.20863646268844604, + "learning_rate": 2.8078666666666666e-05, + "loss": 0.017, + "step": 15768 + }, + { + "epoch": 13.318412162162161, + "grad_norm": 0.2847297787666321, + "learning_rate": 2.8078333333333335e-05, + "loss": 0.0216, + "step": 15769 + }, + { + "epoch": 13.319256756756756, + "grad_norm": 0.5993826985359192, + "learning_rate": 2.8077999999999998e-05, + "loss": 0.017, + "step": 15770 + }, + { + "epoch": 13.32010135135135, + "grad_norm": 0.35385850071907043, + "learning_rate": 2.8077666666666667e-05, + "loss": 0.0163, + "step": 15771 + }, + { + "epoch": 13.320945945945946, + "grad_norm": 0.26875773072242737, + "learning_rate": 2.8077333333333336e-05, + "loss": 0.0232, + "step": 15772 + }, + { + "epoch": 13.32179054054054, + "grad_norm": 0.142966628074646, + "learning_rate": 2.8077e-05, + "loss": 0.0123, + "step": 15773 + }, + { + "epoch": 13.322635135135135, + "grad_norm": 0.2361687421798706, + "learning_rate": 2.807666666666667e-05, + "loss": 0.0218, + "step": 15774 + }, + { + "epoch": 13.32347972972973, + "grad_norm": 0.34851470589637756, + "learning_rate": 2.8076333333333334e-05, + "loss": 0.0139, + "step": 15775 + }, + { + "epoch": 13.324324324324325, + "grad_norm": 0.2294856160879135, + "learning_rate": 2.8076e-05, + "loss": 0.0248, + "step": 15776 + }, + { + "epoch": 13.32516891891892, + "grad_norm": 0.34219276905059814, + "learning_rate": 2.8075666666666666e-05, + "loss": 0.0239, + "step": 15777 + }, + { + "epoch": 13.326013513513514, + "grad_norm": 0.16832390427589417, + "learning_rate": 2.8075333333333335e-05, + "loss": 0.0116, + "step": 15778 + }, + { + "epoch": 13.326858108108109, + "grad_norm": 0.23847253620624542, + "learning_rate": 2.8074999999999998e-05, + "loss": 0.0174, + "step": 15779 + }, + { + "epoch": 13.327702702702704, + "grad_norm": 0.28626489639282227, + "learning_rate": 2.8074666666666667e-05, + "loss": 0.0218, + "step": 15780 + }, + { + "epoch": 13.328547297297296, + "grad_norm": 0.24276621639728546, + "learning_rate": 2.8074333333333337e-05, + "loss": 0.0138, + "step": 15781 + }, + { + "epoch": 13.329391891891891, + "grad_norm": 0.3397562801837921, + "learning_rate": 2.8074e-05, + "loss": 0.0245, + "step": 15782 + }, + { + "epoch": 13.330236486486486, + "grad_norm": 0.16481351852416992, + "learning_rate": 2.807366666666667e-05, + "loss": 0.0125, + "step": 15783 + }, + { + "epoch": 13.33108108108108, + "grad_norm": 0.4146316945552826, + "learning_rate": 2.8073333333333334e-05, + "loss": 0.0238, + "step": 15784 + }, + { + "epoch": 13.331925675675675, + "grad_norm": 0.3756113648414612, + "learning_rate": 2.8073e-05, + "loss": 0.0235, + "step": 15785 + }, + { + "epoch": 13.33277027027027, + "grad_norm": 0.31125009059906006, + "learning_rate": 2.8072666666666666e-05, + "loss": 0.0317, + "step": 15786 + }, + { + "epoch": 13.333614864864865, + "grad_norm": 0.23782943189144135, + "learning_rate": 2.8072333333333336e-05, + "loss": 0.0128, + "step": 15787 + }, + { + "epoch": 13.33445945945946, + "grad_norm": 0.5788887143135071, + "learning_rate": 2.8071999999999998e-05, + "loss": 0.031, + "step": 15788 + }, + { + "epoch": 13.335304054054054, + "grad_norm": 0.41253790259361267, + "learning_rate": 2.8071666666666668e-05, + "loss": 0.0153, + "step": 15789 + }, + { + "epoch": 13.33614864864865, + "grad_norm": 0.2959992289543152, + "learning_rate": 2.8071333333333337e-05, + "loss": 0.0186, + "step": 15790 + }, + { + "epoch": 13.336993243243244, + "grad_norm": 0.6146981716156006, + "learning_rate": 2.8071e-05, + "loss": 0.0369, + "step": 15791 + }, + { + "epoch": 13.337837837837839, + "grad_norm": 0.7833359241485596, + "learning_rate": 2.807066666666667e-05, + "loss": 0.0502, + "step": 15792 + }, + { + "epoch": 13.338682432432432, + "grad_norm": 0.7876502275466919, + "learning_rate": 2.8070333333333335e-05, + "loss": 0.211, + "step": 15793 + }, + { + "epoch": 13.339527027027026, + "grad_norm": 0.5553175806999207, + "learning_rate": 2.807e-05, + "loss": 0.1283, + "step": 15794 + }, + { + "epoch": 13.340371621621621, + "grad_norm": 0.372678279876709, + "learning_rate": 2.8069666666666667e-05, + "loss": 0.0583, + "step": 15795 + }, + { + "epoch": 13.341216216216216, + "grad_norm": 0.34685924649238586, + "learning_rate": 2.8069333333333332e-05, + "loss": 0.0409, + "step": 15796 + }, + { + "epoch": 13.34206081081081, + "grad_norm": 0.16936302185058594, + "learning_rate": 2.8069000000000002e-05, + "loss": 0.0183, + "step": 15797 + }, + { + "epoch": 13.342905405405405, + "grad_norm": 0.27574166655540466, + "learning_rate": 2.8068666666666668e-05, + "loss": 0.015, + "step": 15798 + }, + { + "epoch": 13.34375, + "grad_norm": 0.31280240416526794, + "learning_rate": 2.8068333333333334e-05, + "loss": 0.0254, + "step": 15799 + }, + { + "epoch": 13.344594594594595, + "grad_norm": 0.2248414009809494, + "learning_rate": 2.8068e-05, + "loss": 0.0177, + "step": 15800 + }, + { + "epoch": 13.34543918918919, + "grad_norm": 0.5426284670829773, + "learning_rate": 2.806766666666667e-05, + "loss": 0.0239, + "step": 15801 + }, + { + "epoch": 13.346283783783784, + "grad_norm": 0.3496185839176178, + "learning_rate": 2.806733333333333e-05, + "loss": 0.0295, + "step": 15802 + }, + { + "epoch": 13.347128378378379, + "grad_norm": 0.26863712072372437, + "learning_rate": 2.8067e-05, + "loss": 0.0235, + "step": 15803 + }, + { + "epoch": 13.347972972972974, + "grad_norm": 0.24237439036369324, + "learning_rate": 2.8066666666666667e-05, + "loss": 0.0176, + "step": 15804 + }, + { + "epoch": 13.348817567567568, + "grad_norm": 0.39121922850608826, + "learning_rate": 2.8066333333333333e-05, + "loss": 0.0219, + "step": 15805 + }, + { + "epoch": 13.349662162162161, + "grad_norm": 0.43245795369148254, + "learning_rate": 2.8066000000000002e-05, + "loss": 0.0304, + "step": 15806 + }, + { + "epoch": 13.350506756756756, + "grad_norm": 0.23310743272304535, + "learning_rate": 2.8065666666666668e-05, + "loss": 0.0227, + "step": 15807 + }, + { + "epoch": 13.35135135135135, + "grad_norm": 0.21627368032932281, + "learning_rate": 2.8065333333333334e-05, + "loss": 0.0158, + "step": 15808 + }, + { + "epoch": 13.352195945945946, + "grad_norm": 0.5145116448402405, + "learning_rate": 2.8065e-05, + "loss": 0.0188, + "step": 15809 + }, + { + "epoch": 13.35304054054054, + "grad_norm": 0.2748757600784302, + "learning_rate": 2.806466666666667e-05, + "loss": 0.0173, + "step": 15810 + }, + { + "epoch": 13.353885135135135, + "grad_norm": 0.25924229621887207, + "learning_rate": 2.806433333333333e-05, + "loss": 0.0141, + "step": 15811 + }, + { + "epoch": 13.35472972972973, + "grad_norm": 0.8484669923782349, + "learning_rate": 2.8064e-05, + "loss": 0.0264, + "step": 15812 + }, + { + "epoch": 13.355574324324325, + "grad_norm": 0.7167550325393677, + "learning_rate": 2.8063666666666667e-05, + "loss": 0.0188, + "step": 15813 + }, + { + "epoch": 13.35641891891892, + "grad_norm": 0.6130907535552979, + "learning_rate": 2.8063333333333333e-05, + "loss": 0.0162, + "step": 15814 + }, + { + "epoch": 13.357263513513514, + "grad_norm": 0.8860721588134766, + "learning_rate": 2.8063000000000002e-05, + "loss": 0.0247, + "step": 15815 + }, + { + "epoch": 13.358108108108109, + "grad_norm": 0.6097942590713501, + "learning_rate": 2.8062666666666668e-05, + "loss": 0.0164, + "step": 15816 + }, + { + "epoch": 13.358952702702704, + "grad_norm": 0.29240867495536804, + "learning_rate": 2.8062333333333334e-05, + "loss": 0.0173, + "step": 15817 + }, + { + "epoch": 13.359797297297296, + "grad_norm": 0.39573216438293457, + "learning_rate": 2.8062e-05, + "loss": 0.0221, + "step": 15818 + }, + { + "epoch": 13.360641891891891, + "grad_norm": 0.47419989109039307, + "learning_rate": 2.806166666666667e-05, + "loss": 0.0166, + "step": 15819 + }, + { + "epoch": 13.361486486486486, + "grad_norm": 0.2588688135147095, + "learning_rate": 2.8061333333333332e-05, + "loss": 0.0166, + "step": 15820 + }, + { + "epoch": 13.36233108108108, + "grad_norm": 0.12169445306062698, + "learning_rate": 2.8061e-05, + "loss": 0.0095, + "step": 15821 + }, + { + "epoch": 13.363175675675675, + "grad_norm": 1.782505750656128, + "learning_rate": 2.806066666666667e-05, + "loss": 0.021, + "step": 15822 + }, + { + "epoch": 13.36402027027027, + "grad_norm": 0.343132346868515, + "learning_rate": 2.8060333333333333e-05, + "loss": 0.0144, + "step": 15823 + }, + { + "epoch": 13.364864864864865, + "grad_norm": 0.45178622007369995, + "learning_rate": 2.8060000000000002e-05, + "loss": 0.0332, + "step": 15824 + }, + { + "epoch": 13.36570945945946, + "grad_norm": 0.37548452615737915, + "learning_rate": 2.805966666666667e-05, + "loss": 0.0191, + "step": 15825 + }, + { + "epoch": 13.366554054054054, + "grad_norm": 0.384745717048645, + "learning_rate": 2.8059333333333334e-05, + "loss": 0.0258, + "step": 15826 + }, + { + "epoch": 13.36739864864865, + "grad_norm": 0.3594878613948822, + "learning_rate": 2.8059e-05, + "loss": 0.0215, + "step": 15827 + }, + { + "epoch": 13.368243243243244, + "grad_norm": 0.41764557361602783, + "learning_rate": 2.8058666666666666e-05, + "loss": 0.0226, + "step": 15828 + }, + { + "epoch": 13.369087837837839, + "grad_norm": 0.4326656758785248, + "learning_rate": 2.8058333333333332e-05, + "loss": 0.0298, + "step": 15829 + }, + { + "epoch": 13.369932432432432, + "grad_norm": 0.3866240978240967, + "learning_rate": 2.8058e-05, + "loss": 0.0196, + "step": 15830 + }, + { + "epoch": 13.370777027027026, + "grad_norm": 0.3585021197795868, + "learning_rate": 2.8057666666666667e-05, + "loss": 0.0269, + "step": 15831 + }, + { + "epoch": 13.371621621621621, + "grad_norm": 0.2342132031917572, + "learning_rate": 2.8057333333333333e-05, + "loss": 0.0174, + "step": 15832 + }, + { + "epoch": 13.372466216216216, + "grad_norm": 0.31453147530555725, + "learning_rate": 2.8057000000000003e-05, + "loss": 0.0243, + "step": 15833 + }, + { + "epoch": 13.37331081081081, + "grad_norm": 0.28458070755004883, + "learning_rate": 2.8056666666666665e-05, + "loss": 0.0174, + "step": 15834 + }, + { + "epoch": 13.374155405405405, + "grad_norm": 0.6153767108917236, + "learning_rate": 2.8056333333333334e-05, + "loss": 0.0244, + "step": 15835 + }, + { + "epoch": 13.375, + "grad_norm": Infinity, + "learning_rate": 2.8056333333333334e-05, + "loss": 0.017, + "step": 15836 + }, + { + "epoch": 13.375844594594595, + "grad_norm": 0.3033629357814789, + "learning_rate": 2.8056e-05, + "loss": 0.0214, + "step": 15837 + }, + { + "epoch": 13.37668918918919, + "grad_norm": 0.3339632451534271, + "learning_rate": 2.8055666666666666e-05, + "loss": 0.0194, + "step": 15838 + }, + { + "epoch": 13.377533783783784, + "grad_norm": 0.3379512131214142, + "learning_rate": 2.8055333333333332e-05, + "loss": 0.016, + "step": 15839 + }, + { + "epoch": 13.378378378378379, + "grad_norm": 1.3842644691467285, + "learning_rate": 2.8055e-05, + "loss": 0.0314, + "step": 15840 + }, + { + "epoch": 13.379222972972974, + "grad_norm": 0.5606817603111267, + "learning_rate": 2.8054666666666668e-05, + "loss": 0.0326, + "step": 15841 + }, + { + "epoch": 13.380067567567568, + "grad_norm": 0.5512973666191101, + "learning_rate": 2.8054333333333333e-05, + "loss": 0.0354, + "step": 15842 + }, + { + "epoch": 13.380912162162161, + "grad_norm": 1.099107027053833, + "learning_rate": 2.8054000000000003e-05, + "loss": 0.2969, + "step": 15843 + }, + { + "epoch": 13.381756756756756, + "grad_norm": 0.7202876806259155, + "learning_rate": 2.8053666666666665e-05, + "loss": 0.1139, + "step": 15844 + }, + { + "epoch": 13.38260135135135, + "grad_norm": 0.41879743337631226, + "learning_rate": 2.8053333333333335e-05, + "loss": 0.0668, + "step": 15845 + }, + { + "epoch": 13.383445945945946, + "grad_norm": 0.32022765278816223, + "learning_rate": 2.8053e-05, + "loss": 0.0312, + "step": 15846 + }, + { + "epoch": 13.38429054054054, + "grad_norm": 0.3028301000595093, + "learning_rate": 2.8052666666666667e-05, + "loss": 0.0339, + "step": 15847 + }, + { + "epoch": 13.385135135135135, + "grad_norm": 0.29945096373558044, + "learning_rate": 2.8052333333333336e-05, + "loss": 0.0272, + "step": 15848 + }, + { + "epoch": 13.38597972972973, + "grad_norm": 0.2781507670879364, + "learning_rate": 2.8052000000000002e-05, + "loss": 0.0215, + "step": 15849 + }, + { + "epoch": 13.386824324324325, + "grad_norm": 0.2578343451023102, + "learning_rate": 2.8051666666666668e-05, + "loss": 0.0197, + "step": 15850 + }, + { + "epoch": 13.38766891891892, + "grad_norm": 0.2342921644449234, + "learning_rate": 2.8051333333333334e-05, + "loss": 0.0301, + "step": 15851 + }, + { + "epoch": 13.388513513513514, + "grad_norm": 0.31224992871284485, + "learning_rate": 2.8051000000000003e-05, + "loss": 0.0261, + "step": 15852 + }, + { + "epoch": 13.389358108108109, + "grad_norm": 0.47933971881866455, + "learning_rate": 2.8050666666666666e-05, + "loss": 0.0234, + "step": 15853 + }, + { + "epoch": 13.390202702702704, + "grad_norm": 0.2567388415336609, + "learning_rate": 2.8050333333333335e-05, + "loss": 0.0151, + "step": 15854 + }, + { + "epoch": 13.391047297297296, + "grad_norm": 0.21783213317394257, + "learning_rate": 2.805e-05, + "loss": 0.0143, + "step": 15855 + }, + { + "epoch": 13.391891891891891, + "grad_norm": 0.30288082361221313, + "learning_rate": 2.8049666666666667e-05, + "loss": 0.03, + "step": 15856 + }, + { + "epoch": 13.392736486486486, + "grad_norm": 0.1402735412120819, + "learning_rate": 2.8049333333333336e-05, + "loss": 0.0101, + "step": 15857 + }, + { + "epoch": 13.39358108108108, + "grad_norm": 0.19213424623012543, + "learning_rate": 2.8049e-05, + "loss": 0.0101, + "step": 15858 + }, + { + "epoch": 13.394425675675675, + "grad_norm": 0.2592262625694275, + "learning_rate": 2.8048666666666668e-05, + "loss": 0.0208, + "step": 15859 + }, + { + "epoch": 13.39527027027027, + "grad_norm": 0.14904752373695374, + "learning_rate": 2.8048333333333334e-05, + "loss": 0.0098, + "step": 15860 + }, + { + "epoch": 13.396114864864865, + "grad_norm": 0.259682834148407, + "learning_rate": 2.8048e-05, + "loss": 0.0241, + "step": 15861 + }, + { + "epoch": 13.39695945945946, + "grad_norm": 0.18433617055416107, + "learning_rate": 2.8047666666666666e-05, + "loss": 0.0141, + "step": 15862 + }, + { + "epoch": 13.397804054054054, + "grad_norm": 0.28621140122413635, + "learning_rate": 2.8047333333333335e-05, + "loss": 0.0133, + "step": 15863 + }, + { + "epoch": 13.39864864864865, + "grad_norm": 0.25246739387512207, + "learning_rate": 2.8047e-05, + "loss": 0.0213, + "step": 15864 + }, + { + "epoch": 13.399493243243244, + "grad_norm": 1.501596450805664, + "learning_rate": 2.8046666666666667e-05, + "loss": 0.0181, + "step": 15865 + }, + { + "epoch": 13.400337837837839, + "grad_norm": 0.41335219144821167, + "learning_rate": 2.8046333333333336e-05, + "loss": 0.0314, + "step": 15866 + }, + { + "epoch": 13.401182432432432, + "grad_norm": 0.14715160429477692, + "learning_rate": 2.8046e-05, + "loss": 0.0079, + "step": 15867 + }, + { + "epoch": 13.402027027027026, + "grad_norm": 0.2574227452278137, + "learning_rate": 2.8045666666666668e-05, + "loss": 0.019, + "step": 15868 + }, + { + "epoch": 13.402871621621621, + "grad_norm": 0.24100260436534882, + "learning_rate": 2.8045333333333334e-05, + "loss": 0.0207, + "step": 15869 + }, + { + "epoch": 13.403716216216216, + "grad_norm": 0.1749485433101654, + "learning_rate": 2.8045e-05, + "loss": 0.0079, + "step": 15870 + }, + { + "epoch": 13.40456081081081, + "grad_norm": 0.5209947824478149, + "learning_rate": 2.8044666666666666e-05, + "loss": 0.0241, + "step": 15871 + }, + { + "epoch": 13.405405405405405, + "grad_norm": 0.42238885164260864, + "learning_rate": 2.8044333333333335e-05, + "loss": 0.0187, + "step": 15872 + }, + { + "epoch": 13.40625, + "grad_norm": 0.10823004692792892, + "learning_rate": 2.8044e-05, + "loss": 0.0105, + "step": 15873 + }, + { + "epoch": 13.407094594594595, + "grad_norm": 0.3769637644290924, + "learning_rate": 2.8043666666666667e-05, + "loss": 0.023, + "step": 15874 + }, + { + "epoch": 13.40793918918919, + "grad_norm": 0.30468621850013733, + "learning_rate": 2.8043333333333336e-05, + "loss": 0.0141, + "step": 15875 + }, + { + "epoch": 13.408783783783784, + "grad_norm": 0.21145501732826233, + "learning_rate": 2.8043e-05, + "loss": 0.0133, + "step": 15876 + }, + { + "epoch": 13.409628378378379, + "grad_norm": 0.2056773155927658, + "learning_rate": 2.804266666666667e-05, + "loss": 0.0143, + "step": 15877 + }, + { + "epoch": 13.410472972972974, + "grad_norm": 0.42717134952545166, + "learning_rate": 2.8042333333333334e-05, + "loss": 0.0204, + "step": 15878 + }, + { + "epoch": 13.411317567567568, + "grad_norm": 0.32787176966667175, + "learning_rate": 2.8042e-05, + "loss": 0.0333, + "step": 15879 + }, + { + "epoch": 13.412162162162161, + "grad_norm": 0.31616440415382385, + "learning_rate": 2.8041666666666666e-05, + "loss": 0.0172, + "step": 15880 + }, + { + "epoch": 13.413006756756756, + "grad_norm": 0.41253119707107544, + "learning_rate": 2.8041333333333335e-05, + "loss": 0.0213, + "step": 15881 + }, + { + "epoch": 13.41385135135135, + "grad_norm": 0.33500054478645325, + "learning_rate": 2.8041e-05, + "loss": 0.0123, + "step": 15882 + }, + { + "epoch": 13.414695945945946, + "grad_norm": 0.37680956721305847, + "learning_rate": 2.8040666666666667e-05, + "loss": 0.0273, + "step": 15883 + }, + { + "epoch": 13.41554054054054, + "grad_norm": 0.30040085315704346, + "learning_rate": 2.8040333333333337e-05, + "loss": 0.0125, + "step": 15884 + }, + { + "epoch": 13.416385135135135, + "grad_norm": 0.46416419744491577, + "learning_rate": 2.804e-05, + "loss": 0.0223, + "step": 15885 + }, + { + "epoch": 13.41722972972973, + "grad_norm": 0.29360079765319824, + "learning_rate": 2.803966666666667e-05, + "loss": 0.0194, + "step": 15886 + }, + { + "epoch": 13.418074324324325, + "grad_norm": 0.7771422266960144, + "learning_rate": 2.8039333333333334e-05, + "loss": 0.0235, + "step": 15887 + }, + { + "epoch": 13.41891891891892, + "grad_norm": 0.3236735165119171, + "learning_rate": 2.8039e-05, + "loss": 0.0152, + "step": 15888 + }, + { + "epoch": 13.419763513513514, + "grad_norm": 0.33648020029067993, + "learning_rate": 2.8038666666666666e-05, + "loss": 0.0152, + "step": 15889 + }, + { + "epoch": 13.420608108108109, + "grad_norm": 0.3577655255794525, + "learning_rate": 2.8038333333333332e-05, + "loss": 0.022, + "step": 15890 + }, + { + "epoch": 13.421452702702704, + "grad_norm": 0.26166272163391113, + "learning_rate": 2.8038e-05, + "loss": 0.0265, + "step": 15891 + }, + { + "epoch": 13.422297297297296, + "grad_norm": 0.5046490430831909, + "learning_rate": 2.8037666666666668e-05, + "loss": 0.023, + "step": 15892 + }, + { + "epoch": 13.423141891891891, + "grad_norm": 1.2953095436096191, + "learning_rate": 2.8037333333333333e-05, + "loss": 0.2636, + "step": 15893 + }, + { + "epoch": 13.423986486486486, + "grad_norm": 0.513836681842804, + "learning_rate": 2.8037e-05, + "loss": 0.1439, + "step": 15894 + }, + { + "epoch": 13.42483108108108, + "grad_norm": 0.5229511260986328, + "learning_rate": 2.803666666666667e-05, + "loss": 0.0814, + "step": 15895 + }, + { + "epoch": 13.425675675675675, + "grad_norm": 0.47215336561203003, + "learning_rate": 2.803633333333333e-05, + "loss": 0.0481, + "step": 15896 + }, + { + "epoch": 13.42652027027027, + "grad_norm": 0.444011390209198, + "learning_rate": 2.8036e-05, + "loss": 0.0239, + "step": 15897 + }, + { + "epoch": 13.427364864864865, + "grad_norm": 0.3579530417919159, + "learning_rate": 2.803566666666667e-05, + "loss": 0.025, + "step": 15898 + }, + { + "epoch": 13.42820945945946, + "grad_norm": 0.29309722781181335, + "learning_rate": 2.8035333333333332e-05, + "loss": 0.0235, + "step": 15899 + }, + { + "epoch": 13.429054054054054, + "grad_norm": 0.28301024436950684, + "learning_rate": 2.8035000000000002e-05, + "loss": 0.0219, + "step": 15900 + }, + { + "epoch": 13.42989864864865, + "grad_norm": 0.2777114510536194, + "learning_rate": 2.8034666666666668e-05, + "loss": 0.0297, + "step": 15901 + }, + { + "epoch": 13.430743243243244, + "grad_norm": 0.24403829872608185, + "learning_rate": 2.8034333333333334e-05, + "loss": 0.0188, + "step": 15902 + }, + { + "epoch": 13.431587837837839, + "grad_norm": 0.22522829473018646, + "learning_rate": 2.8034e-05, + "loss": 0.0128, + "step": 15903 + }, + { + "epoch": 13.432432432432432, + "grad_norm": 1.0977365970611572, + "learning_rate": 2.803366666666667e-05, + "loss": 0.0259, + "step": 15904 + }, + { + "epoch": 13.433277027027026, + "grad_norm": 0.3862834870815277, + "learning_rate": 2.803333333333333e-05, + "loss": 0.032, + "step": 15905 + }, + { + "epoch": 13.434121621621621, + "grad_norm": 0.30463576316833496, + "learning_rate": 2.8033e-05, + "loss": 0.0242, + "step": 15906 + }, + { + "epoch": 13.434966216216216, + "grad_norm": 0.23443979024887085, + "learning_rate": 2.803266666666667e-05, + "loss": 0.0175, + "step": 15907 + }, + { + "epoch": 13.43581081081081, + "grad_norm": 0.520993709564209, + "learning_rate": 2.8032333333333333e-05, + "loss": 0.0315, + "step": 15908 + }, + { + "epoch": 13.436655405405405, + "grad_norm": 0.2935575246810913, + "learning_rate": 2.8032000000000002e-05, + "loss": 0.0124, + "step": 15909 + }, + { + "epoch": 13.4375, + "grad_norm": 0.16902193427085876, + "learning_rate": 2.8031666666666668e-05, + "loss": 0.0133, + "step": 15910 + }, + { + "epoch": 13.438344594594595, + "grad_norm": 0.2577638030052185, + "learning_rate": 2.8031333333333334e-05, + "loss": 0.0205, + "step": 15911 + }, + { + "epoch": 13.43918918918919, + "grad_norm": 0.1557370275259018, + "learning_rate": 2.8031e-05, + "loss": 0.0103, + "step": 15912 + }, + { + "epoch": 13.440033783783784, + "grad_norm": 0.574243426322937, + "learning_rate": 2.803066666666667e-05, + "loss": 0.0261, + "step": 15913 + }, + { + "epoch": 13.440878378378379, + "grad_norm": 0.2773253917694092, + "learning_rate": 2.8030333333333335e-05, + "loss": 0.0231, + "step": 15914 + }, + { + "epoch": 13.441722972972974, + "grad_norm": 0.21950122714042664, + "learning_rate": 2.803e-05, + "loss": 0.0151, + "step": 15915 + }, + { + "epoch": 13.442567567567568, + "grad_norm": 0.5102481245994568, + "learning_rate": 2.802966666666667e-05, + "loss": 0.0175, + "step": 15916 + }, + { + "epoch": 13.443412162162161, + "grad_norm": 0.33044102787971497, + "learning_rate": 2.8029333333333333e-05, + "loss": 0.0148, + "step": 15917 + }, + { + "epoch": 13.444256756756756, + "grad_norm": 0.2075614184141159, + "learning_rate": 2.8029000000000002e-05, + "loss": 0.0154, + "step": 15918 + }, + { + "epoch": 13.44510135135135, + "grad_norm": 0.3652046024799347, + "learning_rate": 2.8028666666666665e-05, + "loss": 0.0265, + "step": 15919 + }, + { + "epoch": 13.445945945945946, + "grad_norm": 0.12148845940828323, + "learning_rate": 2.8028333333333334e-05, + "loss": 0.0104, + "step": 15920 + }, + { + "epoch": 13.44679054054054, + "grad_norm": 0.37719815969467163, + "learning_rate": 2.8028e-05, + "loss": 0.0157, + "step": 15921 + }, + { + "epoch": 13.447635135135135, + "grad_norm": 0.30667176842689514, + "learning_rate": 2.8027666666666666e-05, + "loss": 0.0192, + "step": 15922 + }, + { + "epoch": 13.44847972972973, + "grad_norm": 0.26899394392967224, + "learning_rate": 2.8027333333333335e-05, + "loss": 0.0171, + "step": 15923 + }, + { + "epoch": 13.449324324324325, + "grad_norm": 0.304757684469223, + "learning_rate": 2.8027e-05, + "loss": 0.0315, + "step": 15924 + }, + { + "epoch": 13.45016891891892, + "grad_norm": 0.2763625383377075, + "learning_rate": 2.8026666666666667e-05, + "loss": 0.0147, + "step": 15925 + }, + { + "epoch": 13.451013513513514, + "grad_norm": 0.20623615384101868, + "learning_rate": 2.8026333333333333e-05, + "loss": 0.0168, + "step": 15926 + }, + { + "epoch": 13.451858108108109, + "grad_norm": 0.24682652950286865, + "learning_rate": 2.8026000000000002e-05, + "loss": 0.0206, + "step": 15927 + }, + { + "epoch": 13.452702702702704, + "grad_norm": 0.30096760392189026, + "learning_rate": 2.8025666666666665e-05, + "loss": 0.0117, + "step": 15928 + }, + { + "epoch": 13.453547297297296, + "grad_norm": 0.19246794283390045, + "learning_rate": 2.8025333333333334e-05, + "loss": 0.0249, + "step": 15929 + }, + { + "epoch": 13.454391891891891, + "grad_norm": 0.175059512257576, + "learning_rate": 2.8025e-05, + "loss": 0.0118, + "step": 15930 + }, + { + "epoch": 13.455236486486486, + "grad_norm": 0.5044370889663696, + "learning_rate": 2.8024666666666666e-05, + "loss": 0.0186, + "step": 15931 + }, + { + "epoch": 13.45608108108108, + "grad_norm": 0.3186465799808502, + "learning_rate": 2.8024333333333335e-05, + "loss": 0.019, + "step": 15932 + }, + { + "epoch": 13.456925675675675, + "grad_norm": 0.23437438905239105, + "learning_rate": 2.8024e-05, + "loss": 0.0158, + "step": 15933 + }, + { + "epoch": 13.45777027027027, + "grad_norm": 1.0068541765213013, + "learning_rate": 2.8023666666666667e-05, + "loss": 0.017, + "step": 15934 + }, + { + "epoch": 13.458614864864865, + "grad_norm": 0.4983313977718353, + "learning_rate": 2.8023333333333333e-05, + "loss": 0.0206, + "step": 15935 + }, + { + "epoch": 13.45945945945946, + "grad_norm": 0.2374364286661148, + "learning_rate": 2.8023000000000003e-05, + "loss": 0.0163, + "step": 15936 + }, + { + "epoch": 13.460304054054054, + "grad_norm": 1.5763933658599854, + "learning_rate": 2.8022666666666665e-05, + "loss": 0.0312, + "step": 15937 + }, + { + "epoch": 13.46114864864865, + "grad_norm": 0.2614799439907074, + "learning_rate": 2.8022333333333334e-05, + "loss": 0.0151, + "step": 15938 + }, + { + "epoch": 13.461993243243244, + "grad_norm": 0.21156303584575653, + "learning_rate": 2.8022e-05, + "loss": 0.0147, + "step": 15939 + }, + { + "epoch": 13.462837837837839, + "grad_norm": 0.31666529178619385, + "learning_rate": 2.8021666666666666e-05, + "loss": 0.0092, + "step": 15940 + }, + { + "epoch": 13.463682432432432, + "grad_norm": 0.3699331283569336, + "learning_rate": 2.8021333333333336e-05, + "loss": 0.0321, + "step": 15941 + }, + { + "epoch": 13.464527027027026, + "grad_norm": 1.387213945388794, + "learning_rate": 2.8021e-05, + "loss": 0.0381, + "step": 15942 + }, + { + "epoch": 13.465371621621621, + "grad_norm": 0.6554827690124512, + "learning_rate": 2.8020666666666668e-05, + "loss": 0.1753, + "step": 15943 + }, + { + "epoch": 13.466216216216216, + "grad_norm": 0.5293251872062683, + "learning_rate": 2.8020333333333333e-05, + "loss": 0.1164, + "step": 15944 + }, + { + "epoch": 13.46706081081081, + "grad_norm": 0.5293063521385193, + "learning_rate": 2.8020000000000003e-05, + "loss": 0.0595, + "step": 15945 + }, + { + "epoch": 13.467905405405405, + "grad_norm": 0.38901880383491516, + "learning_rate": 2.8019666666666665e-05, + "loss": 0.0431, + "step": 15946 + }, + { + "epoch": 13.46875, + "grad_norm": 0.3262193500995636, + "learning_rate": 2.8019333333333335e-05, + "loss": 0.0319, + "step": 15947 + }, + { + "epoch": 13.469594594594595, + "grad_norm": 0.23774799704551697, + "learning_rate": 2.8019e-05, + "loss": 0.0202, + "step": 15948 + }, + { + "epoch": 13.47043918918919, + "grad_norm": 1.247719168663025, + "learning_rate": 2.8018666666666667e-05, + "loss": 0.0208, + "step": 15949 + }, + { + "epoch": 13.471283783783784, + "grad_norm": 0.3388686180114746, + "learning_rate": 2.8018333333333336e-05, + "loss": 0.0286, + "step": 15950 + }, + { + "epoch": 13.472128378378379, + "grad_norm": 0.36610859632492065, + "learning_rate": 2.8018e-05, + "loss": 0.0234, + "step": 15951 + }, + { + "epoch": 13.472972972972974, + "grad_norm": 0.34056541323661804, + "learning_rate": 2.8017666666666668e-05, + "loss": 0.0249, + "step": 15952 + }, + { + "epoch": 13.473817567567568, + "grad_norm": 0.19162370264530182, + "learning_rate": 2.8017333333333334e-05, + "loss": 0.0098, + "step": 15953 + }, + { + "epoch": 13.474662162162161, + "grad_norm": 0.32372573018074036, + "learning_rate": 2.8017e-05, + "loss": 0.0244, + "step": 15954 + }, + { + "epoch": 13.475506756756756, + "grad_norm": 0.21705840528011322, + "learning_rate": 2.8016666666666666e-05, + "loss": 0.0162, + "step": 15955 + }, + { + "epoch": 13.47635135135135, + "grad_norm": 0.23259836435317993, + "learning_rate": 2.8016333333333335e-05, + "loss": 0.0216, + "step": 15956 + }, + { + "epoch": 13.477195945945946, + "grad_norm": 0.7517135143280029, + "learning_rate": 2.8016e-05, + "loss": 0.0328, + "step": 15957 + }, + { + "epoch": 13.47804054054054, + "grad_norm": 0.2384955734014511, + "learning_rate": 2.8015666666666667e-05, + "loss": 0.0223, + "step": 15958 + }, + { + "epoch": 13.478885135135135, + "grad_norm": 0.2222321778535843, + "learning_rate": 2.8015333333333336e-05, + "loss": 0.0166, + "step": 15959 + }, + { + "epoch": 13.47972972972973, + "grad_norm": 0.21458636224269867, + "learning_rate": 2.8015e-05, + "loss": 0.0218, + "step": 15960 + }, + { + "epoch": 13.480574324324325, + "grad_norm": 0.2540131211280823, + "learning_rate": 2.8014666666666668e-05, + "loss": 0.0152, + "step": 15961 + }, + { + "epoch": 13.48141891891892, + "grad_norm": 0.25307008624076843, + "learning_rate": 2.8014333333333334e-05, + "loss": 0.0187, + "step": 15962 + }, + { + "epoch": 13.482263513513514, + "grad_norm": 0.2831110656261444, + "learning_rate": 2.8014e-05, + "loss": 0.0252, + "step": 15963 + }, + { + "epoch": 13.483108108108109, + "grad_norm": 0.3030589818954468, + "learning_rate": 2.801366666666667e-05, + "loss": 0.0234, + "step": 15964 + }, + { + "epoch": 13.483952702702704, + "grad_norm": 0.20983311533927917, + "learning_rate": 2.8013333333333335e-05, + "loss": 0.0151, + "step": 15965 + }, + { + "epoch": 13.484797297297296, + "grad_norm": 0.35577473044395447, + "learning_rate": 2.8013e-05, + "loss": 0.0196, + "step": 15966 + }, + { + "epoch": 13.485641891891891, + "grad_norm": 0.3784704804420471, + "learning_rate": 2.8012666666666667e-05, + "loss": 0.0166, + "step": 15967 + }, + { + "epoch": 13.486486486486486, + "grad_norm": 0.16807548701763153, + "learning_rate": 2.8012333333333336e-05, + "loss": 0.0146, + "step": 15968 + }, + { + "epoch": 13.48733108108108, + "grad_norm": 0.3482373356819153, + "learning_rate": 2.8012e-05, + "loss": 0.0182, + "step": 15969 + }, + { + "epoch": 13.488175675675675, + "grad_norm": 0.2799760103225708, + "learning_rate": 2.8011666666666668e-05, + "loss": 0.0209, + "step": 15970 + }, + { + "epoch": 13.48902027027027, + "grad_norm": 0.29469600319862366, + "learning_rate": 2.8011333333333334e-05, + "loss": 0.0253, + "step": 15971 + }, + { + "epoch": 13.489864864864865, + "grad_norm": 0.25128912925720215, + "learning_rate": 2.8011e-05, + "loss": 0.0173, + "step": 15972 + }, + { + "epoch": 13.49070945945946, + "grad_norm": 0.30592888593673706, + "learning_rate": 2.801066666666667e-05, + "loss": 0.0189, + "step": 15973 + }, + { + "epoch": 13.491554054054054, + "grad_norm": 0.3856469392776489, + "learning_rate": 2.8010333333333335e-05, + "loss": 0.0219, + "step": 15974 + }, + { + "epoch": 13.49239864864865, + "grad_norm": 0.6156687140464783, + "learning_rate": 2.801e-05, + "loss": 0.0145, + "step": 15975 + }, + { + "epoch": 13.493243243243244, + "grad_norm": 0.3346096873283386, + "learning_rate": 2.8009666666666667e-05, + "loss": 0.0162, + "step": 15976 + }, + { + "epoch": 13.494087837837839, + "grad_norm": 0.2713703215122223, + "learning_rate": 2.8009333333333336e-05, + "loss": 0.0163, + "step": 15977 + }, + { + "epoch": 13.494932432432432, + "grad_norm": 0.5152941346168518, + "learning_rate": 2.8009e-05, + "loss": 0.016, + "step": 15978 + }, + { + "epoch": 13.495777027027026, + "grad_norm": 0.48143377900123596, + "learning_rate": 2.800866666666667e-05, + "loss": 0.0351, + "step": 15979 + }, + { + "epoch": 13.496621621621621, + "grad_norm": 0.2966633141040802, + "learning_rate": 2.800833333333333e-05, + "loss": 0.0168, + "step": 15980 + }, + { + "epoch": 13.497466216216216, + "grad_norm": 0.19878457486629486, + "learning_rate": 2.8008e-05, + "loss": 0.0122, + "step": 15981 + }, + { + "epoch": 13.49831081081081, + "grad_norm": 0.25037094950675964, + "learning_rate": 2.800766666666667e-05, + "loss": 0.0211, + "step": 15982 + }, + { + "epoch": 13.499155405405405, + "grad_norm": 0.6049090623855591, + "learning_rate": 2.8007333333333332e-05, + "loss": 0.0227, + "step": 15983 + }, + { + "epoch": 13.5, + "grad_norm": 0.2057817131280899, + "learning_rate": 2.8007e-05, + "loss": 0.0106, + "step": 15984 + }, + { + "epoch": 13.500844594594595, + "grad_norm": 0.26811808347702026, + "learning_rate": 2.8006666666666667e-05, + "loss": 0.0209, + "step": 15985 + }, + { + "epoch": 13.50168918918919, + "grad_norm": 0.5422404408454895, + "learning_rate": 2.8006333333333333e-05, + "loss": 0.0173, + "step": 15986 + }, + { + "epoch": 13.502533783783784, + "grad_norm": 0.45195272564888, + "learning_rate": 2.8006e-05, + "loss": 0.0236, + "step": 15987 + }, + { + "epoch": 13.503378378378379, + "grad_norm": 0.27748748660087585, + "learning_rate": 2.800566666666667e-05, + "loss": 0.0233, + "step": 15988 + }, + { + "epoch": 13.504222972972974, + "grad_norm": 0.25121620297431946, + "learning_rate": 2.8005333333333334e-05, + "loss": 0.0148, + "step": 15989 + }, + { + "epoch": 13.505067567567568, + "grad_norm": 0.5760290026664734, + "learning_rate": 2.8005e-05, + "loss": 0.0246, + "step": 15990 + }, + { + "epoch": 13.505912162162161, + "grad_norm": 0.6721733212471008, + "learning_rate": 2.800466666666667e-05, + "loss": 0.074, + "step": 15991 + }, + { + "epoch": 13.506756756756756, + "grad_norm": 0.638089120388031, + "learning_rate": 2.8004333333333332e-05, + "loss": 0.0331, + "step": 15992 + }, + { + "epoch": 13.50760135135135, + "grad_norm": 0.5420500040054321, + "learning_rate": 2.8004e-05, + "loss": 0.1668, + "step": 15993 + }, + { + "epoch": 13.508445945945946, + "grad_norm": 0.607776403427124, + "learning_rate": 2.8003666666666668e-05, + "loss": 0.0931, + "step": 15994 + }, + { + "epoch": 13.50929054054054, + "grad_norm": 0.4453822076320648, + "learning_rate": 2.8003333333333333e-05, + "loss": 0.091, + "step": 15995 + }, + { + "epoch": 13.510135135135135, + "grad_norm": 0.25887250900268555, + "learning_rate": 2.8003e-05, + "loss": 0.0459, + "step": 15996 + }, + { + "epoch": 13.51097972972973, + "grad_norm": 0.273031085729599, + "learning_rate": 2.800266666666667e-05, + "loss": 0.0347, + "step": 15997 + }, + { + "epoch": 13.511824324324325, + "grad_norm": 0.2720393240451813, + "learning_rate": 2.8002333333333335e-05, + "loss": 0.0276, + "step": 15998 + }, + { + "epoch": 13.51266891891892, + "grad_norm": 0.2003958374261856, + "learning_rate": 2.8002e-05, + "loss": 0.0194, + "step": 15999 + }, + { + "epoch": 13.513513513513514, + "grad_norm": 0.24102897942066193, + "learning_rate": 2.800166666666667e-05, + "loss": 0.0201, + "step": 16000 + }, + { + "epoch": 13.513513513513514, + "eval_cer": 0.08963135491125211, + "eval_loss": 0.24186615645885468, + "eval_runtime": 8.7654, + "eval_samples_per_second": 11.523, + "eval_steps_per_second": 0.456, + "eval_wer": 0.29012345679012347, + "step": 16000 + }, + { + "epoch": 13.514358108108109, + "grad_norm": 0.28118976950645447, + "learning_rate": 2.8001333333333332e-05, + "loss": 0.0236, + "step": 16001 + }, + { + "epoch": 13.515202702702704, + "grad_norm": 0.24090246856212616, + "learning_rate": 2.8001000000000002e-05, + "loss": 0.0193, + "step": 16002 + }, + { + "epoch": 13.516047297297296, + "grad_norm": 0.35285425186157227, + "learning_rate": 2.8000666666666668e-05, + "loss": 0.026, + "step": 16003 + }, + { + "epoch": 13.516891891891891, + "grad_norm": 0.1905587762594223, + "learning_rate": 2.8000333333333334e-05, + "loss": 0.0286, + "step": 16004 + }, + { + "epoch": 13.517736486486486, + "grad_norm": 0.17955787479877472, + "learning_rate": 2.8e-05, + "loss": 0.0149, + "step": 16005 + }, + { + "epoch": 13.51858108108108, + "grad_norm": 0.3948311507701874, + "learning_rate": 2.799966666666667e-05, + "loss": 0.0214, + "step": 16006 + }, + { + "epoch": 13.519425675675675, + "grad_norm": 0.1958095133304596, + "learning_rate": 2.7999333333333335e-05, + "loss": 0.0081, + "step": 16007 + }, + { + "epoch": 13.52027027027027, + "grad_norm": 0.2375270277261734, + "learning_rate": 2.7999e-05, + "loss": 0.0131, + "step": 16008 + }, + { + "epoch": 13.521114864864865, + "grad_norm": 0.1871577501296997, + "learning_rate": 2.7998666666666667e-05, + "loss": 0.014, + "step": 16009 + }, + { + "epoch": 13.52195945945946, + "grad_norm": 0.1917562186717987, + "learning_rate": 2.7998333333333333e-05, + "loss": 0.0164, + "step": 16010 + }, + { + "epoch": 13.522804054054054, + "grad_norm": 0.3881136178970337, + "learning_rate": 2.7998000000000002e-05, + "loss": 0.0198, + "step": 16011 + }, + { + "epoch": 13.52364864864865, + "grad_norm": 0.24354225397109985, + "learning_rate": 2.7997666666666665e-05, + "loss": 0.017, + "step": 16012 + }, + { + "epoch": 13.524493243243244, + "grad_norm": 0.23031441867351532, + "learning_rate": 2.7997333333333334e-05, + "loss": 0.0189, + "step": 16013 + }, + { + "epoch": 13.525337837837839, + "grad_norm": 0.17870500683784485, + "learning_rate": 2.7997000000000003e-05, + "loss": 0.0212, + "step": 16014 + }, + { + "epoch": 13.526182432432432, + "grad_norm": 0.3112698495388031, + "learning_rate": 2.7996666666666666e-05, + "loss": 0.0148, + "step": 16015 + }, + { + "epoch": 13.527027027027026, + "grad_norm": 0.32912471890449524, + "learning_rate": 2.7996333333333335e-05, + "loss": 0.0228, + "step": 16016 + }, + { + "epoch": 13.527871621621621, + "grad_norm": 0.2432004064321518, + "learning_rate": 2.7996e-05, + "loss": 0.0183, + "step": 16017 + }, + { + "epoch": 13.528716216216216, + "grad_norm": 0.3988374173641205, + "learning_rate": 2.7995666666666667e-05, + "loss": 0.0228, + "step": 16018 + }, + { + "epoch": 13.52956081081081, + "grad_norm": 0.35744601488113403, + "learning_rate": 2.7995333333333333e-05, + "loss": 0.0183, + "step": 16019 + }, + { + "epoch": 13.530405405405405, + "grad_norm": 0.36073461174964905, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0116, + "step": 16020 + }, + { + "epoch": 13.53125, + "grad_norm": 0.2210366278886795, + "learning_rate": 2.7994666666666665e-05, + "loss": 0.0163, + "step": 16021 + }, + { + "epoch": 13.532094594594595, + "grad_norm": 0.2605687975883484, + "learning_rate": 2.7994333333333334e-05, + "loss": 0.0125, + "step": 16022 + }, + { + "epoch": 13.53293918918919, + "grad_norm": 0.2411152869462967, + "learning_rate": 2.7994000000000003e-05, + "loss": 0.0169, + "step": 16023 + }, + { + "epoch": 13.533783783783784, + "grad_norm": 0.1308658868074417, + "learning_rate": 2.7993666666666666e-05, + "loss": 0.0122, + "step": 16024 + }, + { + "epoch": 13.534628378378379, + "grad_norm": 0.3256608247756958, + "learning_rate": 2.7993333333333335e-05, + "loss": 0.0202, + "step": 16025 + }, + { + "epoch": 13.535472972972974, + "grad_norm": 0.23726503551006317, + "learning_rate": 2.7993e-05, + "loss": 0.0151, + "step": 16026 + }, + { + "epoch": 13.536317567567568, + "grad_norm": 0.48837989568710327, + "learning_rate": 2.7992666666666667e-05, + "loss": 0.0184, + "step": 16027 + }, + { + "epoch": 13.537162162162161, + "grad_norm": 0.5965899229049683, + "learning_rate": 2.7992333333333333e-05, + "loss": 0.0138, + "step": 16028 + }, + { + "epoch": 13.538006756756756, + "grad_norm": 0.46017754077911377, + "learning_rate": 2.7992000000000002e-05, + "loss": 0.0278, + "step": 16029 + }, + { + "epoch": 13.53885135135135, + "grad_norm": 0.3340423107147217, + "learning_rate": 2.7991666666666665e-05, + "loss": 0.0222, + "step": 16030 + }, + { + "epoch": 13.539695945945946, + "grad_norm": 0.28969240188598633, + "learning_rate": 2.7991333333333334e-05, + "loss": 0.0107, + "step": 16031 + }, + { + "epoch": 13.54054054054054, + "grad_norm": 0.32978400588035583, + "learning_rate": 2.7991000000000004e-05, + "loss": 0.014, + "step": 16032 + }, + { + "epoch": 13.541385135135135, + "grad_norm": 0.20642498135566711, + "learning_rate": 2.7990666666666666e-05, + "loss": 0.0129, + "step": 16033 + }, + { + "epoch": 13.54222972972973, + "grad_norm": 0.39687538146972656, + "learning_rate": 2.7990333333333335e-05, + "loss": 0.0222, + "step": 16034 + }, + { + "epoch": 13.543074324324325, + "grad_norm": 0.2541005611419678, + "learning_rate": 2.799e-05, + "loss": 0.0128, + "step": 16035 + }, + { + "epoch": 13.54391891891892, + "grad_norm": 0.4156322479248047, + "learning_rate": 2.7989666666666667e-05, + "loss": 0.0285, + "step": 16036 + }, + { + "epoch": 13.544763513513514, + "grad_norm": 0.19782698154449463, + "learning_rate": 2.7989333333333333e-05, + "loss": 0.0089, + "step": 16037 + }, + { + "epoch": 13.545608108108109, + "grad_norm": 0.21181634068489075, + "learning_rate": 2.7989000000000003e-05, + "loss": 0.0138, + "step": 16038 + }, + { + "epoch": 13.546452702702704, + "grad_norm": 0.2956286370754242, + "learning_rate": 2.798866666666667e-05, + "loss": 0.0102, + "step": 16039 + }, + { + "epoch": 13.547297297297296, + "grad_norm": 0.32102540135383606, + "learning_rate": 2.7988333333333334e-05, + "loss": 0.0231, + "step": 16040 + }, + { + "epoch": 13.548141891891891, + "grad_norm": 1.1356409788131714, + "learning_rate": 2.7988e-05, + "loss": 0.0456, + "step": 16041 + }, + { + "epoch": 13.548986486486486, + "grad_norm": 0.31069764494895935, + "learning_rate": 2.7987666666666666e-05, + "loss": 0.0156, + "step": 16042 + }, + { + "epoch": 13.54983108108108, + "grad_norm": 0.6338212490081787, + "learning_rate": 2.7987333333333336e-05, + "loss": 0.1698, + "step": 16043 + }, + { + "epoch": 13.550675675675675, + "grad_norm": 0.5322991609573364, + "learning_rate": 2.7986999999999998e-05, + "loss": 0.1114, + "step": 16044 + }, + { + "epoch": 13.55152027027027, + "grad_norm": 0.43210068345069885, + "learning_rate": 2.7986666666666668e-05, + "loss": 0.066, + "step": 16045 + }, + { + "epoch": 13.552364864864865, + "grad_norm": 0.2535063624382019, + "learning_rate": 2.7986333333333333e-05, + "loss": 0.0325, + "step": 16046 + }, + { + "epoch": 13.55320945945946, + "grad_norm": 0.5330315232276917, + "learning_rate": 2.7986e-05, + "loss": 0.0264, + "step": 16047 + }, + { + "epoch": 13.554054054054054, + "grad_norm": 0.24987584352493286, + "learning_rate": 2.798566666666667e-05, + "loss": 0.0322, + "step": 16048 + }, + { + "epoch": 13.55489864864865, + "grad_norm": 0.28933969140052795, + "learning_rate": 2.7985333333333335e-05, + "loss": 0.0158, + "step": 16049 + }, + { + "epoch": 13.555743243243244, + "grad_norm": 0.29727086424827576, + "learning_rate": 2.7985e-05, + "loss": 0.0132, + "step": 16050 + }, + { + "epoch": 13.556587837837839, + "grad_norm": 0.4633854925632477, + "learning_rate": 2.7984666666666667e-05, + "loss": 0.0325, + "step": 16051 + }, + { + "epoch": 13.557432432432432, + "grad_norm": 0.2891639471054077, + "learning_rate": 2.7984333333333336e-05, + "loss": 0.0211, + "step": 16052 + }, + { + "epoch": 13.558277027027026, + "grad_norm": 0.36358198523521423, + "learning_rate": 2.7984e-05, + "loss": 0.0185, + "step": 16053 + }, + { + "epoch": 13.559121621621621, + "grad_norm": 0.3763450086116791, + "learning_rate": 2.7983666666666668e-05, + "loss": 0.0295, + "step": 16054 + }, + { + "epoch": 13.559966216216216, + "grad_norm": 0.2891330420970917, + "learning_rate": 2.7983333333333334e-05, + "loss": 0.0196, + "step": 16055 + }, + { + "epoch": 13.56081081081081, + "grad_norm": 0.508021354675293, + "learning_rate": 2.7983e-05, + "loss": 0.0241, + "step": 16056 + }, + { + "epoch": 13.561655405405405, + "grad_norm": 0.26123955845832825, + "learning_rate": 2.798266666666667e-05, + "loss": 0.0237, + "step": 16057 + }, + { + "epoch": 13.5625, + "grad_norm": 0.390531063079834, + "learning_rate": 2.7982333333333335e-05, + "loss": 0.0149, + "step": 16058 + }, + { + "epoch": 13.563344594594595, + "grad_norm": 0.3059220016002655, + "learning_rate": 2.7982e-05, + "loss": 0.039, + "step": 16059 + }, + { + "epoch": 13.56418918918919, + "grad_norm": 0.2032761424779892, + "learning_rate": 2.7981666666666667e-05, + "loss": 0.0138, + "step": 16060 + }, + { + "epoch": 13.565033783783784, + "grad_norm": 0.263435959815979, + "learning_rate": 2.7981333333333336e-05, + "loss": 0.016, + "step": 16061 + }, + { + "epoch": 13.565878378378379, + "grad_norm": 0.24344953894615173, + "learning_rate": 2.7981e-05, + "loss": 0.0151, + "step": 16062 + }, + { + "epoch": 13.566722972972974, + "grad_norm": 0.3280199468135834, + "learning_rate": 2.7980666666666668e-05, + "loss": 0.0126, + "step": 16063 + }, + { + "epoch": 13.567567567567568, + "grad_norm": 0.4216371178627014, + "learning_rate": 2.7980333333333337e-05, + "loss": 0.01, + "step": 16064 + }, + { + "epoch": 13.568412162162161, + "grad_norm": 0.3306649625301361, + "learning_rate": 2.798e-05, + "loss": 0.0154, + "step": 16065 + }, + { + "epoch": 13.569256756756756, + "grad_norm": 0.45393672585487366, + "learning_rate": 2.797966666666667e-05, + "loss": 0.0245, + "step": 16066 + }, + { + "epoch": 13.57010135135135, + "grad_norm": 0.26452502608299255, + "learning_rate": 2.7979333333333335e-05, + "loss": 0.013, + "step": 16067 + }, + { + "epoch": 13.570945945945946, + "grad_norm": 0.2455521523952484, + "learning_rate": 2.7979e-05, + "loss": 0.014, + "step": 16068 + }, + { + "epoch": 13.57179054054054, + "grad_norm": 0.28875502943992615, + "learning_rate": 2.7978666666666667e-05, + "loss": 0.0229, + "step": 16069 + }, + { + "epoch": 13.572635135135135, + "grad_norm": 0.23758448660373688, + "learning_rate": 2.7978333333333333e-05, + "loss": 0.0124, + "step": 16070 + }, + { + "epoch": 13.57347972972973, + "grad_norm": 0.1980307400226593, + "learning_rate": 2.7978e-05, + "loss": 0.0092, + "step": 16071 + }, + { + "epoch": 13.574324324324325, + "grad_norm": 0.48063474893569946, + "learning_rate": 2.7977666666666668e-05, + "loss": 0.0253, + "step": 16072 + }, + { + "epoch": 13.57516891891892, + "grad_norm": 0.11407266557216644, + "learning_rate": 2.7977333333333334e-05, + "loss": 0.0088, + "step": 16073 + }, + { + "epoch": 13.576013513513514, + "grad_norm": 0.2801843583583832, + "learning_rate": 2.7977e-05, + "loss": 0.0132, + "step": 16074 + }, + { + "epoch": 13.576858108108109, + "grad_norm": 0.16500519216060638, + "learning_rate": 2.797666666666667e-05, + "loss": 0.0098, + "step": 16075 + }, + { + "epoch": 13.577702702702704, + "grad_norm": 0.18287807703018188, + "learning_rate": 2.7976333333333332e-05, + "loss": 0.0124, + "step": 16076 + }, + { + "epoch": 13.578547297297296, + "grad_norm": 0.43711674213409424, + "learning_rate": 2.7976e-05, + "loss": 0.0568, + "step": 16077 + }, + { + "epoch": 13.579391891891891, + "grad_norm": 0.5011450052261353, + "learning_rate": 2.7975666666666667e-05, + "loss": 0.0294, + "step": 16078 + }, + { + "epoch": 13.580236486486486, + "grad_norm": 0.5462875366210938, + "learning_rate": 2.7975333333333333e-05, + "loss": 0.0199, + "step": 16079 + }, + { + "epoch": 13.58108108108108, + "grad_norm": 0.6077708601951599, + "learning_rate": 2.7975e-05, + "loss": 0.0247, + "step": 16080 + }, + { + "epoch": 13.581925675675675, + "grad_norm": 0.3436647951602936, + "learning_rate": 2.7974666666666668e-05, + "loss": 0.0153, + "step": 16081 + }, + { + "epoch": 13.58277027027027, + "grad_norm": 0.5164973735809326, + "learning_rate": 2.7974333333333334e-05, + "loss": 0.0479, + "step": 16082 + }, + { + "epoch": 13.583614864864865, + "grad_norm": 0.25710049271583557, + "learning_rate": 2.7974e-05, + "loss": 0.02, + "step": 16083 + }, + { + "epoch": 13.58445945945946, + "grad_norm": 0.3438482880592346, + "learning_rate": 2.797366666666667e-05, + "loss": 0.0119, + "step": 16084 + }, + { + "epoch": 13.585304054054054, + "grad_norm": 0.25579866766929626, + "learning_rate": 2.7973333333333332e-05, + "loss": 0.0132, + "step": 16085 + }, + { + "epoch": 13.58614864864865, + "grad_norm": 0.34683898091316223, + "learning_rate": 2.7973e-05, + "loss": 0.0173, + "step": 16086 + }, + { + "epoch": 13.586993243243244, + "grad_norm": 1.504738211631775, + "learning_rate": 2.7972666666666667e-05, + "loss": 0.0237, + "step": 16087 + }, + { + "epoch": 13.587837837837839, + "grad_norm": 0.37335285544395447, + "learning_rate": 2.7972333333333333e-05, + "loss": 0.0289, + "step": 16088 + }, + { + "epoch": 13.588682432432432, + "grad_norm": 0.23316630721092224, + "learning_rate": 2.7972000000000003e-05, + "loss": 0.0142, + "step": 16089 + }, + { + "epoch": 13.589527027027026, + "grad_norm": 0.2771911919116974, + "learning_rate": 2.797166666666667e-05, + "loss": 0.0172, + "step": 16090 + }, + { + "epoch": 13.590371621621621, + "grad_norm": 0.4143233895301819, + "learning_rate": 2.7971333333333334e-05, + "loss": 0.03, + "step": 16091 + }, + { + "epoch": 13.591216216216216, + "grad_norm": 0.46627378463745117, + "learning_rate": 2.7971e-05, + "loss": 0.0204, + "step": 16092 + }, + { + "epoch": 13.59206081081081, + "grad_norm": 0.9400261044502258, + "learning_rate": 2.797066666666667e-05, + "loss": 0.1841, + "step": 16093 + }, + { + "epoch": 13.592905405405405, + "grad_norm": 0.48623889684677124, + "learning_rate": 2.7970333333333332e-05, + "loss": 0.1226, + "step": 16094 + }, + { + "epoch": 13.59375, + "grad_norm": 0.5774137377738953, + "learning_rate": 2.797e-05, + "loss": 0.0865, + "step": 16095 + }, + { + "epoch": 13.594594594594595, + "grad_norm": 0.30347970128059387, + "learning_rate": 2.7969666666666667e-05, + "loss": 0.0453, + "step": 16096 + }, + { + "epoch": 13.59543918918919, + "grad_norm": 0.34901323914527893, + "learning_rate": 2.7969333333333333e-05, + "loss": 0.0208, + "step": 16097 + }, + { + "epoch": 13.596283783783784, + "grad_norm": 0.2708856761455536, + "learning_rate": 2.7969000000000003e-05, + "loss": 0.0268, + "step": 16098 + }, + { + "epoch": 13.597128378378379, + "grad_norm": 0.269328773021698, + "learning_rate": 2.7968666666666665e-05, + "loss": 0.0371, + "step": 16099 + }, + { + "epoch": 13.597972972972974, + "grad_norm": 0.26366618275642395, + "learning_rate": 2.7968333333333335e-05, + "loss": 0.0295, + "step": 16100 + }, + { + "epoch": 13.598817567567568, + "grad_norm": 0.2571895122528076, + "learning_rate": 2.7968e-05, + "loss": 0.0172, + "step": 16101 + }, + { + "epoch": 13.599662162162161, + "grad_norm": 0.22257141768932343, + "learning_rate": 2.7967666666666666e-05, + "loss": 0.0217, + "step": 16102 + }, + { + "epoch": 13.600506756756756, + "grad_norm": 0.4113406538963318, + "learning_rate": 2.7967333333333332e-05, + "loss": 0.0327, + "step": 16103 + }, + { + "epoch": 13.60135135135135, + "grad_norm": 0.20906469225883484, + "learning_rate": 2.7967000000000002e-05, + "loss": 0.0165, + "step": 16104 + }, + { + "epoch": 13.602195945945946, + "grad_norm": 0.15035849809646606, + "learning_rate": 2.7966666666666664e-05, + "loss": 0.0126, + "step": 16105 + }, + { + "epoch": 13.60304054054054, + "grad_norm": 0.3147140145301819, + "learning_rate": 2.7966333333333334e-05, + "loss": 0.014, + "step": 16106 + }, + { + "epoch": 13.603885135135135, + "grad_norm": 0.28991371393203735, + "learning_rate": 2.7966000000000003e-05, + "loss": 0.0206, + "step": 16107 + }, + { + "epoch": 13.60472972972973, + "grad_norm": 0.1983553171157837, + "learning_rate": 2.7965666666666666e-05, + "loss": 0.0145, + "step": 16108 + }, + { + "epoch": 13.605574324324325, + "grad_norm": 0.25702330470085144, + "learning_rate": 2.7965333333333335e-05, + "loss": 0.0158, + "step": 16109 + }, + { + "epoch": 13.60641891891892, + "grad_norm": 0.24102221429347992, + "learning_rate": 2.7965e-05, + "loss": 0.0268, + "step": 16110 + }, + { + "epoch": 13.607263513513514, + "grad_norm": 0.6137382388114929, + "learning_rate": 2.7964666666666667e-05, + "loss": 0.0234, + "step": 16111 + }, + { + "epoch": 13.608108108108109, + "grad_norm": 0.4503389596939087, + "learning_rate": 2.7964333333333333e-05, + "loss": 0.0857, + "step": 16112 + }, + { + "epoch": 13.608952702702704, + "grad_norm": 0.36376211047172546, + "learning_rate": 2.7964000000000002e-05, + "loss": 0.0203, + "step": 16113 + }, + { + "epoch": 13.609797297297296, + "grad_norm": 0.21583901345729828, + "learning_rate": 2.7963666666666668e-05, + "loss": 0.011, + "step": 16114 + }, + { + "epoch": 13.610641891891891, + "grad_norm": 0.6331133842468262, + "learning_rate": 2.7963333333333334e-05, + "loss": 0.0195, + "step": 16115 + }, + { + "epoch": 13.611486486486486, + "grad_norm": 0.350558340549469, + "learning_rate": 2.7963000000000003e-05, + "loss": 0.026, + "step": 16116 + }, + { + "epoch": 13.61233108108108, + "grad_norm": 0.37298741936683655, + "learning_rate": 2.7962666666666666e-05, + "loss": 0.0319, + "step": 16117 + }, + { + "epoch": 13.613175675675675, + "grad_norm": 0.8836919665336609, + "learning_rate": 2.7962333333333335e-05, + "loss": 0.0225, + "step": 16118 + }, + { + "epoch": 13.61402027027027, + "grad_norm": 0.2660037577152252, + "learning_rate": 2.7962e-05, + "loss": 0.0167, + "step": 16119 + }, + { + "epoch": 13.614864864864865, + "grad_norm": 0.5365929007530212, + "learning_rate": 2.7961666666666667e-05, + "loss": 0.0166, + "step": 16120 + }, + { + "epoch": 13.61570945945946, + "grad_norm": 0.2320263534784317, + "learning_rate": 2.7961333333333333e-05, + "loss": 0.0168, + "step": 16121 + }, + { + "epoch": 13.616554054054054, + "grad_norm": 0.26137775182724, + "learning_rate": 2.7961000000000002e-05, + "loss": 0.0179, + "step": 16122 + }, + { + "epoch": 13.61739864864865, + "grad_norm": 0.25162234902381897, + "learning_rate": 2.7960666666666668e-05, + "loss": 0.02, + "step": 16123 + }, + { + "epoch": 13.618243243243244, + "grad_norm": 0.18946611881256104, + "learning_rate": 2.7960333333333334e-05, + "loss": 0.0201, + "step": 16124 + }, + { + "epoch": 13.619087837837839, + "grad_norm": 0.28485244512557983, + "learning_rate": 2.7960000000000003e-05, + "loss": 0.0178, + "step": 16125 + }, + { + "epoch": 13.619932432432432, + "grad_norm": 0.44955384731292725, + "learning_rate": 2.7959666666666666e-05, + "loss": 0.0229, + "step": 16126 + }, + { + "epoch": 13.620777027027026, + "grad_norm": 0.2346201241016388, + "learning_rate": 2.7959333333333335e-05, + "loss": 0.0201, + "step": 16127 + }, + { + "epoch": 13.621621621621621, + "grad_norm": 0.2849792242050171, + "learning_rate": 2.7959e-05, + "loss": 0.0207, + "step": 16128 + }, + { + "epoch": 13.622466216216216, + "grad_norm": 0.39643099904060364, + "learning_rate": 2.7958666666666667e-05, + "loss": 0.026, + "step": 16129 + }, + { + "epoch": 13.62331081081081, + "grad_norm": 0.8220615983009338, + "learning_rate": 2.7958333333333333e-05, + "loss": 0.0127, + "step": 16130 + }, + { + "epoch": 13.624155405405405, + "grad_norm": 0.33367857336997986, + "learning_rate": 2.7958e-05, + "loss": 0.0221, + "step": 16131 + }, + { + "epoch": 13.625, + "grad_norm": 0.3529558777809143, + "learning_rate": 2.7957666666666668e-05, + "loss": 0.0209, + "step": 16132 + }, + { + "epoch": 13.625844594594595, + "grad_norm": 0.21926873922348022, + "learning_rate": 2.7957333333333334e-05, + "loss": 0.0194, + "step": 16133 + }, + { + "epoch": 13.62668918918919, + "grad_norm": 0.3206616938114166, + "learning_rate": 2.7957e-05, + "loss": 0.0164, + "step": 16134 + }, + { + "epoch": 13.627533783783784, + "grad_norm": 0.23268482089042664, + "learning_rate": 2.7956666666666666e-05, + "loss": 0.0192, + "step": 16135 + }, + { + "epoch": 13.628378378378379, + "grad_norm": 0.2927619218826294, + "learning_rate": 2.7956333333333335e-05, + "loss": 0.0137, + "step": 16136 + }, + { + "epoch": 13.629222972972974, + "grad_norm": 1.0943926572799683, + "learning_rate": 2.7955999999999998e-05, + "loss": 0.0256, + "step": 16137 + }, + { + "epoch": 13.630067567567568, + "grad_norm": 0.3237069547176361, + "learning_rate": 2.7955666666666667e-05, + "loss": 0.0144, + "step": 16138 + }, + { + "epoch": 13.630912162162161, + "grad_norm": 0.20200566947460175, + "learning_rate": 2.7955333333333337e-05, + "loss": 0.0164, + "step": 16139 + }, + { + "epoch": 13.631756756756756, + "grad_norm": 0.4090876281261444, + "learning_rate": 2.7955e-05, + "loss": 0.0138, + "step": 16140 + }, + { + "epoch": 13.63260135135135, + "grad_norm": 0.6381800174713135, + "learning_rate": 2.795466666666667e-05, + "loss": 0.0243, + "step": 16141 + }, + { + "epoch": 13.633445945945946, + "grad_norm": 0.87196284532547, + "learning_rate": 2.7954333333333334e-05, + "loss": 0.0349, + "step": 16142 + }, + { + "epoch": 13.63429054054054, + "grad_norm": 0.8558905720710754, + "learning_rate": 2.7954e-05, + "loss": 0.2081, + "step": 16143 + }, + { + "epoch": 13.635135135135135, + "grad_norm": 0.45146963000297546, + "learning_rate": 2.7953666666666666e-05, + "loss": 0.0924, + "step": 16144 + }, + { + "epoch": 13.63597972972973, + "grad_norm": 0.3165004849433899, + "learning_rate": 2.7953333333333336e-05, + "loss": 0.0481, + "step": 16145 + }, + { + "epoch": 13.636824324324325, + "grad_norm": 0.20628519356250763, + "learning_rate": 2.7952999999999998e-05, + "loss": 0.0239, + "step": 16146 + }, + { + "epoch": 13.63766891891892, + "grad_norm": 0.3397146165370941, + "learning_rate": 2.7952666666666667e-05, + "loss": 0.0234, + "step": 16147 + }, + { + "epoch": 13.638513513513514, + "grad_norm": 0.2486695647239685, + "learning_rate": 2.7952333333333337e-05, + "loss": 0.0187, + "step": 16148 + }, + { + "epoch": 13.639358108108109, + "grad_norm": 0.2180568426847458, + "learning_rate": 2.7952e-05, + "loss": 0.0189, + "step": 16149 + }, + { + "epoch": 13.640202702702704, + "grad_norm": 0.57770836353302, + "learning_rate": 2.795166666666667e-05, + "loss": 0.0284, + "step": 16150 + }, + { + "epoch": 13.641047297297296, + "grad_norm": 0.46390923857688904, + "learning_rate": 2.7951333333333335e-05, + "loss": 0.0258, + "step": 16151 + }, + { + "epoch": 13.641891891891891, + "grad_norm": 0.2707643210887909, + "learning_rate": 2.7951e-05, + "loss": 0.0142, + "step": 16152 + }, + { + "epoch": 13.642736486486486, + "grad_norm": 0.5660954713821411, + "learning_rate": 2.7950666666666666e-05, + "loss": 0.0246, + "step": 16153 + }, + { + "epoch": 13.64358108108108, + "grad_norm": 0.34675663709640503, + "learning_rate": 2.7950333333333336e-05, + "loss": 0.0229, + "step": 16154 + }, + { + "epoch": 13.644425675675675, + "grad_norm": 0.5929794907569885, + "learning_rate": 2.795e-05, + "loss": 0.0226, + "step": 16155 + }, + { + "epoch": 13.64527027027027, + "grad_norm": 0.28700441122055054, + "learning_rate": 2.7949666666666668e-05, + "loss": 0.0184, + "step": 16156 + }, + { + "epoch": 13.646114864864865, + "grad_norm": 0.2807639241218567, + "learning_rate": 2.7949333333333337e-05, + "loss": 0.0198, + "step": 16157 + }, + { + "epoch": 13.64695945945946, + "grad_norm": 0.18382245302200317, + "learning_rate": 2.7949e-05, + "loss": 0.0114, + "step": 16158 + }, + { + "epoch": 13.647804054054054, + "grad_norm": 0.2687004804611206, + "learning_rate": 2.794866666666667e-05, + "loss": 0.0183, + "step": 16159 + }, + { + "epoch": 13.64864864864865, + "grad_norm": 0.28107166290283203, + "learning_rate": 2.794833333333333e-05, + "loss": 0.0182, + "step": 16160 + }, + { + "epoch": 13.649493243243244, + "grad_norm": 0.3429552912712097, + "learning_rate": 2.7948e-05, + "loss": 0.0265, + "step": 16161 + }, + { + "epoch": 13.650337837837839, + "grad_norm": 0.2560051381587982, + "learning_rate": 2.7947666666666667e-05, + "loss": 0.0251, + "step": 16162 + }, + { + "epoch": 13.651182432432432, + "grad_norm": 0.22387094795703888, + "learning_rate": 2.7947333333333333e-05, + "loss": 0.0143, + "step": 16163 + }, + { + "epoch": 13.652027027027026, + "grad_norm": 0.2843548357486725, + "learning_rate": 2.7947000000000002e-05, + "loss": 0.0193, + "step": 16164 + }, + { + "epoch": 13.652871621621621, + "grad_norm": 0.27611473202705383, + "learning_rate": 2.7946666666666668e-05, + "loss": 0.0166, + "step": 16165 + }, + { + "epoch": 13.653716216216216, + "grad_norm": 1.0911877155303955, + "learning_rate": 2.7946333333333334e-05, + "loss": 0.0127, + "step": 16166 + }, + { + "epoch": 13.65456081081081, + "grad_norm": 0.3417032063007355, + "learning_rate": 2.7946e-05, + "loss": 0.0118, + "step": 16167 + }, + { + "epoch": 13.655405405405405, + "grad_norm": 0.30825313925743103, + "learning_rate": 2.794566666666667e-05, + "loss": 0.0184, + "step": 16168 + }, + { + "epoch": 13.65625, + "grad_norm": 0.36506080627441406, + "learning_rate": 2.794533333333333e-05, + "loss": 0.026, + "step": 16169 + }, + { + "epoch": 13.657094594594595, + "grad_norm": 0.3961329162120819, + "learning_rate": 2.7945e-05, + "loss": 0.0099, + "step": 16170 + }, + { + "epoch": 13.65793918918919, + "grad_norm": 0.10872790962457657, + "learning_rate": 2.7944666666666667e-05, + "loss": 0.0079, + "step": 16171 + }, + { + "epoch": 13.658783783783784, + "grad_norm": 0.4850275218486786, + "learning_rate": 2.7944333333333333e-05, + "loss": 0.0156, + "step": 16172 + }, + { + "epoch": 13.659628378378379, + "grad_norm": 0.2704842686653137, + "learning_rate": 2.7944000000000002e-05, + "loss": 0.0124, + "step": 16173 + }, + { + "epoch": 13.660472972972974, + "grad_norm": 0.2780350148677826, + "learning_rate": 2.7943666666666668e-05, + "loss": 0.017, + "step": 16174 + }, + { + "epoch": 13.661317567567568, + "grad_norm": 0.3968185782432556, + "learning_rate": 2.7943333333333334e-05, + "loss": 0.0171, + "step": 16175 + }, + { + "epoch": 13.662162162162161, + "grad_norm": 0.29291749000549316, + "learning_rate": 2.7943e-05, + "loss": 0.0165, + "step": 16176 + }, + { + "epoch": 13.663006756756756, + "grad_norm": 0.21574944257736206, + "learning_rate": 2.794266666666667e-05, + "loss": 0.0239, + "step": 16177 + }, + { + "epoch": 13.66385135135135, + "grad_norm": 0.35672447085380554, + "learning_rate": 2.7942333333333332e-05, + "loss": 0.0156, + "step": 16178 + }, + { + "epoch": 13.664695945945946, + "grad_norm": 0.3102535605430603, + "learning_rate": 2.7942e-05, + "loss": 0.0647, + "step": 16179 + }, + { + "epoch": 13.66554054054054, + "grad_norm": 0.2318553626537323, + "learning_rate": 2.7941666666666667e-05, + "loss": 0.0155, + "step": 16180 + }, + { + "epoch": 13.666385135135135, + "grad_norm": 0.24022643268108368, + "learning_rate": 2.7941333333333333e-05, + "loss": 0.0121, + "step": 16181 + }, + { + "epoch": 13.66722972972973, + "grad_norm": 0.4336265027523041, + "learning_rate": 2.7941000000000002e-05, + "loss": 0.0307, + "step": 16182 + }, + { + "epoch": 13.668074324324325, + "grad_norm": 0.3433748483657837, + "learning_rate": 2.7940666666666668e-05, + "loss": 0.0165, + "step": 16183 + }, + { + "epoch": 13.66891891891892, + "grad_norm": 0.2240857481956482, + "learning_rate": 2.7940333333333334e-05, + "loss": 0.0154, + "step": 16184 + }, + { + "epoch": 13.669763513513514, + "grad_norm": 0.3372250199317932, + "learning_rate": 2.794e-05, + "loss": 0.0172, + "step": 16185 + }, + { + "epoch": 13.670608108108109, + "grad_norm": 0.37924596667289734, + "learning_rate": 2.793966666666667e-05, + "loss": 0.0176, + "step": 16186 + }, + { + "epoch": 13.671452702702704, + "grad_norm": 0.4843960702419281, + "learning_rate": 2.7939333333333332e-05, + "loss": 0.0177, + "step": 16187 + }, + { + "epoch": 13.672297297297296, + "grad_norm": 0.2680512070655823, + "learning_rate": 2.7939e-05, + "loss": 0.0164, + "step": 16188 + }, + { + "epoch": 13.673141891891891, + "grad_norm": 0.564902126789093, + "learning_rate": 2.793866666666667e-05, + "loss": 0.0202, + "step": 16189 + }, + { + "epoch": 13.673986486486486, + "grad_norm": 0.8080333471298218, + "learning_rate": 2.7938333333333333e-05, + "loss": 0.0255, + "step": 16190 + }, + { + "epoch": 13.67483108108108, + "grad_norm": 0.5140545964241028, + "learning_rate": 2.7938000000000003e-05, + "loss": 0.0186, + "step": 16191 + }, + { + "epoch": 13.675675675675675, + "grad_norm": 0.40320879220962524, + "learning_rate": 2.7937666666666665e-05, + "loss": 0.0187, + "step": 16192 + }, + { + "epoch": 13.67652027027027, + "grad_norm": 0.6941066384315491, + "learning_rate": 2.7937333333333334e-05, + "loss": 0.1799, + "step": 16193 + }, + { + "epoch": 13.677364864864865, + "grad_norm": 0.5597035884857178, + "learning_rate": 2.7937e-05, + "loss": 0.1479, + "step": 16194 + }, + { + "epoch": 13.67820945945946, + "grad_norm": 0.519451379776001, + "learning_rate": 2.7936666666666666e-05, + "loss": 0.0838, + "step": 16195 + }, + { + "epoch": 13.679054054054054, + "grad_norm": 0.24742349982261658, + "learning_rate": 2.7936333333333332e-05, + "loss": 0.0307, + "step": 16196 + }, + { + "epoch": 13.67989864864865, + "grad_norm": 0.3188227415084839, + "learning_rate": 2.7936e-05, + "loss": 0.0777, + "step": 16197 + }, + { + "epoch": 13.680743243243244, + "grad_norm": 0.2752598226070404, + "learning_rate": 2.7935666666666667e-05, + "loss": 0.0285, + "step": 16198 + }, + { + "epoch": 13.681587837837839, + "grad_norm": 0.2805325388908386, + "learning_rate": 2.7935333333333333e-05, + "loss": 0.0259, + "step": 16199 + }, + { + "epoch": 13.682432432432432, + "grad_norm": 0.48723655939102173, + "learning_rate": 2.7935000000000003e-05, + "loss": 0.0203, + "step": 16200 + }, + { + "epoch": 13.683277027027026, + "grad_norm": 0.23157483339309692, + "learning_rate": 2.7934666666666665e-05, + "loss": 0.0196, + "step": 16201 + }, + { + "epoch": 13.684121621621621, + "grad_norm": 0.2920168340206146, + "learning_rate": 2.7934333333333335e-05, + "loss": 0.0239, + "step": 16202 + }, + { + "epoch": 13.684966216216216, + "grad_norm": 0.22813764214515686, + "learning_rate": 2.7934e-05, + "loss": 0.0171, + "step": 16203 + }, + { + "epoch": 13.68581081081081, + "grad_norm": 0.21814996004104614, + "learning_rate": 2.7933666666666666e-05, + "loss": 0.0172, + "step": 16204 + }, + { + "epoch": 13.686655405405405, + "grad_norm": 0.7567907571792603, + "learning_rate": 2.7933333333333332e-05, + "loss": 0.027, + "step": 16205 + }, + { + "epoch": 13.6875, + "grad_norm": 0.17678846418857574, + "learning_rate": 2.7933000000000002e-05, + "loss": 0.0136, + "step": 16206 + }, + { + "epoch": 13.688344594594595, + "grad_norm": 0.2612326145172119, + "learning_rate": 2.7932666666666668e-05, + "loss": 0.0299, + "step": 16207 + }, + { + "epoch": 13.68918918918919, + "grad_norm": 0.16971899569034576, + "learning_rate": 2.7932333333333334e-05, + "loss": 0.0134, + "step": 16208 + }, + { + "epoch": 13.690033783783784, + "grad_norm": 0.22107505798339844, + "learning_rate": 2.7932000000000003e-05, + "loss": 0.0239, + "step": 16209 + }, + { + "epoch": 13.690878378378379, + "grad_norm": 0.3723365366458893, + "learning_rate": 2.7931666666666665e-05, + "loss": 0.0232, + "step": 16210 + }, + { + "epoch": 13.691722972972974, + "grad_norm": 0.24996089935302734, + "learning_rate": 2.7931333333333335e-05, + "loss": 0.0177, + "step": 16211 + }, + { + "epoch": 13.692567567567568, + "grad_norm": 0.2072087824344635, + "learning_rate": 2.7931e-05, + "loss": 0.012, + "step": 16212 + }, + { + "epoch": 13.693412162162161, + "grad_norm": 0.23141039907932281, + "learning_rate": 2.7930666666666667e-05, + "loss": 0.0146, + "step": 16213 + }, + { + "epoch": 13.694256756756756, + "grad_norm": 0.37296709418296814, + "learning_rate": 2.7930333333333336e-05, + "loss": 0.0124, + "step": 16214 + }, + { + "epoch": 13.69510135135135, + "grad_norm": 0.3773694336414337, + "learning_rate": 2.7930000000000002e-05, + "loss": 0.0208, + "step": 16215 + }, + { + "epoch": 13.695945945945946, + "grad_norm": 0.3122127652168274, + "learning_rate": 2.7929666666666668e-05, + "loss": 0.0238, + "step": 16216 + }, + { + "epoch": 13.69679054054054, + "grad_norm": 0.2991800606250763, + "learning_rate": 2.7929333333333334e-05, + "loss": 0.0356, + "step": 16217 + }, + { + "epoch": 13.697635135135135, + "grad_norm": 0.24168775975704193, + "learning_rate": 2.7929000000000003e-05, + "loss": 0.028, + "step": 16218 + }, + { + "epoch": 13.69847972972973, + "grad_norm": 0.2542780041694641, + "learning_rate": 2.7928666666666666e-05, + "loss": 0.0179, + "step": 16219 + }, + { + "epoch": 13.699324324324325, + "grad_norm": 0.32478225231170654, + "learning_rate": 2.7928333333333335e-05, + "loss": 0.0168, + "step": 16220 + }, + { + "epoch": 13.70016891891892, + "grad_norm": 0.19491377472877502, + "learning_rate": 2.7927999999999998e-05, + "loss": 0.0181, + "step": 16221 + }, + { + "epoch": 13.701013513513514, + "grad_norm": 0.5506712198257446, + "learning_rate": 2.7927666666666667e-05, + "loss": 0.0201, + "step": 16222 + }, + { + "epoch": 13.701858108108109, + "grad_norm": 0.21501491963863373, + "learning_rate": 2.7927333333333336e-05, + "loss": 0.0156, + "step": 16223 + }, + { + "epoch": 13.702702702702704, + "grad_norm": 0.3294637203216553, + "learning_rate": 2.7927e-05, + "loss": 0.0153, + "step": 16224 + }, + { + "epoch": 13.703547297297296, + "grad_norm": 0.21108423173427582, + "learning_rate": 2.7926666666666668e-05, + "loss": 0.0113, + "step": 16225 + }, + { + "epoch": 13.704391891891891, + "grad_norm": 1.071264386177063, + "learning_rate": 2.7926333333333334e-05, + "loss": 0.0374, + "step": 16226 + }, + { + "epoch": 13.705236486486486, + "grad_norm": 0.7850456833839417, + "learning_rate": 2.7926e-05, + "loss": 0.0317, + "step": 16227 + }, + { + "epoch": 13.70608108108108, + "grad_norm": 0.2449038326740265, + "learning_rate": 2.7925666666666666e-05, + "loss": 0.0103, + "step": 16228 + }, + { + "epoch": 13.706925675675675, + "grad_norm": 0.42406970262527466, + "learning_rate": 2.7925333333333335e-05, + "loss": 0.0221, + "step": 16229 + }, + { + "epoch": 13.70777027027027, + "grad_norm": 0.22506509721279144, + "learning_rate": 2.7924999999999998e-05, + "loss": 0.0125, + "step": 16230 + }, + { + "epoch": 13.708614864864865, + "grad_norm": 0.32747992873191833, + "learning_rate": 2.7924666666666667e-05, + "loss": 0.0104, + "step": 16231 + }, + { + "epoch": 13.70945945945946, + "grad_norm": 0.34341028332710266, + "learning_rate": 2.7924333333333336e-05, + "loss": 0.0176, + "step": 16232 + }, + { + "epoch": 13.710304054054054, + "grad_norm": 0.3037194609642029, + "learning_rate": 2.7924e-05, + "loss": 0.0189, + "step": 16233 + }, + { + "epoch": 13.71114864864865, + "grad_norm": 0.3672603666782379, + "learning_rate": 2.7923666666666668e-05, + "loss": 0.0235, + "step": 16234 + }, + { + "epoch": 13.711993243243244, + "grad_norm": 0.3431810438632965, + "learning_rate": 2.7923333333333334e-05, + "loss": 0.015, + "step": 16235 + }, + { + "epoch": 13.712837837837839, + "grad_norm": 0.36421746015548706, + "learning_rate": 2.7923e-05, + "loss": 0.018, + "step": 16236 + }, + { + "epoch": 13.713682432432432, + "grad_norm": 0.507582426071167, + "learning_rate": 2.7922666666666666e-05, + "loss": 0.014, + "step": 16237 + }, + { + "epoch": 13.714527027027026, + "grad_norm": 0.3458860218524933, + "learning_rate": 2.7922333333333335e-05, + "loss": 0.0247, + "step": 16238 + }, + { + "epoch": 13.715371621621621, + "grad_norm": 0.3126634955406189, + "learning_rate": 2.7922e-05, + "loss": 0.0198, + "step": 16239 + }, + { + "epoch": 13.716216216216216, + "grad_norm": 0.46177491545677185, + "learning_rate": 2.7921666666666667e-05, + "loss": 0.0195, + "step": 16240 + }, + { + "epoch": 13.71706081081081, + "grad_norm": 0.7091250419616699, + "learning_rate": 2.7921333333333337e-05, + "loss": 0.0134, + "step": 16241 + }, + { + "epoch": 13.717905405405405, + "grad_norm": 0.6432844996452332, + "learning_rate": 2.7921e-05, + "loss": 0.0402, + "step": 16242 + }, + { + "epoch": 13.71875, + "grad_norm": 0.579963743686676, + "learning_rate": 2.792066666666667e-05, + "loss": 0.2291, + "step": 16243 + }, + { + "epoch": 13.719594594594595, + "grad_norm": 0.8189860582351685, + "learning_rate": 2.7920333333333334e-05, + "loss": 0.1417, + "step": 16244 + }, + { + "epoch": 13.72043918918919, + "grad_norm": 0.31418633460998535, + "learning_rate": 2.792e-05, + "loss": 0.0357, + "step": 16245 + }, + { + "epoch": 13.721283783783784, + "grad_norm": 0.2399778664112091, + "learning_rate": 2.7919666666666666e-05, + "loss": 0.0239, + "step": 16246 + }, + { + "epoch": 13.722128378378379, + "grad_norm": 0.4997183680534363, + "learning_rate": 2.7919333333333336e-05, + "loss": 0.0363, + "step": 16247 + }, + { + "epoch": 13.722972972972974, + "grad_norm": 0.3407302498817444, + "learning_rate": 2.7919e-05, + "loss": 0.0266, + "step": 16248 + }, + { + "epoch": 13.723817567567568, + "grad_norm": 0.20504413545131683, + "learning_rate": 2.7918666666666667e-05, + "loss": 0.0143, + "step": 16249 + }, + { + "epoch": 13.724662162162161, + "grad_norm": 0.19237355887889862, + "learning_rate": 2.7918333333333333e-05, + "loss": 0.0122, + "step": 16250 + }, + { + "epoch": 13.725506756756756, + "grad_norm": 0.19945786893367767, + "learning_rate": 2.7918e-05, + "loss": 0.0109, + "step": 16251 + }, + { + "epoch": 13.72635135135135, + "grad_norm": 0.6108988523483276, + "learning_rate": 2.791766666666667e-05, + "loss": 0.0256, + "step": 16252 + }, + { + "epoch": 13.727195945945946, + "grad_norm": 0.34279876947402954, + "learning_rate": 2.791733333333333e-05, + "loss": 0.0377, + "step": 16253 + }, + { + "epoch": 13.72804054054054, + "grad_norm": 0.18351033329963684, + "learning_rate": 2.7917e-05, + "loss": 0.0136, + "step": 16254 + }, + { + "epoch": 13.728885135135135, + "grad_norm": 0.2379746288061142, + "learning_rate": 2.7916666666666666e-05, + "loss": 0.0201, + "step": 16255 + }, + { + "epoch": 13.72972972972973, + "grad_norm": 0.4793849289417267, + "learning_rate": 2.7916333333333332e-05, + "loss": 0.035, + "step": 16256 + }, + { + "epoch": 13.730574324324325, + "grad_norm": 0.26705870032310486, + "learning_rate": 2.7916000000000002e-05, + "loss": 0.0252, + "step": 16257 + }, + { + "epoch": 13.73141891891892, + "grad_norm": 0.2728113532066345, + "learning_rate": 2.7915666666666668e-05, + "loss": 0.018, + "step": 16258 + }, + { + "epoch": 13.732263513513514, + "grad_norm": 0.30112335085868835, + "learning_rate": 2.7915333333333334e-05, + "loss": 0.0194, + "step": 16259 + }, + { + "epoch": 13.733108108108109, + "grad_norm": 0.13658392429351807, + "learning_rate": 2.7915e-05, + "loss": 0.0093, + "step": 16260 + }, + { + "epoch": 13.733952702702704, + "grad_norm": 0.2129296213388443, + "learning_rate": 2.791466666666667e-05, + "loss": 0.0145, + "step": 16261 + }, + { + "epoch": 13.734797297297296, + "grad_norm": 0.23696133494377136, + "learning_rate": 2.791433333333333e-05, + "loss": 0.0128, + "step": 16262 + }, + { + "epoch": 13.735641891891891, + "grad_norm": 0.29210835695266724, + "learning_rate": 2.7914e-05, + "loss": 0.0208, + "step": 16263 + }, + { + "epoch": 13.736486486486486, + "grad_norm": 0.18757319450378418, + "learning_rate": 2.791366666666667e-05, + "loss": 0.013, + "step": 16264 + }, + { + "epoch": 13.73733108108108, + "grad_norm": 0.42002347111701965, + "learning_rate": 2.7913333333333333e-05, + "loss": 0.0215, + "step": 16265 + }, + { + "epoch": 13.738175675675675, + "grad_norm": 0.2050369530916214, + "learning_rate": 2.7913000000000002e-05, + "loss": 0.0187, + "step": 16266 + }, + { + "epoch": 13.73902027027027, + "grad_norm": 0.2789708971977234, + "learning_rate": 2.7912666666666668e-05, + "loss": 0.0205, + "step": 16267 + }, + { + "epoch": 13.739864864864865, + "grad_norm": 0.7184935212135315, + "learning_rate": 2.7912333333333334e-05, + "loss": 0.02, + "step": 16268 + }, + { + "epoch": 13.74070945945946, + "grad_norm": 0.3642340898513794, + "learning_rate": 2.7912e-05, + "loss": 0.0272, + "step": 16269 + }, + { + "epoch": 13.741554054054054, + "grad_norm": 0.22940000891685486, + "learning_rate": 2.791166666666667e-05, + "loss": 0.0091, + "step": 16270 + }, + { + "epoch": 13.74239864864865, + "grad_norm": 0.30085456371307373, + "learning_rate": 2.791133333333333e-05, + "loss": 0.0245, + "step": 16271 + }, + { + "epoch": 13.743243243243244, + "grad_norm": 0.2861683666706085, + "learning_rate": 2.7911e-05, + "loss": 0.0194, + "step": 16272 + }, + { + "epoch": 13.744087837837839, + "grad_norm": 0.21140605211257935, + "learning_rate": 2.791066666666667e-05, + "loss": 0.0131, + "step": 16273 + }, + { + "epoch": 13.744932432432432, + "grad_norm": 0.530722975730896, + "learning_rate": 2.7910333333333333e-05, + "loss": 0.0325, + "step": 16274 + }, + { + "epoch": 13.745777027027026, + "grad_norm": 0.4252815246582031, + "learning_rate": 2.7910000000000002e-05, + "loss": 0.0238, + "step": 16275 + }, + { + "epoch": 13.746621621621621, + "grad_norm": 0.48341071605682373, + "learning_rate": 2.7909666666666668e-05, + "loss": 0.0246, + "step": 16276 + }, + { + "epoch": 13.747466216216216, + "grad_norm": 0.1690625250339508, + "learning_rate": 2.7909333333333334e-05, + "loss": 0.0122, + "step": 16277 + }, + { + "epoch": 13.74831081081081, + "grad_norm": 0.26102301478385925, + "learning_rate": 2.7909e-05, + "loss": 0.0109, + "step": 16278 + }, + { + "epoch": 13.749155405405405, + "grad_norm": 0.45835572481155396, + "learning_rate": 2.790866666666667e-05, + "loss": 0.0199, + "step": 16279 + }, + { + "epoch": 13.75, + "grad_norm": 0.1464633196592331, + "learning_rate": 2.7908333333333332e-05, + "loss": 0.0096, + "step": 16280 + }, + { + "epoch": 13.750844594594595, + "grad_norm": 0.3070089519023895, + "learning_rate": 2.7908e-05, + "loss": 0.0111, + "step": 16281 + }, + { + "epoch": 13.75168918918919, + "grad_norm": 0.23310552537441254, + "learning_rate": 2.7907666666666667e-05, + "loss": 0.0142, + "step": 16282 + }, + { + "epoch": 13.752533783783784, + "grad_norm": 0.44125279784202576, + "learning_rate": 2.7907333333333333e-05, + "loss": 0.0238, + "step": 16283 + }, + { + "epoch": 13.753378378378379, + "grad_norm": 0.23385083675384521, + "learning_rate": 2.7907000000000002e-05, + "loss": 0.0153, + "step": 16284 + }, + { + "epoch": 13.754222972972974, + "grad_norm": 0.3277779817581177, + "learning_rate": 2.7906666666666665e-05, + "loss": 0.0141, + "step": 16285 + }, + { + "epoch": 13.755067567567568, + "grad_norm": 0.21009129285812378, + "learning_rate": 2.7906333333333334e-05, + "loss": 0.0552, + "step": 16286 + }, + { + "epoch": 13.755912162162161, + "grad_norm": 0.6330579519271851, + "learning_rate": 2.7906e-05, + "loss": 0.0295, + "step": 16287 + }, + { + "epoch": 13.756756756756756, + "grad_norm": 0.4431682527065277, + "learning_rate": 2.7905666666666666e-05, + "loss": 0.0194, + "step": 16288 + }, + { + "epoch": 13.75760135135135, + "grad_norm": 0.5118160247802734, + "learning_rate": 2.7905333333333335e-05, + "loss": 0.0241, + "step": 16289 + }, + { + "epoch": 13.758445945945946, + "grad_norm": 0.47147396206855774, + "learning_rate": 2.7905e-05, + "loss": 0.0261, + "step": 16290 + }, + { + "epoch": 13.75929054054054, + "grad_norm": 0.4484524428844452, + "learning_rate": 2.7904666666666667e-05, + "loss": 0.0169, + "step": 16291 + }, + { + "epoch": 13.760135135135135, + "grad_norm": 0.8231729865074158, + "learning_rate": 2.7904333333333333e-05, + "loss": 0.0413, + "step": 16292 + }, + { + "epoch": 13.76097972972973, + "grad_norm": 1.0626955032348633, + "learning_rate": 2.7904000000000003e-05, + "loss": 0.2038, + "step": 16293 + }, + { + "epoch": 13.761824324324325, + "grad_norm": 0.6230788826942444, + "learning_rate": 2.7903666666666665e-05, + "loss": 0.1471, + "step": 16294 + }, + { + "epoch": 13.76266891891892, + "grad_norm": 0.3360411822795868, + "learning_rate": 2.7903333333333334e-05, + "loss": 0.0584, + "step": 16295 + }, + { + "epoch": 13.763513513513514, + "grad_norm": 0.2587604820728302, + "learning_rate": 2.7903e-05, + "loss": 0.0397, + "step": 16296 + }, + { + "epoch": 13.764358108108109, + "grad_norm": 0.39169108867645264, + "learning_rate": 2.7902666666666666e-05, + "loss": 0.0249, + "step": 16297 + }, + { + "epoch": 13.765202702702704, + "grad_norm": 0.32336708903312683, + "learning_rate": 2.7902333333333336e-05, + "loss": 0.0395, + "step": 16298 + }, + { + "epoch": 13.766047297297296, + "grad_norm": 0.520719051361084, + "learning_rate": 2.7902e-05, + "loss": 0.0314, + "step": 16299 + }, + { + "epoch": 13.766891891891891, + "grad_norm": 0.2857978045940399, + "learning_rate": 2.7901666666666667e-05, + "loss": 0.0228, + "step": 16300 + }, + { + "epoch": 13.767736486486486, + "grad_norm": 0.2097986489534378, + "learning_rate": 2.7901333333333333e-05, + "loss": 0.0116, + "step": 16301 + }, + { + "epoch": 13.76858108108108, + "grad_norm": 0.24643433094024658, + "learning_rate": 2.7901000000000003e-05, + "loss": 0.0243, + "step": 16302 + }, + { + "epoch": 13.769425675675675, + "grad_norm": 0.44852083921432495, + "learning_rate": 2.7900666666666665e-05, + "loss": 0.0197, + "step": 16303 + }, + { + "epoch": 13.77027027027027, + "grad_norm": 0.26399943232536316, + "learning_rate": 2.7900333333333335e-05, + "loss": 0.0226, + "step": 16304 + }, + { + "epoch": 13.771114864864865, + "grad_norm": 0.22228769958019257, + "learning_rate": 2.79e-05, + "loss": 0.0276, + "step": 16305 + }, + { + "epoch": 13.77195945945946, + "grad_norm": 0.4150683879852295, + "learning_rate": 2.7899666666666666e-05, + "loss": 0.0367, + "step": 16306 + }, + { + "epoch": 13.772804054054054, + "grad_norm": 0.2590288817882538, + "learning_rate": 2.7899333333333336e-05, + "loss": 0.0162, + "step": 16307 + }, + { + "epoch": 13.77364864864865, + "grad_norm": 0.36314767599105835, + "learning_rate": 2.7899000000000002e-05, + "loss": 0.0132, + "step": 16308 + }, + { + "epoch": 13.774493243243244, + "grad_norm": 0.27060654759407043, + "learning_rate": 2.7898666666666668e-05, + "loss": 0.0189, + "step": 16309 + }, + { + "epoch": 13.775337837837839, + "grad_norm": 0.28437432646751404, + "learning_rate": 2.7898333333333334e-05, + "loss": 0.0179, + "step": 16310 + }, + { + "epoch": 13.776182432432432, + "grad_norm": 0.41839489340782166, + "learning_rate": 2.7898e-05, + "loss": 0.0318, + "step": 16311 + }, + { + "epoch": 13.777027027027026, + "grad_norm": 0.7703118920326233, + "learning_rate": 2.7897666666666665e-05, + "loss": 0.0414, + "step": 16312 + }, + { + "epoch": 13.777871621621621, + "grad_norm": 0.22051502764225006, + "learning_rate": 2.7897333333333335e-05, + "loss": 0.012, + "step": 16313 + }, + { + "epoch": 13.778716216216216, + "grad_norm": 0.23403850197792053, + "learning_rate": 2.7897e-05, + "loss": 0.0167, + "step": 16314 + }, + { + "epoch": 13.77956081081081, + "grad_norm": 0.2604449391365051, + "learning_rate": 2.7896666666666667e-05, + "loss": 0.0178, + "step": 16315 + }, + { + "epoch": 13.780405405405405, + "grad_norm": 0.28347522020339966, + "learning_rate": 2.7896333333333336e-05, + "loss": 0.018, + "step": 16316 + }, + { + "epoch": 13.78125, + "grad_norm": 0.5283024907112122, + "learning_rate": 2.7896e-05, + "loss": 0.0122, + "step": 16317 + }, + { + "epoch": 13.782094594594595, + "grad_norm": 0.3999168872833252, + "learning_rate": 2.7895666666666668e-05, + "loss": 0.0212, + "step": 16318 + }, + { + "epoch": 13.78293918918919, + "grad_norm": 0.3815142810344696, + "learning_rate": 2.7895333333333334e-05, + "loss": 0.0264, + "step": 16319 + }, + { + "epoch": 13.783783783783784, + "grad_norm": 0.18605183064937592, + "learning_rate": 2.7895e-05, + "loss": 0.0115, + "step": 16320 + }, + { + "epoch": 13.784628378378379, + "grad_norm": 0.17007844150066376, + "learning_rate": 2.7894666666666666e-05, + "loss": 0.0137, + "step": 16321 + }, + { + "epoch": 13.785472972972974, + "grad_norm": 0.4649750888347626, + "learning_rate": 2.7894333333333335e-05, + "loss": 0.0297, + "step": 16322 + }, + { + "epoch": 13.786317567567568, + "grad_norm": 0.1356547474861145, + "learning_rate": 2.7894e-05, + "loss": 0.0086, + "step": 16323 + }, + { + "epoch": 13.787162162162161, + "grad_norm": 0.46447351574897766, + "learning_rate": 2.7893666666666667e-05, + "loss": 0.0183, + "step": 16324 + }, + { + "epoch": 13.788006756756756, + "grad_norm": 0.28638535737991333, + "learning_rate": 2.7893333333333336e-05, + "loss": 0.0205, + "step": 16325 + }, + { + "epoch": 13.78885135135135, + "grad_norm": 0.2329179346561432, + "learning_rate": 2.7893e-05, + "loss": 0.0126, + "step": 16326 + }, + { + "epoch": 13.789695945945946, + "grad_norm": 0.31326475739479065, + "learning_rate": 2.7892666666666668e-05, + "loss": 0.0231, + "step": 16327 + }, + { + "epoch": 13.79054054054054, + "grad_norm": 0.3220471441745758, + "learning_rate": 2.7892333333333334e-05, + "loss": 0.016, + "step": 16328 + }, + { + "epoch": 13.791385135135135, + "grad_norm": 0.1815711259841919, + "learning_rate": 2.7892e-05, + "loss": 0.0106, + "step": 16329 + }, + { + "epoch": 13.79222972972973, + "grad_norm": 0.27914708852767944, + "learning_rate": 2.7891666666666666e-05, + "loss": 0.013, + "step": 16330 + }, + { + "epoch": 13.793074324324325, + "grad_norm": 0.3658321499824524, + "learning_rate": 2.7891333333333335e-05, + "loss": 0.0217, + "step": 16331 + }, + { + "epoch": 13.79391891891892, + "grad_norm": 0.6371797323226929, + "learning_rate": 2.7891e-05, + "loss": 0.041, + "step": 16332 + }, + { + "epoch": 13.794763513513514, + "grad_norm": 0.3036729097366333, + "learning_rate": 2.7890666666666667e-05, + "loss": 0.0315, + "step": 16333 + }, + { + "epoch": 13.795608108108109, + "grad_norm": 0.24197570979595184, + "learning_rate": 2.7890333333333336e-05, + "loss": 0.0164, + "step": 16334 + }, + { + "epoch": 13.796452702702704, + "grad_norm": 0.45276954770088196, + "learning_rate": 2.789e-05, + "loss": 0.0235, + "step": 16335 + }, + { + "epoch": 13.797297297297296, + "grad_norm": 0.49708688259124756, + "learning_rate": 2.7889666666666668e-05, + "loss": 0.0373, + "step": 16336 + }, + { + "epoch": 13.798141891891891, + "grad_norm": 0.30248674750328064, + "learning_rate": 2.7889333333333334e-05, + "loss": 0.0174, + "step": 16337 + }, + { + "epoch": 13.798986486486486, + "grad_norm": 0.3208949565887451, + "learning_rate": 2.7889e-05, + "loss": 0.0165, + "step": 16338 + }, + { + "epoch": 13.79983108108108, + "grad_norm": 0.26550528407096863, + "learning_rate": 2.788866666666667e-05, + "loss": 0.0162, + "step": 16339 + }, + { + "epoch": 13.800675675675675, + "grad_norm": 0.7945832014083862, + "learning_rate": 2.7888333333333335e-05, + "loss": 0.0364, + "step": 16340 + }, + { + "epoch": 13.80152027027027, + "grad_norm": 0.2904093861579895, + "learning_rate": 2.7888e-05, + "loss": 0.0219, + "step": 16341 + }, + { + "epoch": 13.802364864864865, + "grad_norm": 0.3872789442539215, + "learning_rate": 2.7887666666666667e-05, + "loss": 0.0249, + "step": 16342 + }, + { + "epoch": 13.80320945945946, + "grad_norm": 0.593412458896637, + "learning_rate": 2.7887333333333333e-05, + "loss": 0.1762, + "step": 16343 + }, + { + "epoch": 13.804054054054054, + "grad_norm": 0.45857545733451843, + "learning_rate": 2.7887e-05, + "loss": 0.0966, + "step": 16344 + }, + { + "epoch": 13.80489864864865, + "grad_norm": 0.33011776208877563, + "learning_rate": 2.788666666666667e-05, + "loss": 0.036, + "step": 16345 + }, + { + "epoch": 13.805743243243244, + "grad_norm": 0.4200292229652405, + "learning_rate": 2.788633333333333e-05, + "loss": 0.0365, + "step": 16346 + }, + { + "epoch": 13.806587837837839, + "grad_norm": 0.27185264229774475, + "learning_rate": 2.7886e-05, + "loss": 0.0384, + "step": 16347 + }, + { + "epoch": 13.807432432432432, + "grad_norm": 0.4814414381980896, + "learning_rate": 2.788566666666667e-05, + "loss": 0.0347, + "step": 16348 + }, + { + "epoch": 13.808277027027026, + "grad_norm": 0.3981855511665344, + "learning_rate": 2.7885333333333332e-05, + "loss": 0.0292, + "step": 16349 + }, + { + "epoch": 13.809121621621621, + "grad_norm": 0.2552737295627594, + "learning_rate": 2.7885e-05, + "loss": 0.0142, + "step": 16350 + }, + { + "epoch": 13.809966216216216, + "grad_norm": 0.5579147338867188, + "learning_rate": 2.7884666666666667e-05, + "loss": 0.0198, + "step": 16351 + }, + { + "epoch": 13.81081081081081, + "grad_norm": 0.5021296143531799, + "learning_rate": 2.7884333333333333e-05, + "loss": 0.0167, + "step": 16352 + }, + { + "epoch": 13.811655405405405, + "grad_norm": 0.2883318364620209, + "learning_rate": 2.7884e-05, + "loss": 0.0274, + "step": 16353 + }, + { + "epoch": 13.8125, + "grad_norm": 0.35167771577835083, + "learning_rate": 2.788366666666667e-05, + "loss": 0.0248, + "step": 16354 + }, + { + "epoch": 13.813344594594595, + "grad_norm": 0.3994517922401428, + "learning_rate": 2.788333333333333e-05, + "loss": 0.0276, + "step": 16355 + }, + { + "epoch": 13.81418918918919, + "grad_norm": 0.32065197825431824, + "learning_rate": 2.7883e-05, + "loss": 0.0236, + "step": 16356 + }, + { + "epoch": 13.815033783783784, + "grad_norm": 0.30914145708084106, + "learning_rate": 2.788266666666667e-05, + "loss": 0.0229, + "step": 16357 + }, + { + "epoch": 13.815878378378379, + "grad_norm": 0.1832132637500763, + "learning_rate": 2.7882333333333332e-05, + "loss": 0.0151, + "step": 16358 + }, + { + "epoch": 13.816722972972974, + "grad_norm": 0.2939929962158203, + "learning_rate": 2.7882000000000002e-05, + "loss": 0.0132, + "step": 16359 + }, + { + "epoch": 13.817567567567568, + "grad_norm": 0.3880079686641693, + "learning_rate": 2.7881666666666668e-05, + "loss": 0.0217, + "step": 16360 + }, + { + "epoch": 13.818412162162161, + "grad_norm": 0.19021375477313995, + "learning_rate": 2.7881333333333334e-05, + "loss": 0.0185, + "step": 16361 + }, + { + "epoch": 13.819256756756756, + "grad_norm": 0.2762802541255951, + "learning_rate": 2.7881e-05, + "loss": 0.0154, + "step": 16362 + }, + { + "epoch": 13.82010135135135, + "grad_norm": 0.24415458738803864, + "learning_rate": 2.788066666666667e-05, + "loss": 0.0237, + "step": 16363 + }, + { + "epoch": 13.820945945945946, + "grad_norm": 0.2533993721008301, + "learning_rate": 2.7880333333333335e-05, + "loss": 0.0129, + "step": 16364 + }, + { + "epoch": 13.82179054054054, + "grad_norm": 0.23968371748924255, + "learning_rate": 2.788e-05, + "loss": 0.0191, + "step": 16365 + }, + { + "epoch": 13.822635135135135, + "grad_norm": 0.3078904449939728, + "learning_rate": 2.787966666666667e-05, + "loss": 0.0182, + "step": 16366 + }, + { + "epoch": 13.82347972972973, + "grad_norm": 0.2781379520893097, + "learning_rate": 2.7879333333333333e-05, + "loss": 0.0191, + "step": 16367 + }, + { + "epoch": 13.824324324324325, + "grad_norm": 0.20753099024295807, + "learning_rate": 2.7879000000000002e-05, + "loss": 0.0104, + "step": 16368 + }, + { + "epoch": 13.82516891891892, + "grad_norm": 0.20400960743427277, + "learning_rate": 2.7878666666666668e-05, + "loss": 0.0212, + "step": 16369 + }, + { + "epoch": 13.826013513513514, + "grad_norm": 0.3964596688747406, + "learning_rate": 2.7878333333333334e-05, + "loss": 0.018, + "step": 16370 + }, + { + "epoch": 13.826858108108109, + "grad_norm": 0.26933208107948303, + "learning_rate": 2.7878e-05, + "loss": 0.0113, + "step": 16371 + }, + { + "epoch": 13.827702702702704, + "grad_norm": 0.41719260811805725, + "learning_rate": 2.7877666666666666e-05, + "loss": 0.0229, + "step": 16372 + }, + { + "epoch": 13.828547297297296, + "grad_norm": 0.3600572645664215, + "learning_rate": 2.7877333333333335e-05, + "loss": 0.0168, + "step": 16373 + }, + { + "epoch": 13.829391891891891, + "grad_norm": 0.2602625787258148, + "learning_rate": 2.7877e-05, + "loss": 0.0238, + "step": 16374 + }, + { + "epoch": 13.830236486486486, + "grad_norm": 0.32622888684272766, + "learning_rate": 2.7876666666666667e-05, + "loss": 0.0144, + "step": 16375 + }, + { + "epoch": 13.83108108108108, + "grad_norm": 0.3193621039390564, + "learning_rate": 2.7876333333333333e-05, + "loss": 0.0195, + "step": 16376 + }, + { + "epoch": 13.831925675675675, + "grad_norm": 0.27214694023132324, + "learning_rate": 2.7876000000000002e-05, + "loss": 0.0157, + "step": 16377 + }, + { + "epoch": 13.83277027027027, + "grad_norm": 0.20872297883033752, + "learning_rate": 2.7875666666666665e-05, + "loss": 0.0063, + "step": 16378 + }, + { + "epoch": 13.833614864864865, + "grad_norm": 0.6356427073478699, + "learning_rate": 2.7875333333333334e-05, + "loss": 0.0187, + "step": 16379 + }, + { + "epoch": 13.83445945945946, + "grad_norm": 0.7682353258132935, + "learning_rate": 2.7875e-05, + "loss": 0.0252, + "step": 16380 + }, + { + "epoch": 13.835304054054054, + "grad_norm": 0.33024537563323975, + "learning_rate": 2.7874666666666666e-05, + "loss": 0.0203, + "step": 16381 + }, + { + "epoch": 13.83614864864865, + "grad_norm": 0.3408967852592468, + "learning_rate": 2.7874333333333335e-05, + "loss": 0.0133, + "step": 16382 + }, + { + "epoch": 13.836993243243244, + "grad_norm": 0.45348331332206726, + "learning_rate": 2.7874e-05, + "loss": 0.0266, + "step": 16383 + }, + { + "epoch": 13.837837837837839, + "grad_norm": 0.30448806285858154, + "learning_rate": 2.7873666666666667e-05, + "loss": 0.0252, + "step": 16384 + }, + { + "epoch": 13.838682432432432, + "grad_norm": 0.23834432661533356, + "learning_rate": 2.7873333333333333e-05, + "loss": 0.0173, + "step": 16385 + }, + { + "epoch": 13.839527027027026, + "grad_norm": 0.5443035960197449, + "learning_rate": 2.7873000000000002e-05, + "loss": 0.0278, + "step": 16386 + }, + { + "epoch": 13.840371621621621, + "grad_norm": 0.3382824957370758, + "learning_rate": 2.7872666666666665e-05, + "loss": 0.0263, + "step": 16387 + }, + { + "epoch": 13.841216216216216, + "grad_norm": 0.45583197474479675, + "learning_rate": 2.7872333333333334e-05, + "loss": 0.0305, + "step": 16388 + }, + { + "epoch": 13.84206081081081, + "grad_norm": 1.0205776691436768, + "learning_rate": 2.7872000000000004e-05, + "loss": 0.0248, + "step": 16389 + }, + { + "epoch": 13.842905405405405, + "grad_norm": 0.7155402302742004, + "learning_rate": 2.7871666666666666e-05, + "loss": 0.0249, + "step": 16390 + }, + { + "epoch": 13.84375, + "grad_norm": 0.43325814604759216, + "learning_rate": 2.7871333333333335e-05, + "loss": 0.0157, + "step": 16391 + }, + { + "epoch": 13.844594594594595, + "grad_norm": 0.29458773136138916, + "learning_rate": 2.7871e-05, + "loss": 0.0274, + "step": 16392 + }, + { + "epoch": 13.84543918918919, + "grad_norm": 0.7301111221313477, + "learning_rate": 2.7870666666666667e-05, + "loss": 0.1676, + "step": 16393 + }, + { + "epoch": 13.846283783783784, + "grad_norm": 0.7294097542762756, + "learning_rate": 2.7870333333333333e-05, + "loss": 0.1091, + "step": 16394 + }, + { + "epoch": 13.847128378378379, + "grad_norm": 0.2435464709997177, + "learning_rate": 2.7870000000000003e-05, + "loss": 0.0414, + "step": 16395 + }, + { + "epoch": 13.847972972972974, + "grad_norm": 0.33471453189849854, + "learning_rate": 2.7869666666666665e-05, + "loss": 0.0517, + "step": 16396 + }, + { + "epoch": 13.848817567567568, + "grad_norm": 0.2854982316493988, + "learning_rate": 2.7869333333333334e-05, + "loss": 0.0266, + "step": 16397 + }, + { + "epoch": 13.849662162162161, + "grad_norm": 0.8408147692680359, + "learning_rate": 2.7869000000000004e-05, + "loss": 0.018, + "step": 16398 + }, + { + "epoch": 13.850506756756756, + "grad_norm": 0.5797700881958008, + "learning_rate": 2.7868666666666666e-05, + "loss": 0.0197, + "step": 16399 + }, + { + "epoch": 13.85135135135135, + "grad_norm": 0.42324432730674744, + "learning_rate": 2.7868333333333336e-05, + "loss": 0.0196, + "step": 16400 + }, + { + "epoch": 13.852195945945946, + "grad_norm": 0.1544431746006012, + "learning_rate": 2.7868e-05, + "loss": 0.0141, + "step": 16401 + }, + { + "epoch": 13.85304054054054, + "grad_norm": 0.17866390943527222, + "learning_rate": 2.7867666666666667e-05, + "loss": 0.0179, + "step": 16402 + }, + { + "epoch": 13.853885135135135, + "grad_norm": 0.2973186671733856, + "learning_rate": 2.7867333333333333e-05, + "loss": 0.0197, + "step": 16403 + }, + { + "epoch": 13.85472972972973, + "grad_norm": 0.1772218644618988, + "learning_rate": 2.7867e-05, + "loss": 0.0123, + "step": 16404 + }, + { + "epoch": 13.855574324324325, + "grad_norm": 0.23582841455936432, + "learning_rate": 2.7866666666666665e-05, + "loss": 0.0184, + "step": 16405 + }, + { + "epoch": 13.85641891891892, + "grad_norm": 0.20302744209766388, + "learning_rate": 2.7866333333333335e-05, + "loss": 0.0176, + "step": 16406 + }, + { + "epoch": 13.857263513513514, + "grad_norm": 0.2561096251010895, + "learning_rate": 2.7866e-05, + "loss": 0.0155, + "step": 16407 + }, + { + "epoch": 13.858108108108109, + "grad_norm": 0.2592264413833618, + "learning_rate": 2.7865666666666666e-05, + "loss": 0.0189, + "step": 16408 + }, + { + "epoch": 13.858952702702704, + "grad_norm": 0.21465454995632172, + "learning_rate": 2.7865333333333336e-05, + "loss": 0.0248, + "step": 16409 + }, + { + "epoch": 13.859797297297296, + "grad_norm": 0.2393125593662262, + "learning_rate": 2.7865e-05, + "loss": 0.0134, + "step": 16410 + }, + { + "epoch": 13.860641891891891, + "grad_norm": 0.5599182844161987, + "learning_rate": 2.7864666666666668e-05, + "loss": 0.0163, + "step": 16411 + }, + { + "epoch": 13.861486486486486, + "grad_norm": 0.39050161838531494, + "learning_rate": 2.7864333333333334e-05, + "loss": 0.0116, + "step": 16412 + }, + { + "epoch": 13.86233108108108, + "grad_norm": 0.24680916965007782, + "learning_rate": 2.7864e-05, + "loss": 0.0206, + "step": 16413 + }, + { + "epoch": 13.863175675675675, + "grad_norm": 0.31634074449539185, + "learning_rate": 2.786366666666667e-05, + "loss": 0.035, + "step": 16414 + }, + { + "epoch": 13.86402027027027, + "grad_norm": 0.44420188665390015, + "learning_rate": 2.7863333333333335e-05, + "loss": 0.0192, + "step": 16415 + }, + { + "epoch": 13.864864864864865, + "grad_norm": 0.325219064950943, + "learning_rate": 2.7863e-05, + "loss": 0.0338, + "step": 16416 + }, + { + "epoch": 13.86570945945946, + "grad_norm": 0.35920777916908264, + "learning_rate": 2.7862666666666667e-05, + "loss": 0.0195, + "step": 16417 + }, + { + "epoch": 13.866554054054054, + "grad_norm": 0.2846837043762207, + "learning_rate": 2.7862333333333336e-05, + "loss": 0.0177, + "step": 16418 + }, + { + "epoch": 13.86739864864865, + "grad_norm": 0.25246644020080566, + "learning_rate": 2.7862e-05, + "loss": 0.0147, + "step": 16419 + }, + { + "epoch": 13.868243243243244, + "grad_norm": 0.8791785836219788, + "learning_rate": 2.7861666666666668e-05, + "loss": 0.0187, + "step": 16420 + }, + { + "epoch": 13.869087837837839, + "grad_norm": 0.40739327669143677, + "learning_rate": 2.7861333333333334e-05, + "loss": 0.0174, + "step": 16421 + }, + { + "epoch": 13.869932432432432, + "grad_norm": 0.2185235619544983, + "learning_rate": 2.7861e-05, + "loss": 0.0186, + "step": 16422 + }, + { + "epoch": 13.870777027027026, + "grad_norm": 0.3158319294452667, + "learning_rate": 2.786066666666667e-05, + "loss": 0.0199, + "step": 16423 + }, + { + "epoch": 13.871621621621621, + "grad_norm": 0.2952422499656677, + "learning_rate": 2.7860333333333335e-05, + "loss": 0.0223, + "step": 16424 + }, + { + "epoch": 13.872466216216216, + "grad_norm": 0.3006342053413391, + "learning_rate": 2.786e-05, + "loss": 0.0195, + "step": 16425 + }, + { + "epoch": 13.87331081081081, + "grad_norm": 0.3940091133117676, + "learning_rate": 2.7859666666666667e-05, + "loss": 0.0232, + "step": 16426 + }, + { + "epoch": 13.874155405405405, + "grad_norm": 0.3903473913669586, + "learning_rate": 2.7859333333333336e-05, + "loss": 0.0192, + "step": 16427 + }, + { + "epoch": 13.875, + "grad_norm": 0.3657788038253784, + "learning_rate": 2.7859e-05, + "loss": 0.02, + "step": 16428 + }, + { + "epoch": 13.875844594594595, + "grad_norm": 0.361594557762146, + "learning_rate": 2.7858666666666668e-05, + "loss": 0.0251, + "step": 16429 + }, + { + "epoch": 13.87668918918919, + "grad_norm": 0.6060013771057129, + "learning_rate": 2.7858333333333334e-05, + "loss": 0.0161, + "step": 16430 + }, + { + "epoch": 13.877533783783784, + "grad_norm": 0.22765490412712097, + "learning_rate": 2.7858e-05, + "loss": 0.0142, + "step": 16431 + }, + { + "epoch": 13.878378378378379, + "grad_norm": 0.16792422533035278, + "learning_rate": 2.785766666666667e-05, + "loss": 0.0142, + "step": 16432 + }, + { + "epoch": 13.879222972972974, + "grad_norm": 0.4182423949241638, + "learning_rate": 2.7857333333333332e-05, + "loss": 0.0255, + "step": 16433 + }, + { + "epoch": 13.880067567567568, + "grad_norm": 0.5492784380912781, + "learning_rate": 2.7857e-05, + "loss": 0.0281, + "step": 16434 + }, + { + "epoch": 13.880912162162161, + "grad_norm": 0.49810168147087097, + "learning_rate": 2.7856666666666667e-05, + "loss": 0.0194, + "step": 16435 + }, + { + "epoch": 13.881756756756756, + "grad_norm": 0.48105064034461975, + "learning_rate": 2.7856333333333333e-05, + "loss": 0.037, + "step": 16436 + }, + { + "epoch": 13.88260135135135, + "grad_norm": 0.3321689963340759, + "learning_rate": 2.7856e-05, + "loss": 0.0151, + "step": 16437 + }, + { + "epoch": 13.883445945945946, + "grad_norm": 0.7610068321228027, + "learning_rate": 2.7855666666666668e-05, + "loss": 0.0343, + "step": 16438 + }, + { + "epoch": 13.88429054054054, + "grad_norm": 0.18701933324337006, + "learning_rate": 2.7855333333333334e-05, + "loss": 0.0116, + "step": 16439 + }, + { + "epoch": 13.885135135135135, + "grad_norm": 0.23267680406570435, + "learning_rate": 2.7855e-05, + "loss": 0.0182, + "step": 16440 + }, + { + "epoch": 13.88597972972973, + "grad_norm": 0.7282441854476929, + "learning_rate": 2.785466666666667e-05, + "loss": 0.0301, + "step": 16441 + }, + { + "epoch": 13.886824324324325, + "grad_norm": 0.5785070657730103, + "learning_rate": 2.7854333333333332e-05, + "loss": 0.0333, + "step": 16442 + }, + { + "epoch": 13.88766891891892, + "grad_norm": 0.6477532982826233, + "learning_rate": 2.7854e-05, + "loss": 0.1856, + "step": 16443 + }, + { + "epoch": 13.888513513513514, + "grad_norm": 0.5089526772499084, + "learning_rate": 2.7853666666666667e-05, + "loss": 0.1212, + "step": 16444 + }, + { + "epoch": 13.889358108108109, + "grad_norm": 0.42625585198402405, + "learning_rate": 2.7853333333333333e-05, + "loss": 0.0695, + "step": 16445 + }, + { + "epoch": 13.890202702702704, + "grad_norm": 0.40709981322288513, + "learning_rate": 2.7853e-05, + "loss": 0.0603, + "step": 16446 + }, + { + "epoch": 13.891047297297296, + "grad_norm": 0.3051796853542328, + "learning_rate": 2.785266666666667e-05, + "loss": 0.0289, + "step": 16447 + }, + { + "epoch": 13.891891891891891, + "grad_norm": 0.26647037267684937, + "learning_rate": 2.7852333333333334e-05, + "loss": 0.0308, + "step": 16448 + }, + { + "epoch": 13.892736486486486, + "grad_norm": 0.25560256838798523, + "learning_rate": 2.7852e-05, + "loss": 0.0241, + "step": 16449 + }, + { + "epoch": 13.89358108108108, + "grad_norm": 0.32891932129859924, + "learning_rate": 2.785166666666667e-05, + "loss": 0.0208, + "step": 16450 + }, + { + "epoch": 13.894425675675675, + "grad_norm": 0.31709158420562744, + "learning_rate": 2.7851333333333332e-05, + "loss": 0.0317, + "step": 16451 + }, + { + "epoch": 13.89527027027027, + "grad_norm": 0.1923564225435257, + "learning_rate": 2.7851e-05, + "loss": 0.0142, + "step": 16452 + }, + { + "epoch": 13.896114864864865, + "grad_norm": 0.2698856294155121, + "learning_rate": 2.7850666666666667e-05, + "loss": 0.0186, + "step": 16453 + }, + { + "epoch": 13.89695945945946, + "grad_norm": 0.20698516070842743, + "learning_rate": 2.7850333333333333e-05, + "loss": 0.0267, + "step": 16454 + }, + { + "epoch": 13.897804054054054, + "grad_norm": 0.16820867359638214, + "learning_rate": 2.7850000000000003e-05, + "loss": 0.0133, + "step": 16455 + }, + { + "epoch": 13.89864864864865, + "grad_norm": 0.29229670763015747, + "learning_rate": 2.784966666666667e-05, + "loss": 0.0289, + "step": 16456 + }, + { + "epoch": 13.899493243243244, + "grad_norm": 0.2971605658531189, + "learning_rate": 2.7849333333333335e-05, + "loss": 0.0157, + "step": 16457 + }, + { + "epoch": 13.900337837837839, + "grad_norm": 0.23787671327590942, + "learning_rate": 2.7849e-05, + "loss": 0.0202, + "step": 16458 + }, + { + "epoch": 13.901182432432432, + "grad_norm": 0.3977454602718353, + "learning_rate": 2.784866666666667e-05, + "loss": 0.0293, + "step": 16459 + }, + { + "epoch": 13.902027027027026, + "grad_norm": 0.2220098376274109, + "learning_rate": 2.7848333333333332e-05, + "loss": 0.0221, + "step": 16460 + }, + { + "epoch": 13.902871621621621, + "grad_norm": 0.345988005399704, + "learning_rate": 2.7848000000000002e-05, + "loss": 0.016, + "step": 16461 + }, + { + "epoch": 13.903716216216216, + "grad_norm": 0.3537645637989044, + "learning_rate": 2.7847666666666664e-05, + "loss": 0.0226, + "step": 16462 + }, + { + "epoch": 13.90456081081081, + "grad_norm": 0.21794359385967255, + "learning_rate": 2.7847333333333334e-05, + "loss": 0.0167, + "step": 16463 + }, + { + "epoch": 13.905405405405405, + "grad_norm": 0.22157040238380432, + "learning_rate": 2.7847000000000003e-05, + "loss": 0.022, + "step": 16464 + }, + { + "epoch": 13.90625, + "grad_norm": 0.36817511916160583, + "learning_rate": 2.7846666666666665e-05, + "loss": 0.0325, + "step": 16465 + }, + { + "epoch": 13.907094594594595, + "grad_norm": 0.2903040051460266, + "learning_rate": 2.7846333333333335e-05, + "loss": 0.0263, + "step": 16466 + }, + { + "epoch": 13.90793918918919, + "grad_norm": 0.4608272314071655, + "learning_rate": 2.7846e-05, + "loss": 0.0122, + "step": 16467 + }, + { + "epoch": 13.908783783783784, + "grad_norm": 0.15033137798309326, + "learning_rate": 2.7845666666666667e-05, + "loss": 0.0124, + "step": 16468 + }, + { + "epoch": 13.909628378378379, + "grad_norm": 0.20699633657932281, + "learning_rate": 2.7845333333333333e-05, + "loss": 0.0136, + "step": 16469 + }, + { + "epoch": 13.910472972972974, + "grad_norm": 0.732410728931427, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0442, + "step": 16470 + }, + { + "epoch": 13.911317567567568, + "grad_norm": 0.1756887137889862, + "learning_rate": 2.7844666666666664e-05, + "loss": 0.0118, + "step": 16471 + }, + { + "epoch": 13.912162162162161, + "grad_norm": 0.2204255610704422, + "learning_rate": 2.7844333333333334e-05, + "loss": 0.0127, + "step": 16472 + }, + { + "epoch": 13.913006756756756, + "grad_norm": 0.4169881045818329, + "learning_rate": 2.7844000000000003e-05, + "loss": 0.0126, + "step": 16473 + }, + { + "epoch": 13.91385135135135, + "grad_norm": 0.2187182903289795, + "learning_rate": 2.7843666666666666e-05, + "loss": 0.0224, + "step": 16474 + }, + { + "epoch": 13.914695945945946, + "grad_norm": 0.24558156728744507, + "learning_rate": 2.7843333333333335e-05, + "loss": 0.0188, + "step": 16475 + }, + { + "epoch": 13.91554054054054, + "grad_norm": 0.3517301678657532, + "learning_rate": 2.7843e-05, + "loss": 0.0191, + "step": 16476 + }, + { + "epoch": 13.916385135135135, + "grad_norm": 0.37373778223991394, + "learning_rate": 2.7842666666666667e-05, + "loss": 0.0163, + "step": 16477 + }, + { + "epoch": 13.91722972972973, + "grad_norm": 0.4622776508331299, + "learning_rate": 2.7842333333333333e-05, + "loss": 0.0278, + "step": 16478 + }, + { + "epoch": 13.918074324324325, + "grad_norm": 0.24472002685070038, + "learning_rate": 2.7842000000000002e-05, + "loss": 0.0154, + "step": 16479 + }, + { + "epoch": 13.91891891891892, + "grad_norm": 0.2854034900665283, + "learning_rate": 2.7841666666666668e-05, + "loss": 0.0535, + "step": 16480 + }, + { + "epoch": 13.919763513513514, + "grad_norm": 0.3162023425102234, + "learning_rate": 2.7841333333333334e-05, + "loss": 0.0157, + "step": 16481 + }, + { + "epoch": 13.920608108108109, + "grad_norm": 0.21152593195438385, + "learning_rate": 2.7841000000000003e-05, + "loss": 0.0104, + "step": 16482 + }, + { + "epoch": 13.921452702702704, + "grad_norm": 0.330634742975235, + "learning_rate": 2.7840666666666666e-05, + "loss": 0.0169, + "step": 16483 + }, + { + "epoch": 13.922297297297296, + "grad_norm": 0.4391671419143677, + "learning_rate": 2.7840333333333335e-05, + "loss": 0.012, + "step": 16484 + }, + { + "epoch": 13.923141891891891, + "grad_norm": 0.2502332627773285, + "learning_rate": 2.784e-05, + "loss": 0.0165, + "step": 16485 + }, + { + "epoch": 13.923986486486486, + "grad_norm": 0.27992087602615356, + "learning_rate": 2.7839666666666667e-05, + "loss": 0.0192, + "step": 16486 + }, + { + "epoch": 13.92483108108108, + "grad_norm": 0.38158929347991943, + "learning_rate": 2.7839333333333333e-05, + "loss": 0.0209, + "step": 16487 + }, + { + "epoch": 13.925675675675675, + "grad_norm": 0.4432036876678467, + "learning_rate": 2.7839000000000002e-05, + "loss": 0.0175, + "step": 16488 + }, + { + "epoch": 13.92652027027027, + "grad_norm": 0.39853280782699585, + "learning_rate": 2.7838666666666668e-05, + "loss": 0.0172, + "step": 16489 + }, + { + "epoch": 13.927364864864865, + "grad_norm": 0.2490987777709961, + "learning_rate": 2.7838333333333334e-05, + "loss": 0.0176, + "step": 16490 + }, + { + "epoch": 13.92820945945946, + "grad_norm": 0.3715716302394867, + "learning_rate": 2.7838000000000004e-05, + "loss": 0.0246, + "step": 16491 + }, + { + "epoch": 13.929054054054054, + "grad_norm": 0.6950526833534241, + "learning_rate": 2.7837666666666666e-05, + "loss": 0.0882, + "step": 16492 + }, + { + "epoch": 13.92989864864865, + "grad_norm": 0.54353928565979, + "learning_rate": 2.7837333333333335e-05, + "loss": 0.171, + "step": 16493 + }, + { + "epoch": 13.930743243243244, + "grad_norm": 0.5349200367927551, + "learning_rate": 2.7836999999999998e-05, + "loss": 0.0931, + "step": 16494 + }, + { + "epoch": 13.931587837837839, + "grad_norm": 0.4290359914302826, + "learning_rate": 2.7836666666666667e-05, + "loss": 0.0799, + "step": 16495 + }, + { + "epoch": 13.932432432432432, + "grad_norm": 0.2832987308502197, + "learning_rate": 2.7836333333333333e-05, + "loss": 0.0314, + "step": 16496 + }, + { + "epoch": 13.933277027027026, + "grad_norm": 0.2523161768913269, + "learning_rate": 2.7836e-05, + "loss": 0.0194, + "step": 16497 + }, + { + "epoch": 13.934121621621621, + "grad_norm": 0.22724615037441254, + "learning_rate": 2.783566666666667e-05, + "loss": 0.0143, + "step": 16498 + }, + { + "epoch": 13.934966216216216, + "grad_norm": 0.19244861602783203, + "learning_rate": 2.7835333333333334e-05, + "loss": 0.016, + "step": 16499 + }, + { + "epoch": 13.93581081081081, + "grad_norm": 0.16857299208641052, + "learning_rate": 2.7835e-05, + "loss": 0.0183, + "step": 16500 + }, + { + "epoch": 13.936655405405405, + "grad_norm": 0.31323525309562683, + "learning_rate": 2.7834666666666666e-05, + "loss": 0.0274, + "step": 16501 + }, + { + "epoch": 13.9375, + "grad_norm": 0.1969996839761734, + "learning_rate": 2.7834333333333336e-05, + "loss": 0.0137, + "step": 16502 + }, + { + "epoch": 13.938344594594595, + "grad_norm": 0.33328136801719666, + "learning_rate": 2.7833999999999998e-05, + "loss": 0.0209, + "step": 16503 + }, + { + "epoch": 13.93918918918919, + "grad_norm": 0.25873300433158875, + "learning_rate": 2.7833666666666667e-05, + "loss": 0.0226, + "step": 16504 + }, + { + "epoch": 13.940033783783784, + "grad_norm": 0.19333812594413757, + "learning_rate": 2.7833333333333337e-05, + "loss": 0.0129, + "step": 16505 + }, + { + "epoch": 13.940878378378379, + "grad_norm": 0.4304805099964142, + "learning_rate": 2.7833e-05, + "loss": 0.0324, + "step": 16506 + }, + { + "epoch": 13.941722972972974, + "grad_norm": 0.4337078928947449, + "learning_rate": 2.783266666666667e-05, + "loss": 0.0184, + "step": 16507 + }, + { + "epoch": 13.942567567567568, + "grad_norm": 0.5029478669166565, + "learning_rate": 2.7832333333333335e-05, + "loss": 0.0209, + "step": 16508 + }, + { + "epoch": 13.943412162162161, + "grad_norm": 0.7297004461288452, + "learning_rate": 2.7832e-05, + "loss": 0.022, + "step": 16509 + }, + { + "epoch": 13.944256756756756, + "grad_norm": 0.23096539080142975, + "learning_rate": 2.7831666666666666e-05, + "loss": 0.015, + "step": 16510 + }, + { + "epoch": 13.94510135135135, + "grad_norm": 0.3705713748931885, + "learning_rate": 2.7831333333333336e-05, + "loss": 0.0193, + "step": 16511 + }, + { + "epoch": 13.945945945945946, + "grad_norm": 0.2002715915441513, + "learning_rate": 2.7831e-05, + "loss": 0.0202, + "step": 16512 + }, + { + "epoch": 13.94679054054054, + "grad_norm": 0.1884133219718933, + "learning_rate": 2.7830666666666668e-05, + "loss": 0.013, + "step": 16513 + }, + { + "epoch": 13.947635135135135, + "grad_norm": 0.2618867754936218, + "learning_rate": 2.7830333333333337e-05, + "loss": 0.018, + "step": 16514 + }, + { + "epoch": 13.94847972972973, + "grad_norm": 0.3222339451313019, + "learning_rate": 2.783e-05, + "loss": 0.0226, + "step": 16515 + }, + { + "epoch": 13.949324324324325, + "grad_norm": 0.4623001217842102, + "learning_rate": 2.782966666666667e-05, + "loss": 0.017, + "step": 16516 + }, + { + "epoch": 13.95016891891892, + "grad_norm": 0.3449246287345886, + "learning_rate": 2.7829333333333335e-05, + "loss": 0.0271, + "step": 16517 + }, + { + "epoch": 13.951013513513514, + "grad_norm": 0.3724271357059479, + "learning_rate": 2.7829e-05, + "loss": 0.0228, + "step": 16518 + }, + { + "epoch": 13.951858108108109, + "grad_norm": 0.2711343765258789, + "learning_rate": 2.7828666666666667e-05, + "loss": 0.0277, + "step": 16519 + }, + { + "epoch": 13.952702702702704, + "grad_norm": 0.24176263809204102, + "learning_rate": 2.7828333333333336e-05, + "loss": 0.0136, + "step": 16520 + }, + { + "epoch": 13.953547297297296, + "grad_norm": 0.2151598036289215, + "learning_rate": 2.7828e-05, + "loss": 0.0142, + "step": 16521 + }, + { + "epoch": 13.954391891891891, + "grad_norm": 0.20020736753940582, + "learning_rate": 2.7827666666666668e-05, + "loss": 0.013, + "step": 16522 + }, + { + "epoch": 13.955236486486486, + "grad_norm": 0.19705788791179657, + "learning_rate": 2.7827333333333334e-05, + "loss": 0.0129, + "step": 16523 + }, + { + "epoch": 13.95608108108108, + "grad_norm": 0.7717267274856567, + "learning_rate": 2.7827e-05, + "loss": 0.0238, + "step": 16524 + }, + { + "epoch": 13.956925675675675, + "grad_norm": 0.32624056935310364, + "learning_rate": 2.782666666666667e-05, + "loss": 0.0221, + "step": 16525 + }, + { + "epoch": 13.95777027027027, + "grad_norm": 0.36193782091140747, + "learning_rate": 2.782633333333333e-05, + "loss": 0.074, + "step": 16526 + }, + { + "epoch": 13.958614864864865, + "grad_norm": 0.21093696355819702, + "learning_rate": 2.7826e-05, + "loss": 0.0149, + "step": 16527 + }, + { + "epoch": 13.95945945945946, + "grad_norm": 0.3671296238899231, + "learning_rate": 2.7825666666666667e-05, + "loss": 0.0214, + "step": 16528 + }, + { + "epoch": 13.960304054054054, + "grad_norm": 0.21283140778541565, + "learning_rate": 2.7825333333333333e-05, + "loss": 0.0164, + "step": 16529 + }, + { + "epoch": 13.96114864864865, + "grad_norm": 0.3292774558067322, + "learning_rate": 2.7825000000000002e-05, + "loss": 0.0176, + "step": 16530 + }, + { + "epoch": 13.961993243243244, + "grad_norm": 1.1141867637634277, + "learning_rate": 2.7824666666666668e-05, + "loss": 0.0221, + "step": 16531 + }, + { + "epoch": 13.962837837837839, + "grad_norm": 0.17885422706604004, + "learning_rate": 2.7824333333333334e-05, + "loss": 0.0123, + "step": 16532 + }, + { + "epoch": 13.963682432432432, + "grad_norm": 0.6158616542816162, + "learning_rate": 2.7824e-05, + "loss": 0.015, + "step": 16533 + }, + { + "epoch": 13.964527027027026, + "grad_norm": 0.25860223174095154, + "learning_rate": 2.782366666666667e-05, + "loss": 0.0124, + "step": 16534 + }, + { + "epoch": 13.965371621621621, + "grad_norm": 0.23255892097949982, + "learning_rate": 2.7823333333333332e-05, + "loss": 0.0115, + "step": 16535 + }, + { + "epoch": 13.966216216216216, + "grad_norm": 0.4588927626609802, + "learning_rate": 2.7823e-05, + "loss": 0.0273, + "step": 16536 + }, + { + "epoch": 13.96706081081081, + "grad_norm": 0.27588725090026855, + "learning_rate": 2.7822666666666667e-05, + "loss": 0.0131, + "step": 16537 + }, + { + "epoch": 13.967905405405405, + "grad_norm": 0.31615421175956726, + "learning_rate": 2.7822333333333333e-05, + "loss": 0.0134, + "step": 16538 + }, + { + "epoch": 13.96875, + "grad_norm": 0.43354812264442444, + "learning_rate": 2.7822000000000002e-05, + "loss": 0.0234, + "step": 16539 + }, + { + "epoch": 13.969594594594595, + "grad_norm": 0.2870362401008606, + "learning_rate": 2.7821666666666668e-05, + "loss": 0.0235, + "step": 16540 + }, + { + "epoch": 13.97043918918919, + "grad_norm": 1.4268897771835327, + "learning_rate": 2.7821333333333334e-05, + "loss": 0.1283, + "step": 16541 + }, + { + "epoch": 13.971283783783784, + "grad_norm": 0.9676016569137573, + "learning_rate": 2.7821e-05, + "loss": 0.0487, + "step": 16542 + }, + { + "epoch": 13.972128378378379, + "grad_norm": 0.4834692180156708, + "learning_rate": 2.782066666666667e-05, + "loss": 0.1572, + "step": 16543 + }, + { + "epoch": 13.972972972972974, + "grad_norm": 0.39899179339408875, + "learning_rate": 2.7820333333333332e-05, + "loss": 0.0548, + "step": 16544 + }, + { + "epoch": 13.973817567567568, + "grad_norm": 0.3752460777759552, + "learning_rate": 2.782e-05, + "loss": 0.0353, + "step": 16545 + }, + { + "epoch": 13.974662162162161, + "grad_norm": 0.24666963517665863, + "learning_rate": 2.7819666666666667e-05, + "loss": 0.0236, + "step": 16546 + }, + { + "epoch": 13.975506756756756, + "grad_norm": 0.306460976600647, + "learning_rate": 2.7819333333333333e-05, + "loss": 0.0346, + "step": 16547 + }, + { + "epoch": 13.97635135135135, + "grad_norm": 0.18432050943374634, + "learning_rate": 2.7819000000000002e-05, + "loss": 0.0158, + "step": 16548 + }, + { + "epoch": 13.977195945945946, + "grad_norm": 0.24342644214630127, + "learning_rate": 2.781866666666667e-05, + "loss": 0.0198, + "step": 16549 + }, + { + "epoch": 13.97804054054054, + "grad_norm": 0.3145102262496948, + "learning_rate": 2.7818333333333334e-05, + "loss": 0.0213, + "step": 16550 + }, + { + "epoch": 13.978885135135135, + "grad_norm": 0.4171375632286072, + "learning_rate": 2.7818e-05, + "loss": 0.0446, + "step": 16551 + }, + { + "epoch": 13.97972972972973, + "grad_norm": 0.19073984026908875, + "learning_rate": 2.781766666666667e-05, + "loss": 0.0149, + "step": 16552 + }, + { + "epoch": 13.980574324324325, + "grad_norm": 0.2205214947462082, + "learning_rate": 2.7817333333333332e-05, + "loss": 0.0146, + "step": 16553 + }, + { + "epoch": 13.98141891891892, + "grad_norm": 0.3280463218688965, + "learning_rate": 2.7817e-05, + "loss": 0.0194, + "step": 16554 + }, + { + "epoch": 13.982263513513514, + "grad_norm": 0.32288873195648193, + "learning_rate": 2.7816666666666667e-05, + "loss": 0.0221, + "step": 16555 + }, + { + "epoch": 13.983108108108109, + "grad_norm": 0.26199883222579956, + "learning_rate": 2.7816333333333333e-05, + "loss": 0.019, + "step": 16556 + }, + { + "epoch": 13.983952702702704, + "grad_norm": 0.37894511222839355, + "learning_rate": 2.7816000000000003e-05, + "loss": 0.03, + "step": 16557 + }, + { + "epoch": 13.984797297297296, + "grad_norm": 0.5325700044631958, + "learning_rate": 2.7815666666666665e-05, + "loss": 0.026, + "step": 16558 + }, + { + "epoch": 13.985641891891891, + "grad_norm": 0.31742313504219055, + "learning_rate": 2.7815333333333335e-05, + "loss": 0.0209, + "step": 16559 + }, + { + "epoch": 13.986486486486486, + "grad_norm": 0.2883126735687256, + "learning_rate": 2.7815e-05, + "loss": 0.0182, + "step": 16560 + }, + { + "epoch": 13.98733108108108, + "grad_norm": 0.24327126145362854, + "learning_rate": 2.7814666666666666e-05, + "loss": 0.0211, + "step": 16561 + }, + { + "epoch": 13.988175675675675, + "grad_norm": 0.5712132453918457, + "learning_rate": 2.7814333333333332e-05, + "loss": 0.0182, + "step": 16562 + }, + { + "epoch": 13.98902027027027, + "grad_norm": 0.2567342519760132, + "learning_rate": 2.7814000000000002e-05, + "loss": 0.0133, + "step": 16563 + }, + { + "epoch": 13.989864864864865, + "grad_norm": 0.19528712332248688, + "learning_rate": 2.7813666666666668e-05, + "loss": 0.0159, + "step": 16564 + }, + { + "epoch": 13.99070945945946, + "grad_norm": 0.2888202965259552, + "learning_rate": 2.7813333333333334e-05, + "loss": 0.0253, + "step": 16565 + }, + { + "epoch": 13.991554054054054, + "grad_norm": 0.3678304851055145, + "learning_rate": 2.7813000000000003e-05, + "loss": 0.0303, + "step": 16566 + }, + { + "epoch": 13.99239864864865, + "grad_norm": 0.34476205706596375, + "learning_rate": 2.7812666666666665e-05, + "loss": 0.0195, + "step": 16567 + }, + { + "epoch": 13.993243243243244, + "grad_norm": 0.15539239346981049, + "learning_rate": 2.7812333333333335e-05, + "loss": 0.0127, + "step": 16568 + }, + { + "epoch": 13.994087837837839, + "grad_norm": 0.6066041588783264, + "learning_rate": 2.7812e-05, + "loss": 0.0207, + "step": 16569 + }, + { + "epoch": 13.994932432432432, + "grad_norm": 0.41007959842681885, + "learning_rate": 2.7811666666666667e-05, + "loss": 0.0171, + "step": 16570 + }, + { + "epoch": 13.995777027027026, + "grad_norm": 0.33945539593696594, + "learning_rate": 2.7811333333333333e-05, + "loss": 0.018, + "step": 16571 + }, + { + "epoch": 13.996621621621621, + "grad_norm": 0.31185850501060486, + "learning_rate": 2.7811000000000002e-05, + "loss": 0.0263, + "step": 16572 + }, + { + "epoch": 13.997466216216216, + "grad_norm": 0.3344400227069855, + "learning_rate": 2.7810666666666668e-05, + "loss": 0.0212, + "step": 16573 + }, + { + "epoch": 13.99831081081081, + "grad_norm": 0.5941693186759949, + "learning_rate": 2.7810333333333334e-05, + "loss": 0.0339, + "step": 16574 + }, + { + "epoch": 13.999155405405405, + "grad_norm": 0.45331257581710815, + "learning_rate": 2.7810000000000003e-05, + "loss": 0.0223, + "step": 16575 + }, + { + "epoch": 14.0, + "grad_norm": 0.5235801935195923, + "learning_rate": 2.7809666666666666e-05, + "loss": 0.0268, + "step": 16576 + }, + { + "epoch": 14.000844594594595, + "grad_norm": 0.5366891026496887, + "learning_rate": 2.7809333333333335e-05, + "loss": 0.1553, + "step": 16577 + }, + { + "epoch": 14.00168918918919, + "grad_norm": 0.49941229820251465, + "learning_rate": 2.7809e-05, + "loss": 0.0764, + "step": 16578 + }, + { + "epoch": 14.002533783783784, + "grad_norm": 0.3976868987083435, + "learning_rate": 2.7808666666666667e-05, + "loss": 0.0389, + "step": 16579 + }, + { + "epoch": 14.003378378378379, + "grad_norm": 0.21835951507091522, + "learning_rate": 2.7808333333333336e-05, + "loss": 0.0243, + "step": 16580 + }, + { + "epoch": 14.004222972972974, + "grad_norm": 0.268140584230423, + "learning_rate": 2.7808000000000002e-05, + "loss": 0.0206, + "step": 16581 + }, + { + "epoch": 14.005067567567568, + "grad_norm": 0.5024497509002686, + "learning_rate": 2.7807666666666668e-05, + "loss": 0.0299, + "step": 16582 + }, + { + "epoch": 14.005912162162161, + "grad_norm": 0.2959395945072174, + "learning_rate": 2.7807333333333334e-05, + "loss": 0.0296, + "step": 16583 + }, + { + "epoch": 14.006756756756756, + "grad_norm": 0.2292509824037552, + "learning_rate": 2.7807e-05, + "loss": 0.0191, + "step": 16584 + }, + { + "epoch": 14.00760135135135, + "grad_norm": 0.3273006081581116, + "learning_rate": 2.7806666666666666e-05, + "loss": 0.0191, + "step": 16585 + }, + { + "epoch": 14.008445945945946, + "grad_norm": 0.21103116869926453, + "learning_rate": 2.7806333333333335e-05, + "loss": 0.0108, + "step": 16586 + }, + { + "epoch": 14.00929054054054, + "grad_norm": 0.19579359889030457, + "learning_rate": 2.7805999999999998e-05, + "loss": 0.0139, + "step": 16587 + }, + { + "epoch": 14.010135135135135, + "grad_norm": 0.2125360667705536, + "learning_rate": 2.7805666666666667e-05, + "loss": 0.0157, + "step": 16588 + }, + { + "epoch": 14.01097972972973, + "grad_norm": 0.2009190022945404, + "learning_rate": 2.7805333333333336e-05, + "loss": 0.0171, + "step": 16589 + }, + { + "epoch": 14.011824324324325, + "grad_norm": 0.7529754042625427, + "learning_rate": 2.7805e-05, + "loss": 0.0204, + "step": 16590 + }, + { + "epoch": 14.01266891891892, + "grad_norm": 0.36038923263549805, + "learning_rate": 2.7804666666666668e-05, + "loss": 0.0171, + "step": 16591 + }, + { + "epoch": 14.013513513513514, + "grad_norm": 0.29701897501945496, + "learning_rate": 2.7804333333333334e-05, + "loss": 0.0112, + "step": 16592 + }, + { + "epoch": 14.014358108108109, + "grad_norm": 0.3058828115463257, + "learning_rate": 2.7804e-05, + "loss": 0.0173, + "step": 16593 + }, + { + "epoch": 14.015202702702704, + "grad_norm": 0.24160751700401306, + "learning_rate": 2.7803666666666666e-05, + "loss": 0.0099, + "step": 16594 + }, + { + "epoch": 14.016047297297296, + "grad_norm": 0.11518795043230057, + "learning_rate": 2.7803333333333335e-05, + "loss": 0.01, + "step": 16595 + }, + { + "epoch": 14.016891891891891, + "grad_norm": 0.31528639793395996, + "learning_rate": 2.7802999999999998e-05, + "loss": 0.0174, + "step": 16596 + }, + { + "epoch": 14.017736486486486, + "grad_norm": 0.27631354331970215, + "learning_rate": 2.7802666666666667e-05, + "loss": 0.02, + "step": 16597 + }, + { + "epoch": 14.01858108108108, + "grad_norm": 0.3556373417377472, + "learning_rate": 2.7802333333333337e-05, + "loss": 0.0148, + "step": 16598 + }, + { + "epoch": 14.019425675675675, + "grad_norm": 0.20790855586528778, + "learning_rate": 2.7802e-05, + "loss": 0.0259, + "step": 16599 + }, + { + "epoch": 14.02027027027027, + "grad_norm": 0.28622812032699585, + "learning_rate": 2.780166666666667e-05, + "loss": 0.0224, + "step": 16600 + }, + { + "epoch": 14.021114864864865, + "grad_norm": 0.8461205363273621, + "learning_rate": 2.7801333333333334e-05, + "loss": 0.0362, + "step": 16601 + }, + { + "epoch": 14.02195945945946, + "grad_norm": 0.3069656789302826, + "learning_rate": 2.7801e-05, + "loss": 0.0183, + "step": 16602 + }, + { + "epoch": 14.022804054054054, + "grad_norm": 0.3357972204685211, + "learning_rate": 2.7800666666666666e-05, + "loss": 0.0195, + "step": 16603 + }, + { + "epoch": 14.02364864864865, + "grad_norm": 0.23826086521148682, + "learning_rate": 2.7800333333333336e-05, + "loss": 0.0163, + "step": 16604 + }, + { + "epoch": 14.024493243243244, + "grad_norm": 0.2114565223455429, + "learning_rate": 2.78e-05, + "loss": 0.0125, + "step": 16605 + }, + { + "epoch": 14.025337837837839, + "grad_norm": 0.15665000677108765, + "learning_rate": 2.7799666666666667e-05, + "loss": 0.0102, + "step": 16606 + }, + { + "epoch": 14.026182432432432, + "grad_norm": 0.22447890043258667, + "learning_rate": 2.7799333333333337e-05, + "loss": 0.0097, + "step": 16607 + }, + { + "epoch": 14.027027027027026, + "grad_norm": 0.24029377102851868, + "learning_rate": 2.7799e-05, + "loss": 0.0161, + "step": 16608 + }, + { + "epoch": 14.027871621621621, + "grad_norm": 0.2618573307991028, + "learning_rate": 2.779866666666667e-05, + "loss": 0.0148, + "step": 16609 + }, + { + "epoch": 14.028716216216216, + "grad_norm": 0.5456849932670593, + "learning_rate": 2.7798333333333335e-05, + "loss": 0.0226, + "step": 16610 + }, + { + "epoch": 14.02956081081081, + "grad_norm": 0.3822503387928009, + "learning_rate": 2.7798e-05, + "loss": 0.0251, + "step": 16611 + }, + { + "epoch": 14.030405405405405, + "grad_norm": 0.16187605261802673, + "learning_rate": 2.7797666666666666e-05, + "loss": 0.0132, + "step": 16612 + }, + { + "epoch": 14.03125, + "grad_norm": 0.1637832224369049, + "learning_rate": 2.7797333333333332e-05, + "loss": 0.0117, + "step": 16613 + }, + { + "epoch": 14.032094594594595, + "grad_norm": 0.1836278885602951, + "learning_rate": 2.7797e-05, + "loss": 0.0157, + "step": 16614 + }, + { + "epoch": 14.03293918918919, + "grad_norm": 0.8699482679367065, + "learning_rate": 2.7796666666666668e-05, + "loss": 0.0153, + "step": 16615 + }, + { + "epoch": 14.033783783783784, + "grad_norm": 0.15969789028167725, + "learning_rate": 2.7796333333333334e-05, + "loss": 0.0117, + "step": 16616 + }, + { + "epoch": 14.034628378378379, + "grad_norm": 0.29477524757385254, + "learning_rate": 2.7796e-05, + "loss": 0.0182, + "step": 16617 + }, + { + "epoch": 14.035472972972974, + "grad_norm": 0.29605454206466675, + "learning_rate": 2.779566666666667e-05, + "loss": 0.0132, + "step": 16618 + }, + { + "epoch": 14.036317567567568, + "grad_norm": 1.5214446783065796, + "learning_rate": 2.779533333333333e-05, + "loss": 0.0205, + "step": 16619 + }, + { + "epoch": 14.037162162162161, + "grad_norm": 0.2045123130083084, + "learning_rate": 2.7795e-05, + "loss": 0.0135, + "step": 16620 + }, + { + "epoch": 14.038006756756756, + "grad_norm": 0.20675189793109894, + "learning_rate": 2.7794666666666667e-05, + "loss": 0.01, + "step": 16621 + }, + { + "epoch": 14.03885135135135, + "grad_norm": 0.2522714138031006, + "learning_rate": 2.7794333333333333e-05, + "loss": 0.0146, + "step": 16622 + }, + { + "epoch": 14.039695945945946, + "grad_norm": 0.4750326871871948, + "learning_rate": 2.7794000000000002e-05, + "loss": 0.0217, + "step": 16623 + }, + { + "epoch": 14.04054054054054, + "grad_norm": 0.30142927169799805, + "learning_rate": 2.7793666666666668e-05, + "loss": 0.022, + "step": 16624 + }, + { + "epoch": 14.041385135135135, + "grad_norm": 1.1280089616775513, + "learning_rate": 2.7793333333333334e-05, + "loss": 0.0224, + "step": 16625 + }, + { + "epoch": 14.04222972972973, + "grad_norm": 0.6369203329086304, + "learning_rate": 2.7793e-05, + "loss": 0.0553, + "step": 16626 + }, + { + "epoch": 14.043074324324325, + "grad_norm": 0.46994125843048096, + "learning_rate": 2.779266666666667e-05, + "loss": 0.1311, + "step": 16627 + }, + { + "epoch": 14.04391891891892, + "grad_norm": 0.44133055210113525, + "learning_rate": 2.779233333333333e-05, + "loss": 0.1107, + "step": 16628 + }, + { + "epoch": 14.044763513513514, + "grad_norm": 0.27017655968666077, + "learning_rate": 2.7792e-05, + "loss": 0.0456, + "step": 16629 + }, + { + "epoch": 14.045608108108109, + "grad_norm": 0.3060626685619354, + "learning_rate": 2.779166666666667e-05, + "loss": 0.0258, + "step": 16630 + }, + { + "epoch": 14.046452702702704, + "grad_norm": 0.5677992701530457, + "learning_rate": 2.7791333333333333e-05, + "loss": 0.0309, + "step": 16631 + }, + { + "epoch": 14.047297297297296, + "grad_norm": 0.2097408026456833, + "learning_rate": 2.7791000000000002e-05, + "loss": 0.0184, + "step": 16632 + }, + { + "epoch": 14.048141891891891, + "grad_norm": 0.34846270084381104, + "learning_rate": 2.7790666666666668e-05, + "loss": 0.0351, + "step": 16633 + }, + { + "epoch": 14.048986486486486, + "grad_norm": 0.23156698048114777, + "learning_rate": 2.7790333333333334e-05, + "loss": 0.0133, + "step": 16634 + }, + { + "epoch": 14.04983108108108, + "grad_norm": 0.22609320282936096, + "learning_rate": 2.779e-05, + "loss": 0.0207, + "step": 16635 + }, + { + "epoch": 14.050675675675675, + "grad_norm": 0.182684987783432, + "learning_rate": 2.778966666666667e-05, + "loss": 0.0154, + "step": 16636 + }, + { + "epoch": 14.05152027027027, + "grad_norm": 0.30837586522102356, + "learning_rate": 2.7789333333333332e-05, + "loss": 0.0211, + "step": 16637 + }, + { + "epoch": 14.052364864864865, + "grad_norm": 0.26348963379859924, + "learning_rate": 2.7789e-05, + "loss": 0.0126, + "step": 16638 + }, + { + "epoch": 14.05320945945946, + "grad_norm": 0.16325809061527252, + "learning_rate": 2.778866666666667e-05, + "loss": 0.0129, + "step": 16639 + }, + { + "epoch": 14.054054054054054, + "grad_norm": 0.20587468147277832, + "learning_rate": 2.7788333333333333e-05, + "loss": 0.0115, + "step": 16640 + }, + { + "epoch": 14.05489864864865, + "grad_norm": 0.2812146544456482, + "learning_rate": 2.7788000000000002e-05, + "loss": 0.0229, + "step": 16641 + }, + { + "epoch": 14.055743243243244, + "grad_norm": 0.22080373764038086, + "learning_rate": 2.7787666666666668e-05, + "loss": 0.022, + "step": 16642 + }, + { + "epoch": 14.056587837837839, + "grad_norm": 0.15266399085521698, + "learning_rate": 2.7787333333333334e-05, + "loss": 0.0103, + "step": 16643 + }, + { + "epoch": 14.057432432432432, + "grad_norm": 0.35341396927833557, + "learning_rate": 2.7787e-05, + "loss": 0.0271, + "step": 16644 + }, + { + "epoch": 14.058277027027026, + "grad_norm": 0.2734980285167694, + "learning_rate": 2.7786666666666666e-05, + "loss": 0.0152, + "step": 16645 + }, + { + "epoch": 14.059121621621621, + "grad_norm": 0.1703265905380249, + "learning_rate": 2.7786333333333332e-05, + "loss": 0.0168, + "step": 16646 + }, + { + "epoch": 14.059966216216216, + "grad_norm": 0.3028947412967682, + "learning_rate": 2.7786e-05, + "loss": 0.0169, + "step": 16647 + }, + { + "epoch": 14.06081081081081, + "grad_norm": 0.2914745509624481, + "learning_rate": 2.7785666666666667e-05, + "loss": 0.0163, + "step": 16648 + }, + { + "epoch": 14.061655405405405, + "grad_norm": 0.1443714052438736, + "learning_rate": 2.7785333333333333e-05, + "loss": 0.0068, + "step": 16649 + }, + { + "epoch": 14.0625, + "grad_norm": 0.24004094302654266, + "learning_rate": 2.7785000000000002e-05, + "loss": 0.0118, + "step": 16650 + }, + { + "epoch": 14.063344594594595, + "grad_norm": 0.26096153259277344, + "learning_rate": 2.7784666666666665e-05, + "loss": 0.0189, + "step": 16651 + }, + { + "epoch": 14.06418918918919, + "grad_norm": 0.4273219704627991, + "learning_rate": 2.7784333333333334e-05, + "loss": 0.0199, + "step": 16652 + }, + { + "epoch": 14.065033783783784, + "grad_norm": 0.15072378516197205, + "learning_rate": 2.7784e-05, + "loss": 0.0094, + "step": 16653 + }, + { + "epoch": 14.065878378378379, + "grad_norm": 0.34429222345352173, + "learning_rate": 2.7783666666666666e-05, + "loss": 0.0203, + "step": 16654 + }, + { + "epoch": 14.066722972972974, + "grad_norm": 0.18291044235229492, + "learning_rate": 2.7783333333333336e-05, + "loss": 0.0117, + "step": 16655 + }, + { + "epoch": 14.067567567567568, + "grad_norm": 1.573058009147644, + "learning_rate": 2.7783e-05, + "loss": 0.0232, + "step": 16656 + }, + { + "epoch": 14.068412162162161, + "grad_norm": 0.6787816882133484, + "learning_rate": 2.7782666666666667e-05, + "loss": 0.0186, + "step": 16657 + }, + { + "epoch": 14.069256756756756, + "grad_norm": 0.5020928382873535, + "learning_rate": 2.7782333333333333e-05, + "loss": 0.0308, + "step": 16658 + }, + { + "epoch": 14.07010135135135, + "grad_norm": 0.2697597146034241, + "learning_rate": 2.7782000000000003e-05, + "loss": 0.0145, + "step": 16659 + }, + { + "epoch": 14.070945945945946, + "grad_norm": 0.35033607482910156, + "learning_rate": 2.7781666666666665e-05, + "loss": 0.0196, + "step": 16660 + }, + { + "epoch": 14.07179054054054, + "grad_norm": 0.6853422522544861, + "learning_rate": 2.7781333333333335e-05, + "loss": 0.0249, + "step": 16661 + }, + { + "epoch": 14.072635135135135, + "grad_norm": 0.16235558688640594, + "learning_rate": 2.7781e-05, + "loss": 0.0189, + "step": 16662 + }, + { + "epoch": 14.07347972972973, + "grad_norm": 0.3711273670196533, + "learning_rate": 2.7780666666666666e-05, + "loss": 0.0191, + "step": 16663 + }, + { + "epoch": 14.074324324324325, + "grad_norm": 0.31115812063217163, + "learning_rate": 2.7780333333333336e-05, + "loss": 0.0165, + "step": 16664 + }, + { + "epoch": 14.07516891891892, + "grad_norm": 0.28142446279525757, + "learning_rate": 2.778e-05, + "loss": 0.0102, + "step": 16665 + }, + { + "epoch": 14.076013513513514, + "grad_norm": 0.28771138191223145, + "learning_rate": 2.7779666666666668e-05, + "loss": 0.0169, + "step": 16666 + }, + { + "epoch": 14.076858108108109, + "grad_norm": 0.315487265586853, + "learning_rate": 2.7779333333333334e-05, + "loss": 0.0168, + "step": 16667 + }, + { + "epoch": 14.077702702702704, + "grad_norm": 0.32767927646636963, + "learning_rate": 2.7779000000000003e-05, + "loss": 0.0181, + "step": 16668 + }, + { + "epoch": 14.078547297297296, + "grad_norm": 0.22785501182079315, + "learning_rate": 2.7778666666666665e-05, + "loss": 0.017, + "step": 16669 + }, + { + "epoch": 14.079391891891891, + "grad_norm": 0.22429530322551727, + "learning_rate": 2.7778333333333335e-05, + "loss": 0.0096, + "step": 16670 + }, + { + "epoch": 14.080236486486486, + "grad_norm": 0.2499316781759262, + "learning_rate": 2.7778e-05, + "loss": 0.0117, + "step": 16671 + }, + { + "epoch": 14.08108108108108, + "grad_norm": 1.0205352306365967, + "learning_rate": 2.7777666666666667e-05, + "loss": 0.0338, + "step": 16672 + }, + { + "epoch": 14.081925675675675, + "grad_norm": 0.28016915917396545, + "learning_rate": 2.7777333333333336e-05, + "loss": 0.0157, + "step": 16673 + }, + { + "epoch": 14.08277027027027, + "grad_norm": 0.5825987458229065, + "learning_rate": 2.7777e-05, + "loss": 0.0187, + "step": 16674 + }, + { + "epoch": 14.083614864864865, + "grad_norm": 1.818872332572937, + "learning_rate": 2.7776666666666668e-05, + "loss": 0.0645, + "step": 16675 + }, + { + "epoch": 14.08445945945946, + "grad_norm": 0.729606032371521, + "learning_rate": 2.7776333333333334e-05, + "loss": 0.033, + "step": 16676 + }, + { + "epoch": 14.085304054054054, + "grad_norm": 0.6142212748527527, + "learning_rate": 2.7776e-05, + "loss": 0.1832, + "step": 16677 + }, + { + "epoch": 14.08614864864865, + "grad_norm": 0.5493701696395874, + "learning_rate": 2.7775666666666666e-05, + "loss": 0.0893, + "step": 16678 + }, + { + "epoch": 14.086993243243244, + "grad_norm": 0.33987778425216675, + "learning_rate": 2.7775333333333335e-05, + "loss": 0.0498, + "step": 16679 + }, + { + "epoch": 14.087837837837839, + "grad_norm": 0.34479251503944397, + "learning_rate": 2.7775e-05, + "loss": 0.0418, + "step": 16680 + }, + { + "epoch": 14.088682432432432, + "grad_norm": 0.43314048647880554, + "learning_rate": 2.7774666666666667e-05, + "loss": 0.0429, + "step": 16681 + }, + { + "epoch": 14.089527027027026, + "grad_norm": 0.33604779839515686, + "learning_rate": 2.7774333333333336e-05, + "loss": 0.0284, + "step": 16682 + }, + { + "epoch": 14.090371621621621, + "grad_norm": 0.38032063841819763, + "learning_rate": 2.7774e-05, + "loss": 0.0241, + "step": 16683 + }, + { + "epoch": 14.091216216216216, + "grad_norm": 0.3343351483345032, + "learning_rate": 2.7773666666666668e-05, + "loss": 0.019, + "step": 16684 + }, + { + "epoch": 14.09206081081081, + "grad_norm": 0.48857465386390686, + "learning_rate": 2.7773333333333334e-05, + "loss": 0.0451, + "step": 16685 + }, + { + "epoch": 14.092905405405405, + "grad_norm": 0.28926122188568115, + "learning_rate": 2.7773e-05, + "loss": 0.0206, + "step": 16686 + }, + { + "epoch": 14.09375, + "grad_norm": 0.2400895208120346, + "learning_rate": 2.7772666666666666e-05, + "loss": 0.0125, + "step": 16687 + }, + { + "epoch": 14.094594594594595, + "grad_norm": 0.1832740157842636, + "learning_rate": 2.7772333333333335e-05, + "loss": 0.0183, + "step": 16688 + }, + { + "epoch": 14.09543918918919, + "grad_norm": 0.18697093427181244, + "learning_rate": 2.7772e-05, + "loss": 0.014, + "step": 16689 + }, + { + "epoch": 14.096283783783784, + "grad_norm": 0.2382945567369461, + "learning_rate": 2.7771666666666667e-05, + "loss": 0.0301, + "step": 16690 + }, + { + "epoch": 14.097128378378379, + "grad_norm": 0.2995310425758362, + "learning_rate": 2.7771333333333336e-05, + "loss": 0.0246, + "step": 16691 + }, + { + "epoch": 14.097972972972974, + "grad_norm": 0.5400676131248474, + "learning_rate": 2.7771e-05, + "loss": 0.0183, + "step": 16692 + }, + { + "epoch": 14.098817567567568, + "grad_norm": 0.19638684391975403, + "learning_rate": 2.7770666666666668e-05, + "loss": 0.0189, + "step": 16693 + }, + { + "epoch": 14.099662162162161, + "grad_norm": 0.2542787492275238, + "learning_rate": 2.7770333333333334e-05, + "loss": 0.0185, + "step": 16694 + }, + { + "epoch": 14.100506756756756, + "grad_norm": 0.4941234290599823, + "learning_rate": 2.777e-05, + "loss": 0.0176, + "step": 16695 + }, + { + "epoch": 14.10135135135135, + "grad_norm": 0.2207653522491455, + "learning_rate": 2.7769666666666666e-05, + "loss": 0.0155, + "step": 16696 + }, + { + "epoch": 14.102195945945946, + "grad_norm": 0.24570409953594208, + "learning_rate": 2.7769333333333335e-05, + "loss": 0.0187, + "step": 16697 + }, + { + "epoch": 14.10304054054054, + "grad_norm": 0.21265827119350433, + "learning_rate": 2.7769e-05, + "loss": 0.0136, + "step": 16698 + }, + { + "epoch": 14.103885135135135, + "grad_norm": 0.2628410756587982, + "learning_rate": 2.7768666666666667e-05, + "loss": 0.0105, + "step": 16699 + }, + { + "epoch": 14.10472972972973, + "grad_norm": 0.2540550231933594, + "learning_rate": 2.7768333333333337e-05, + "loss": 0.0128, + "step": 16700 + }, + { + "epoch": 14.105574324324325, + "grad_norm": 0.30402159690856934, + "learning_rate": 2.7768e-05, + "loss": 0.015, + "step": 16701 + }, + { + "epoch": 14.10641891891892, + "grad_norm": 0.3181631863117218, + "learning_rate": 2.776766666666667e-05, + "loss": 0.0252, + "step": 16702 + }, + { + "epoch": 14.107263513513514, + "grad_norm": 0.37911590933799744, + "learning_rate": 2.7767333333333334e-05, + "loss": 0.0156, + "step": 16703 + }, + { + "epoch": 14.108108108108109, + "grad_norm": 0.160770446062088, + "learning_rate": 2.7767e-05, + "loss": 0.0105, + "step": 16704 + }, + { + "epoch": 14.108952702702704, + "grad_norm": 0.43262213468551636, + "learning_rate": 2.776666666666667e-05, + "loss": 0.0166, + "step": 16705 + }, + { + "epoch": 14.109797297297296, + "grad_norm": 0.24170057475566864, + "learning_rate": 2.7766333333333332e-05, + "loss": 0.0138, + "step": 16706 + }, + { + "epoch": 14.110641891891891, + "grad_norm": 0.1840374618768692, + "learning_rate": 2.7766e-05, + "loss": 0.0194, + "step": 16707 + }, + { + "epoch": 14.111486486486486, + "grad_norm": 0.255342572927475, + "learning_rate": 2.7765666666666667e-05, + "loss": 0.0136, + "step": 16708 + }, + { + "epoch": 14.11233108108108, + "grad_norm": 0.264224648475647, + "learning_rate": 2.7765333333333333e-05, + "loss": 0.0193, + "step": 16709 + }, + { + "epoch": 14.113175675675675, + "grad_norm": 0.27180400490760803, + "learning_rate": 2.7765e-05, + "loss": 0.0104, + "step": 16710 + }, + { + "epoch": 14.11402027027027, + "grad_norm": 0.9001285433769226, + "learning_rate": 2.776466666666667e-05, + "loss": 0.0171, + "step": 16711 + }, + { + "epoch": 14.114864864864865, + "grad_norm": 0.3278326094150543, + "learning_rate": 2.776433333333333e-05, + "loss": 0.0129, + "step": 16712 + }, + { + "epoch": 14.11570945945946, + "grad_norm": 0.2035312056541443, + "learning_rate": 2.7764e-05, + "loss": 0.0125, + "step": 16713 + }, + { + "epoch": 14.116554054054054, + "grad_norm": 0.3166108727455139, + "learning_rate": 2.776366666666667e-05, + "loss": 0.0249, + "step": 16714 + }, + { + "epoch": 14.11739864864865, + "grad_norm": 0.19582657516002655, + "learning_rate": 2.7763333333333332e-05, + "loss": 0.0085, + "step": 16715 + }, + { + "epoch": 14.118243243243244, + "grad_norm": 0.2785060703754425, + "learning_rate": 2.7763e-05, + "loss": 0.015, + "step": 16716 + }, + { + "epoch": 14.119087837837839, + "grad_norm": 0.3645496666431427, + "learning_rate": 2.7762666666666668e-05, + "loss": 0.0136, + "step": 16717 + }, + { + "epoch": 14.119932432432432, + "grad_norm": 0.3948225975036621, + "learning_rate": 2.7762333333333334e-05, + "loss": 0.0151, + "step": 16718 + }, + { + "epoch": 14.120777027027026, + "grad_norm": 2.206538200378418, + "learning_rate": 2.7762e-05, + "loss": 0.0245, + "step": 16719 + }, + { + "epoch": 14.121621621621621, + "grad_norm": 0.2514978349208832, + "learning_rate": 2.776166666666667e-05, + "loss": 0.018, + "step": 16720 + }, + { + "epoch": 14.122466216216216, + "grad_norm": 0.39817458391189575, + "learning_rate": 2.776133333333333e-05, + "loss": 0.0192, + "step": 16721 + }, + { + "epoch": 14.12331081081081, + "grad_norm": 0.3590584397315979, + "learning_rate": 2.7761e-05, + "loss": 0.0146, + "step": 16722 + }, + { + "epoch": 14.124155405405405, + "grad_norm": 0.6634851098060608, + "learning_rate": 2.776066666666667e-05, + "loss": 0.0268, + "step": 16723 + }, + { + "epoch": 14.125, + "grad_norm": 0.5031072497367859, + "learning_rate": 2.7760333333333333e-05, + "loss": 0.0221, + "step": 16724 + }, + { + "epoch": 14.125844594594595, + "grad_norm": 0.36605116724967957, + "learning_rate": 2.7760000000000002e-05, + "loss": 0.0129, + "step": 16725 + }, + { + "epoch": 14.12668918918919, + "grad_norm": 0.7100427746772766, + "learning_rate": 2.7759666666666668e-05, + "loss": 0.0275, + "step": 16726 + }, + { + "epoch": 14.127533783783784, + "grad_norm": 0.49093765020370483, + "learning_rate": 2.7759333333333334e-05, + "loss": 0.1544, + "step": 16727 + }, + { + "epoch": 14.128378378378379, + "grad_norm": 0.5944860577583313, + "learning_rate": 2.7759e-05, + "loss": 0.1063, + "step": 16728 + }, + { + "epoch": 14.129222972972974, + "grad_norm": 0.6154860258102417, + "learning_rate": 2.775866666666667e-05, + "loss": 0.0805, + "step": 16729 + }, + { + "epoch": 14.130067567567568, + "grad_norm": 0.45393800735473633, + "learning_rate": 2.7758333333333335e-05, + "loss": 0.0497, + "step": 16730 + }, + { + "epoch": 14.130912162162161, + "grad_norm": 0.1528303176164627, + "learning_rate": 2.7758e-05, + "loss": 0.0181, + "step": 16731 + }, + { + "epoch": 14.131756756756756, + "grad_norm": 0.5125111937522888, + "learning_rate": 2.775766666666667e-05, + "loss": 0.0278, + "step": 16732 + }, + { + "epoch": 14.13260135135135, + "grad_norm": 0.22212572395801544, + "learning_rate": 2.7757333333333333e-05, + "loss": 0.0147, + "step": 16733 + }, + { + "epoch": 14.133445945945946, + "grad_norm": 0.2953188419342041, + "learning_rate": 2.7757000000000002e-05, + "loss": 0.0241, + "step": 16734 + }, + { + "epoch": 14.13429054054054, + "grad_norm": 0.2660309672355652, + "learning_rate": 2.7756666666666665e-05, + "loss": 0.0173, + "step": 16735 + }, + { + "epoch": 14.135135135135135, + "grad_norm": 0.19204068183898926, + "learning_rate": 2.7756333333333334e-05, + "loss": 0.0144, + "step": 16736 + }, + { + "epoch": 14.13597972972973, + "grad_norm": 0.3544107675552368, + "learning_rate": 2.7756e-05, + "loss": 0.0273, + "step": 16737 + }, + { + "epoch": 14.136824324324325, + "grad_norm": 0.27764904499053955, + "learning_rate": 2.7755666666666666e-05, + "loss": 0.0277, + "step": 16738 + }, + { + "epoch": 14.13766891891892, + "grad_norm": 0.41592055559158325, + "learning_rate": 2.7755333333333335e-05, + "loss": 0.029, + "step": 16739 + }, + { + "epoch": 14.138513513513514, + "grad_norm": 0.18703027069568634, + "learning_rate": 2.7755e-05, + "loss": 0.0138, + "step": 16740 + }, + { + "epoch": 14.139358108108109, + "grad_norm": 0.21569092571735382, + "learning_rate": 2.7754666666666667e-05, + "loss": 0.0199, + "step": 16741 + }, + { + "epoch": 14.140202702702704, + "grad_norm": 0.16648276150226593, + "learning_rate": 2.7754333333333333e-05, + "loss": 0.0107, + "step": 16742 + }, + { + "epoch": 14.141047297297296, + "grad_norm": 0.15555638074874878, + "learning_rate": 2.7754000000000002e-05, + "loss": 0.0156, + "step": 16743 + }, + { + "epoch": 14.141891891891891, + "grad_norm": 0.2574484646320343, + "learning_rate": 2.7753666666666665e-05, + "loss": 0.0193, + "step": 16744 + }, + { + "epoch": 14.142736486486486, + "grad_norm": 0.23271608352661133, + "learning_rate": 2.7753333333333334e-05, + "loss": 0.0161, + "step": 16745 + }, + { + "epoch": 14.14358108108108, + "grad_norm": 0.30382490158081055, + "learning_rate": 2.7753e-05, + "loss": 0.0169, + "step": 16746 + }, + { + "epoch": 14.144425675675675, + "grad_norm": 0.20635327696800232, + "learning_rate": 2.7752666666666666e-05, + "loss": 0.0168, + "step": 16747 + }, + { + "epoch": 14.14527027027027, + "grad_norm": 0.2729405462741852, + "learning_rate": 2.7752333333333335e-05, + "loss": 0.0137, + "step": 16748 + }, + { + "epoch": 14.146114864864865, + "grad_norm": 0.359629362821579, + "learning_rate": 2.7752e-05, + "loss": 0.0176, + "step": 16749 + }, + { + "epoch": 14.14695945945946, + "grad_norm": 0.15744474530220032, + "learning_rate": 2.7751666666666667e-05, + "loss": 0.0082, + "step": 16750 + }, + { + "epoch": 14.147804054054054, + "grad_norm": 0.1990240067243576, + "learning_rate": 2.7751333333333333e-05, + "loss": 0.0172, + "step": 16751 + }, + { + "epoch": 14.14864864864865, + "grad_norm": 0.28586748242378235, + "learning_rate": 2.7751000000000002e-05, + "loss": 0.0109, + "step": 16752 + }, + { + "epoch": 14.149493243243244, + "grad_norm": 0.38503947854042053, + "learning_rate": 2.7750666666666665e-05, + "loss": 0.0193, + "step": 16753 + }, + { + "epoch": 14.150337837837839, + "grad_norm": 0.3252883851528168, + "learning_rate": 2.7750333333333334e-05, + "loss": 0.0232, + "step": 16754 + }, + { + "epoch": 14.151182432432432, + "grad_norm": 0.33495888113975525, + "learning_rate": 2.7750000000000004e-05, + "loss": 0.0159, + "step": 16755 + }, + { + "epoch": 14.152027027027026, + "grad_norm": 0.2847669720649719, + "learning_rate": 2.7749666666666666e-05, + "loss": 0.0131, + "step": 16756 + }, + { + "epoch": 14.152871621621621, + "grad_norm": 0.17127002775669098, + "learning_rate": 2.7749333333333336e-05, + "loss": 0.0121, + "step": 16757 + }, + { + "epoch": 14.153716216216216, + "grad_norm": 0.25911274552345276, + "learning_rate": 2.7749e-05, + "loss": 0.0126, + "step": 16758 + }, + { + "epoch": 14.15456081081081, + "grad_norm": 0.22811023890972137, + "learning_rate": 2.7748666666666667e-05, + "loss": 0.0149, + "step": 16759 + }, + { + "epoch": 14.155405405405405, + "grad_norm": 0.6857486367225647, + "learning_rate": 2.7748333333333333e-05, + "loss": 0.0156, + "step": 16760 + }, + { + "epoch": 14.15625, + "grad_norm": 0.21591660380363464, + "learning_rate": 2.7748000000000003e-05, + "loss": 0.0093, + "step": 16761 + }, + { + "epoch": 14.157094594594595, + "grad_norm": 0.23406249284744263, + "learning_rate": 2.7747666666666665e-05, + "loss": 0.0135, + "step": 16762 + }, + { + "epoch": 14.15793918918919, + "grad_norm": 0.10288596153259277, + "learning_rate": 2.7747333333333335e-05, + "loss": 0.0081, + "step": 16763 + }, + { + "epoch": 14.158783783783784, + "grad_norm": 0.522186279296875, + "learning_rate": 2.7747000000000004e-05, + "loss": 0.1026, + "step": 16764 + }, + { + "epoch": 14.159628378378379, + "grad_norm": 1.2612570524215698, + "learning_rate": 2.7746666666666666e-05, + "loss": 0.0166, + "step": 16765 + }, + { + "epoch": 14.160472972972974, + "grad_norm": 0.5079571604728699, + "learning_rate": 2.7746333333333336e-05, + "loss": 0.0141, + "step": 16766 + }, + { + "epoch": 14.161317567567568, + "grad_norm": 1.067360520362854, + "learning_rate": 2.7745999999999998e-05, + "loss": 0.0155, + "step": 16767 + }, + { + "epoch": 14.162162162162161, + "grad_norm": 2.2404227256774902, + "learning_rate": 2.7745666666666668e-05, + "loss": 0.0285, + "step": 16768 + }, + { + "epoch": 14.163006756756756, + "grad_norm": 0.37463125586509705, + "learning_rate": 2.7745333333333334e-05, + "loss": 0.018, + "step": 16769 + }, + { + "epoch": 14.16385135135135, + "grad_norm": 0.2566549777984619, + "learning_rate": 2.7745e-05, + "loss": 0.014, + "step": 16770 + }, + { + "epoch": 14.164695945945946, + "grad_norm": 0.23287859559059143, + "learning_rate": 2.7744666666666665e-05, + "loss": 0.0163, + "step": 16771 + }, + { + "epoch": 14.16554054054054, + "grad_norm": 0.3567855656147003, + "learning_rate": 2.7744333333333335e-05, + "loss": 0.0176, + "step": 16772 + }, + { + "epoch": 14.166385135135135, + "grad_norm": 0.488803893327713, + "learning_rate": 2.7744e-05, + "loss": 0.0163, + "step": 16773 + }, + { + "epoch": 14.16722972972973, + "grad_norm": 0.8537649512290955, + "learning_rate": 2.7743666666666667e-05, + "loss": 0.0181, + "step": 16774 + }, + { + "epoch": 14.168074324324325, + "grad_norm": 0.7348847389221191, + "learning_rate": 2.7743333333333336e-05, + "loss": 0.0233, + "step": 16775 + }, + { + "epoch": 14.16891891891892, + "grad_norm": 0.5997647643089294, + "learning_rate": 2.7743e-05, + "loss": 0.0267, + "step": 16776 + }, + { + "epoch": 14.169763513513514, + "grad_norm": 0.5091172456741333, + "learning_rate": 2.7742666666666668e-05, + "loss": 0.153, + "step": 16777 + }, + { + "epoch": 14.170608108108109, + "grad_norm": 0.9664973616600037, + "learning_rate": 2.7742333333333334e-05, + "loss": 0.088, + "step": 16778 + }, + { + "epoch": 14.171452702702704, + "grad_norm": 0.39073240756988525, + "learning_rate": 2.7742e-05, + "loss": 0.0635, + "step": 16779 + }, + { + "epoch": 14.172297297297296, + "grad_norm": 0.325545996427536, + "learning_rate": 2.774166666666667e-05, + "loss": 0.0428, + "step": 16780 + }, + { + "epoch": 14.173141891891891, + "grad_norm": 0.18963447213172913, + "learning_rate": 2.7741333333333335e-05, + "loss": 0.0207, + "step": 16781 + }, + { + "epoch": 14.173986486486486, + "grad_norm": 0.24108624458312988, + "learning_rate": 2.7741e-05, + "loss": 0.0157, + "step": 16782 + }, + { + "epoch": 14.17483108108108, + "grad_norm": 0.18530705571174622, + "learning_rate": 2.7740666666666667e-05, + "loss": 0.0136, + "step": 16783 + }, + { + "epoch": 14.175675675675675, + "grad_norm": 0.17553392052650452, + "learning_rate": 2.7740333333333336e-05, + "loss": 0.0116, + "step": 16784 + }, + { + "epoch": 14.17652027027027, + "grad_norm": 0.21965496242046356, + "learning_rate": 2.774e-05, + "loss": 0.0103, + "step": 16785 + }, + { + "epoch": 14.177364864864865, + "grad_norm": 0.24858543276786804, + "learning_rate": 2.7739666666666668e-05, + "loss": 0.0154, + "step": 16786 + }, + { + "epoch": 14.17820945945946, + "grad_norm": 0.228207528591156, + "learning_rate": 2.7739333333333334e-05, + "loss": 0.0122, + "step": 16787 + }, + { + "epoch": 14.179054054054054, + "grad_norm": 0.24676834046840668, + "learning_rate": 2.7739e-05, + "loss": 0.0179, + "step": 16788 + }, + { + "epoch": 14.17989864864865, + "grad_norm": 0.5641473531723022, + "learning_rate": 2.773866666666667e-05, + "loss": 0.0186, + "step": 16789 + }, + { + "epoch": 14.180743243243244, + "grad_norm": 0.1932879537343979, + "learning_rate": 2.7738333333333335e-05, + "loss": 0.0229, + "step": 16790 + }, + { + "epoch": 14.181587837837839, + "grad_norm": 0.22104015946388245, + "learning_rate": 2.7738e-05, + "loss": 0.0139, + "step": 16791 + }, + { + "epoch": 14.182432432432432, + "grad_norm": 0.4572620987892151, + "learning_rate": 2.7737666666666667e-05, + "loss": 0.017, + "step": 16792 + }, + { + "epoch": 14.183277027027026, + "grad_norm": 0.3665553629398346, + "learning_rate": 2.7737333333333336e-05, + "loss": 0.0247, + "step": 16793 + }, + { + "epoch": 14.184121621621621, + "grad_norm": 0.8636660575866699, + "learning_rate": 2.7737e-05, + "loss": 0.0201, + "step": 16794 + }, + { + "epoch": 14.184966216216216, + "grad_norm": 0.23400337994098663, + "learning_rate": 2.7736666666666668e-05, + "loss": 0.0185, + "step": 16795 + }, + { + "epoch": 14.18581081081081, + "grad_norm": 0.24582120776176453, + "learning_rate": 2.773633333333333e-05, + "loss": 0.0171, + "step": 16796 + }, + { + "epoch": 14.186655405405405, + "grad_norm": 0.2261250764131546, + "learning_rate": 2.7736e-05, + "loss": 0.011, + "step": 16797 + }, + { + "epoch": 14.1875, + "grad_norm": 0.2084290087223053, + "learning_rate": 2.773566666666667e-05, + "loss": 0.0141, + "step": 16798 + }, + { + "epoch": 14.188344594594595, + "grad_norm": 0.17812137305736542, + "learning_rate": 2.7735333333333332e-05, + "loss": 0.0162, + "step": 16799 + }, + { + "epoch": 14.18918918918919, + "grad_norm": 0.2667568027973175, + "learning_rate": 2.7735e-05, + "loss": 0.0147, + "step": 16800 + }, + { + "epoch": 14.190033783783784, + "grad_norm": 0.24301695823669434, + "learning_rate": 2.7734666666666667e-05, + "loss": 0.0243, + "step": 16801 + }, + { + "epoch": 14.190878378378379, + "grad_norm": 0.6726135611534119, + "learning_rate": 2.7734333333333333e-05, + "loss": 0.0163, + "step": 16802 + }, + { + "epoch": 14.191722972972974, + "grad_norm": 0.34160947799682617, + "learning_rate": 2.7734e-05, + "loss": 0.0283, + "step": 16803 + }, + { + "epoch": 14.192567567567568, + "grad_norm": 0.22049106657505035, + "learning_rate": 2.773366666666667e-05, + "loss": 0.0149, + "step": 16804 + }, + { + "epoch": 14.193412162162161, + "grad_norm": 0.5054778456687927, + "learning_rate": 2.7733333333333334e-05, + "loss": 0.0148, + "step": 16805 + }, + { + "epoch": 14.194256756756756, + "grad_norm": 0.285353422164917, + "learning_rate": 2.7733e-05, + "loss": 0.0181, + "step": 16806 + }, + { + "epoch": 14.19510135135135, + "grad_norm": 0.3409714996814728, + "learning_rate": 2.773266666666667e-05, + "loss": 0.0205, + "step": 16807 + }, + { + "epoch": 14.195945945945946, + "grad_norm": 0.5992918610572815, + "learning_rate": 2.7732333333333332e-05, + "loss": 0.018, + "step": 16808 + }, + { + "epoch": 14.19679054054054, + "grad_norm": 0.5573692917823792, + "learning_rate": 2.7732e-05, + "loss": 0.0265, + "step": 16809 + }, + { + "epoch": 14.197635135135135, + "grad_norm": 0.202254518866539, + "learning_rate": 2.7731666666666667e-05, + "loss": 0.0159, + "step": 16810 + }, + { + "epoch": 14.19847972972973, + "grad_norm": 0.21649615466594696, + "learning_rate": 2.7731333333333333e-05, + "loss": 0.0216, + "step": 16811 + }, + { + "epoch": 14.199324324324325, + "grad_norm": 0.5250161290168762, + "learning_rate": 2.7731e-05, + "loss": 0.0171, + "step": 16812 + }, + { + "epoch": 14.20016891891892, + "grad_norm": 0.341309517621994, + "learning_rate": 2.773066666666667e-05, + "loss": 0.0289, + "step": 16813 + }, + { + "epoch": 14.201013513513514, + "grad_norm": 0.16878226399421692, + "learning_rate": 2.7730333333333335e-05, + "loss": 0.0115, + "step": 16814 + }, + { + "epoch": 14.201858108108109, + "grad_norm": 0.2933872640132904, + "learning_rate": 2.773e-05, + "loss": 0.0161, + "step": 16815 + }, + { + "epoch": 14.202702702702704, + "grad_norm": 0.2146281599998474, + "learning_rate": 2.772966666666667e-05, + "loss": 0.009, + "step": 16816 + }, + { + "epoch": 14.203547297297296, + "grad_norm": 0.42624571919441223, + "learning_rate": 2.7729333333333332e-05, + "loss": 0.02, + "step": 16817 + }, + { + "epoch": 14.204391891891891, + "grad_norm": 0.25034356117248535, + "learning_rate": 2.7729e-05, + "loss": 0.0121, + "step": 16818 + }, + { + "epoch": 14.205236486486486, + "grad_norm": 0.2429601103067398, + "learning_rate": 2.7728666666666668e-05, + "loss": 0.0091, + "step": 16819 + }, + { + "epoch": 14.20608108108108, + "grad_norm": 0.27349960803985596, + "learning_rate": 2.7728333333333334e-05, + "loss": 0.0124, + "step": 16820 + }, + { + "epoch": 14.206925675675675, + "grad_norm": 0.4322678744792938, + "learning_rate": 2.7728e-05, + "loss": 0.0214, + "step": 16821 + }, + { + "epoch": 14.20777027027027, + "grad_norm": 0.13890713453292847, + "learning_rate": 2.772766666666667e-05, + "loss": 0.0104, + "step": 16822 + }, + { + "epoch": 14.208614864864865, + "grad_norm": 0.3288232386112213, + "learning_rate": 2.7727333333333335e-05, + "loss": 0.027, + "step": 16823 + }, + { + "epoch": 14.20945945945946, + "grad_norm": 0.3112126886844635, + "learning_rate": 2.7727e-05, + "loss": 0.0238, + "step": 16824 + }, + { + "epoch": 14.210304054054054, + "grad_norm": 0.23634184896945953, + "learning_rate": 2.7726666666666667e-05, + "loss": 0.0185, + "step": 16825 + }, + { + "epoch": 14.21114864864865, + "grad_norm": 0.8125497102737427, + "learning_rate": 2.7726333333333333e-05, + "loss": 0.0201, + "step": 16826 + }, + { + "epoch": 14.211993243243244, + "grad_norm": 0.5439503788948059, + "learning_rate": 2.7726000000000002e-05, + "loss": 0.1521, + "step": 16827 + }, + { + "epoch": 14.212837837837839, + "grad_norm": 0.8527184128761292, + "learning_rate": 2.7725666666666664e-05, + "loss": 0.1288, + "step": 16828 + }, + { + "epoch": 14.213682432432432, + "grad_norm": 0.30238252878189087, + "learning_rate": 2.7725333333333334e-05, + "loss": 0.0479, + "step": 16829 + }, + { + "epoch": 14.214527027027026, + "grad_norm": 0.24511076509952545, + "learning_rate": 2.7725000000000003e-05, + "loss": 0.0259, + "step": 16830 + }, + { + "epoch": 14.215371621621621, + "grad_norm": 0.22531385719776154, + "learning_rate": 2.7724666666666666e-05, + "loss": 0.0218, + "step": 16831 + }, + { + "epoch": 14.216216216216216, + "grad_norm": 0.21326391398906708, + "learning_rate": 2.7724333333333335e-05, + "loss": 0.0229, + "step": 16832 + }, + { + "epoch": 14.21706081081081, + "grad_norm": 1.3538769483566284, + "learning_rate": 2.7724e-05, + "loss": 0.0262, + "step": 16833 + }, + { + "epoch": 14.217905405405405, + "grad_norm": 0.4225586950778961, + "learning_rate": 2.7723666666666667e-05, + "loss": 0.0331, + "step": 16834 + }, + { + "epoch": 14.21875, + "grad_norm": 0.3026675283908844, + "learning_rate": 2.7723333333333333e-05, + "loss": 0.0174, + "step": 16835 + }, + { + "epoch": 14.219594594594595, + "grad_norm": 0.4825049638748169, + "learning_rate": 2.7723000000000002e-05, + "loss": 0.0247, + "step": 16836 + }, + { + "epoch": 14.22043918918919, + "grad_norm": 0.15786154568195343, + "learning_rate": 2.7722666666666665e-05, + "loss": 0.0191, + "step": 16837 + }, + { + "epoch": 14.221283783783784, + "grad_norm": 0.30366408824920654, + "learning_rate": 2.7722333333333334e-05, + "loss": 0.0225, + "step": 16838 + }, + { + "epoch": 14.222128378378379, + "grad_norm": 0.18912822008132935, + "learning_rate": 2.7722000000000003e-05, + "loss": 0.0162, + "step": 16839 + }, + { + "epoch": 14.222972972972974, + "grad_norm": 0.17569619417190552, + "learning_rate": 2.7721666666666666e-05, + "loss": 0.0167, + "step": 16840 + }, + { + "epoch": 14.223817567567568, + "grad_norm": 0.2055077999830246, + "learning_rate": 2.7721333333333335e-05, + "loss": 0.0199, + "step": 16841 + }, + { + "epoch": 14.224662162162161, + "grad_norm": 0.2003905028104782, + "learning_rate": 2.7721e-05, + "loss": 0.016, + "step": 16842 + }, + { + "epoch": 14.225506756756756, + "grad_norm": 0.3042210042476654, + "learning_rate": 2.7720666666666667e-05, + "loss": 0.0197, + "step": 16843 + }, + { + "epoch": 14.22635135135135, + "grad_norm": 0.25275084376335144, + "learning_rate": 2.7720333333333333e-05, + "loss": 0.0301, + "step": 16844 + }, + { + "epoch": 14.227195945945946, + "grad_norm": 0.4040714502334595, + "learning_rate": 2.7720000000000002e-05, + "loss": 0.0222, + "step": 16845 + }, + { + "epoch": 14.22804054054054, + "grad_norm": 0.2292325347661972, + "learning_rate": 2.7719666666666665e-05, + "loss": 0.0135, + "step": 16846 + }, + { + "epoch": 14.228885135135135, + "grad_norm": 0.15660855174064636, + "learning_rate": 2.7719333333333334e-05, + "loss": 0.0207, + "step": 16847 + }, + { + "epoch": 14.22972972972973, + "grad_norm": 0.2064787894487381, + "learning_rate": 2.7719000000000003e-05, + "loss": 0.0143, + "step": 16848 + }, + { + "epoch": 14.230574324324325, + "grad_norm": 0.4261897802352905, + "learning_rate": 2.7718666666666666e-05, + "loss": 0.0212, + "step": 16849 + }, + { + "epoch": 14.23141891891892, + "grad_norm": 0.20962119102478027, + "learning_rate": 2.7718333333333335e-05, + "loss": 0.01, + "step": 16850 + }, + { + "epoch": 14.232263513513514, + "grad_norm": 0.3168581426143646, + "learning_rate": 2.7718e-05, + "loss": 0.0162, + "step": 16851 + }, + { + "epoch": 14.233108108108109, + "grad_norm": 0.29137086868286133, + "learning_rate": 2.7717666666666667e-05, + "loss": 0.0217, + "step": 16852 + }, + { + "epoch": 14.233952702702704, + "grad_norm": 0.4070902466773987, + "learning_rate": 2.7717333333333333e-05, + "loss": 0.0215, + "step": 16853 + }, + { + "epoch": 14.234797297297296, + "grad_norm": 0.1866527795791626, + "learning_rate": 2.7717000000000002e-05, + "loss": 0.0079, + "step": 16854 + }, + { + "epoch": 14.235641891891891, + "grad_norm": 0.1472235769033432, + "learning_rate": 2.771666666666667e-05, + "loss": 0.0119, + "step": 16855 + }, + { + "epoch": 14.236486486486486, + "grad_norm": 0.34133660793304443, + "learning_rate": 2.7716333333333334e-05, + "loss": 0.0155, + "step": 16856 + }, + { + "epoch": 14.23733108108108, + "grad_norm": 0.2622823119163513, + "learning_rate": 2.7716e-05, + "loss": 0.0193, + "step": 16857 + }, + { + "epoch": 14.238175675675675, + "grad_norm": 0.31747710704803467, + "learning_rate": 2.7715666666666666e-05, + "loss": 0.0088, + "step": 16858 + }, + { + "epoch": 14.23902027027027, + "grad_norm": 0.2870628237724304, + "learning_rate": 2.7715333333333336e-05, + "loss": 0.0102, + "step": 16859 + }, + { + "epoch": 14.239864864864865, + "grad_norm": 0.24436959624290466, + "learning_rate": 2.7714999999999998e-05, + "loss": 0.0111, + "step": 16860 + }, + { + "epoch": 14.24070945945946, + "grad_norm": 0.25728487968444824, + "learning_rate": 2.7714666666666667e-05, + "loss": 0.0161, + "step": 16861 + }, + { + "epoch": 14.241554054054054, + "grad_norm": 0.22716641426086426, + "learning_rate": 2.7714333333333333e-05, + "loss": 0.0161, + "step": 16862 + }, + { + "epoch": 14.24239864864865, + "grad_norm": 0.2873261570930481, + "learning_rate": 2.7714e-05, + "loss": 0.0169, + "step": 16863 + }, + { + "epoch": 14.243243243243244, + "grad_norm": 0.2535189092159271, + "learning_rate": 2.771366666666667e-05, + "loss": 0.0187, + "step": 16864 + }, + { + "epoch": 14.244087837837839, + "grad_norm": 0.27000749111175537, + "learning_rate": 2.7713333333333335e-05, + "loss": 0.0201, + "step": 16865 + }, + { + "epoch": 14.244932432432432, + "grad_norm": 0.34305045008659363, + "learning_rate": 2.7713e-05, + "loss": 0.0193, + "step": 16866 + }, + { + "epoch": 14.245777027027026, + "grad_norm": 0.3312833905220032, + "learning_rate": 2.7712666666666666e-05, + "loss": 0.0168, + "step": 16867 + }, + { + "epoch": 14.246621621621621, + "grad_norm": 0.41106143593788147, + "learning_rate": 2.7712333333333336e-05, + "loss": 0.0112, + "step": 16868 + }, + { + "epoch": 14.247466216216216, + "grad_norm": 0.9210276007652283, + "learning_rate": 2.7711999999999998e-05, + "loss": 0.0247, + "step": 16869 + }, + { + "epoch": 14.24831081081081, + "grad_norm": 0.3101535141468048, + "learning_rate": 2.7711666666666668e-05, + "loss": 0.0129, + "step": 16870 + }, + { + "epoch": 14.249155405405405, + "grad_norm": 0.48340490460395813, + "learning_rate": 2.7711333333333334e-05, + "loss": 0.0141, + "step": 16871 + }, + { + "epoch": 14.25, + "grad_norm": 1.472981333732605, + "learning_rate": 2.7711e-05, + "loss": 0.032, + "step": 16872 + }, + { + "epoch": 14.250844594594595, + "grad_norm": 0.5213685035705566, + "learning_rate": 2.771066666666667e-05, + "loss": 0.0147, + "step": 16873 + }, + { + "epoch": 14.25168918918919, + "grad_norm": 0.30871468782424927, + "learning_rate": 2.7710333333333335e-05, + "loss": 0.0191, + "step": 16874 + }, + { + "epoch": 14.252533783783784, + "grad_norm": 0.6083167791366577, + "learning_rate": 2.771e-05, + "loss": 0.0256, + "step": 16875 + }, + { + "epoch": 14.253378378378379, + "grad_norm": 0.31882816553115845, + "learning_rate": 2.7709666666666667e-05, + "loss": 0.0171, + "step": 16876 + }, + { + "epoch": 14.254222972972974, + "grad_norm": 0.5428602695465088, + "learning_rate": 2.7709333333333336e-05, + "loss": 0.151, + "step": 16877 + }, + { + "epoch": 14.255067567567568, + "grad_norm": 0.41756463050842285, + "learning_rate": 2.7709e-05, + "loss": 0.0888, + "step": 16878 + }, + { + "epoch": 14.255912162162161, + "grad_norm": 0.34416642785072327, + "learning_rate": 2.7708666666666668e-05, + "loss": 0.0411, + "step": 16879 + }, + { + "epoch": 14.256756756756756, + "grad_norm": 0.2823604941368103, + "learning_rate": 2.7708333333333337e-05, + "loss": 0.0342, + "step": 16880 + }, + { + "epoch": 14.25760135135135, + "grad_norm": 0.2777057886123657, + "learning_rate": 2.7708e-05, + "loss": 0.0306, + "step": 16881 + }, + { + "epoch": 14.258445945945946, + "grad_norm": 0.5329310297966003, + "learning_rate": 2.770766666666667e-05, + "loss": 0.0315, + "step": 16882 + }, + { + "epoch": 14.25929054054054, + "grad_norm": 0.3605317771434784, + "learning_rate": 2.7707333333333335e-05, + "loss": 0.0317, + "step": 16883 + }, + { + "epoch": 14.260135135135135, + "grad_norm": 0.43108031153678894, + "learning_rate": 2.7707e-05, + "loss": 0.0278, + "step": 16884 + }, + { + "epoch": 14.26097972972973, + "grad_norm": 0.26612982153892517, + "learning_rate": 2.7706666666666667e-05, + "loss": 0.0216, + "step": 16885 + }, + { + "epoch": 14.261824324324325, + "grad_norm": 0.42497938871383667, + "learning_rate": 2.7706333333333333e-05, + "loss": 0.0239, + "step": 16886 + }, + { + "epoch": 14.26266891891892, + "grad_norm": 0.28410157561302185, + "learning_rate": 2.7706e-05, + "loss": 0.0222, + "step": 16887 + }, + { + "epoch": 14.263513513513514, + "grad_norm": 0.27433115243911743, + "learning_rate": 2.7705666666666668e-05, + "loss": 0.0215, + "step": 16888 + }, + { + "epoch": 14.264358108108109, + "grad_norm": 0.6576905846595764, + "learning_rate": 2.7705333333333334e-05, + "loss": 0.033, + "step": 16889 + }, + { + "epoch": 14.265202702702704, + "grad_norm": 0.3059101402759552, + "learning_rate": 2.7705e-05, + "loss": 0.0292, + "step": 16890 + }, + { + "epoch": 14.266047297297296, + "grad_norm": 0.2636149227619171, + "learning_rate": 2.770466666666667e-05, + "loss": 0.0207, + "step": 16891 + }, + { + "epoch": 14.266891891891891, + "grad_norm": 0.15473562479019165, + "learning_rate": 2.7704333333333332e-05, + "loss": 0.014, + "step": 16892 + }, + { + "epoch": 14.267736486486486, + "grad_norm": 0.3406480550765991, + "learning_rate": 2.7704e-05, + "loss": 0.0219, + "step": 16893 + }, + { + "epoch": 14.26858108108108, + "grad_norm": 0.2811708152294159, + "learning_rate": 2.7703666666666667e-05, + "loss": 0.0165, + "step": 16894 + }, + { + "epoch": 14.269425675675675, + "grad_norm": 0.23636142909526825, + "learning_rate": 2.7703333333333333e-05, + "loss": 0.0184, + "step": 16895 + }, + { + "epoch": 14.27027027027027, + "grad_norm": 0.37725815176963806, + "learning_rate": 2.7703e-05, + "loss": 0.0244, + "step": 16896 + }, + { + "epoch": 14.271114864864865, + "grad_norm": 0.2167038917541504, + "learning_rate": 2.7702666666666668e-05, + "loss": 0.0169, + "step": 16897 + }, + { + "epoch": 14.27195945945946, + "grad_norm": 0.23866382241249084, + "learning_rate": 2.7702333333333334e-05, + "loss": 0.016, + "step": 16898 + }, + { + "epoch": 14.272804054054054, + "grad_norm": 0.4243312180042267, + "learning_rate": 2.7702e-05, + "loss": 0.0161, + "step": 16899 + }, + { + "epoch": 14.27364864864865, + "grad_norm": 0.2276887744665146, + "learning_rate": 2.770166666666667e-05, + "loss": 0.0139, + "step": 16900 + }, + { + "epoch": 14.274493243243244, + "grad_norm": 0.22379183769226074, + "learning_rate": 2.7701333333333332e-05, + "loss": 0.0156, + "step": 16901 + }, + { + "epoch": 14.275337837837839, + "grad_norm": 0.3886987268924713, + "learning_rate": 2.7701e-05, + "loss": 0.0287, + "step": 16902 + }, + { + "epoch": 14.276182432432432, + "grad_norm": 0.2339458465576172, + "learning_rate": 2.7700666666666667e-05, + "loss": 0.0167, + "step": 16903 + }, + { + "epoch": 14.277027027027026, + "grad_norm": 0.25251781940460205, + "learning_rate": 2.7700333333333333e-05, + "loss": 0.0117, + "step": 16904 + }, + { + "epoch": 14.277871621621621, + "grad_norm": 0.2475978583097458, + "learning_rate": 2.7700000000000002e-05, + "loss": 0.0102, + "step": 16905 + }, + { + "epoch": 14.278716216216216, + "grad_norm": 1.0065226554870605, + "learning_rate": 2.769966666666667e-05, + "loss": 0.0315, + "step": 16906 + }, + { + "epoch": 14.27956081081081, + "grad_norm": 0.17300035059452057, + "learning_rate": 2.7699333333333334e-05, + "loss": 0.0112, + "step": 16907 + }, + { + "epoch": 14.280405405405405, + "grad_norm": 0.37155380845069885, + "learning_rate": 2.7699e-05, + "loss": 0.0205, + "step": 16908 + }, + { + "epoch": 14.28125, + "grad_norm": 0.2861500382423401, + "learning_rate": 2.769866666666667e-05, + "loss": 0.0141, + "step": 16909 + }, + { + "epoch": 14.282094594594595, + "grad_norm": 0.267474502325058, + "learning_rate": 2.7698333333333332e-05, + "loss": 0.0165, + "step": 16910 + }, + { + "epoch": 14.28293918918919, + "grad_norm": 0.39542272686958313, + "learning_rate": 2.7698e-05, + "loss": 0.0181, + "step": 16911 + }, + { + "epoch": 14.283783783783784, + "grad_norm": 0.7029929757118225, + "learning_rate": 2.7697666666666667e-05, + "loss": 0.0141, + "step": 16912 + }, + { + "epoch": 14.284628378378379, + "grad_norm": 0.31730568408966064, + "learning_rate": 2.7697333333333333e-05, + "loss": 0.0098, + "step": 16913 + }, + { + "epoch": 14.285472972972974, + "grad_norm": 0.1680157333612442, + "learning_rate": 2.7697000000000003e-05, + "loss": 0.0091, + "step": 16914 + }, + { + "epoch": 14.286317567567568, + "grad_norm": 1.6397277116775513, + "learning_rate": 2.769666666666667e-05, + "loss": 0.0321, + "step": 16915 + }, + { + "epoch": 14.287162162162161, + "grad_norm": 0.17597705125808716, + "learning_rate": 2.7696333333333335e-05, + "loss": 0.0134, + "step": 16916 + }, + { + "epoch": 14.288006756756756, + "grad_norm": 0.45820102095603943, + "learning_rate": 2.7696e-05, + "loss": 0.0191, + "step": 16917 + }, + { + "epoch": 14.28885135135135, + "grad_norm": 0.24875541031360626, + "learning_rate": 2.7695666666666666e-05, + "loss": 0.013, + "step": 16918 + }, + { + "epoch": 14.289695945945946, + "grad_norm": 0.6132169365882874, + "learning_rate": 2.7695333333333332e-05, + "loss": 0.0211, + "step": 16919 + }, + { + "epoch": 14.29054054054054, + "grad_norm": 0.2855378985404968, + "learning_rate": 2.7695e-05, + "loss": 0.0199, + "step": 16920 + }, + { + "epoch": 14.291385135135135, + "grad_norm": 0.18560034036636353, + "learning_rate": 2.7694666666666668e-05, + "loss": 0.0111, + "step": 16921 + }, + { + "epoch": 14.29222972972973, + "grad_norm": 0.4279244840145111, + "learning_rate": 2.7694333333333334e-05, + "loss": 0.0155, + "step": 16922 + }, + { + "epoch": 14.293074324324325, + "grad_norm": 0.31923791766166687, + "learning_rate": 2.7694000000000003e-05, + "loss": 0.0246, + "step": 16923 + }, + { + "epoch": 14.29391891891892, + "grad_norm": 0.7861078381538391, + "learning_rate": 2.7693666666666665e-05, + "loss": 0.0281, + "step": 16924 + }, + { + "epoch": 14.294763513513514, + "grad_norm": 0.5151810050010681, + "learning_rate": 2.7693333333333335e-05, + "loss": 0.0217, + "step": 16925 + }, + { + "epoch": 14.295608108108109, + "grad_norm": 0.5698148608207703, + "learning_rate": 2.7693e-05, + "loss": 0.0324, + "step": 16926 + }, + { + "epoch": 14.296452702702704, + "grad_norm": 0.7379234433174133, + "learning_rate": 2.7692666666666667e-05, + "loss": 0.1471, + "step": 16927 + }, + { + "epoch": 14.297297297297296, + "grad_norm": 0.6147665977478027, + "learning_rate": 2.7692333333333333e-05, + "loss": 0.1027, + "step": 16928 + }, + { + "epoch": 14.298141891891891, + "grad_norm": 0.5058428049087524, + "learning_rate": 2.7692000000000002e-05, + "loss": 0.0426, + "step": 16929 + }, + { + "epoch": 14.298986486486486, + "grad_norm": 0.23900549113750458, + "learning_rate": 2.7691666666666668e-05, + "loss": 0.0299, + "step": 16930 + }, + { + "epoch": 14.29983108108108, + "grad_norm": 0.22187721729278564, + "learning_rate": 2.7691333333333334e-05, + "loss": 0.0253, + "step": 16931 + }, + { + "epoch": 14.300675675675675, + "grad_norm": 0.29375454783439636, + "learning_rate": 2.7691000000000003e-05, + "loss": 0.0168, + "step": 16932 + }, + { + "epoch": 14.30152027027027, + "grad_norm": 0.2769779562950134, + "learning_rate": 2.7690666666666666e-05, + "loss": 0.0264, + "step": 16933 + }, + { + "epoch": 14.302364864864865, + "grad_norm": 0.27695998549461365, + "learning_rate": 2.7690333333333335e-05, + "loss": 0.0194, + "step": 16934 + }, + { + "epoch": 14.30320945945946, + "grad_norm": 0.14884339272975922, + "learning_rate": 2.769e-05, + "loss": 0.0098, + "step": 16935 + }, + { + "epoch": 14.304054054054054, + "grad_norm": 0.3467543125152588, + "learning_rate": 2.7689666666666667e-05, + "loss": 0.015, + "step": 16936 + }, + { + "epoch": 14.30489864864865, + "grad_norm": 0.2621203660964966, + "learning_rate": 2.7689333333333333e-05, + "loss": 0.0119, + "step": 16937 + }, + { + "epoch": 14.305743243243244, + "grad_norm": 0.23986217379570007, + "learning_rate": 2.7689000000000002e-05, + "loss": 0.0191, + "step": 16938 + }, + { + "epoch": 14.306587837837839, + "grad_norm": 0.2582884132862091, + "learning_rate": 2.7688666666666668e-05, + "loss": 0.0145, + "step": 16939 + }, + { + "epoch": 14.307432432432432, + "grad_norm": 0.35781416296958923, + "learning_rate": 2.7688333333333334e-05, + "loss": 0.0142, + "step": 16940 + }, + { + "epoch": 14.308277027027026, + "grad_norm": 0.38127997517585754, + "learning_rate": 2.7688000000000003e-05, + "loss": 0.0131, + "step": 16941 + }, + { + "epoch": 14.309121621621621, + "grad_norm": 0.2152596414089203, + "learning_rate": 2.7687666666666666e-05, + "loss": 0.0163, + "step": 16942 + }, + { + "epoch": 14.309966216216216, + "grad_norm": 0.26090648770332336, + "learning_rate": 2.7687333333333335e-05, + "loss": 0.0162, + "step": 16943 + }, + { + "epoch": 14.31081081081081, + "grad_norm": 0.8140245676040649, + "learning_rate": 2.7687e-05, + "loss": 0.0251, + "step": 16944 + }, + { + "epoch": 14.311655405405405, + "grad_norm": 0.15094460546970367, + "learning_rate": 2.7686666666666667e-05, + "loss": 0.0071, + "step": 16945 + }, + { + "epoch": 14.3125, + "grad_norm": 0.31030768156051636, + "learning_rate": 2.7686333333333333e-05, + "loss": 0.0179, + "step": 16946 + }, + { + "epoch": 14.313344594594595, + "grad_norm": 0.2436574548482895, + "learning_rate": 2.7686e-05, + "loss": 0.0143, + "step": 16947 + }, + { + "epoch": 14.31418918918919, + "grad_norm": 0.534746527671814, + "learning_rate": 2.7685666666666668e-05, + "loss": 0.0254, + "step": 16948 + }, + { + "epoch": 14.315033783783784, + "grad_norm": 0.35717660188674927, + "learning_rate": 2.7685333333333334e-05, + "loss": 0.0245, + "step": 16949 + }, + { + "epoch": 14.315878378378379, + "grad_norm": 0.11151223629713058, + "learning_rate": 2.7685e-05, + "loss": 0.0069, + "step": 16950 + }, + { + "epoch": 14.316722972972974, + "grad_norm": 0.14453047513961792, + "learning_rate": 2.7684666666666666e-05, + "loss": 0.0118, + "step": 16951 + }, + { + "epoch": 14.317567567567568, + "grad_norm": 0.661001980304718, + "learning_rate": 2.7684333333333335e-05, + "loss": 0.0308, + "step": 16952 + }, + { + "epoch": 14.318412162162161, + "grad_norm": 0.3466961979866028, + "learning_rate": 2.7683999999999998e-05, + "loss": 0.0286, + "step": 16953 + }, + { + "epoch": 14.319256756756756, + "grad_norm": 0.2373342216014862, + "learning_rate": 2.7683666666666667e-05, + "loss": 0.0171, + "step": 16954 + }, + { + "epoch": 14.32010135135135, + "grad_norm": 0.21355897188186646, + "learning_rate": 2.7683333333333337e-05, + "loss": 0.0147, + "step": 16955 + }, + { + "epoch": 14.320945945945946, + "grad_norm": 0.1512381136417389, + "learning_rate": 2.7683e-05, + "loss": 0.0119, + "step": 16956 + }, + { + "epoch": 14.32179054054054, + "grad_norm": 0.38868576288223267, + "learning_rate": 2.768266666666667e-05, + "loss": 0.0188, + "step": 16957 + }, + { + "epoch": 14.322635135135135, + "grad_norm": 0.7295973896980286, + "learning_rate": 2.7682333333333334e-05, + "loss": 0.0179, + "step": 16958 + }, + { + "epoch": 14.32347972972973, + "grad_norm": 0.21773400902748108, + "learning_rate": 2.7682e-05, + "loss": 0.012, + "step": 16959 + }, + { + "epoch": 14.324324324324325, + "grad_norm": 0.9028536677360535, + "learning_rate": 2.7681666666666666e-05, + "loss": 0.0128, + "step": 16960 + }, + { + "epoch": 14.32516891891892, + "grad_norm": 0.2560582756996155, + "learning_rate": 2.7681333333333336e-05, + "loss": 0.0155, + "step": 16961 + }, + { + "epoch": 14.326013513513514, + "grad_norm": 0.21331068873405457, + "learning_rate": 2.7680999999999998e-05, + "loss": 0.0163, + "step": 16962 + }, + { + "epoch": 14.326858108108109, + "grad_norm": 0.2893608510494232, + "learning_rate": 2.7680666666666667e-05, + "loss": 0.022, + "step": 16963 + }, + { + "epoch": 14.327702702702704, + "grad_norm": 0.21872900426387787, + "learning_rate": 2.7680333333333337e-05, + "loss": 0.0122, + "step": 16964 + }, + { + "epoch": 14.328547297297296, + "grad_norm": 0.19313600659370422, + "learning_rate": 2.768e-05, + "loss": 0.0096, + "step": 16965 + }, + { + "epoch": 14.329391891891891, + "grad_norm": 0.26574262976646423, + "learning_rate": 2.767966666666667e-05, + "loss": 0.0186, + "step": 16966 + }, + { + "epoch": 14.330236486486486, + "grad_norm": 0.30134251713752747, + "learning_rate": 2.7679333333333335e-05, + "loss": 0.0188, + "step": 16967 + }, + { + "epoch": 14.33108108108108, + "grad_norm": 0.2390223890542984, + "learning_rate": 2.7679e-05, + "loss": 0.0156, + "step": 16968 + }, + { + "epoch": 14.331925675675675, + "grad_norm": 0.338974267244339, + "learning_rate": 2.7678666666666666e-05, + "loss": 0.014, + "step": 16969 + }, + { + "epoch": 14.33277027027027, + "grad_norm": 0.41828522086143494, + "learning_rate": 2.7678333333333336e-05, + "loss": 0.0119, + "step": 16970 + }, + { + "epoch": 14.333614864864865, + "grad_norm": 0.29497408866882324, + "learning_rate": 2.7678e-05, + "loss": 0.0163, + "step": 16971 + }, + { + "epoch": 14.33445945945946, + "grad_norm": 1.5011420249938965, + "learning_rate": 2.7677666666666668e-05, + "loss": 0.0206, + "step": 16972 + }, + { + "epoch": 14.335304054054054, + "grad_norm": 0.6693481802940369, + "learning_rate": 2.7677333333333337e-05, + "loss": 0.0294, + "step": 16973 + }, + { + "epoch": 14.33614864864865, + "grad_norm": 0.3987065553665161, + "learning_rate": 2.7677e-05, + "loss": 0.0222, + "step": 16974 + }, + { + "epoch": 14.336993243243244, + "grad_norm": 1.4178189039230347, + "learning_rate": 2.767666666666667e-05, + "loss": 0.0195, + "step": 16975 + }, + { + "epoch": 14.337837837837839, + "grad_norm": 0.3870125710964203, + "learning_rate": 2.767633333333333e-05, + "loss": 0.0208, + "step": 16976 + }, + { + "epoch": 14.338682432432432, + "grad_norm": 0.6711336970329285, + "learning_rate": 2.7676e-05, + "loss": 0.154, + "step": 16977 + }, + { + "epoch": 14.339527027027026, + "grad_norm": 0.5904123187065125, + "learning_rate": 2.7675666666666667e-05, + "loss": 0.0766, + "step": 16978 + }, + { + "epoch": 14.340371621621621, + "grad_norm": 0.40429145097732544, + "learning_rate": 2.7675333333333333e-05, + "loss": 0.0547, + "step": 16979 + }, + { + "epoch": 14.341216216216216, + "grad_norm": 0.31936758756637573, + "learning_rate": 2.7675000000000002e-05, + "loss": 0.0224, + "step": 16980 + }, + { + "epoch": 14.34206081081081, + "grad_norm": 0.31079354882240295, + "learning_rate": 2.7674666666666668e-05, + "loss": 0.0316, + "step": 16981 + }, + { + "epoch": 14.342905405405405, + "grad_norm": 0.2237711101770401, + "learning_rate": 2.7674333333333334e-05, + "loss": 0.0166, + "step": 16982 + }, + { + "epoch": 14.34375, + "grad_norm": 0.30208033323287964, + "learning_rate": 2.7674e-05, + "loss": 0.0177, + "step": 16983 + }, + { + "epoch": 14.344594594594595, + "grad_norm": 0.4432922303676605, + "learning_rate": 2.767366666666667e-05, + "loss": 0.0181, + "step": 16984 + }, + { + "epoch": 14.34543918918919, + "grad_norm": 0.1828324794769287, + "learning_rate": 2.767333333333333e-05, + "loss": 0.0105, + "step": 16985 + }, + { + "epoch": 14.346283783783784, + "grad_norm": 0.369004487991333, + "learning_rate": 2.7673e-05, + "loss": 0.0213, + "step": 16986 + }, + { + "epoch": 14.347128378378379, + "grad_norm": 0.20470936596393585, + "learning_rate": 2.7672666666666667e-05, + "loss": 0.0204, + "step": 16987 + }, + { + "epoch": 14.347972972972974, + "grad_norm": 0.25613877177238464, + "learning_rate": 2.7672333333333333e-05, + "loss": 0.0277, + "step": 16988 + }, + { + "epoch": 14.348817567567568, + "grad_norm": 0.30153149366378784, + "learning_rate": 2.7672000000000002e-05, + "loss": 0.0134, + "step": 16989 + }, + { + "epoch": 14.349662162162161, + "grad_norm": 0.27826106548309326, + "learning_rate": 2.7671666666666668e-05, + "loss": 0.012, + "step": 16990 + }, + { + "epoch": 14.350506756756756, + "grad_norm": 0.38478484749794006, + "learning_rate": 2.7671333333333334e-05, + "loss": 0.0202, + "step": 16991 + }, + { + "epoch": 14.35135135135135, + "grad_norm": 0.27249687910079956, + "learning_rate": 2.7671e-05, + "loss": 0.0128, + "step": 16992 + }, + { + "epoch": 14.352195945945946, + "grad_norm": 0.22098088264465332, + "learning_rate": 2.767066666666667e-05, + "loss": 0.0232, + "step": 16993 + }, + { + "epoch": 14.35304054054054, + "grad_norm": 0.3296721875667572, + "learning_rate": 2.7670333333333332e-05, + "loss": 0.0154, + "step": 16994 + }, + { + "epoch": 14.353885135135135, + "grad_norm": 0.23705504834651947, + "learning_rate": 2.767e-05, + "loss": 0.0179, + "step": 16995 + }, + { + "epoch": 14.35472972972973, + "grad_norm": 0.622748076915741, + "learning_rate": 2.7669666666666667e-05, + "loss": 0.0188, + "step": 16996 + }, + { + "epoch": 14.355574324324325, + "grad_norm": 0.2025245875120163, + "learning_rate": 2.7669333333333333e-05, + "loss": 0.0232, + "step": 16997 + }, + { + "epoch": 14.35641891891892, + "grad_norm": 0.3415968716144562, + "learning_rate": 2.7669000000000002e-05, + "loss": 0.0135, + "step": 16998 + }, + { + "epoch": 14.357263513513514, + "grad_norm": 0.3054976165294647, + "learning_rate": 2.7668666666666668e-05, + "loss": 0.0168, + "step": 16999 + }, + { + "epoch": 14.358108108108109, + "grad_norm": 0.38159269094467163, + "learning_rate": 2.7668333333333334e-05, + "loss": 0.0216, + "step": 17000 + }, + { + "epoch": 14.358108108108109, + "eval_cer": 0.09332583728214601, + "eval_loss": 0.24783213436603546, + "eval_runtime": 8.4038, + "eval_samples_per_second": 12.018, + "eval_steps_per_second": 0.476, + "eval_wer": 0.2795414462081129, + "step": 17000 + }, + { + "epoch": 14.358952702702704, + "grad_norm": 0.3248612880706787, + "learning_rate": 2.7668e-05, + "loss": 0.0135, + "step": 17001 + }, + { + "epoch": 14.359797297297296, + "grad_norm": 0.3459874391555786, + "learning_rate": 2.766766666666667e-05, + "loss": 0.0221, + "step": 17002 + }, + { + "epoch": 14.360641891891891, + "grad_norm": 0.2550729215145111, + "learning_rate": 2.7667333333333332e-05, + "loss": 0.0186, + "step": 17003 + }, + { + "epoch": 14.361486486486486, + "grad_norm": 0.3748936951160431, + "learning_rate": 2.7667e-05, + "loss": 0.0154, + "step": 17004 + }, + { + "epoch": 14.36233108108108, + "grad_norm": 0.17608317732810974, + "learning_rate": 2.766666666666667e-05, + "loss": 0.0162, + "step": 17005 + }, + { + "epoch": 14.363175675675675, + "grad_norm": 0.37461909651756287, + "learning_rate": 2.7666333333333333e-05, + "loss": 0.0232, + "step": 17006 + }, + { + "epoch": 14.36402027027027, + "grad_norm": 0.25672590732574463, + "learning_rate": 2.7666000000000002e-05, + "loss": 0.0175, + "step": 17007 + }, + { + "epoch": 14.364864864864865, + "grad_norm": 0.198264479637146, + "learning_rate": 2.7665666666666665e-05, + "loss": 0.0107, + "step": 17008 + }, + { + "epoch": 14.36570945945946, + "grad_norm": 0.2383527010679245, + "learning_rate": 2.7665333333333334e-05, + "loss": 0.0099, + "step": 17009 + }, + { + "epoch": 14.366554054054054, + "grad_norm": 0.3483014702796936, + "learning_rate": 2.7665e-05, + "loss": 0.0256, + "step": 17010 + }, + { + "epoch": 14.36739864864865, + "grad_norm": 0.2759745717048645, + "learning_rate": 2.7664666666666666e-05, + "loss": 0.0171, + "step": 17011 + }, + { + "epoch": 14.368243243243244, + "grad_norm": 0.3866800367832184, + "learning_rate": 2.7664333333333332e-05, + "loss": 0.0246, + "step": 17012 + }, + { + "epoch": 14.369087837837839, + "grad_norm": 0.3158210813999176, + "learning_rate": 2.7664e-05, + "loss": 0.0151, + "step": 17013 + }, + { + "epoch": 14.369932432432432, + "grad_norm": 0.31969109177589417, + "learning_rate": 2.7663666666666667e-05, + "loss": 0.0128, + "step": 17014 + }, + { + "epoch": 14.370777027027026, + "grad_norm": 0.24995671212673187, + "learning_rate": 2.7663333333333333e-05, + "loss": 0.0189, + "step": 17015 + }, + { + "epoch": 14.371621621621621, + "grad_norm": 0.28797686100006104, + "learning_rate": 2.7663000000000003e-05, + "loss": 0.0113, + "step": 17016 + }, + { + "epoch": 14.372466216216216, + "grad_norm": 0.19471026957035065, + "learning_rate": 2.7662666666666665e-05, + "loss": 0.0144, + "step": 17017 + }, + { + "epoch": 14.37331081081081, + "grad_norm": 0.560161828994751, + "learning_rate": 2.7662333333333335e-05, + "loss": 0.0208, + "step": 17018 + }, + { + "epoch": 14.374155405405405, + "grad_norm": 0.252743661403656, + "learning_rate": 2.7662e-05, + "loss": 0.0157, + "step": 17019 + }, + { + "epoch": 14.375, + "grad_norm": 0.11539154499769211, + "learning_rate": 2.7661666666666666e-05, + "loss": 0.0072, + "step": 17020 + }, + { + "epoch": 14.375844594594595, + "grad_norm": 0.42511311173439026, + "learning_rate": 2.7661333333333336e-05, + "loss": 0.0346, + "step": 17021 + }, + { + "epoch": 14.37668918918919, + "grad_norm": 0.2459031343460083, + "learning_rate": 2.7661e-05, + "loss": 0.0159, + "step": 17022 + }, + { + "epoch": 14.377533783783784, + "grad_norm": 0.31949570775032043, + "learning_rate": 2.7660666666666668e-05, + "loss": 0.02, + "step": 17023 + }, + { + "epoch": 14.378378378378379, + "grad_norm": 0.40940818190574646, + "learning_rate": 2.7660333333333334e-05, + "loss": 0.0238, + "step": 17024 + }, + { + "epoch": 14.379222972972974, + "grad_norm": 0.3937109410762787, + "learning_rate": 2.7660000000000003e-05, + "loss": 0.0118, + "step": 17025 + }, + { + "epoch": 14.380067567567568, + "grad_norm": 0.4404243528842926, + "learning_rate": 2.7659666666666665e-05, + "loss": 0.0175, + "step": 17026 + }, + { + "epoch": 14.380912162162161, + "grad_norm": 0.542253315448761, + "learning_rate": 2.7659333333333335e-05, + "loss": 0.1541, + "step": 17027 + }, + { + "epoch": 14.381756756756756, + "grad_norm": 0.4083004593849182, + "learning_rate": 2.7659e-05, + "loss": 0.093, + "step": 17028 + }, + { + "epoch": 14.38260135135135, + "grad_norm": 0.4617122411727905, + "learning_rate": 2.7658666666666667e-05, + "loss": 0.0417, + "step": 17029 + }, + { + "epoch": 14.383445945945946, + "grad_norm": 0.3068074882030487, + "learning_rate": 2.7658333333333336e-05, + "loss": 0.0267, + "step": 17030 + }, + { + "epoch": 14.38429054054054, + "grad_norm": 0.4298931062221527, + "learning_rate": 2.7658000000000002e-05, + "loss": 0.0315, + "step": 17031 + }, + { + "epoch": 14.385135135135135, + "grad_norm": 0.3058876693248749, + "learning_rate": 2.7657666666666668e-05, + "loss": 0.0136, + "step": 17032 + }, + { + "epoch": 14.38597972972973, + "grad_norm": 0.15482008457183838, + "learning_rate": 2.7657333333333334e-05, + "loss": 0.0096, + "step": 17033 + }, + { + "epoch": 14.386824324324325, + "grad_norm": 0.2411445677280426, + "learning_rate": 2.7657000000000003e-05, + "loss": 0.0143, + "step": 17034 + }, + { + "epoch": 14.38766891891892, + "grad_norm": 0.34897032380104065, + "learning_rate": 2.7656666666666666e-05, + "loss": 0.0223, + "step": 17035 + }, + { + "epoch": 14.388513513513514, + "grad_norm": 0.254623681306839, + "learning_rate": 2.7656333333333335e-05, + "loss": 0.0228, + "step": 17036 + }, + { + "epoch": 14.389358108108109, + "grad_norm": 0.20337672531604767, + "learning_rate": 2.7655999999999997e-05, + "loss": 0.0095, + "step": 17037 + }, + { + "epoch": 14.390202702702704, + "grad_norm": 0.34301599860191345, + "learning_rate": 2.7655666666666667e-05, + "loss": 0.0307, + "step": 17038 + }, + { + "epoch": 14.391047297297296, + "grad_norm": 0.23814615607261658, + "learning_rate": 2.7655333333333336e-05, + "loss": 0.0125, + "step": 17039 + }, + { + "epoch": 14.391891891891891, + "grad_norm": 0.4860275387763977, + "learning_rate": 2.7655e-05, + "loss": 0.0183, + "step": 17040 + }, + { + "epoch": 14.392736486486486, + "grad_norm": 0.2980540990829468, + "learning_rate": 2.7654666666666668e-05, + "loss": 0.0194, + "step": 17041 + }, + { + "epoch": 14.39358108108108, + "grad_norm": 0.1802554577589035, + "learning_rate": 2.7654333333333334e-05, + "loss": 0.0092, + "step": 17042 + }, + { + "epoch": 14.394425675675675, + "grad_norm": 0.20153050124645233, + "learning_rate": 2.7654e-05, + "loss": 0.0168, + "step": 17043 + }, + { + "epoch": 14.39527027027027, + "grad_norm": 0.2181505411863327, + "learning_rate": 2.7653666666666666e-05, + "loss": 0.0204, + "step": 17044 + }, + { + "epoch": 14.396114864864865, + "grad_norm": 0.23392744362354279, + "learning_rate": 2.7653333333333335e-05, + "loss": 0.0193, + "step": 17045 + }, + { + "epoch": 14.39695945945946, + "grad_norm": 0.184937983751297, + "learning_rate": 2.7653e-05, + "loss": 0.0198, + "step": 17046 + }, + { + "epoch": 14.397804054054054, + "grad_norm": 0.2706857919692993, + "learning_rate": 2.7652666666666667e-05, + "loss": 0.0202, + "step": 17047 + }, + { + "epoch": 14.39864864864865, + "grad_norm": 0.18985164165496826, + "learning_rate": 2.7652333333333336e-05, + "loss": 0.0224, + "step": 17048 + }, + { + "epoch": 14.399493243243244, + "grad_norm": 0.38726112246513367, + "learning_rate": 2.7652e-05, + "loss": 0.0229, + "step": 17049 + }, + { + "epoch": 14.400337837837839, + "grad_norm": 0.2834818959236145, + "learning_rate": 2.7651666666666668e-05, + "loss": 0.0126, + "step": 17050 + }, + { + "epoch": 14.401182432432432, + "grad_norm": 0.4509948492050171, + "learning_rate": 2.7651333333333334e-05, + "loss": 0.0165, + "step": 17051 + }, + { + "epoch": 14.402027027027026, + "grad_norm": 0.28735285997390747, + "learning_rate": 2.7651e-05, + "loss": 0.0227, + "step": 17052 + }, + { + "epoch": 14.402871621621621, + "grad_norm": 0.3306596279144287, + "learning_rate": 2.7650666666666666e-05, + "loss": 0.0205, + "step": 17053 + }, + { + "epoch": 14.403716216216216, + "grad_norm": 0.22288769483566284, + "learning_rate": 2.7650333333333335e-05, + "loss": 0.0174, + "step": 17054 + }, + { + "epoch": 14.40456081081081, + "grad_norm": 0.1928974688053131, + "learning_rate": 2.765e-05, + "loss": 0.0122, + "step": 17055 + }, + { + "epoch": 14.405405405405405, + "grad_norm": 0.16393059492111206, + "learning_rate": 2.7649666666666667e-05, + "loss": 0.0107, + "step": 17056 + }, + { + "epoch": 14.40625, + "grad_norm": 0.3742287755012512, + "learning_rate": 2.7649333333333336e-05, + "loss": 0.0135, + "step": 17057 + }, + { + "epoch": 14.407094594594595, + "grad_norm": 0.18377770483493805, + "learning_rate": 2.7649e-05, + "loss": 0.0166, + "step": 17058 + }, + { + "epoch": 14.40793918918919, + "grad_norm": 0.3664066195487976, + "learning_rate": 2.764866666666667e-05, + "loss": 0.0192, + "step": 17059 + }, + { + "epoch": 14.408783783783784, + "grad_norm": 0.6281691789627075, + "learning_rate": 2.7648333333333334e-05, + "loss": 0.0296, + "step": 17060 + }, + { + "epoch": 14.409628378378379, + "grad_norm": 0.49486082792282104, + "learning_rate": 2.7648e-05, + "loss": 0.0183, + "step": 17061 + }, + { + "epoch": 14.410472972972974, + "grad_norm": 0.3731693923473358, + "learning_rate": 2.7647666666666666e-05, + "loss": 0.0656, + "step": 17062 + }, + { + "epoch": 14.411317567567568, + "grad_norm": 0.2897559106349945, + "learning_rate": 2.7647333333333335e-05, + "loss": 0.0199, + "step": 17063 + }, + { + "epoch": 14.412162162162161, + "grad_norm": 0.4249238967895508, + "learning_rate": 2.7647e-05, + "loss": 0.0237, + "step": 17064 + }, + { + "epoch": 14.413006756756756, + "grad_norm": 0.2368403971195221, + "learning_rate": 2.7646666666666667e-05, + "loss": 0.0166, + "step": 17065 + }, + { + "epoch": 14.41385135135135, + "grad_norm": 0.2100604921579361, + "learning_rate": 2.7646333333333337e-05, + "loss": 0.0127, + "step": 17066 + }, + { + "epoch": 14.414695945945946, + "grad_norm": 0.32862037420272827, + "learning_rate": 2.7646e-05, + "loss": 0.0183, + "step": 17067 + }, + { + "epoch": 14.41554054054054, + "grad_norm": 0.25456443428993225, + "learning_rate": 2.764566666666667e-05, + "loss": 0.0172, + "step": 17068 + }, + { + "epoch": 14.416385135135135, + "grad_norm": 0.16917741298675537, + "learning_rate": 2.764533333333333e-05, + "loss": 0.014, + "step": 17069 + }, + { + "epoch": 14.41722972972973, + "grad_norm": 0.2238016426563263, + "learning_rate": 2.7645e-05, + "loss": 0.018, + "step": 17070 + }, + { + "epoch": 14.418074324324325, + "grad_norm": 0.34080085158348083, + "learning_rate": 2.764466666666667e-05, + "loss": 0.0237, + "step": 17071 + }, + { + "epoch": 14.41891891891892, + "grad_norm": 0.3077409565448761, + "learning_rate": 2.7644333333333332e-05, + "loss": 0.0146, + "step": 17072 + }, + { + "epoch": 14.419763513513514, + "grad_norm": 0.21552976965904236, + "learning_rate": 2.7644e-05, + "loss": 0.0149, + "step": 17073 + }, + { + "epoch": 14.420608108108109, + "grad_norm": 0.30211132764816284, + "learning_rate": 2.7643666666666668e-05, + "loss": 0.0171, + "step": 17074 + }, + { + "epoch": 14.421452702702704, + "grad_norm": 0.4525962769985199, + "learning_rate": 2.7643333333333334e-05, + "loss": 0.0259, + "step": 17075 + }, + { + "epoch": 14.422297297297296, + "grad_norm": 0.4253425896167755, + "learning_rate": 2.7643e-05, + "loss": 0.0192, + "step": 17076 + }, + { + "epoch": 14.423141891891891, + "grad_norm": 0.5949749946594238, + "learning_rate": 2.764266666666667e-05, + "loss": 0.1994, + "step": 17077 + }, + { + "epoch": 14.423986486486486, + "grad_norm": 0.5328323245048523, + "learning_rate": 2.764233333333333e-05, + "loss": 0.0698, + "step": 17078 + }, + { + "epoch": 14.42483108108108, + "grad_norm": 0.35433951020240784, + "learning_rate": 2.7642e-05, + "loss": 0.0422, + "step": 17079 + }, + { + "epoch": 14.425675675675675, + "grad_norm": 0.458719402551651, + "learning_rate": 2.764166666666667e-05, + "loss": 0.0704, + "step": 17080 + }, + { + "epoch": 14.42652027027027, + "grad_norm": 0.25156837701797485, + "learning_rate": 2.7641333333333333e-05, + "loss": 0.032, + "step": 17081 + }, + { + "epoch": 14.427364864864865, + "grad_norm": 0.4181310534477234, + "learning_rate": 2.7641000000000002e-05, + "loss": 0.0276, + "step": 17082 + }, + { + "epoch": 14.42820945945946, + "grad_norm": 0.22434002161026, + "learning_rate": 2.7640666666666668e-05, + "loss": 0.014, + "step": 17083 + }, + { + "epoch": 14.429054054054054, + "grad_norm": 0.18897639214992523, + "learning_rate": 2.7640333333333334e-05, + "loss": 0.0134, + "step": 17084 + }, + { + "epoch": 14.42989864864865, + "grad_norm": 0.24573186039924622, + "learning_rate": 2.764e-05, + "loss": 0.0126, + "step": 17085 + }, + { + "epoch": 14.430743243243244, + "grad_norm": 0.3590746223926544, + "learning_rate": 2.763966666666667e-05, + "loss": 0.0167, + "step": 17086 + }, + { + "epoch": 14.431587837837839, + "grad_norm": 0.480375200510025, + "learning_rate": 2.763933333333333e-05, + "loss": 0.0163, + "step": 17087 + }, + { + "epoch": 14.432432432432432, + "grad_norm": 0.3384031057357788, + "learning_rate": 2.7639e-05, + "loss": 0.0152, + "step": 17088 + }, + { + "epoch": 14.433277027027026, + "grad_norm": 0.1970699429512024, + "learning_rate": 2.763866666666667e-05, + "loss": 0.0156, + "step": 17089 + }, + { + "epoch": 14.434121621621621, + "grad_norm": 0.4082813858985901, + "learning_rate": 2.7638333333333333e-05, + "loss": 0.0157, + "step": 17090 + }, + { + "epoch": 14.434966216216216, + "grad_norm": 0.3515951633453369, + "learning_rate": 2.7638000000000002e-05, + "loss": 0.0203, + "step": 17091 + }, + { + "epoch": 14.43581081081081, + "grad_norm": 0.9910027384757996, + "learning_rate": 2.7637666666666668e-05, + "loss": 0.0178, + "step": 17092 + }, + { + "epoch": 14.436655405405405, + "grad_norm": 0.27060362696647644, + "learning_rate": 2.7637333333333334e-05, + "loss": 0.0171, + "step": 17093 + }, + { + "epoch": 14.4375, + "grad_norm": 0.3343617618083954, + "learning_rate": 2.7637e-05, + "loss": 0.0133, + "step": 17094 + }, + { + "epoch": 14.438344594594595, + "grad_norm": 0.23064038157463074, + "learning_rate": 2.763666666666667e-05, + "loss": 0.0184, + "step": 17095 + }, + { + "epoch": 14.43918918918919, + "grad_norm": 0.42511501908302307, + "learning_rate": 2.7636333333333335e-05, + "loss": 0.0252, + "step": 17096 + }, + { + "epoch": 14.440033783783784, + "grad_norm": 0.28860726952552795, + "learning_rate": 2.7636e-05, + "loss": 0.017, + "step": 17097 + }, + { + "epoch": 14.440878378378379, + "grad_norm": 0.2626093327999115, + "learning_rate": 2.7635666666666667e-05, + "loss": 0.0189, + "step": 17098 + }, + { + "epoch": 14.441722972972974, + "grad_norm": 0.296740859746933, + "learning_rate": 2.7635333333333333e-05, + "loss": 0.0172, + "step": 17099 + }, + { + "epoch": 14.442567567567568, + "grad_norm": 0.21234019100666046, + "learning_rate": 2.7635000000000002e-05, + "loss": 0.0094, + "step": 17100 + }, + { + "epoch": 14.443412162162161, + "grad_norm": 0.6871110796928406, + "learning_rate": 2.7634666666666665e-05, + "loss": 0.0286, + "step": 17101 + }, + { + "epoch": 14.444256756756756, + "grad_norm": 0.1676739603281021, + "learning_rate": 2.7634333333333334e-05, + "loss": 0.0147, + "step": 17102 + }, + { + "epoch": 14.44510135135135, + "grad_norm": 0.3312435746192932, + "learning_rate": 2.7634e-05, + "loss": 0.0229, + "step": 17103 + }, + { + "epoch": 14.445945945945946, + "grad_norm": 0.39678457379341125, + "learning_rate": 2.7633666666666666e-05, + "loss": 0.017, + "step": 17104 + }, + { + "epoch": 14.44679054054054, + "grad_norm": 0.23628151416778564, + "learning_rate": 2.7633333333333335e-05, + "loss": 0.021, + "step": 17105 + }, + { + "epoch": 14.447635135135135, + "grad_norm": 0.18886052072048187, + "learning_rate": 2.7633e-05, + "loss": 0.0121, + "step": 17106 + }, + { + "epoch": 14.44847972972973, + "grad_norm": 0.5664435625076294, + "learning_rate": 2.7632666666666667e-05, + "loss": 0.0221, + "step": 17107 + }, + { + "epoch": 14.449324324324325, + "grad_norm": 0.3122120499610901, + "learning_rate": 2.7632333333333333e-05, + "loss": 0.0169, + "step": 17108 + }, + { + "epoch": 14.45016891891892, + "grad_norm": 0.28666356205940247, + "learning_rate": 2.7632000000000002e-05, + "loss": 0.0631, + "step": 17109 + }, + { + "epoch": 14.451013513513514, + "grad_norm": 0.1897261142730713, + "learning_rate": 2.7631666666666665e-05, + "loss": 0.0188, + "step": 17110 + }, + { + "epoch": 14.451858108108109, + "grad_norm": 0.2724250853061676, + "learning_rate": 2.7631333333333334e-05, + "loss": 0.022, + "step": 17111 + }, + { + "epoch": 14.452702702702704, + "grad_norm": 0.2277699112892151, + "learning_rate": 2.7631e-05, + "loss": 0.0141, + "step": 17112 + }, + { + "epoch": 14.453547297297296, + "grad_norm": 0.37523049116134644, + "learning_rate": 2.7630666666666666e-05, + "loss": 0.0143, + "step": 17113 + }, + { + "epoch": 14.454391891891891, + "grad_norm": 0.31068068742752075, + "learning_rate": 2.7630333333333335e-05, + "loss": 0.013, + "step": 17114 + }, + { + "epoch": 14.455236486486486, + "grad_norm": 0.23387637734413147, + "learning_rate": 2.763e-05, + "loss": 0.0163, + "step": 17115 + }, + { + "epoch": 14.45608108108108, + "grad_norm": 0.28965821862220764, + "learning_rate": 2.7629666666666667e-05, + "loss": 0.0175, + "step": 17116 + }, + { + "epoch": 14.456925675675675, + "grad_norm": 0.18720220029354095, + "learning_rate": 2.7629333333333333e-05, + "loss": 0.0102, + "step": 17117 + }, + { + "epoch": 14.45777027027027, + "grad_norm": 0.357986181974411, + "learning_rate": 2.7629000000000003e-05, + "loss": 0.0213, + "step": 17118 + }, + { + "epoch": 14.458614864864865, + "grad_norm": 0.4197021424770355, + "learning_rate": 2.7628666666666665e-05, + "loss": 0.0245, + "step": 17119 + }, + { + "epoch": 14.45945945945946, + "grad_norm": 0.30624425411224365, + "learning_rate": 2.7628333333333334e-05, + "loss": 0.0187, + "step": 17120 + }, + { + "epoch": 14.460304054054054, + "grad_norm": 0.25937125086784363, + "learning_rate": 2.7628000000000004e-05, + "loss": 0.014, + "step": 17121 + }, + { + "epoch": 14.46114864864865, + "grad_norm": 0.3337166905403137, + "learning_rate": 2.7627666666666666e-05, + "loss": 0.0205, + "step": 17122 + }, + { + "epoch": 14.461993243243244, + "grad_norm": 0.48006191849708557, + "learning_rate": 2.7627333333333336e-05, + "loss": 0.0201, + "step": 17123 + }, + { + "epoch": 14.462837837837839, + "grad_norm": 1.8765121698379517, + "learning_rate": 2.7627e-05, + "loss": 0.0437, + "step": 17124 + }, + { + "epoch": 14.463682432432432, + "grad_norm": 0.5273558497428894, + "learning_rate": 2.7626666666666668e-05, + "loss": 0.0304, + "step": 17125 + }, + { + "epoch": 14.464527027027026, + "grad_norm": 1.2363842725753784, + "learning_rate": 2.7626333333333333e-05, + "loss": 0.038, + "step": 17126 + }, + { + "epoch": 14.465371621621621, + "grad_norm": 0.5495951175689697, + "learning_rate": 2.7626e-05, + "loss": 0.1858, + "step": 17127 + }, + { + "epoch": 14.466216216216216, + "grad_norm": 0.5317350625991821, + "learning_rate": 2.7625666666666665e-05, + "loss": 0.1108, + "step": 17128 + }, + { + "epoch": 14.46706081081081, + "grad_norm": 0.37995272874832153, + "learning_rate": 2.7625333333333335e-05, + "loss": 0.0526, + "step": 17129 + }, + { + "epoch": 14.467905405405405, + "grad_norm": 0.38379618525505066, + "learning_rate": 2.7625e-05, + "loss": 0.0247, + "step": 17130 + }, + { + "epoch": 14.46875, + "grad_norm": 0.289315789937973, + "learning_rate": 2.7624666666666667e-05, + "loss": 0.0497, + "step": 17131 + }, + { + "epoch": 14.469594594594595, + "grad_norm": 0.5365801453590393, + "learning_rate": 2.7624333333333336e-05, + "loss": 0.0295, + "step": 17132 + }, + { + "epoch": 14.47043918918919, + "grad_norm": 0.32857745885849, + "learning_rate": 2.7624e-05, + "loss": 0.0159, + "step": 17133 + }, + { + "epoch": 14.471283783783784, + "grad_norm": 0.3251252770423889, + "learning_rate": 2.7623666666666668e-05, + "loss": 0.0244, + "step": 17134 + }, + { + "epoch": 14.472128378378379, + "grad_norm": 0.2352510541677475, + "learning_rate": 2.7623333333333334e-05, + "loss": 0.0109, + "step": 17135 + }, + { + "epoch": 14.472972972972974, + "grad_norm": 0.347474068403244, + "learning_rate": 2.7623e-05, + "loss": 0.0151, + "step": 17136 + }, + { + "epoch": 14.473817567567568, + "grad_norm": 0.208424910902977, + "learning_rate": 2.7622666666666666e-05, + "loss": 0.0239, + "step": 17137 + }, + { + "epoch": 14.474662162162161, + "grad_norm": 0.30793890357017517, + "learning_rate": 2.7622333333333335e-05, + "loss": 0.0201, + "step": 17138 + }, + { + "epoch": 14.475506756756756, + "grad_norm": 0.15446916222572327, + "learning_rate": 2.7622e-05, + "loss": 0.0155, + "step": 17139 + }, + { + "epoch": 14.47635135135135, + "grad_norm": 1.6237114667892456, + "learning_rate": 2.7621666666666667e-05, + "loss": 0.0235, + "step": 17140 + }, + { + "epoch": 14.477195945945946, + "grad_norm": 0.1601257175207138, + "learning_rate": 2.7621333333333336e-05, + "loss": 0.0102, + "step": 17141 + }, + { + "epoch": 14.47804054054054, + "grad_norm": 0.33362171053886414, + "learning_rate": 2.7621e-05, + "loss": 0.0147, + "step": 17142 + }, + { + "epoch": 14.478885135135135, + "grad_norm": 0.4558410048484802, + "learning_rate": 2.7620666666666668e-05, + "loss": 0.0251, + "step": 17143 + }, + { + "epoch": 14.47972972972973, + "grad_norm": 0.3240131437778473, + "learning_rate": 2.7620333333333334e-05, + "loss": 0.0221, + "step": 17144 + }, + { + "epoch": 14.480574324324325, + "grad_norm": 0.26787135004997253, + "learning_rate": 2.762e-05, + "loss": 0.0206, + "step": 17145 + }, + { + "epoch": 14.48141891891892, + "grad_norm": 0.3470001816749573, + "learning_rate": 2.761966666666667e-05, + "loss": 0.028, + "step": 17146 + }, + { + "epoch": 14.482263513513514, + "grad_norm": 0.38091251254081726, + "learning_rate": 2.7619333333333335e-05, + "loss": 0.0137, + "step": 17147 + }, + { + "epoch": 14.483108108108109, + "grad_norm": 0.1719188094139099, + "learning_rate": 2.7619e-05, + "loss": 0.0096, + "step": 17148 + }, + { + "epoch": 14.483952702702704, + "grad_norm": 0.2822992205619812, + "learning_rate": 2.7618666666666667e-05, + "loss": 0.0214, + "step": 17149 + }, + { + "epoch": 14.484797297297296, + "grad_norm": 0.4705917239189148, + "learning_rate": 2.7618333333333336e-05, + "loss": 0.0599, + "step": 17150 + }, + { + "epoch": 14.485641891891891, + "grad_norm": 0.334125280380249, + "learning_rate": 2.7618e-05, + "loss": 0.0228, + "step": 17151 + }, + { + "epoch": 14.486486486486486, + "grad_norm": 0.28045856952667236, + "learning_rate": 2.7617666666666668e-05, + "loss": 0.0129, + "step": 17152 + }, + { + "epoch": 14.48733108108108, + "grad_norm": 0.1563519686460495, + "learning_rate": 2.7617333333333334e-05, + "loss": 0.0134, + "step": 17153 + }, + { + "epoch": 14.488175675675675, + "grad_norm": 0.25821182131767273, + "learning_rate": 2.7617e-05, + "loss": 0.0125, + "step": 17154 + }, + { + "epoch": 14.48902027027027, + "grad_norm": 0.39911597967147827, + "learning_rate": 2.761666666666667e-05, + "loss": 0.0166, + "step": 17155 + }, + { + "epoch": 14.489864864864865, + "grad_norm": 0.3288264870643616, + "learning_rate": 2.7616333333333335e-05, + "loss": 0.0195, + "step": 17156 + }, + { + "epoch": 14.49070945945946, + "grad_norm": 0.5876527428627014, + "learning_rate": 2.7616e-05, + "loss": 0.0188, + "step": 17157 + }, + { + "epoch": 14.491554054054054, + "grad_norm": 0.27031955122947693, + "learning_rate": 2.7615666666666667e-05, + "loss": 0.0281, + "step": 17158 + }, + { + "epoch": 14.49239864864865, + "grad_norm": 0.25191277265548706, + "learning_rate": 2.7615333333333333e-05, + "loss": 0.0159, + "step": 17159 + }, + { + "epoch": 14.493243243243244, + "grad_norm": 0.41572001576423645, + "learning_rate": 2.7615e-05, + "loss": 0.0285, + "step": 17160 + }, + { + "epoch": 14.494087837837839, + "grad_norm": 0.2766179144382477, + "learning_rate": 2.761466666666667e-05, + "loss": 0.0118, + "step": 17161 + }, + { + "epoch": 14.494932432432432, + "grad_norm": 0.32387781143188477, + "learning_rate": 2.761433333333333e-05, + "loss": 0.0152, + "step": 17162 + }, + { + "epoch": 14.495777027027026, + "grad_norm": 0.29841092228889465, + "learning_rate": 2.7614e-05, + "loss": 0.0197, + "step": 17163 + }, + { + "epoch": 14.496621621621621, + "grad_norm": 0.7002884149551392, + "learning_rate": 2.761366666666667e-05, + "loss": 0.0198, + "step": 17164 + }, + { + "epoch": 14.497466216216216, + "grad_norm": 0.3084958493709564, + "learning_rate": 2.7613333333333332e-05, + "loss": 0.0162, + "step": 17165 + }, + { + "epoch": 14.49831081081081, + "grad_norm": 0.4539986252784729, + "learning_rate": 2.7613e-05, + "loss": 0.0396, + "step": 17166 + }, + { + "epoch": 14.499155405405405, + "grad_norm": 0.30652564764022827, + "learning_rate": 2.7612666666666667e-05, + "loss": 0.0161, + "step": 17167 + }, + { + "epoch": 14.5, + "grad_norm": 0.5140907764434814, + "learning_rate": 2.7612333333333333e-05, + "loss": 0.0235, + "step": 17168 + }, + { + "epoch": 14.500844594594595, + "grad_norm": 0.42724519968032837, + "learning_rate": 2.7612e-05, + "loss": 0.0249, + "step": 17169 + }, + { + "epoch": 14.50168918918919, + "grad_norm": 0.43857237696647644, + "learning_rate": 2.761166666666667e-05, + "loss": 0.0236, + "step": 17170 + }, + { + "epoch": 14.502533783783784, + "grad_norm": 0.5415659546852112, + "learning_rate": 2.7611333333333334e-05, + "loss": 0.0261, + "step": 17171 + }, + { + "epoch": 14.503378378378379, + "grad_norm": 0.6345382332801819, + "learning_rate": 2.7611e-05, + "loss": 0.0292, + "step": 17172 + }, + { + "epoch": 14.504222972972974, + "grad_norm": 0.5798710584640503, + "learning_rate": 2.761066666666667e-05, + "loss": 0.0249, + "step": 17173 + }, + { + "epoch": 14.505067567567568, + "grad_norm": 0.6168501973152161, + "learning_rate": 2.7610333333333332e-05, + "loss": 0.0343, + "step": 17174 + }, + { + "epoch": 14.505912162162161, + "grad_norm": 0.3859148323535919, + "learning_rate": 2.761e-05, + "loss": 0.0229, + "step": 17175 + }, + { + "epoch": 14.506756756756756, + "grad_norm": 0.6559233069419861, + "learning_rate": 2.7609666666666668e-05, + "loss": 0.0253, + "step": 17176 + }, + { + "epoch": 14.50760135135135, + "grad_norm": 0.520695686340332, + "learning_rate": 2.7609333333333333e-05, + "loss": 0.1497, + "step": 17177 + }, + { + "epoch": 14.508445945945946, + "grad_norm": 0.6917375922203064, + "learning_rate": 2.7609e-05, + "loss": 0.1392, + "step": 17178 + }, + { + "epoch": 14.50929054054054, + "grad_norm": 0.35469409823417664, + "learning_rate": 2.760866666666667e-05, + "loss": 0.0547, + "step": 17179 + }, + { + "epoch": 14.510135135135135, + "grad_norm": 0.4345727562904358, + "learning_rate": 2.7608333333333335e-05, + "loss": 0.0353, + "step": 17180 + }, + { + "epoch": 14.51097972972973, + "grad_norm": 0.27308475971221924, + "learning_rate": 2.7608e-05, + "loss": 0.0254, + "step": 17181 + }, + { + "epoch": 14.511824324324325, + "grad_norm": 0.6915180683135986, + "learning_rate": 2.760766666666667e-05, + "loss": 0.0302, + "step": 17182 + }, + { + "epoch": 14.51266891891892, + "grad_norm": 0.29120883345603943, + "learning_rate": 2.7607333333333332e-05, + "loss": 0.0136, + "step": 17183 + }, + { + "epoch": 14.513513513513514, + "grad_norm": 0.3618503510951996, + "learning_rate": 2.7607000000000002e-05, + "loss": 0.0236, + "step": 17184 + }, + { + "epoch": 14.514358108108109, + "grad_norm": 0.4160272777080536, + "learning_rate": 2.7606666666666668e-05, + "loss": 0.0163, + "step": 17185 + }, + { + "epoch": 14.515202702702704, + "grad_norm": 0.6133851408958435, + "learning_rate": 2.7606333333333334e-05, + "loss": 0.0244, + "step": 17186 + }, + { + "epoch": 14.516047297297296, + "grad_norm": 0.294141948223114, + "learning_rate": 2.7606e-05, + "loss": 0.017, + "step": 17187 + }, + { + "epoch": 14.516891891891891, + "grad_norm": 0.3507195711135864, + "learning_rate": 2.7605666666666666e-05, + "loss": 0.0172, + "step": 17188 + }, + { + "epoch": 14.517736486486486, + "grad_norm": 0.38997882604599, + "learning_rate": 2.7605333333333335e-05, + "loss": 0.0242, + "step": 17189 + }, + { + "epoch": 14.51858108108108, + "grad_norm": 0.3711439371109009, + "learning_rate": 2.7605e-05, + "loss": 0.0287, + "step": 17190 + }, + { + "epoch": 14.519425675675675, + "grad_norm": 0.2120158076286316, + "learning_rate": 2.7604666666666667e-05, + "loss": 0.0211, + "step": 17191 + }, + { + "epoch": 14.52027027027027, + "grad_norm": 0.5150094628334045, + "learning_rate": 2.7604333333333333e-05, + "loss": 0.0302, + "step": 17192 + }, + { + "epoch": 14.521114864864865, + "grad_norm": 0.4118615388870239, + "learning_rate": 2.7604000000000002e-05, + "loss": 0.0369, + "step": 17193 + }, + { + "epoch": 14.52195945945946, + "grad_norm": 0.30774009227752686, + "learning_rate": 2.7603666666666665e-05, + "loss": 0.0193, + "step": 17194 + }, + { + "epoch": 14.522804054054054, + "grad_norm": 0.29474928975105286, + "learning_rate": 2.7603333333333334e-05, + "loss": 0.0167, + "step": 17195 + }, + { + "epoch": 14.52364864864865, + "grad_norm": 0.31007811427116394, + "learning_rate": 2.7603000000000003e-05, + "loss": 0.015, + "step": 17196 + }, + { + "epoch": 14.524493243243244, + "grad_norm": 0.19651103019714355, + "learning_rate": 2.7602666666666666e-05, + "loss": 0.0166, + "step": 17197 + }, + { + "epoch": 14.525337837837839, + "grad_norm": 0.4018732011318207, + "learning_rate": 2.7602333333333335e-05, + "loss": 0.0167, + "step": 17198 + }, + { + "epoch": 14.526182432432432, + "grad_norm": 0.19503569602966309, + "learning_rate": 2.7602e-05, + "loss": 0.015, + "step": 17199 + }, + { + "epoch": 14.527027027027026, + "grad_norm": 0.34628546237945557, + "learning_rate": 2.7601666666666667e-05, + "loss": 0.0228, + "step": 17200 + }, + { + "epoch": 14.527871621621621, + "grad_norm": 0.35883283615112305, + "learning_rate": 2.7601333333333333e-05, + "loss": 0.0234, + "step": 17201 + }, + { + "epoch": 14.528716216216216, + "grad_norm": 0.15314149856567383, + "learning_rate": 2.7601000000000002e-05, + "loss": 0.0082, + "step": 17202 + }, + { + "epoch": 14.52956081081081, + "grad_norm": 0.30155086517333984, + "learning_rate": 2.7600666666666665e-05, + "loss": 0.033, + "step": 17203 + }, + { + "epoch": 14.530405405405405, + "grad_norm": 0.21244828402996063, + "learning_rate": 2.7600333333333334e-05, + "loss": 0.0186, + "step": 17204 + }, + { + "epoch": 14.53125, + "grad_norm": 0.16538988053798676, + "learning_rate": 2.7600000000000003e-05, + "loss": 0.0087, + "step": 17205 + }, + { + "epoch": 14.532094594594595, + "grad_norm": 0.4069901406764984, + "learning_rate": 2.7599666666666666e-05, + "loss": 0.0253, + "step": 17206 + }, + { + "epoch": 14.53293918918919, + "grad_norm": 0.4694984257221222, + "learning_rate": 2.7599333333333335e-05, + "loss": 0.0196, + "step": 17207 + }, + { + "epoch": 14.533783783783784, + "grad_norm": 0.40024667978286743, + "learning_rate": 2.7599e-05, + "loss": 0.0172, + "step": 17208 + }, + { + "epoch": 14.534628378378379, + "grad_norm": 0.2474476397037506, + "learning_rate": 2.7598666666666667e-05, + "loss": 0.0169, + "step": 17209 + }, + { + "epoch": 14.535472972972974, + "grad_norm": 0.24533939361572266, + "learning_rate": 2.7598333333333333e-05, + "loss": 0.0111, + "step": 17210 + }, + { + "epoch": 14.536317567567568, + "grad_norm": 0.29910749197006226, + "learning_rate": 2.7598000000000002e-05, + "loss": 0.0215, + "step": 17211 + }, + { + "epoch": 14.537162162162161, + "grad_norm": 0.22803419828414917, + "learning_rate": 2.7597666666666665e-05, + "loss": 0.0139, + "step": 17212 + }, + { + "epoch": 14.538006756756756, + "grad_norm": 2.050431489944458, + "learning_rate": 2.7597333333333334e-05, + "loss": 0.0139, + "step": 17213 + }, + { + "epoch": 14.53885135135135, + "grad_norm": 0.4118911623954773, + "learning_rate": 2.7597000000000004e-05, + "loss": 0.0166, + "step": 17214 + }, + { + "epoch": 14.539695945945946, + "grad_norm": 0.23074054718017578, + "learning_rate": 2.7596666666666666e-05, + "loss": 0.0154, + "step": 17215 + }, + { + "epoch": 14.54054054054054, + "grad_norm": 0.4223705530166626, + "learning_rate": 2.7596333333333335e-05, + "loss": 0.0209, + "step": 17216 + }, + { + "epoch": 14.541385135135135, + "grad_norm": 0.48165184259414673, + "learning_rate": 2.7596e-05, + "loss": 0.0291, + "step": 17217 + }, + { + "epoch": 14.54222972972973, + "grad_norm": 0.2090214490890503, + "learning_rate": 2.7595666666666667e-05, + "loss": 0.0202, + "step": 17218 + }, + { + "epoch": 14.543074324324325, + "grad_norm": 0.25740790367126465, + "learning_rate": 2.7595333333333333e-05, + "loss": 0.0167, + "step": 17219 + }, + { + "epoch": 14.54391891891892, + "grad_norm": 0.4711361229419708, + "learning_rate": 2.7595e-05, + "loss": 0.0215, + "step": 17220 + }, + { + "epoch": 14.544763513513514, + "grad_norm": 0.8634582161903381, + "learning_rate": 2.759466666666667e-05, + "loss": 0.0307, + "step": 17221 + }, + { + "epoch": 14.545608108108109, + "grad_norm": 0.5196779370307922, + "learning_rate": 2.7594333333333334e-05, + "loss": 0.0229, + "step": 17222 + }, + { + "epoch": 14.546452702702704, + "grad_norm": 0.17052946984767914, + "learning_rate": 2.7594e-05, + "loss": 0.0127, + "step": 17223 + }, + { + "epoch": 14.547297297297296, + "grad_norm": 0.40486276149749756, + "learning_rate": 2.7593666666666666e-05, + "loss": 0.0173, + "step": 17224 + }, + { + "epoch": 14.548141891891891, + "grad_norm": 1.3055481910705566, + "learning_rate": 2.7593333333333336e-05, + "loss": 0.0219, + "step": 17225 + }, + { + "epoch": 14.548986486486486, + "grad_norm": 0.5728471279144287, + "learning_rate": 2.7592999999999998e-05, + "loss": 0.0399, + "step": 17226 + }, + { + "epoch": 14.54983108108108, + "grad_norm": 0.9042392373085022, + "learning_rate": 2.7592666666666668e-05, + "loss": 0.2108, + "step": 17227 + }, + { + "epoch": 14.550675675675675, + "grad_norm": 0.4352417588233948, + "learning_rate": 2.7592333333333333e-05, + "loss": 0.1227, + "step": 17228 + }, + { + "epoch": 14.55152027027027, + "grad_norm": 0.36972376704216003, + "learning_rate": 2.7592e-05, + "loss": 0.0616, + "step": 17229 + }, + { + "epoch": 14.552364864864865, + "grad_norm": 0.3299458622932434, + "learning_rate": 2.759166666666667e-05, + "loss": 0.03, + "step": 17230 + }, + { + "epoch": 14.55320945945946, + "grad_norm": 0.34949547052383423, + "learning_rate": 2.7591333333333335e-05, + "loss": 0.0269, + "step": 17231 + }, + { + "epoch": 14.554054054054054, + "grad_norm": 0.27991095185279846, + "learning_rate": 2.7591e-05, + "loss": 0.0161, + "step": 17232 + }, + { + "epoch": 14.55489864864865, + "grad_norm": 0.4973297417163849, + "learning_rate": 2.7590666666666667e-05, + "loss": 0.0213, + "step": 17233 + }, + { + "epoch": 14.555743243243244, + "grad_norm": 0.20689596235752106, + "learning_rate": 2.7590333333333336e-05, + "loss": 0.0201, + "step": 17234 + }, + { + "epoch": 14.556587837837839, + "grad_norm": 0.29861605167388916, + "learning_rate": 2.759e-05, + "loss": 0.0223, + "step": 17235 + }, + { + "epoch": 14.557432432432432, + "grad_norm": 0.3414527177810669, + "learning_rate": 2.7589666666666668e-05, + "loss": 0.0164, + "step": 17236 + }, + { + "epoch": 14.558277027027026, + "grad_norm": 0.2227717787027359, + "learning_rate": 2.7589333333333334e-05, + "loss": 0.0303, + "step": 17237 + }, + { + "epoch": 14.559121621621621, + "grad_norm": 0.2233225256204605, + "learning_rate": 2.7589e-05, + "loss": 0.025, + "step": 17238 + }, + { + "epoch": 14.559966216216216, + "grad_norm": 0.24185964465141296, + "learning_rate": 2.758866666666667e-05, + "loss": 0.022, + "step": 17239 + }, + { + "epoch": 14.56081081081081, + "grad_norm": 0.13952137529850006, + "learning_rate": 2.7588333333333335e-05, + "loss": 0.0123, + "step": 17240 + }, + { + "epoch": 14.561655405405405, + "grad_norm": 0.19767692685127258, + "learning_rate": 2.7588e-05, + "loss": 0.0163, + "step": 17241 + }, + { + "epoch": 14.5625, + "grad_norm": 0.2876833379268646, + "learning_rate": 2.7587666666666667e-05, + "loss": 0.012, + "step": 17242 + }, + { + "epoch": 14.563344594594595, + "grad_norm": 0.19179274141788483, + "learning_rate": 2.7587333333333336e-05, + "loss": 0.0123, + "step": 17243 + }, + { + "epoch": 14.56418918918919, + "grad_norm": 0.5204126238822937, + "learning_rate": 2.7587e-05, + "loss": 0.0195, + "step": 17244 + }, + { + "epoch": 14.565033783783784, + "grad_norm": 0.21649979054927826, + "learning_rate": 2.7586666666666668e-05, + "loss": 0.0182, + "step": 17245 + }, + { + "epoch": 14.565878378378379, + "grad_norm": 0.8234124779701233, + "learning_rate": 2.7586333333333337e-05, + "loss": 0.0164, + "step": 17246 + }, + { + "epoch": 14.566722972972974, + "grad_norm": 1.7644121646881104, + "learning_rate": 2.7586e-05, + "loss": 0.026, + "step": 17247 + }, + { + "epoch": 14.567567567567568, + "grad_norm": 0.245303675532341, + "learning_rate": 2.758566666666667e-05, + "loss": 0.0123, + "step": 17248 + }, + { + "epoch": 14.568412162162161, + "grad_norm": 0.3374214470386505, + "learning_rate": 2.758533333333333e-05, + "loss": 0.0196, + "step": 17249 + }, + { + "epoch": 14.569256756756756, + "grad_norm": 0.6126911044120789, + "learning_rate": 2.7585e-05, + "loss": 0.0195, + "step": 17250 + }, + { + "epoch": 14.57010135135135, + "grad_norm": 0.15290942788124084, + "learning_rate": 2.7584666666666667e-05, + "loss": 0.0102, + "step": 17251 + }, + { + "epoch": 14.570945945945946, + "grad_norm": 0.3094507157802582, + "learning_rate": 2.7584333333333333e-05, + "loss": 0.0149, + "step": 17252 + }, + { + "epoch": 14.57179054054054, + "grad_norm": 0.2968059182167053, + "learning_rate": 2.7584e-05, + "loss": 0.0122, + "step": 17253 + }, + { + "epoch": 14.572635135135135, + "grad_norm": 0.2479894906282425, + "learning_rate": 2.7583666666666668e-05, + "loss": 0.0174, + "step": 17254 + }, + { + "epoch": 14.57347972972973, + "grad_norm": 1.0240620374679565, + "learning_rate": 2.7583333333333334e-05, + "loss": 0.0194, + "step": 17255 + }, + { + "epoch": 14.574324324324325, + "grad_norm": 0.4399951994419098, + "learning_rate": 2.7583e-05, + "loss": 0.0187, + "step": 17256 + }, + { + "epoch": 14.57516891891892, + "grad_norm": 0.37971359491348267, + "learning_rate": 2.758266666666667e-05, + "loss": 0.021, + "step": 17257 + }, + { + "epoch": 14.576013513513514, + "grad_norm": 0.26653802394866943, + "learning_rate": 2.7582333333333332e-05, + "loss": 0.0184, + "step": 17258 + }, + { + "epoch": 14.576858108108109, + "grad_norm": 0.33573946356773376, + "learning_rate": 2.7582e-05, + "loss": 0.0191, + "step": 17259 + }, + { + "epoch": 14.577702702702704, + "grad_norm": 0.2228344976902008, + "learning_rate": 2.7581666666666667e-05, + "loss": 0.0089, + "step": 17260 + }, + { + "epoch": 14.578547297297296, + "grad_norm": 0.25912898778915405, + "learning_rate": 2.7581333333333333e-05, + "loss": 0.0158, + "step": 17261 + }, + { + "epoch": 14.579391891891891, + "grad_norm": 0.25870540738105774, + "learning_rate": 2.7581e-05, + "loss": 0.0212, + "step": 17262 + }, + { + "epoch": 14.580236486486486, + "grad_norm": 0.25008323788642883, + "learning_rate": 2.758066666666667e-05, + "loss": 0.0238, + "step": 17263 + }, + { + "epoch": 14.58108108108108, + "grad_norm": 0.2931080162525177, + "learning_rate": 2.7580333333333334e-05, + "loss": 0.0149, + "step": 17264 + }, + { + "epoch": 14.581925675675675, + "grad_norm": 0.4015907645225525, + "learning_rate": 2.758e-05, + "loss": 0.0133, + "step": 17265 + }, + { + "epoch": 14.58277027027027, + "grad_norm": 0.27270835638046265, + "learning_rate": 2.757966666666667e-05, + "loss": 0.0167, + "step": 17266 + }, + { + "epoch": 14.583614864864865, + "grad_norm": 0.2390909045934677, + "learning_rate": 2.7579333333333332e-05, + "loss": 0.0127, + "step": 17267 + }, + { + "epoch": 14.58445945945946, + "grad_norm": 0.2108451873064041, + "learning_rate": 2.7579e-05, + "loss": 0.0107, + "step": 17268 + }, + { + "epoch": 14.585304054054054, + "grad_norm": 0.5885406732559204, + "learning_rate": 2.7578666666666667e-05, + "loss": 0.0238, + "step": 17269 + }, + { + "epoch": 14.58614864864865, + "grad_norm": 0.37110435962677, + "learning_rate": 2.7578333333333333e-05, + "loss": 0.0163, + "step": 17270 + }, + { + "epoch": 14.586993243243244, + "grad_norm": 0.45959171652793884, + "learning_rate": 2.7578000000000003e-05, + "loss": 0.0132, + "step": 17271 + }, + { + "epoch": 14.587837837837839, + "grad_norm": 0.2508629560470581, + "learning_rate": 2.757766666666667e-05, + "loss": 0.0131, + "step": 17272 + }, + { + "epoch": 14.588682432432432, + "grad_norm": 0.22853754460811615, + "learning_rate": 2.7577333333333334e-05, + "loss": 0.0146, + "step": 17273 + }, + { + "epoch": 14.589527027027026, + "grad_norm": 0.28127366304397583, + "learning_rate": 2.7577e-05, + "loss": 0.0178, + "step": 17274 + }, + { + "epoch": 14.590371621621621, + "grad_norm": 0.20090752840042114, + "learning_rate": 2.757666666666667e-05, + "loss": 0.0125, + "step": 17275 + }, + { + "epoch": 14.591216216216216, + "grad_norm": 0.8779267072677612, + "learning_rate": 2.7576333333333332e-05, + "loss": 0.0401, + "step": 17276 + }, + { + "epoch": 14.59206081081081, + "grad_norm": 0.8144975900650024, + "learning_rate": 2.7576e-05, + "loss": 0.2334, + "step": 17277 + }, + { + "epoch": 14.592905405405405, + "grad_norm": 0.5015342235565186, + "learning_rate": 2.7575666666666668e-05, + "loss": 0.1093, + "step": 17278 + }, + { + "epoch": 14.59375, + "grad_norm": 0.38418829441070557, + "learning_rate": 2.7575333333333333e-05, + "loss": 0.0642, + "step": 17279 + }, + { + "epoch": 14.594594594594595, + "grad_norm": 0.3709740936756134, + "learning_rate": 2.7575000000000003e-05, + "loss": 0.0425, + "step": 17280 + }, + { + "epoch": 14.59543918918919, + "grad_norm": 0.44052451848983765, + "learning_rate": 2.7574666666666665e-05, + "loss": 0.0147, + "step": 17281 + }, + { + "epoch": 14.596283783783784, + "grad_norm": 0.30707696080207825, + "learning_rate": 2.7574333333333335e-05, + "loss": 0.026, + "step": 17282 + }, + { + "epoch": 14.597128378378379, + "grad_norm": 0.2610158920288086, + "learning_rate": 2.7574e-05, + "loss": 0.0204, + "step": 17283 + }, + { + "epoch": 14.597972972972974, + "grad_norm": 0.2494453489780426, + "learning_rate": 2.7573666666666667e-05, + "loss": 0.0142, + "step": 17284 + }, + { + "epoch": 14.598817567567568, + "grad_norm": 0.2506932318210602, + "learning_rate": 2.7573333333333332e-05, + "loss": 0.0181, + "step": 17285 + }, + { + "epoch": 14.599662162162161, + "grad_norm": 0.23814332485198975, + "learning_rate": 2.7573000000000002e-05, + "loss": 0.0291, + "step": 17286 + }, + { + "epoch": 14.600506756756756, + "grad_norm": 0.27788808941841125, + "learning_rate": 2.7572666666666664e-05, + "loss": 0.0169, + "step": 17287 + }, + { + "epoch": 14.60135135135135, + "grad_norm": 0.2317017912864685, + "learning_rate": 2.7572333333333334e-05, + "loss": 0.014, + "step": 17288 + }, + { + "epoch": 14.602195945945946, + "grad_norm": 0.19420847296714783, + "learning_rate": 2.7572000000000003e-05, + "loss": 0.0244, + "step": 17289 + }, + { + "epoch": 14.60304054054054, + "grad_norm": 0.29579052329063416, + "learning_rate": 2.7571666666666666e-05, + "loss": 0.0167, + "step": 17290 + }, + { + "epoch": 14.603885135135135, + "grad_norm": 0.21692775189876556, + "learning_rate": 2.7571333333333335e-05, + "loss": 0.0118, + "step": 17291 + }, + { + "epoch": 14.60472972972973, + "grad_norm": 0.5278792381286621, + "learning_rate": 2.7571e-05, + "loss": 0.0257, + "step": 17292 + }, + { + "epoch": 14.605574324324325, + "grad_norm": 0.19108180701732635, + "learning_rate": 2.7570666666666667e-05, + "loss": 0.0164, + "step": 17293 + }, + { + "epoch": 14.60641891891892, + "grad_norm": 0.2607153058052063, + "learning_rate": 2.7570333333333333e-05, + "loss": 0.0167, + "step": 17294 + }, + { + "epoch": 14.607263513513514, + "grad_norm": 0.1489918977022171, + "learning_rate": 2.7570000000000002e-05, + "loss": 0.0077, + "step": 17295 + }, + { + "epoch": 14.608108108108109, + "grad_norm": 0.3449654281139374, + "learning_rate": 2.7569666666666668e-05, + "loss": 0.0347, + "step": 17296 + }, + { + "epoch": 14.608952702702704, + "grad_norm": 0.2083059549331665, + "learning_rate": 2.7569333333333334e-05, + "loss": 0.0141, + "step": 17297 + }, + { + "epoch": 14.609797297297296, + "grad_norm": 0.12328977137804031, + "learning_rate": 2.7569000000000003e-05, + "loss": 0.0104, + "step": 17298 + }, + { + "epoch": 14.610641891891891, + "grad_norm": 0.2636164128780365, + "learning_rate": 2.7568666666666666e-05, + "loss": 0.0175, + "step": 17299 + }, + { + "epoch": 14.611486486486486, + "grad_norm": 0.16386081278324127, + "learning_rate": 2.7568333333333335e-05, + "loss": 0.0149, + "step": 17300 + }, + { + "epoch": 14.61233108108108, + "grad_norm": 0.3551594018936157, + "learning_rate": 2.7568e-05, + "loss": 0.0125, + "step": 17301 + }, + { + "epoch": 14.613175675675675, + "grad_norm": 0.30843618512153625, + "learning_rate": 2.7567666666666667e-05, + "loss": 0.023, + "step": 17302 + }, + { + "epoch": 14.61402027027027, + "grad_norm": 0.3488873541355133, + "learning_rate": 2.7567333333333333e-05, + "loss": 0.0146, + "step": 17303 + }, + { + "epoch": 14.614864864864865, + "grad_norm": 0.28247103095054626, + "learning_rate": 2.7567000000000002e-05, + "loss": 0.0143, + "step": 17304 + }, + { + "epoch": 14.61570945945946, + "grad_norm": 0.26341789960861206, + "learning_rate": 2.7566666666666668e-05, + "loss": 0.0093, + "step": 17305 + }, + { + "epoch": 14.616554054054054, + "grad_norm": 0.28720560669898987, + "learning_rate": 2.7566333333333334e-05, + "loss": 0.0093, + "step": 17306 + }, + { + "epoch": 14.61739864864865, + "grad_norm": 0.29908302426338196, + "learning_rate": 2.7566000000000003e-05, + "loss": 0.0135, + "step": 17307 + }, + { + "epoch": 14.618243243243244, + "grad_norm": 0.2685949504375458, + "learning_rate": 2.7565666666666666e-05, + "loss": 0.0131, + "step": 17308 + }, + { + "epoch": 14.619087837837839, + "grad_norm": 0.39052489399909973, + "learning_rate": 2.7565333333333335e-05, + "loss": 0.0265, + "step": 17309 + }, + { + "epoch": 14.619932432432432, + "grad_norm": 0.23123179376125336, + "learning_rate": 2.7564999999999998e-05, + "loss": 0.0138, + "step": 17310 + }, + { + "epoch": 14.620777027027026, + "grad_norm": 0.3499506413936615, + "learning_rate": 2.7564666666666667e-05, + "loss": 0.019, + "step": 17311 + }, + { + "epoch": 14.621621621621621, + "grad_norm": 0.22031402587890625, + "learning_rate": 2.7564333333333333e-05, + "loss": 0.0162, + "step": 17312 + }, + { + "epoch": 14.622466216216216, + "grad_norm": 0.1709207147359848, + "learning_rate": 2.7564e-05, + "loss": 0.0129, + "step": 17313 + }, + { + "epoch": 14.62331081081081, + "grad_norm": 0.2675308883190155, + "learning_rate": 2.756366666666667e-05, + "loss": 0.0116, + "step": 17314 + }, + { + "epoch": 14.624155405405405, + "grad_norm": 0.3790448009967804, + "learning_rate": 2.7563333333333334e-05, + "loss": 0.0162, + "step": 17315 + }, + { + "epoch": 14.625, + "grad_norm": 0.2674373686313629, + "learning_rate": 2.7563e-05, + "loss": 0.0122, + "step": 17316 + }, + { + "epoch": 14.625844594594595, + "grad_norm": 0.32066699862480164, + "learning_rate": 2.7562666666666666e-05, + "loss": 0.0193, + "step": 17317 + }, + { + "epoch": 14.62668918918919, + "grad_norm": 0.28270700573921204, + "learning_rate": 2.7562333333333335e-05, + "loss": 0.0157, + "step": 17318 + }, + { + "epoch": 14.627533783783784, + "grad_norm": 0.2089986354112625, + "learning_rate": 2.7561999999999998e-05, + "loss": 0.0072, + "step": 17319 + }, + { + "epoch": 14.628378378378379, + "grad_norm": 0.731052815914154, + "learning_rate": 2.7561666666666667e-05, + "loss": 0.0186, + "step": 17320 + }, + { + "epoch": 14.629222972972974, + "grad_norm": 0.31542131304740906, + "learning_rate": 2.7561333333333337e-05, + "loss": 0.0235, + "step": 17321 + }, + { + "epoch": 14.630067567567568, + "grad_norm": 0.6298955082893372, + "learning_rate": 2.7561e-05, + "loss": 0.0211, + "step": 17322 + }, + { + "epoch": 14.630912162162161, + "grad_norm": 0.4530806243419647, + "learning_rate": 2.756066666666667e-05, + "loss": 0.0137, + "step": 17323 + }, + { + "epoch": 14.631756756756756, + "grad_norm": 0.28883442282676697, + "learning_rate": 2.7560333333333334e-05, + "loss": 0.0144, + "step": 17324 + }, + { + "epoch": 14.63260135135135, + "grad_norm": 0.2406596690416336, + "learning_rate": 2.756e-05, + "loss": 0.0133, + "step": 17325 + }, + { + "epoch": 14.633445945945946, + "grad_norm": 0.48386919498443604, + "learning_rate": 2.7559666666666666e-05, + "loss": 0.0251, + "step": 17326 + }, + { + "epoch": 14.63429054054054, + "grad_norm": 0.6890122294425964, + "learning_rate": 2.7559333333333336e-05, + "loss": 0.1465, + "step": 17327 + }, + { + "epoch": 14.635135135135135, + "grad_norm": 0.6326867341995239, + "learning_rate": 2.7558999999999998e-05, + "loss": 0.1183, + "step": 17328 + }, + { + "epoch": 14.63597972972973, + "grad_norm": 0.34072574973106384, + "learning_rate": 2.7558666666666668e-05, + "loss": 0.0543, + "step": 17329 + }, + { + "epoch": 14.636824324324325, + "grad_norm": 0.3867832124233246, + "learning_rate": 2.7558333333333337e-05, + "loss": 0.0252, + "step": 17330 + }, + { + "epoch": 14.63766891891892, + "grad_norm": 0.27825257182121277, + "learning_rate": 2.7558e-05, + "loss": 0.0228, + "step": 17331 + }, + { + "epoch": 14.638513513513514, + "grad_norm": 0.30272045731544495, + "learning_rate": 2.755766666666667e-05, + "loss": 0.0295, + "step": 17332 + }, + { + "epoch": 14.639358108108109, + "grad_norm": 0.30721721053123474, + "learning_rate": 2.7557333333333335e-05, + "loss": 0.0199, + "step": 17333 + }, + { + "epoch": 14.640202702702704, + "grad_norm": 0.23038697242736816, + "learning_rate": 2.7557e-05, + "loss": 0.018, + "step": 17334 + }, + { + "epoch": 14.641047297297296, + "grad_norm": 0.346199631690979, + "learning_rate": 2.7556666666666667e-05, + "loss": 0.0191, + "step": 17335 + }, + { + "epoch": 14.641891891891891, + "grad_norm": 0.2085748165845871, + "learning_rate": 2.7556333333333336e-05, + "loss": 0.0224, + "step": 17336 + }, + { + "epoch": 14.642736486486486, + "grad_norm": 0.1900370866060257, + "learning_rate": 2.7556e-05, + "loss": 0.0154, + "step": 17337 + }, + { + "epoch": 14.64358108108108, + "grad_norm": 0.30152198672294617, + "learning_rate": 2.7555666666666668e-05, + "loss": 0.0236, + "step": 17338 + }, + { + "epoch": 14.644425675675675, + "grad_norm": 0.3884243965148926, + "learning_rate": 2.7555333333333334e-05, + "loss": 0.0193, + "step": 17339 + }, + { + "epoch": 14.64527027027027, + "grad_norm": 0.3583484888076782, + "learning_rate": 2.7555e-05, + "loss": 0.0219, + "step": 17340 + }, + { + "epoch": 14.646114864864865, + "grad_norm": 0.2555959224700928, + "learning_rate": 2.755466666666667e-05, + "loss": 0.0108, + "step": 17341 + }, + { + "epoch": 14.64695945945946, + "grad_norm": 0.2431381642818451, + "learning_rate": 2.755433333333333e-05, + "loss": 0.0113, + "step": 17342 + }, + { + "epoch": 14.647804054054054, + "grad_norm": 0.19573691487312317, + "learning_rate": 2.7554e-05, + "loss": 0.0144, + "step": 17343 + }, + { + "epoch": 14.64864864864865, + "grad_norm": 0.33790066838264465, + "learning_rate": 2.7553666666666667e-05, + "loss": 0.0187, + "step": 17344 + }, + { + "epoch": 14.649493243243244, + "grad_norm": 0.2043997347354889, + "learning_rate": 2.7553333333333333e-05, + "loss": 0.0152, + "step": 17345 + }, + { + "epoch": 14.650337837837839, + "grad_norm": 0.22955502569675446, + "learning_rate": 2.7553000000000002e-05, + "loss": 0.0122, + "step": 17346 + }, + { + "epoch": 14.651182432432432, + "grad_norm": 0.27289003133773804, + "learning_rate": 2.7552666666666668e-05, + "loss": 0.0206, + "step": 17347 + }, + { + "epoch": 14.652027027027026, + "grad_norm": 0.2824997305870056, + "learning_rate": 2.7552333333333334e-05, + "loss": 0.0173, + "step": 17348 + }, + { + "epoch": 14.652871621621621, + "grad_norm": 0.252495676279068, + "learning_rate": 2.7552e-05, + "loss": 0.0194, + "step": 17349 + }, + { + "epoch": 14.653716216216216, + "grad_norm": 0.6460283994674683, + "learning_rate": 2.755166666666667e-05, + "loss": 0.0222, + "step": 17350 + }, + { + "epoch": 14.65456081081081, + "grad_norm": 0.19722123444080353, + "learning_rate": 2.755133333333333e-05, + "loss": 0.0122, + "step": 17351 + }, + { + "epoch": 14.655405405405405, + "grad_norm": 0.18770569562911987, + "learning_rate": 2.7551e-05, + "loss": 0.0085, + "step": 17352 + }, + { + "epoch": 14.65625, + "grad_norm": 0.5555214881896973, + "learning_rate": 2.7550666666666667e-05, + "loss": 0.0209, + "step": 17353 + }, + { + "epoch": 14.657094594594595, + "grad_norm": 0.17884637415409088, + "learning_rate": 2.7550333333333333e-05, + "loss": 0.0101, + "step": 17354 + }, + { + "epoch": 14.65793918918919, + "grad_norm": 0.3500117063522339, + "learning_rate": 2.7550000000000002e-05, + "loss": 0.0149, + "step": 17355 + }, + { + "epoch": 14.658783783783784, + "grad_norm": 0.38195809721946716, + "learning_rate": 2.7549666666666668e-05, + "loss": 0.0172, + "step": 17356 + }, + { + "epoch": 14.659628378378379, + "grad_norm": 0.3501572906970978, + "learning_rate": 2.7549333333333334e-05, + "loss": 0.017, + "step": 17357 + }, + { + "epoch": 14.660472972972974, + "grad_norm": 0.2450983077287674, + "learning_rate": 2.7549e-05, + "loss": 0.0104, + "step": 17358 + }, + { + "epoch": 14.661317567567568, + "grad_norm": 0.2001952975988388, + "learning_rate": 2.754866666666667e-05, + "loss": 0.0161, + "step": 17359 + }, + { + "epoch": 14.662162162162161, + "grad_norm": 0.4321559965610504, + "learning_rate": 2.7548333333333332e-05, + "loss": 0.0157, + "step": 17360 + }, + { + "epoch": 14.663006756756756, + "grad_norm": 0.5713538527488708, + "learning_rate": 2.7548e-05, + "loss": 0.0186, + "step": 17361 + }, + { + "epoch": 14.66385135135135, + "grad_norm": 0.5811508893966675, + "learning_rate": 2.7547666666666667e-05, + "loss": 0.0224, + "step": 17362 + }, + { + "epoch": 14.664695945945946, + "grad_norm": 0.283601850271225, + "learning_rate": 2.7547333333333333e-05, + "loss": 0.017, + "step": 17363 + }, + { + "epoch": 14.66554054054054, + "grad_norm": 0.2783445715904236, + "learning_rate": 2.7547000000000002e-05, + "loss": 0.0143, + "step": 17364 + }, + { + "epoch": 14.666385135135135, + "grad_norm": 0.23463626205921173, + "learning_rate": 2.754666666666667e-05, + "loss": 0.0173, + "step": 17365 + }, + { + "epoch": 14.66722972972973, + "grad_norm": 0.2886278033256531, + "learning_rate": 2.7546333333333334e-05, + "loss": 0.0115, + "step": 17366 + }, + { + "epoch": 14.668074324324325, + "grad_norm": 0.3071713447570801, + "learning_rate": 2.7546e-05, + "loss": 0.0195, + "step": 17367 + }, + { + "epoch": 14.66891891891892, + "grad_norm": 0.2390311360359192, + "learning_rate": 2.754566666666667e-05, + "loss": 0.0103, + "step": 17368 + }, + { + "epoch": 14.669763513513514, + "grad_norm": 0.23544658720493317, + "learning_rate": 2.7545333333333332e-05, + "loss": 0.0164, + "step": 17369 + }, + { + "epoch": 14.670608108108109, + "grad_norm": 0.5429546236991882, + "learning_rate": 2.7545e-05, + "loss": 0.0134, + "step": 17370 + }, + { + "epoch": 14.671452702702704, + "grad_norm": 0.3113456964492798, + "learning_rate": 2.7544666666666667e-05, + "loss": 0.0179, + "step": 17371 + }, + { + "epoch": 14.672297297297296, + "grad_norm": 0.1672033667564392, + "learning_rate": 2.7544333333333333e-05, + "loss": 0.0114, + "step": 17372 + }, + { + "epoch": 14.673141891891891, + "grad_norm": 1.0350817441940308, + "learning_rate": 2.7544000000000003e-05, + "loss": 0.0121, + "step": 17373 + }, + { + "epoch": 14.673986486486486, + "grad_norm": 0.6108446717262268, + "learning_rate": 2.7543666666666665e-05, + "loss": 0.0234, + "step": 17374 + }, + { + "epoch": 14.67483108108108, + "grad_norm": 0.597007691860199, + "learning_rate": 2.7543333333333334e-05, + "loss": 0.0296, + "step": 17375 + }, + { + "epoch": 14.675675675675675, + "grad_norm": 0.48598945140838623, + "learning_rate": 2.7543e-05, + "loss": 0.0243, + "step": 17376 + }, + { + "epoch": 14.67652027027027, + "grad_norm": 0.6418631076812744, + "learning_rate": 2.7542666666666666e-05, + "loss": 0.19, + "step": 17377 + }, + { + "epoch": 14.677364864864865, + "grad_norm": 0.5475075840950012, + "learning_rate": 2.7542333333333332e-05, + "loss": 0.1407, + "step": 17378 + }, + { + "epoch": 14.67820945945946, + "grad_norm": 0.7295931577682495, + "learning_rate": 2.7542e-05, + "loss": 0.0833, + "step": 17379 + }, + { + "epoch": 14.679054054054054, + "grad_norm": 0.30371546745300293, + "learning_rate": 2.7541666666666668e-05, + "loss": 0.0382, + "step": 17380 + }, + { + "epoch": 14.67989864864865, + "grad_norm": 0.700351357460022, + "learning_rate": 2.7541333333333333e-05, + "loss": 0.0262, + "step": 17381 + }, + { + "epoch": 14.680743243243244, + "grad_norm": 0.23111334443092346, + "learning_rate": 2.7541000000000003e-05, + "loss": 0.0285, + "step": 17382 + }, + { + "epoch": 14.681587837837839, + "grad_norm": 0.21256013214588165, + "learning_rate": 2.7540666666666665e-05, + "loss": 0.0182, + "step": 17383 + }, + { + "epoch": 14.682432432432432, + "grad_norm": 0.22296875715255737, + "learning_rate": 2.7540333333333335e-05, + "loss": 0.0196, + "step": 17384 + }, + { + "epoch": 14.683277027027026, + "grad_norm": 0.18962378799915314, + "learning_rate": 2.754e-05, + "loss": 0.017, + "step": 17385 + }, + { + "epoch": 14.684121621621621, + "grad_norm": 0.36484062671661377, + "learning_rate": 2.7539666666666667e-05, + "loss": 0.0246, + "step": 17386 + }, + { + "epoch": 14.684966216216216, + "grad_norm": 0.19895024597644806, + "learning_rate": 2.7539333333333332e-05, + "loss": 0.014, + "step": 17387 + }, + { + "epoch": 14.68581081081081, + "grad_norm": 0.44795385003089905, + "learning_rate": 2.7539000000000002e-05, + "loss": 0.0172, + "step": 17388 + }, + { + "epoch": 14.686655405405405, + "grad_norm": 0.14998306334018707, + "learning_rate": 2.7538666666666668e-05, + "loss": 0.0103, + "step": 17389 + }, + { + "epoch": 14.6875, + "grad_norm": 0.3774251937866211, + "learning_rate": 2.7538333333333334e-05, + "loss": 0.027, + "step": 17390 + }, + { + "epoch": 14.688344594594595, + "grad_norm": 0.31412115693092346, + "learning_rate": 2.7538000000000003e-05, + "loss": 0.0212, + "step": 17391 + }, + { + "epoch": 14.68918918918919, + "grad_norm": 0.29554152488708496, + "learning_rate": 2.7537666666666666e-05, + "loss": 0.0149, + "step": 17392 + }, + { + "epoch": 14.690033783783784, + "grad_norm": 0.1829446256160736, + "learning_rate": 2.7537333333333335e-05, + "loss": 0.0174, + "step": 17393 + }, + { + "epoch": 14.690878378378379, + "grad_norm": 0.14105287194252014, + "learning_rate": 2.7537e-05, + "loss": 0.0138, + "step": 17394 + }, + { + "epoch": 14.691722972972974, + "grad_norm": 0.6449616551399231, + "learning_rate": 2.7536666666666667e-05, + "loss": 0.0241, + "step": 17395 + }, + { + "epoch": 14.692567567567568, + "grad_norm": 0.19182823598384857, + "learning_rate": 2.7536333333333336e-05, + "loss": 0.0149, + "step": 17396 + }, + { + "epoch": 14.693412162162161, + "grad_norm": 0.14210504293441772, + "learning_rate": 2.7536000000000002e-05, + "loss": 0.0141, + "step": 17397 + }, + { + "epoch": 14.694256756756756, + "grad_norm": 0.2292264848947525, + "learning_rate": 2.7535666666666668e-05, + "loss": 0.0162, + "step": 17398 + }, + { + "epoch": 14.69510135135135, + "grad_norm": 0.29086998105049133, + "learning_rate": 2.7535333333333334e-05, + "loss": 0.0233, + "step": 17399 + }, + { + "epoch": 14.695945945945946, + "grad_norm": 0.23271134495735168, + "learning_rate": 2.7535e-05, + "loss": 0.0164, + "step": 17400 + }, + { + "epoch": 14.69679054054054, + "grad_norm": 0.17242905497550964, + "learning_rate": 2.7534666666666666e-05, + "loss": 0.0134, + "step": 17401 + }, + { + "epoch": 14.697635135135135, + "grad_norm": 0.3042043447494507, + "learning_rate": 2.7534333333333335e-05, + "loss": 0.0178, + "step": 17402 + }, + { + "epoch": 14.69847972972973, + "grad_norm": 0.5489065051078796, + "learning_rate": 2.7533999999999998e-05, + "loss": 0.017, + "step": 17403 + }, + { + "epoch": 14.699324324324325, + "grad_norm": 0.1614181250333786, + "learning_rate": 2.7533666666666667e-05, + "loss": 0.0094, + "step": 17404 + }, + { + "epoch": 14.70016891891892, + "grad_norm": 0.34526652097702026, + "learning_rate": 2.7533333333333336e-05, + "loss": 0.0207, + "step": 17405 + }, + { + "epoch": 14.701013513513514, + "grad_norm": 0.40604594349861145, + "learning_rate": 2.7533e-05, + "loss": 0.0223, + "step": 17406 + }, + { + "epoch": 14.701858108108109, + "grad_norm": 0.17020906507968903, + "learning_rate": 2.7532666666666668e-05, + "loss": 0.0091, + "step": 17407 + }, + { + "epoch": 14.702702702702704, + "grad_norm": 0.3219987154006958, + "learning_rate": 2.7532333333333334e-05, + "loss": 0.0213, + "step": 17408 + }, + { + "epoch": 14.703547297297296, + "grad_norm": 0.14047524333000183, + "learning_rate": 2.7532e-05, + "loss": 0.0117, + "step": 17409 + }, + { + "epoch": 14.704391891891891, + "grad_norm": 0.28333377838134766, + "learning_rate": 2.7531666666666666e-05, + "loss": 0.0113, + "step": 17410 + }, + { + "epoch": 14.705236486486486, + "grad_norm": 0.30655214190483093, + "learning_rate": 2.7531333333333335e-05, + "loss": 0.0141, + "step": 17411 + }, + { + "epoch": 14.70608108108108, + "grad_norm": 0.30615952610969543, + "learning_rate": 2.7531e-05, + "loss": 0.0261, + "step": 17412 + }, + { + "epoch": 14.706925675675675, + "grad_norm": 0.3375077247619629, + "learning_rate": 2.7530666666666667e-05, + "loss": 0.0156, + "step": 17413 + }, + { + "epoch": 14.70777027027027, + "grad_norm": 0.3754187822341919, + "learning_rate": 2.7530333333333336e-05, + "loss": 0.0228, + "step": 17414 + }, + { + "epoch": 14.708614864864865, + "grad_norm": 0.1843884140253067, + "learning_rate": 2.753e-05, + "loss": 0.0119, + "step": 17415 + }, + { + "epoch": 14.70945945945946, + "grad_norm": 0.25504618883132935, + "learning_rate": 2.752966666666667e-05, + "loss": 0.0125, + "step": 17416 + }, + { + "epoch": 14.710304054054054, + "grad_norm": 0.5755109190940857, + "learning_rate": 2.7529333333333334e-05, + "loss": 0.0278, + "step": 17417 + }, + { + "epoch": 14.71114864864865, + "grad_norm": 0.23901325464248657, + "learning_rate": 2.7529e-05, + "loss": 0.0126, + "step": 17418 + }, + { + "epoch": 14.711993243243244, + "grad_norm": 0.19979803264141083, + "learning_rate": 2.7528666666666666e-05, + "loss": 0.0099, + "step": 17419 + }, + { + "epoch": 14.712837837837839, + "grad_norm": 0.23772640526294708, + "learning_rate": 2.7528333333333335e-05, + "loss": 0.0161, + "step": 17420 + }, + { + "epoch": 14.713682432432432, + "grad_norm": 0.2900755703449249, + "learning_rate": 2.7528e-05, + "loss": 0.0627, + "step": 17421 + }, + { + "epoch": 14.714527027027026, + "grad_norm": 0.23382896184921265, + "learning_rate": 2.7527666666666667e-05, + "loss": 0.0164, + "step": 17422 + }, + { + "epoch": 14.715371621621621, + "grad_norm": 0.6140313148498535, + "learning_rate": 2.7527333333333337e-05, + "loss": 0.032, + "step": 17423 + }, + { + "epoch": 14.716216216216216, + "grad_norm": 0.3168325424194336, + "learning_rate": 2.7527e-05, + "loss": 0.0162, + "step": 17424 + }, + { + "epoch": 14.71706081081081, + "grad_norm": 1.4266654253005981, + "learning_rate": 2.752666666666667e-05, + "loss": 0.0232, + "step": 17425 + }, + { + "epoch": 14.717905405405405, + "grad_norm": 0.19808050990104675, + "learning_rate": 2.7526333333333334e-05, + "loss": 0.0134, + "step": 17426 + }, + { + "epoch": 14.71875, + "grad_norm": 0.6369650959968567, + "learning_rate": 2.7526e-05, + "loss": 0.1956, + "step": 17427 + }, + { + "epoch": 14.719594594594595, + "grad_norm": 0.5854957699775696, + "learning_rate": 2.7525666666666666e-05, + "loss": 0.1092, + "step": 17428 + }, + { + "epoch": 14.72043918918919, + "grad_norm": 0.5133712291717529, + "learning_rate": 2.7525333333333336e-05, + "loss": 0.067, + "step": 17429 + }, + { + "epoch": 14.721283783783784, + "grad_norm": 0.3247067630290985, + "learning_rate": 2.7525e-05, + "loss": 0.042, + "step": 17430 + }, + { + "epoch": 14.722128378378379, + "grad_norm": 0.2681865096092224, + "learning_rate": 2.7524666666666668e-05, + "loss": 0.0263, + "step": 17431 + }, + { + "epoch": 14.722972972972974, + "grad_norm": 0.41702768206596375, + "learning_rate": 2.7524333333333333e-05, + "loss": 0.0312, + "step": 17432 + }, + { + "epoch": 14.723817567567568, + "grad_norm": 0.37287667393684387, + "learning_rate": 2.7524e-05, + "loss": 0.0176, + "step": 17433 + }, + { + "epoch": 14.724662162162161, + "grad_norm": 0.2583152949810028, + "learning_rate": 2.752366666666667e-05, + "loss": 0.0232, + "step": 17434 + }, + { + "epoch": 14.725506756756756, + "grad_norm": 0.36361271142959595, + "learning_rate": 2.752333333333333e-05, + "loss": 0.0191, + "step": 17435 + }, + { + "epoch": 14.72635135135135, + "grad_norm": 0.36092400550842285, + "learning_rate": 2.7523e-05, + "loss": 0.0161, + "step": 17436 + }, + { + "epoch": 14.727195945945946, + "grad_norm": 0.32054635882377625, + "learning_rate": 2.7522666666666667e-05, + "loss": 0.0225, + "step": 17437 + }, + { + "epoch": 14.72804054054054, + "grad_norm": 0.3143829107284546, + "learning_rate": 2.7522333333333332e-05, + "loss": 0.0184, + "step": 17438 + }, + { + "epoch": 14.728885135135135, + "grad_norm": 0.13959811627864838, + "learning_rate": 2.7522000000000002e-05, + "loss": 0.0095, + "step": 17439 + }, + { + "epoch": 14.72972972972973, + "grad_norm": 0.38448840379714966, + "learning_rate": 2.7521666666666668e-05, + "loss": 0.026, + "step": 17440 + }, + { + "epoch": 14.730574324324325, + "grad_norm": 0.22701868414878845, + "learning_rate": 2.7521333333333334e-05, + "loss": 0.011, + "step": 17441 + }, + { + "epoch": 14.73141891891892, + "grad_norm": 0.18031767010688782, + "learning_rate": 2.7521e-05, + "loss": 0.0153, + "step": 17442 + }, + { + "epoch": 14.732263513513514, + "grad_norm": 0.1779181808233261, + "learning_rate": 2.752066666666667e-05, + "loss": 0.0135, + "step": 17443 + }, + { + "epoch": 14.733108108108109, + "grad_norm": 0.20346702635288239, + "learning_rate": 2.752033333333333e-05, + "loss": 0.0126, + "step": 17444 + }, + { + "epoch": 14.733952702702704, + "grad_norm": 0.32966044545173645, + "learning_rate": 2.752e-05, + "loss": 0.0159, + "step": 17445 + }, + { + "epoch": 14.734797297297296, + "grad_norm": 0.13632993400096893, + "learning_rate": 2.751966666666667e-05, + "loss": 0.0123, + "step": 17446 + }, + { + "epoch": 14.735641891891891, + "grad_norm": 0.2642545998096466, + "learning_rate": 2.7519333333333333e-05, + "loss": 0.018, + "step": 17447 + }, + { + "epoch": 14.736486486486486, + "grad_norm": 0.1901707500219345, + "learning_rate": 2.7519000000000002e-05, + "loss": 0.0094, + "step": 17448 + }, + { + "epoch": 14.73733108108108, + "grad_norm": 0.2100408524274826, + "learning_rate": 2.7518666666666668e-05, + "loss": 0.0073, + "step": 17449 + }, + { + "epoch": 14.738175675675675, + "grad_norm": 0.2650509774684906, + "learning_rate": 2.7518333333333334e-05, + "loss": 0.0151, + "step": 17450 + }, + { + "epoch": 14.73902027027027, + "grad_norm": 0.29264503717422485, + "learning_rate": 2.7518e-05, + "loss": 0.0172, + "step": 17451 + }, + { + "epoch": 14.739864864864865, + "grad_norm": 0.2155633568763733, + "learning_rate": 2.751766666666667e-05, + "loss": 0.0169, + "step": 17452 + }, + { + "epoch": 14.74070945945946, + "grad_norm": 0.1224869042634964, + "learning_rate": 2.751733333333333e-05, + "loss": 0.012, + "step": 17453 + }, + { + "epoch": 14.741554054054054, + "grad_norm": 0.18731512129306793, + "learning_rate": 2.7517e-05, + "loss": 0.0125, + "step": 17454 + }, + { + "epoch": 14.74239864864865, + "grad_norm": 0.17756475508213043, + "learning_rate": 2.751666666666667e-05, + "loss": 0.0073, + "step": 17455 + }, + { + "epoch": 14.743243243243244, + "grad_norm": 0.44217947125434875, + "learning_rate": 2.7516333333333333e-05, + "loss": 0.0218, + "step": 17456 + }, + { + "epoch": 14.744087837837839, + "grad_norm": 0.28676801919937134, + "learning_rate": 2.7516000000000002e-05, + "loss": 0.0175, + "step": 17457 + }, + { + "epoch": 14.744932432432432, + "grad_norm": 0.4878830015659332, + "learning_rate": 2.7515666666666668e-05, + "loss": 0.0212, + "step": 17458 + }, + { + "epoch": 14.745777027027026, + "grad_norm": 0.3247388005256653, + "learning_rate": 2.7515333333333334e-05, + "loss": 0.0149, + "step": 17459 + }, + { + "epoch": 14.746621621621621, + "grad_norm": 0.3813655376434326, + "learning_rate": 2.7515e-05, + "loss": 0.0122, + "step": 17460 + }, + { + "epoch": 14.747466216216216, + "grad_norm": 0.3035840690135956, + "learning_rate": 2.7514666666666666e-05, + "loss": 0.0151, + "step": 17461 + }, + { + "epoch": 14.74831081081081, + "grad_norm": 0.5401944518089294, + "learning_rate": 2.7514333333333335e-05, + "loss": 0.0095, + "step": 17462 + }, + { + "epoch": 14.749155405405405, + "grad_norm": 0.2643114924430847, + "learning_rate": 2.7514e-05, + "loss": 0.0207, + "step": 17463 + }, + { + "epoch": 14.75, + "grad_norm": 0.9187915921211243, + "learning_rate": 2.7513666666666667e-05, + "loss": 0.0148, + "step": 17464 + }, + { + "epoch": 14.750844594594595, + "grad_norm": 0.2696629762649536, + "learning_rate": 2.7513333333333333e-05, + "loss": 0.0191, + "step": 17465 + }, + { + "epoch": 14.75168918918919, + "grad_norm": 0.2526063621044159, + "learning_rate": 2.7513000000000002e-05, + "loss": 0.0128, + "step": 17466 + }, + { + "epoch": 14.752533783783784, + "grad_norm": 0.27101969718933105, + "learning_rate": 2.7512666666666665e-05, + "loss": 0.0185, + "step": 17467 + }, + { + "epoch": 14.753378378378379, + "grad_norm": 0.45199620723724365, + "learning_rate": 2.7512333333333334e-05, + "loss": 0.0156, + "step": 17468 + }, + { + "epoch": 14.754222972972974, + "grad_norm": 0.23332558572292328, + "learning_rate": 2.7512e-05, + "loss": 0.0188, + "step": 17469 + }, + { + "epoch": 14.755067567567568, + "grad_norm": 0.327723890542984, + "learning_rate": 2.7511666666666666e-05, + "loss": 0.014, + "step": 17470 + }, + { + "epoch": 14.755912162162161, + "grad_norm": 0.34489932656288147, + "learning_rate": 2.7511333333333335e-05, + "loss": 0.0131, + "step": 17471 + }, + { + "epoch": 14.756756756756756, + "grad_norm": 0.2476990669965744, + "learning_rate": 2.7511e-05, + "loss": 0.0141, + "step": 17472 + }, + { + "epoch": 14.75760135135135, + "grad_norm": 0.2759331166744232, + "learning_rate": 2.7510666666666667e-05, + "loss": 0.0112, + "step": 17473 + }, + { + "epoch": 14.758445945945946, + "grad_norm": 0.13766519725322723, + "learning_rate": 2.7510333333333333e-05, + "loss": 0.0088, + "step": 17474 + }, + { + "epoch": 14.75929054054054, + "grad_norm": 0.7797513008117676, + "learning_rate": 2.7510000000000003e-05, + "loss": 0.0212, + "step": 17475 + }, + { + "epoch": 14.760135135135135, + "grad_norm": 0.7414463758468628, + "learning_rate": 2.7509666666666665e-05, + "loss": 0.0504, + "step": 17476 + }, + { + "epoch": 14.76097972972973, + "grad_norm": 0.6415631175041199, + "learning_rate": 2.7509333333333334e-05, + "loss": 0.1876, + "step": 17477 + }, + { + "epoch": 14.761824324324325, + "grad_norm": 0.3572717010974884, + "learning_rate": 2.7509e-05, + "loss": 0.0775, + "step": 17478 + }, + { + "epoch": 14.76266891891892, + "grad_norm": 0.6662834286689758, + "learning_rate": 2.7508666666666666e-05, + "loss": 0.044, + "step": 17479 + }, + { + "epoch": 14.763513513513514, + "grad_norm": 0.33951133489608765, + "learning_rate": 2.7508333333333336e-05, + "loss": 0.0252, + "step": 17480 + }, + { + "epoch": 14.764358108108109, + "grad_norm": 0.2600175738334656, + "learning_rate": 2.7508e-05, + "loss": 0.0343, + "step": 17481 + }, + { + "epoch": 14.765202702702704, + "grad_norm": 0.44407880306243896, + "learning_rate": 2.7507666666666668e-05, + "loss": 0.0461, + "step": 17482 + }, + { + "epoch": 14.766047297297296, + "grad_norm": 0.14586633443832397, + "learning_rate": 2.7507333333333333e-05, + "loss": 0.0156, + "step": 17483 + }, + { + "epoch": 14.766891891891891, + "grad_norm": 0.4909445345401764, + "learning_rate": 2.7507000000000003e-05, + "loss": 0.0188, + "step": 17484 + }, + { + "epoch": 14.767736486486486, + "grad_norm": 0.48850739002227783, + "learning_rate": 2.7506666666666665e-05, + "loss": 0.0232, + "step": 17485 + }, + { + "epoch": 14.76858108108108, + "grad_norm": 0.33442068099975586, + "learning_rate": 2.7506333333333335e-05, + "loss": 0.0272, + "step": 17486 + }, + { + "epoch": 14.769425675675675, + "grad_norm": 0.25854814052581787, + "learning_rate": 2.7506e-05, + "loss": 0.017, + "step": 17487 + }, + { + "epoch": 14.77027027027027, + "grad_norm": 0.16947467625141144, + "learning_rate": 2.7505666666666667e-05, + "loss": 0.0126, + "step": 17488 + }, + { + "epoch": 14.771114864864865, + "grad_norm": 0.3015094995498657, + "learning_rate": 2.7505333333333336e-05, + "loss": 0.0183, + "step": 17489 + }, + { + "epoch": 14.77195945945946, + "grad_norm": 0.51774662733078, + "learning_rate": 2.7505e-05, + "loss": 0.0191, + "step": 17490 + }, + { + "epoch": 14.772804054054054, + "grad_norm": 0.3257923424243927, + "learning_rate": 2.7504666666666668e-05, + "loss": 0.024, + "step": 17491 + }, + { + "epoch": 14.77364864864865, + "grad_norm": 0.1698167473077774, + "learning_rate": 2.7504333333333334e-05, + "loss": 0.013, + "step": 17492 + }, + { + "epoch": 14.774493243243244, + "grad_norm": 0.18621715903282166, + "learning_rate": 2.7504e-05, + "loss": 0.0185, + "step": 17493 + }, + { + "epoch": 14.775337837837839, + "grad_norm": 0.24730387330055237, + "learning_rate": 2.7503666666666666e-05, + "loss": 0.0209, + "step": 17494 + }, + { + "epoch": 14.776182432432432, + "grad_norm": 0.472044974565506, + "learning_rate": 2.7503333333333335e-05, + "loss": 0.0179, + "step": 17495 + }, + { + "epoch": 14.777027027027026, + "grad_norm": 0.4553891122341156, + "learning_rate": 2.7503e-05, + "loss": 0.0164, + "step": 17496 + }, + { + "epoch": 14.777871621621621, + "grad_norm": 0.26184266805648804, + "learning_rate": 2.7502666666666667e-05, + "loss": 0.0285, + "step": 17497 + }, + { + "epoch": 14.778716216216216, + "grad_norm": 0.28561654686927795, + "learning_rate": 2.7502333333333336e-05, + "loss": 0.0133, + "step": 17498 + }, + { + "epoch": 14.77956081081081, + "grad_norm": 0.18062421679496765, + "learning_rate": 2.7502e-05, + "loss": 0.0108, + "step": 17499 + }, + { + "epoch": 14.780405405405405, + "grad_norm": 0.5064212679862976, + "learning_rate": 2.7501666666666668e-05, + "loss": 0.0316, + "step": 17500 + }, + { + "epoch": 14.78125, + "grad_norm": 0.2810537815093994, + "learning_rate": 2.7501333333333334e-05, + "loss": 0.0177, + "step": 17501 + }, + { + "epoch": 14.782094594594595, + "grad_norm": 0.3189161717891693, + "learning_rate": 2.7501e-05, + "loss": 0.0213, + "step": 17502 + }, + { + "epoch": 14.78293918918919, + "grad_norm": 0.21608760952949524, + "learning_rate": 2.7500666666666666e-05, + "loss": 0.0132, + "step": 17503 + }, + { + "epoch": 14.783783783783784, + "grad_norm": 0.24607285857200623, + "learning_rate": 2.7500333333333335e-05, + "loss": 0.0118, + "step": 17504 + }, + { + "epoch": 14.784628378378379, + "grad_norm": 0.18619759380817413, + "learning_rate": 2.75e-05, + "loss": 0.0116, + "step": 17505 + }, + { + "epoch": 14.785472972972974, + "grad_norm": 0.164974644780159, + "learning_rate": 2.7499666666666667e-05, + "loss": 0.0089, + "step": 17506 + }, + { + "epoch": 14.786317567567568, + "grad_norm": 0.2595199644565582, + "learning_rate": 2.7499333333333336e-05, + "loss": 0.0237, + "step": 17507 + }, + { + "epoch": 14.787162162162161, + "grad_norm": 0.17146140336990356, + "learning_rate": 2.7499e-05, + "loss": 0.0077, + "step": 17508 + }, + { + "epoch": 14.788006756756756, + "grad_norm": 0.23090142011642456, + "learning_rate": 2.7498666666666668e-05, + "loss": 0.0126, + "step": 17509 + }, + { + "epoch": 14.78885135135135, + "grad_norm": 0.33803412318229675, + "learning_rate": 2.7498333333333334e-05, + "loss": 0.0151, + "step": 17510 + }, + { + "epoch": 14.789695945945946, + "grad_norm": 0.15055546164512634, + "learning_rate": 2.7498e-05, + "loss": 0.0104, + "step": 17511 + }, + { + "epoch": 14.79054054054054, + "grad_norm": 0.29068347811698914, + "learning_rate": 2.749766666666667e-05, + "loss": 0.0135, + "step": 17512 + }, + { + "epoch": 14.791385135135135, + "grad_norm": 0.22533150017261505, + "learning_rate": 2.7497333333333335e-05, + "loss": 0.01, + "step": 17513 + }, + { + "epoch": 14.79222972972973, + "grad_norm": 0.32666638493537903, + "learning_rate": 2.7497e-05, + "loss": 0.0158, + "step": 17514 + }, + { + "epoch": 14.793074324324325, + "grad_norm": 0.42954108119010925, + "learning_rate": 2.7496666666666667e-05, + "loss": 0.0152, + "step": 17515 + }, + { + "epoch": 14.79391891891892, + "grad_norm": 0.3208874464035034, + "learning_rate": 2.7496333333333336e-05, + "loss": 0.0287, + "step": 17516 + }, + { + "epoch": 14.794763513513514, + "grad_norm": 0.2406112402677536, + "learning_rate": 2.7496e-05, + "loss": 0.0132, + "step": 17517 + }, + { + "epoch": 14.795608108108109, + "grad_norm": 0.22908920049667358, + "learning_rate": 2.749566666666667e-05, + "loss": 0.0097, + "step": 17518 + }, + { + "epoch": 14.796452702702704, + "grad_norm": 0.39872339367866516, + "learning_rate": 2.7495333333333334e-05, + "loss": 0.0308, + "step": 17519 + }, + { + "epoch": 14.797297297297296, + "grad_norm": 0.48224252462387085, + "learning_rate": 2.7495e-05, + "loss": 0.0199, + "step": 17520 + }, + { + "epoch": 14.798141891891891, + "grad_norm": 0.20730337500572205, + "learning_rate": 2.749466666666667e-05, + "loss": 0.0079, + "step": 17521 + }, + { + "epoch": 14.798986486486486, + "grad_norm": 0.17181158065795898, + "learning_rate": 2.7494333333333332e-05, + "loss": 0.0154, + "step": 17522 + }, + { + "epoch": 14.79983108108108, + "grad_norm": 0.31140729784965515, + "learning_rate": 2.7494e-05, + "loss": 0.014, + "step": 17523 + }, + { + "epoch": 14.800675675675675, + "grad_norm": 0.26683884859085083, + "learning_rate": 2.7493666666666667e-05, + "loss": 0.0139, + "step": 17524 + }, + { + "epoch": 14.80152027027027, + "grad_norm": 1.0467662811279297, + "learning_rate": 2.7493333333333333e-05, + "loss": 0.0406, + "step": 17525 + }, + { + "epoch": 14.802364864864865, + "grad_norm": 0.4089495837688446, + "learning_rate": 2.7493e-05, + "loss": 0.0123, + "step": 17526 + }, + { + "epoch": 14.80320945945946, + "grad_norm": 0.642394483089447, + "learning_rate": 2.749266666666667e-05, + "loss": 0.1736, + "step": 17527 + }, + { + "epoch": 14.804054054054054, + "grad_norm": 0.6268061995506287, + "learning_rate": 2.749233333333333e-05, + "loss": 0.1175, + "step": 17528 + }, + { + "epoch": 14.80489864864865, + "grad_norm": 0.28387337923049927, + "learning_rate": 2.7492e-05, + "loss": 0.0516, + "step": 17529 + }, + { + "epoch": 14.805743243243244, + "grad_norm": 0.4151756167411804, + "learning_rate": 2.749166666666667e-05, + "loss": 0.039, + "step": 17530 + }, + { + "epoch": 14.806587837837839, + "grad_norm": 0.25291404128074646, + "learning_rate": 2.7491333333333332e-05, + "loss": 0.0177, + "step": 17531 + }, + { + "epoch": 14.807432432432432, + "grad_norm": 0.3108617067337036, + "learning_rate": 2.7491e-05, + "loss": 0.0246, + "step": 17532 + }, + { + "epoch": 14.808277027027026, + "grad_norm": 0.3046988546848297, + "learning_rate": 2.7490666666666668e-05, + "loss": 0.0306, + "step": 17533 + }, + { + "epoch": 14.809121621621621, + "grad_norm": 0.24340969324111938, + "learning_rate": 2.7490333333333333e-05, + "loss": 0.0187, + "step": 17534 + }, + { + "epoch": 14.809966216216216, + "grad_norm": 0.35564061999320984, + "learning_rate": 2.749e-05, + "loss": 0.0267, + "step": 17535 + }, + { + "epoch": 14.81081081081081, + "grad_norm": 0.3011179566383362, + "learning_rate": 2.748966666666667e-05, + "loss": 0.0167, + "step": 17536 + }, + { + "epoch": 14.811655405405405, + "grad_norm": 0.2584680914878845, + "learning_rate": 2.7489333333333335e-05, + "loss": 0.027, + "step": 17537 + }, + { + "epoch": 14.8125, + "grad_norm": 0.1395886093378067, + "learning_rate": 2.7489e-05, + "loss": 0.0101, + "step": 17538 + }, + { + "epoch": 14.813344594594595, + "grad_norm": 0.19924794137477875, + "learning_rate": 2.748866666666667e-05, + "loss": 0.0173, + "step": 17539 + }, + { + "epoch": 14.81418918918919, + "grad_norm": 0.1627795398235321, + "learning_rate": 2.7488333333333332e-05, + "loss": 0.0144, + "step": 17540 + }, + { + "epoch": 14.815033783783784, + "grad_norm": 0.11563704162836075, + "learning_rate": 2.7488000000000002e-05, + "loss": 0.0079, + "step": 17541 + }, + { + "epoch": 14.815878378378379, + "grad_norm": 0.22725945711135864, + "learning_rate": 2.7487666666666668e-05, + "loss": 0.0124, + "step": 17542 + }, + { + "epoch": 14.816722972972974, + "grad_norm": 0.19689609110355377, + "learning_rate": 2.7487333333333334e-05, + "loss": 0.0164, + "step": 17543 + }, + { + "epoch": 14.817567567567568, + "grad_norm": 0.41922542452812195, + "learning_rate": 2.7487e-05, + "loss": 0.0188, + "step": 17544 + }, + { + "epoch": 14.818412162162161, + "grad_norm": 0.22084861993789673, + "learning_rate": 2.748666666666667e-05, + "loss": 0.015, + "step": 17545 + }, + { + "epoch": 14.819256756756756, + "grad_norm": 0.18684498965740204, + "learning_rate": 2.7486333333333335e-05, + "loss": 0.0135, + "step": 17546 + }, + { + "epoch": 14.82010135135135, + "grad_norm": 0.19921888411045074, + "learning_rate": 2.7486e-05, + "loss": 0.0091, + "step": 17547 + }, + { + "epoch": 14.820945945945946, + "grad_norm": 0.29373258352279663, + "learning_rate": 2.748566666666667e-05, + "loss": 0.0194, + "step": 17548 + }, + { + "epoch": 14.82179054054054, + "grad_norm": 0.3264840543270111, + "learning_rate": 2.7485333333333333e-05, + "loss": 0.0194, + "step": 17549 + }, + { + "epoch": 14.822635135135135, + "grad_norm": 0.2769629657268524, + "learning_rate": 2.7485000000000002e-05, + "loss": 0.0322, + "step": 17550 + }, + { + "epoch": 14.82347972972973, + "grad_norm": 0.12638334929943085, + "learning_rate": 2.7484666666666665e-05, + "loss": 0.0112, + "step": 17551 + }, + { + "epoch": 14.824324324324325, + "grad_norm": 0.27538156509399414, + "learning_rate": 2.7484333333333334e-05, + "loss": 0.017, + "step": 17552 + }, + { + "epoch": 14.82516891891892, + "grad_norm": 0.35543712973594666, + "learning_rate": 2.7484e-05, + "loss": 0.0205, + "step": 17553 + }, + { + "epoch": 14.826013513513514, + "grad_norm": 0.1547991782426834, + "learning_rate": 2.7483666666666666e-05, + "loss": 0.0119, + "step": 17554 + }, + { + "epoch": 14.826858108108109, + "grad_norm": 0.22246816754341125, + "learning_rate": 2.7483333333333335e-05, + "loss": 0.0114, + "step": 17555 + }, + { + "epoch": 14.827702702702704, + "grad_norm": 0.25220614671707153, + "learning_rate": 2.7483e-05, + "loss": 0.0134, + "step": 17556 + }, + { + "epoch": 14.828547297297296, + "grad_norm": 0.2254861444234848, + "learning_rate": 2.7482666666666667e-05, + "loss": 0.0117, + "step": 17557 + }, + { + "epoch": 14.829391891891891, + "grad_norm": 0.179210364818573, + "learning_rate": 2.7482333333333333e-05, + "loss": 0.0187, + "step": 17558 + }, + { + "epoch": 14.830236486486486, + "grad_norm": 0.40272200107574463, + "learning_rate": 2.7482000000000002e-05, + "loss": 0.0195, + "step": 17559 + }, + { + "epoch": 14.83108108108108, + "grad_norm": 0.25543132424354553, + "learning_rate": 2.7481666666666665e-05, + "loss": 0.0203, + "step": 17560 + }, + { + "epoch": 14.831925675675675, + "grad_norm": 0.540328323841095, + "learning_rate": 2.7481333333333334e-05, + "loss": 0.0166, + "step": 17561 + }, + { + "epoch": 14.83277027027027, + "grad_norm": 0.21304771304130554, + "learning_rate": 2.7481000000000003e-05, + "loss": 0.0107, + "step": 17562 + }, + { + "epoch": 14.833614864864865, + "grad_norm": 0.2305658608675003, + "learning_rate": 2.7480666666666666e-05, + "loss": 0.0154, + "step": 17563 + }, + { + "epoch": 14.83445945945946, + "grad_norm": 0.2239355444908142, + "learning_rate": 2.7480333333333335e-05, + "loss": 0.0134, + "step": 17564 + }, + { + "epoch": 14.835304054054054, + "grad_norm": 0.410264253616333, + "learning_rate": 2.748e-05, + "loss": 0.031, + "step": 17565 + }, + { + "epoch": 14.83614864864865, + "grad_norm": 0.795661211013794, + "learning_rate": 2.7479666666666667e-05, + "loss": 0.012, + "step": 17566 + }, + { + "epoch": 14.836993243243244, + "grad_norm": 0.6371428370475769, + "learning_rate": 2.7479333333333333e-05, + "loss": 0.0207, + "step": 17567 + }, + { + "epoch": 14.837837837837839, + "grad_norm": 0.4255951941013336, + "learning_rate": 2.7479000000000002e-05, + "loss": 0.0251, + "step": 17568 + }, + { + "epoch": 14.838682432432432, + "grad_norm": 0.25268498063087463, + "learning_rate": 2.7478666666666665e-05, + "loss": 0.0157, + "step": 17569 + }, + { + "epoch": 14.839527027027026, + "grad_norm": 0.8362874388694763, + "learning_rate": 2.7478333333333334e-05, + "loss": 0.044, + "step": 17570 + }, + { + "epoch": 14.840371621621621, + "grad_norm": 1.413928747177124, + "learning_rate": 2.7478000000000004e-05, + "loss": 0.0286, + "step": 17571 + }, + { + "epoch": 14.841216216216216, + "grad_norm": 0.39171910285949707, + "learning_rate": 2.7477666666666666e-05, + "loss": 0.0254, + "step": 17572 + }, + { + "epoch": 14.84206081081081, + "grad_norm": 0.19797615706920624, + "learning_rate": 2.7477333333333335e-05, + "loss": 0.0094, + "step": 17573 + }, + { + "epoch": 14.842905405405405, + "grad_norm": 0.19817085564136505, + "learning_rate": 2.7477e-05, + "loss": 0.0136, + "step": 17574 + }, + { + "epoch": 14.84375, + "grad_norm": 0.27441325783729553, + "learning_rate": 2.7476666666666667e-05, + "loss": 0.0144, + "step": 17575 + }, + { + "epoch": 14.844594594594595, + "grad_norm": 1.2005798816680908, + "learning_rate": 2.7476333333333333e-05, + "loss": 0.0369, + "step": 17576 + }, + { + "epoch": 14.84543918918919, + "grad_norm": 0.614607036113739, + "learning_rate": 2.7476000000000003e-05, + "loss": 0.1689, + "step": 17577 + }, + { + "epoch": 14.846283783783784, + "grad_norm": 0.467265784740448, + "learning_rate": 2.7475666666666665e-05, + "loss": 0.1177, + "step": 17578 + }, + { + "epoch": 14.847128378378379, + "grad_norm": 0.3578373193740845, + "learning_rate": 2.7475333333333334e-05, + "loss": 0.0559, + "step": 17579 + }, + { + "epoch": 14.847972972972974, + "grad_norm": 0.3351316452026367, + "learning_rate": 2.7475000000000004e-05, + "loss": 0.0331, + "step": 17580 + }, + { + "epoch": 14.848817567567568, + "grad_norm": 0.3547561466693878, + "learning_rate": 2.7474666666666666e-05, + "loss": 0.0362, + "step": 17581 + }, + { + "epoch": 14.849662162162161, + "grad_norm": 0.1827349215745926, + "learning_rate": 2.7474333333333336e-05, + "loss": 0.0152, + "step": 17582 + }, + { + "epoch": 14.850506756756756, + "grad_norm": 0.6190364956855774, + "learning_rate": 2.7473999999999998e-05, + "loss": 0.0467, + "step": 17583 + }, + { + "epoch": 14.85135135135135, + "grad_norm": 0.1599547415971756, + "learning_rate": 2.7473666666666668e-05, + "loss": 0.0118, + "step": 17584 + }, + { + "epoch": 14.852195945945946, + "grad_norm": 0.20677433907985687, + "learning_rate": 2.7473333333333333e-05, + "loss": 0.0173, + "step": 17585 + }, + { + "epoch": 14.85304054054054, + "grad_norm": 0.25052353739738464, + "learning_rate": 2.7473e-05, + "loss": 0.0224, + "step": 17586 + }, + { + "epoch": 14.853885135135135, + "grad_norm": 0.1579350382089615, + "learning_rate": 2.747266666666667e-05, + "loss": 0.0128, + "step": 17587 + }, + { + "epoch": 14.85472972972973, + "grad_norm": 0.2733592092990875, + "learning_rate": 2.7472333333333335e-05, + "loss": 0.0273, + "step": 17588 + }, + { + "epoch": 14.855574324324325, + "grad_norm": 0.24102500081062317, + "learning_rate": 2.7472e-05, + "loss": 0.0176, + "step": 17589 + }, + { + "epoch": 14.85641891891892, + "grad_norm": 1.6215885877609253, + "learning_rate": 2.7471666666666667e-05, + "loss": 0.0212, + "step": 17590 + }, + { + "epoch": 14.857263513513514, + "grad_norm": 0.7661364674568176, + "learning_rate": 2.7471333333333336e-05, + "loss": 0.0329, + "step": 17591 + }, + { + "epoch": 14.858108108108109, + "grad_norm": 0.2582935094833374, + "learning_rate": 2.7471e-05, + "loss": 0.0177, + "step": 17592 + }, + { + "epoch": 14.858952702702704, + "grad_norm": 0.2055194079875946, + "learning_rate": 2.7470666666666668e-05, + "loss": 0.0173, + "step": 17593 + }, + { + "epoch": 14.859797297297296, + "grad_norm": 0.18989287316799164, + "learning_rate": 2.7470333333333334e-05, + "loss": 0.0132, + "step": 17594 + }, + { + "epoch": 14.860641891891891, + "grad_norm": 0.2611903250217438, + "learning_rate": 2.747e-05, + "loss": 0.0111, + "step": 17595 + }, + { + "epoch": 14.861486486486486, + "grad_norm": 0.30075085163116455, + "learning_rate": 2.746966666666667e-05, + "loss": 0.0192, + "step": 17596 + }, + { + "epoch": 14.86233108108108, + "grad_norm": 0.2103220671415329, + "learning_rate": 2.7469333333333335e-05, + "loss": 0.0179, + "step": 17597 + }, + { + "epoch": 14.863175675675675, + "grad_norm": 0.24709452688694, + "learning_rate": 2.7469e-05, + "loss": 0.018, + "step": 17598 + }, + { + "epoch": 14.86402027027027, + "grad_norm": 0.33775052428245544, + "learning_rate": 2.7468666666666667e-05, + "loss": 0.0217, + "step": 17599 + }, + { + "epoch": 14.864864864864865, + "grad_norm": 0.20594516396522522, + "learning_rate": 2.7468333333333336e-05, + "loss": 0.0176, + "step": 17600 + }, + { + "epoch": 14.86570945945946, + "grad_norm": 0.398640513420105, + "learning_rate": 2.7468e-05, + "loss": 0.0226, + "step": 17601 + }, + { + "epoch": 14.866554054054054, + "grad_norm": 0.30193856358528137, + "learning_rate": 2.7467666666666668e-05, + "loss": 0.0154, + "step": 17602 + }, + { + "epoch": 14.86739864864865, + "grad_norm": 0.2912801504135132, + "learning_rate": 2.7467333333333334e-05, + "loss": 0.0186, + "step": 17603 + }, + { + "epoch": 14.868243243243244, + "grad_norm": 0.5333725810050964, + "learning_rate": 2.7467e-05, + "loss": 0.0119, + "step": 17604 + }, + { + "epoch": 14.869087837837839, + "grad_norm": 0.4019011855125427, + "learning_rate": 2.746666666666667e-05, + "loss": 0.0187, + "step": 17605 + }, + { + "epoch": 14.869932432432432, + "grad_norm": 0.3304581642150879, + "learning_rate": 2.7466333333333335e-05, + "loss": 0.0275, + "step": 17606 + }, + { + "epoch": 14.870777027027026, + "grad_norm": 0.3839356005191803, + "learning_rate": 2.7466e-05, + "loss": 0.0235, + "step": 17607 + }, + { + "epoch": 14.871621621621621, + "grad_norm": 0.30737006664276123, + "learning_rate": 2.7465666666666667e-05, + "loss": 0.0213, + "step": 17608 + }, + { + "epoch": 14.872466216216216, + "grad_norm": 1.099241852760315, + "learning_rate": 2.7465333333333336e-05, + "loss": 0.0135, + "step": 17609 + }, + { + "epoch": 14.87331081081081, + "grad_norm": 0.22766166925430298, + "learning_rate": 2.7465e-05, + "loss": 0.0189, + "step": 17610 + }, + { + "epoch": 14.874155405405405, + "grad_norm": 0.20895130932331085, + "learning_rate": 2.7464666666666668e-05, + "loss": 0.0126, + "step": 17611 + }, + { + "epoch": 14.875, + "grad_norm": 0.1848423033952713, + "learning_rate": 2.7464333333333334e-05, + "loss": 0.0155, + "step": 17612 + }, + { + "epoch": 14.875844594594595, + "grad_norm": 0.30462172627449036, + "learning_rate": 2.7464e-05, + "loss": 0.0195, + "step": 17613 + }, + { + "epoch": 14.87668918918919, + "grad_norm": 0.34097644686698914, + "learning_rate": 2.746366666666667e-05, + "loss": 0.0214, + "step": 17614 + }, + { + "epoch": 14.877533783783784, + "grad_norm": 0.8221174478530884, + "learning_rate": 2.7463333333333332e-05, + "loss": 0.0209, + "step": 17615 + }, + { + "epoch": 14.878378378378379, + "grad_norm": 0.6123266220092773, + "learning_rate": 2.7463e-05, + "loss": 0.034, + "step": 17616 + }, + { + "epoch": 14.879222972972974, + "grad_norm": 0.4231208562850952, + "learning_rate": 2.7462666666666667e-05, + "loss": 0.0147, + "step": 17617 + }, + { + "epoch": 14.880067567567568, + "grad_norm": 0.3224676847457886, + "learning_rate": 2.7462333333333333e-05, + "loss": 0.0163, + "step": 17618 + }, + { + "epoch": 14.880912162162161, + "grad_norm": 0.19646748900413513, + "learning_rate": 2.7462e-05, + "loss": 0.0099, + "step": 17619 + }, + { + "epoch": 14.881756756756756, + "grad_norm": 0.43470683693885803, + "learning_rate": 2.7461666666666668e-05, + "loss": 0.0197, + "step": 17620 + }, + { + "epoch": 14.88260135135135, + "grad_norm": 0.30351507663726807, + "learning_rate": 2.7461333333333334e-05, + "loss": 0.0141, + "step": 17621 + }, + { + "epoch": 14.883445945945946, + "grad_norm": 0.41376516222953796, + "learning_rate": 2.7461e-05, + "loss": 0.0213, + "step": 17622 + }, + { + "epoch": 14.88429054054054, + "grad_norm": 0.4416322410106659, + "learning_rate": 2.746066666666667e-05, + "loss": 0.0146, + "step": 17623 + }, + { + "epoch": 14.885135135135135, + "grad_norm": 0.5596340894699097, + "learning_rate": 2.7460333333333332e-05, + "loss": 0.0301, + "step": 17624 + }, + { + "epoch": 14.88597972972973, + "grad_norm": 0.3675873875617981, + "learning_rate": 2.746e-05, + "loss": 0.0287, + "step": 17625 + }, + { + "epoch": 14.886824324324325, + "grad_norm": 0.5287912487983704, + "learning_rate": 2.7459666666666667e-05, + "loss": 0.0203, + "step": 17626 + }, + { + "epoch": 14.88766891891892, + "grad_norm": 0.5866295099258423, + "learning_rate": 2.7459333333333333e-05, + "loss": 0.189, + "step": 17627 + }, + { + "epoch": 14.888513513513514, + "grad_norm": 0.6216700077056885, + "learning_rate": 2.7459e-05, + "loss": 0.1095, + "step": 17628 + }, + { + "epoch": 14.889358108108109, + "grad_norm": 0.3281610906124115, + "learning_rate": 2.745866666666667e-05, + "loss": 0.0426, + "step": 17629 + }, + { + "epoch": 14.890202702702704, + "grad_norm": 0.2264558970928192, + "learning_rate": 2.7458333333333334e-05, + "loss": 0.0272, + "step": 17630 + }, + { + "epoch": 14.891047297297296, + "grad_norm": 0.32763826847076416, + "learning_rate": 2.7458e-05, + "loss": 0.0253, + "step": 17631 + }, + { + "epoch": 14.891891891891891, + "grad_norm": 0.1877337098121643, + "learning_rate": 2.745766666666667e-05, + "loss": 0.015, + "step": 17632 + }, + { + "epoch": 14.892736486486486, + "grad_norm": 0.7113193869590759, + "learning_rate": 2.7457333333333332e-05, + "loss": 0.0382, + "step": 17633 + }, + { + "epoch": 14.89358108108108, + "grad_norm": 0.15033495426177979, + "learning_rate": 2.7457e-05, + "loss": 0.0116, + "step": 17634 + }, + { + "epoch": 14.894425675675675, + "grad_norm": 0.4501470923423767, + "learning_rate": 2.7456666666666667e-05, + "loss": 0.0352, + "step": 17635 + }, + { + "epoch": 14.89527027027027, + "grad_norm": 0.20506520569324493, + "learning_rate": 2.7456333333333333e-05, + "loss": 0.0134, + "step": 17636 + }, + { + "epoch": 14.896114864864865, + "grad_norm": 0.314855694770813, + "learning_rate": 2.7456000000000003e-05, + "loss": 0.0189, + "step": 17637 + }, + { + "epoch": 14.89695945945946, + "grad_norm": 0.24812310934066772, + "learning_rate": 2.745566666666667e-05, + "loss": 0.0196, + "step": 17638 + }, + { + "epoch": 14.897804054054054, + "grad_norm": 1.175771951675415, + "learning_rate": 2.7455333333333335e-05, + "loss": 0.0164, + "step": 17639 + }, + { + "epoch": 14.89864864864865, + "grad_norm": 0.3231068551540375, + "learning_rate": 2.7455e-05, + "loss": 0.0342, + "step": 17640 + }, + { + "epoch": 14.899493243243244, + "grad_norm": 0.28202322125434875, + "learning_rate": 2.745466666666667e-05, + "loss": 0.0225, + "step": 17641 + }, + { + "epoch": 14.900337837837839, + "grad_norm": 0.6419926881790161, + "learning_rate": 2.7454333333333332e-05, + "loss": 0.0279, + "step": 17642 + }, + { + "epoch": 14.901182432432432, + "grad_norm": 0.29272252321243286, + "learning_rate": 2.7454000000000002e-05, + "loss": 0.0218, + "step": 17643 + }, + { + "epoch": 14.902027027027026, + "grad_norm": 0.20209185779094696, + "learning_rate": 2.7453666666666664e-05, + "loss": 0.0122, + "step": 17644 + }, + { + "epoch": 14.902871621621621, + "grad_norm": 0.22207865118980408, + "learning_rate": 2.7453333333333334e-05, + "loss": 0.0139, + "step": 17645 + }, + { + "epoch": 14.903716216216216, + "grad_norm": 0.25113388895988464, + "learning_rate": 2.7453000000000003e-05, + "loss": 0.0139, + "step": 17646 + }, + { + "epoch": 14.90456081081081, + "grad_norm": 0.3360274136066437, + "learning_rate": 2.7452666666666666e-05, + "loss": 0.0144, + "step": 17647 + }, + { + "epoch": 14.905405405405405, + "grad_norm": 0.28324753046035767, + "learning_rate": 2.7452333333333335e-05, + "loss": 0.0173, + "step": 17648 + }, + { + "epoch": 14.90625, + "grad_norm": 0.29547786712646484, + "learning_rate": 2.7452e-05, + "loss": 0.0169, + "step": 17649 + }, + { + "epoch": 14.907094594594595, + "grad_norm": 0.22144518792629242, + "learning_rate": 2.7451666666666667e-05, + "loss": 0.0201, + "step": 17650 + }, + { + "epoch": 14.90793918918919, + "grad_norm": 0.3689109981060028, + "learning_rate": 2.7451333333333333e-05, + "loss": 0.0143, + "step": 17651 + }, + { + "epoch": 14.908783783783784, + "grad_norm": 0.2356109619140625, + "learning_rate": 2.7451000000000002e-05, + "loss": 0.0161, + "step": 17652 + }, + { + "epoch": 14.909628378378379, + "grad_norm": 0.13267205655574799, + "learning_rate": 2.7450666666666665e-05, + "loss": 0.0075, + "step": 17653 + }, + { + "epoch": 14.910472972972974, + "grad_norm": 0.6976618766784668, + "learning_rate": 2.7450333333333334e-05, + "loss": 0.0147, + "step": 17654 + }, + { + "epoch": 14.911317567567568, + "grad_norm": 0.34871554374694824, + "learning_rate": 2.7450000000000003e-05, + "loss": 0.0162, + "step": 17655 + }, + { + "epoch": 14.912162162162161, + "grad_norm": 0.3539850115776062, + "learning_rate": 2.7449666666666666e-05, + "loss": 0.0224, + "step": 17656 + }, + { + "epoch": 14.913006756756756, + "grad_norm": 0.2260473519563675, + "learning_rate": 2.7449333333333335e-05, + "loss": 0.0163, + "step": 17657 + }, + { + "epoch": 14.91385135135135, + "grad_norm": 0.25602543354034424, + "learning_rate": 2.7449e-05, + "loss": 0.0203, + "step": 17658 + }, + { + "epoch": 14.914695945945946, + "grad_norm": 0.12294454872608185, + "learning_rate": 2.7448666666666667e-05, + "loss": 0.0115, + "step": 17659 + }, + { + "epoch": 14.91554054054054, + "grad_norm": 0.14209851622581482, + "learning_rate": 2.7448333333333333e-05, + "loss": 0.0101, + "step": 17660 + }, + { + "epoch": 14.916385135135135, + "grad_norm": 0.22784997522830963, + "learning_rate": 2.7448000000000002e-05, + "loss": 0.0182, + "step": 17661 + }, + { + "epoch": 14.91722972972973, + "grad_norm": 0.2655850946903229, + "learning_rate": 2.7447666666666668e-05, + "loss": 0.023, + "step": 17662 + }, + { + "epoch": 14.918074324324325, + "grad_norm": 0.41055870056152344, + "learning_rate": 2.7447333333333334e-05, + "loss": 0.0181, + "step": 17663 + }, + { + "epoch": 14.91891891891892, + "grad_norm": 0.5593407154083252, + "learning_rate": 2.7447000000000003e-05, + "loss": 0.022, + "step": 17664 + }, + { + "epoch": 14.919763513513514, + "grad_norm": 0.39068832993507385, + "learning_rate": 2.7446666666666666e-05, + "loss": 0.0127, + "step": 17665 + }, + { + "epoch": 14.920608108108109, + "grad_norm": 0.37154853343963623, + "learning_rate": 2.7446333333333335e-05, + "loss": 0.0186, + "step": 17666 + }, + { + "epoch": 14.921452702702704, + "grad_norm": 0.48798948526382446, + "learning_rate": 2.7446e-05, + "loss": 0.0248, + "step": 17667 + }, + { + "epoch": 14.922297297297296, + "grad_norm": 0.6619752049446106, + "learning_rate": 2.7445666666666667e-05, + "loss": 0.0185, + "step": 17668 + }, + { + "epoch": 14.923141891891891, + "grad_norm": 0.4992196559906006, + "learning_rate": 2.7445333333333333e-05, + "loss": 0.0248, + "step": 17669 + }, + { + "epoch": 14.923986486486486, + "grad_norm": 0.23566028475761414, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0102, + "step": 17670 + }, + { + "epoch": 14.92483108108108, + "grad_norm": 0.3288363218307495, + "learning_rate": 2.7444666666666668e-05, + "loss": 0.0192, + "step": 17671 + }, + { + "epoch": 14.925675675675675, + "grad_norm": 0.304519385099411, + "learning_rate": 2.7444333333333334e-05, + "loss": 0.0119, + "step": 17672 + }, + { + "epoch": 14.92652027027027, + "grad_norm": 0.38109156489372253, + "learning_rate": 2.7444e-05, + "loss": 0.0178, + "step": 17673 + }, + { + "epoch": 14.927364864864865, + "grad_norm": 0.5715152025222778, + "learning_rate": 2.7443666666666666e-05, + "loss": 0.0276, + "step": 17674 + }, + { + "epoch": 14.92820945945946, + "grad_norm": 0.3616698384284973, + "learning_rate": 2.7443333333333335e-05, + "loss": 0.0191, + "step": 17675 + }, + { + "epoch": 14.929054054054054, + "grad_norm": 1.328487515449524, + "learning_rate": 2.7442999999999998e-05, + "loss": 0.014, + "step": 17676 + }, + { + "epoch": 14.92989864864865, + "grad_norm": 0.6267707943916321, + "learning_rate": 2.7442666666666667e-05, + "loss": 0.1669, + "step": 17677 + }, + { + "epoch": 14.930743243243244, + "grad_norm": 0.49388906359672546, + "learning_rate": 2.7442333333333333e-05, + "loss": 0.0914, + "step": 17678 + }, + { + "epoch": 14.931587837837839, + "grad_norm": 0.28424444794654846, + "learning_rate": 2.7442e-05, + "loss": 0.0406, + "step": 17679 + }, + { + "epoch": 14.932432432432432, + "grad_norm": 0.2548215091228485, + "learning_rate": 2.744166666666667e-05, + "loss": 0.0217, + "step": 17680 + }, + { + "epoch": 14.933277027027026, + "grad_norm": 0.3257817327976227, + "learning_rate": 2.7441333333333334e-05, + "loss": 0.0263, + "step": 17681 + }, + { + "epoch": 14.934121621621621, + "grad_norm": 0.5279668569564819, + "learning_rate": 2.7441e-05, + "loss": 0.0304, + "step": 17682 + }, + { + "epoch": 14.934966216216216, + "grad_norm": 0.22201496362686157, + "learning_rate": 2.7440666666666666e-05, + "loss": 0.0099, + "step": 17683 + }, + { + "epoch": 14.93581081081081, + "grad_norm": 0.37126877903938293, + "learning_rate": 2.7440333333333336e-05, + "loss": 0.0203, + "step": 17684 + }, + { + "epoch": 14.936655405405405, + "grad_norm": 0.6675005555152893, + "learning_rate": 2.7439999999999998e-05, + "loss": 0.0256, + "step": 17685 + }, + { + "epoch": 14.9375, + "grad_norm": 0.4349510967731476, + "learning_rate": 2.7439666666666667e-05, + "loss": 0.0168, + "step": 17686 + }, + { + "epoch": 14.938344594594595, + "grad_norm": 0.25268977880477905, + "learning_rate": 2.7439333333333337e-05, + "loss": 0.0225, + "step": 17687 + }, + { + "epoch": 14.93918918918919, + "grad_norm": 0.4392460584640503, + "learning_rate": 2.7439e-05, + "loss": 0.0209, + "step": 17688 + }, + { + "epoch": 14.940033783783784, + "grad_norm": 0.32475757598876953, + "learning_rate": 2.743866666666667e-05, + "loss": 0.016, + "step": 17689 + }, + { + "epoch": 14.940878378378379, + "grad_norm": 0.19082976877689362, + "learning_rate": 2.7438333333333335e-05, + "loss": 0.0209, + "step": 17690 + }, + { + "epoch": 14.941722972972974, + "grad_norm": 0.30986180901527405, + "learning_rate": 2.7438e-05, + "loss": 0.0135, + "step": 17691 + }, + { + "epoch": 14.942567567567568, + "grad_norm": 0.32135313749313354, + "learning_rate": 2.7437666666666666e-05, + "loss": 0.0252, + "step": 17692 + }, + { + "epoch": 14.943412162162161, + "grad_norm": 0.609009325504303, + "learning_rate": 2.7437333333333336e-05, + "loss": 0.0197, + "step": 17693 + }, + { + "epoch": 14.944256756756756, + "grad_norm": 0.24327391386032104, + "learning_rate": 2.7437e-05, + "loss": 0.0159, + "step": 17694 + }, + { + "epoch": 14.94510135135135, + "grad_norm": 0.15849871933460236, + "learning_rate": 2.7436666666666668e-05, + "loss": 0.0113, + "step": 17695 + }, + { + "epoch": 14.945945945945946, + "grad_norm": 0.5809922218322754, + "learning_rate": 2.7436333333333337e-05, + "loss": 0.0771, + "step": 17696 + }, + { + "epoch": 14.94679054054054, + "grad_norm": 0.1934364289045334, + "learning_rate": 2.7436e-05, + "loss": 0.0159, + "step": 17697 + }, + { + "epoch": 14.947635135135135, + "grad_norm": 1.0075629949569702, + "learning_rate": 2.743566666666667e-05, + "loss": 0.0242, + "step": 17698 + }, + { + "epoch": 14.94847972972973, + "grad_norm": 0.2939158082008362, + "learning_rate": 2.7435333333333335e-05, + "loss": 0.0136, + "step": 17699 + }, + { + "epoch": 14.949324324324325, + "grad_norm": 0.5140507221221924, + "learning_rate": 2.7435e-05, + "loss": 0.024, + "step": 17700 + }, + { + "epoch": 14.95016891891892, + "grad_norm": 0.36375492811203003, + "learning_rate": 2.7434666666666667e-05, + "loss": 0.0244, + "step": 17701 + }, + { + "epoch": 14.951013513513514, + "grad_norm": 0.29163384437561035, + "learning_rate": 2.7434333333333333e-05, + "loss": 0.0112, + "step": 17702 + }, + { + "epoch": 14.951858108108109, + "grad_norm": 0.2252618968486786, + "learning_rate": 2.7434e-05, + "loss": 0.0137, + "step": 17703 + }, + { + "epoch": 14.952702702702704, + "grad_norm": 0.2694896459579468, + "learning_rate": 2.7433666666666668e-05, + "loss": 0.0143, + "step": 17704 + }, + { + "epoch": 14.953547297297296, + "grad_norm": 0.9254451990127563, + "learning_rate": 2.7433333333333334e-05, + "loss": 0.0189, + "step": 17705 + }, + { + "epoch": 14.954391891891891, + "grad_norm": 0.46640509366989136, + "learning_rate": 2.7433e-05, + "loss": 0.0259, + "step": 17706 + }, + { + "epoch": 14.955236486486486, + "grad_norm": 0.8182923793792725, + "learning_rate": 2.743266666666667e-05, + "loss": 0.0288, + "step": 17707 + }, + { + "epoch": 14.95608108108108, + "grad_norm": 0.26914501190185547, + "learning_rate": 2.743233333333333e-05, + "loss": 0.0121, + "step": 17708 + }, + { + "epoch": 14.956925675675675, + "grad_norm": 0.2650911808013916, + "learning_rate": 2.7432e-05, + "loss": 0.0187, + "step": 17709 + }, + { + "epoch": 14.95777027027027, + "grad_norm": 0.30512192845344543, + "learning_rate": 2.7431666666666667e-05, + "loss": 0.0171, + "step": 17710 + }, + { + "epoch": 14.958614864864865, + "grad_norm": 0.35458865761756897, + "learning_rate": 2.7431333333333333e-05, + "loss": 0.012, + "step": 17711 + }, + { + "epoch": 14.95945945945946, + "grad_norm": 0.1948443502187729, + "learning_rate": 2.7431000000000002e-05, + "loss": 0.0119, + "step": 17712 + }, + { + "epoch": 14.960304054054054, + "grad_norm": 0.7796433568000793, + "learning_rate": 2.7430666666666668e-05, + "loss": 0.0177, + "step": 17713 + }, + { + "epoch": 14.96114864864865, + "grad_norm": 0.37436121702194214, + "learning_rate": 2.7430333333333334e-05, + "loss": 0.0201, + "step": 17714 + }, + { + "epoch": 14.961993243243244, + "grad_norm": 0.2616153359413147, + "learning_rate": 2.743e-05, + "loss": 0.0131, + "step": 17715 + }, + { + "epoch": 14.962837837837839, + "grad_norm": 1.3489972352981567, + "learning_rate": 2.742966666666667e-05, + "loss": 0.0358, + "step": 17716 + }, + { + "epoch": 14.963682432432432, + "grad_norm": 0.2125919908285141, + "learning_rate": 2.7429333333333332e-05, + "loss": 0.014, + "step": 17717 + }, + { + "epoch": 14.964527027027026, + "grad_norm": 0.27377843856811523, + "learning_rate": 2.7429e-05, + "loss": 0.0137, + "step": 17718 + }, + { + "epoch": 14.965371621621621, + "grad_norm": 0.24960191547870636, + "learning_rate": 2.7428666666666667e-05, + "loss": 0.013, + "step": 17719 + }, + { + "epoch": 14.966216216216216, + "grad_norm": 0.34668228030204773, + "learning_rate": 2.7428333333333333e-05, + "loss": 0.0166, + "step": 17720 + }, + { + "epoch": 14.96706081081081, + "grad_norm": 0.21583634614944458, + "learning_rate": 2.7428000000000002e-05, + "loss": 0.0113, + "step": 17721 + }, + { + "epoch": 14.967905405405405, + "grad_norm": 0.3090130090713501, + "learning_rate": 2.7427666666666668e-05, + "loss": 0.0195, + "step": 17722 + }, + { + "epoch": 14.96875, + "grad_norm": 0.21136082708835602, + "learning_rate": 2.7427333333333334e-05, + "loss": 0.0099, + "step": 17723 + }, + { + "epoch": 14.969594594594595, + "grad_norm": 0.6375191807746887, + "learning_rate": 2.7427e-05, + "loss": 0.014, + "step": 17724 + }, + { + "epoch": 14.97043918918919, + "grad_norm": 0.6183659434318542, + "learning_rate": 2.742666666666667e-05, + "loss": 0.0411, + "step": 17725 + }, + { + "epoch": 14.971283783783784, + "grad_norm": 1.153014898300171, + "learning_rate": 2.7426333333333332e-05, + "loss": 0.1144, + "step": 17726 + }, + { + "epoch": 14.972128378378379, + "grad_norm": 0.49556198716163635, + "learning_rate": 2.7426e-05, + "loss": 0.1601, + "step": 17727 + }, + { + "epoch": 14.972972972972974, + "grad_norm": 0.33828574419021606, + "learning_rate": 2.7425666666666667e-05, + "loss": 0.0674, + "step": 17728 + }, + { + "epoch": 14.973817567567568, + "grad_norm": 0.3323119878768921, + "learning_rate": 2.7425333333333333e-05, + "loss": 0.0331, + "step": 17729 + }, + { + "epoch": 14.974662162162161, + "grad_norm": 0.32128289341926575, + "learning_rate": 2.7425000000000003e-05, + "loss": 0.0319, + "step": 17730 + }, + { + "epoch": 14.975506756756756, + "grad_norm": 0.2243398278951645, + "learning_rate": 2.742466666666667e-05, + "loss": 0.0157, + "step": 17731 + }, + { + "epoch": 14.97635135135135, + "grad_norm": 0.2365753948688507, + "learning_rate": 2.7424333333333334e-05, + "loss": 0.0217, + "step": 17732 + }, + { + "epoch": 14.977195945945946, + "grad_norm": 0.2926297187805176, + "learning_rate": 2.7424e-05, + "loss": 0.0185, + "step": 17733 + }, + { + "epoch": 14.97804054054054, + "grad_norm": 0.17996901273727417, + "learning_rate": 2.7423666666666666e-05, + "loss": 0.0107, + "step": 17734 + }, + { + "epoch": 14.978885135135135, + "grad_norm": 0.38347670435905457, + "learning_rate": 2.7423333333333332e-05, + "loss": 0.0196, + "step": 17735 + }, + { + "epoch": 14.97972972972973, + "grad_norm": 0.39469069242477417, + "learning_rate": 2.7423e-05, + "loss": 0.0361, + "step": 17736 + }, + { + "epoch": 14.980574324324325, + "grad_norm": 0.4916832447052002, + "learning_rate": 2.7422666666666667e-05, + "loss": 0.0186, + "step": 17737 + }, + { + "epoch": 14.98141891891892, + "grad_norm": 0.14290931820869446, + "learning_rate": 2.7422333333333333e-05, + "loss": 0.0121, + "step": 17738 + }, + { + "epoch": 14.982263513513514, + "grad_norm": 0.2096690535545349, + "learning_rate": 2.7422000000000003e-05, + "loss": 0.0163, + "step": 17739 + }, + { + "epoch": 14.983108108108109, + "grad_norm": 0.36780086159706116, + "learning_rate": 2.7421666666666665e-05, + "loss": 0.0223, + "step": 17740 + }, + { + "epoch": 14.983952702702704, + "grad_norm": 0.27499642968177795, + "learning_rate": 2.7421333333333335e-05, + "loss": 0.0171, + "step": 17741 + }, + { + "epoch": 14.984797297297296, + "grad_norm": 0.15016800165176392, + "learning_rate": 2.7421e-05, + "loss": 0.0085, + "step": 17742 + }, + { + "epoch": 14.985641891891891, + "grad_norm": 0.1922166496515274, + "learning_rate": 2.7420666666666666e-05, + "loss": 0.0123, + "step": 17743 + }, + { + "epoch": 14.986486486486486, + "grad_norm": 0.2794208228588104, + "learning_rate": 2.7420333333333332e-05, + "loss": 0.0184, + "step": 17744 + }, + { + "epoch": 14.98733108108108, + "grad_norm": 0.2551141381263733, + "learning_rate": 2.7420000000000002e-05, + "loss": 0.0123, + "step": 17745 + }, + { + "epoch": 14.988175675675675, + "grad_norm": 0.23211251199245453, + "learning_rate": 2.7419666666666668e-05, + "loss": 0.0158, + "step": 17746 + }, + { + "epoch": 14.98902027027027, + "grad_norm": 0.1688227653503418, + "learning_rate": 2.7419333333333334e-05, + "loss": 0.0087, + "step": 17747 + }, + { + "epoch": 14.989864864864865, + "grad_norm": 0.8807157874107361, + "learning_rate": 2.7419000000000003e-05, + "loss": 0.0274, + "step": 17748 + }, + { + "epoch": 14.99070945945946, + "grad_norm": 0.36007604002952576, + "learning_rate": 2.7418666666666665e-05, + "loss": 0.0154, + "step": 17749 + }, + { + "epoch": 14.991554054054054, + "grad_norm": 0.3639102578163147, + "learning_rate": 2.7418333333333335e-05, + "loss": 0.0187, + "step": 17750 + }, + { + "epoch": 14.99239864864865, + "grad_norm": 0.19586606323719025, + "learning_rate": 2.7418e-05, + "loss": 0.0078, + "step": 17751 + }, + { + "epoch": 14.993243243243244, + "grad_norm": 0.8419461250305176, + "learning_rate": 2.7417666666666667e-05, + "loss": 0.0203, + "step": 17752 + }, + { + "epoch": 14.994087837837839, + "grad_norm": 0.46233949065208435, + "learning_rate": 2.7417333333333333e-05, + "loss": 0.0181, + "step": 17753 + }, + { + "epoch": 14.994932432432432, + "grad_norm": 0.25799980759620667, + "learning_rate": 2.7417000000000002e-05, + "loss": 0.0103, + "step": 17754 + }, + { + "epoch": 14.995777027027026, + "grad_norm": 0.2140788584947586, + "learning_rate": 2.7416666666666668e-05, + "loss": 0.012, + "step": 17755 + }, + { + "epoch": 14.996621621621621, + "grad_norm": 0.43426910042762756, + "learning_rate": 2.7416333333333334e-05, + "loss": 0.0135, + "step": 17756 + }, + { + "epoch": 14.997466216216216, + "grad_norm": 0.4623633623123169, + "learning_rate": 2.7416000000000003e-05, + "loss": 0.0121, + "step": 17757 + }, + { + "epoch": 14.99831081081081, + "grad_norm": 0.5217733383178711, + "learning_rate": 2.7415666666666666e-05, + "loss": 0.018, + "step": 17758 + }, + { + "epoch": 14.999155405405405, + "grad_norm": 0.6287567615509033, + "learning_rate": 2.7415333333333335e-05, + "loss": 0.029, + "step": 17759 + }, + { + "epoch": 15.0, + "grad_norm": 0.8148454427719116, + "learning_rate": 2.7415e-05, + "loss": 0.0452, + "step": 17760 + }, + { + "epoch": 15.000844594594595, + "grad_norm": 0.5032569169998169, + "learning_rate": 2.7414666666666667e-05, + "loss": 0.1756, + "step": 17761 + }, + { + "epoch": 15.00168918918919, + "grad_norm": 0.5208267569541931, + "learning_rate": 2.7414333333333336e-05, + "loss": 0.1274, + "step": 17762 + }, + { + "epoch": 15.002533783783784, + "grad_norm": 0.4059546887874603, + "learning_rate": 2.7414e-05, + "loss": 0.0431, + "step": 17763 + }, + { + "epoch": 15.003378378378379, + "grad_norm": 0.38586604595184326, + "learning_rate": 2.7413666666666668e-05, + "loss": 0.0233, + "step": 17764 + }, + { + "epoch": 15.004222972972974, + "grad_norm": 0.233872652053833, + "learning_rate": 2.7413333333333334e-05, + "loss": 0.0241, + "step": 17765 + }, + { + "epoch": 15.005067567567568, + "grad_norm": 0.228112131357193, + "learning_rate": 2.7413e-05, + "loss": 0.0235, + "step": 17766 + }, + { + "epoch": 15.005912162162161, + "grad_norm": 0.3856860399246216, + "learning_rate": 2.7412666666666666e-05, + "loss": 0.0211, + "step": 17767 + }, + { + "epoch": 15.006756756756756, + "grad_norm": 0.25394323468208313, + "learning_rate": 2.7412333333333335e-05, + "loss": 0.0132, + "step": 17768 + }, + { + "epoch": 15.00760135135135, + "grad_norm": 0.26148298382759094, + "learning_rate": 2.7411999999999998e-05, + "loss": 0.0195, + "step": 17769 + }, + { + "epoch": 15.008445945945946, + "grad_norm": 0.22444087266921997, + "learning_rate": 2.7411666666666667e-05, + "loss": 0.0175, + "step": 17770 + }, + { + "epoch": 15.00929054054054, + "grad_norm": 0.27129054069519043, + "learning_rate": 2.7411333333333336e-05, + "loss": 0.0188, + "step": 17771 + }, + { + "epoch": 15.010135135135135, + "grad_norm": 0.36476704478263855, + "learning_rate": 2.7411e-05, + "loss": 0.0108, + "step": 17772 + }, + { + "epoch": 15.01097972972973, + "grad_norm": 0.5422179698944092, + "learning_rate": 2.7410666666666668e-05, + "loss": 0.0206, + "step": 17773 + }, + { + "epoch": 15.011824324324325, + "grad_norm": 0.1802860051393509, + "learning_rate": 2.7410333333333334e-05, + "loss": 0.0111, + "step": 17774 + }, + { + "epoch": 15.01266891891892, + "grad_norm": 0.18882052600383759, + "learning_rate": 2.741e-05, + "loss": 0.0254, + "step": 17775 + }, + { + "epoch": 15.013513513513514, + "grad_norm": 0.243893563747406, + "learning_rate": 2.7409666666666666e-05, + "loss": 0.0157, + "step": 17776 + }, + { + "epoch": 15.014358108108109, + "grad_norm": 0.26985758543014526, + "learning_rate": 2.7409333333333335e-05, + "loss": 0.0159, + "step": 17777 + }, + { + "epoch": 15.015202702702704, + "grad_norm": 0.33246859908103943, + "learning_rate": 2.7408999999999998e-05, + "loss": 0.0139, + "step": 17778 + }, + { + "epoch": 15.016047297297296, + "grad_norm": 0.2664749324321747, + "learning_rate": 2.7408666666666667e-05, + "loss": 0.0168, + "step": 17779 + }, + { + "epoch": 15.016891891891891, + "grad_norm": 0.26637211441993713, + "learning_rate": 2.7408333333333337e-05, + "loss": 0.0183, + "step": 17780 + }, + { + "epoch": 15.017736486486486, + "grad_norm": 0.24230951070785522, + "learning_rate": 2.7408e-05, + "loss": 0.0102, + "step": 17781 + }, + { + "epoch": 15.01858108108108, + "grad_norm": 0.0908738225698471, + "learning_rate": 2.740766666666667e-05, + "loss": 0.0066, + "step": 17782 + }, + { + "epoch": 15.019425675675675, + "grad_norm": 0.17157644033432007, + "learning_rate": 2.7407333333333334e-05, + "loss": 0.0115, + "step": 17783 + }, + { + "epoch": 15.02027027027027, + "grad_norm": 0.1851910799741745, + "learning_rate": 2.7407e-05, + "loss": 0.0162, + "step": 17784 + }, + { + "epoch": 15.021114864864865, + "grad_norm": 0.1804533749818802, + "learning_rate": 2.7406666666666666e-05, + "loss": 0.0136, + "step": 17785 + }, + { + "epoch": 15.02195945945946, + "grad_norm": 0.15672919154167175, + "learning_rate": 2.7406333333333336e-05, + "loss": 0.0067, + "step": 17786 + }, + { + "epoch": 15.022804054054054, + "grad_norm": 0.233239084482193, + "learning_rate": 2.7406e-05, + "loss": 0.0142, + "step": 17787 + }, + { + "epoch": 15.02364864864865, + "grad_norm": 0.25374120473861694, + "learning_rate": 2.7405666666666667e-05, + "loss": 0.0142, + "step": 17788 + }, + { + "epoch": 15.024493243243244, + "grad_norm": 0.15914849936962128, + "learning_rate": 2.7405333333333337e-05, + "loss": 0.0105, + "step": 17789 + }, + { + "epoch": 15.025337837837839, + "grad_norm": 0.25468748807907104, + "learning_rate": 2.7405e-05, + "loss": 0.0137, + "step": 17790 + }, + { + "epoch": 15.026182432432432, + "grad_norm": 0.3040173649787903, + "learning_rate": 2.740466666666667e-05, + "loss": 0.012, + "step": 17791 + }, + { + "epoch": 15.027027027027026, + "grad_norm": 0.1803482174873352, + "learning_rate": 2.7404333333333335e-05, + "loss": 0.0108, + "step": 17792 + }, + { + "epoch": 15.027871621621621, + "grad_norm": 0.3277655839920044, + "learning_rate": 2.7404e-05, + "loss": 0.019, + "step": 17793 + }, + { + "epoch": 15.028716216216216, + "grad_norm": 0.5858389735221863, + "learning_rate": 2.7403666666666666e-05, + "loss": 0.0129, + "step": 17794 + }, + { + "epoch": 15.02956081081081, + "grad_norm": 0.30318087339401245, + "learning_rate": 2.7403333333333332e-05, + "loss": 0.0092, + "step": 17795 + }, + { + "epoch": 15.030405405405405, + "grad_norm": 0.44646647572517395, + "learning_rate": 2.7403000000000002e-05, + "loss": 0.0106, + "step": 17796 + }, + { + "epoch": 15.03125, + "grad_norm": 0.14338061213493347, + "learning_rate": 2.7402666666666668e-05, + "loss": 0.0076, + "step": 17797 + }, + { + "epoch": 15.032094594594595, + "grad_norm": 0.7523435354232788, + "learning_rate": 2.7402333333333334e-05, + "loss": 0.0633, + "step": 17798 + }, + { + "epoch": 15.03293918918919, + "grad_norm": 0.21822434663772583, + "learning_rate": 2.7402e-05, + "loss": 0.0137, + "step": 17799 + }, + { + "epoch": 15.033783783783784, + "grad_norm": 0.19263312220573425, + "learning_rate": 2.740166666666667e-05, + "loss": 0.0123, + "step": 17800 + }, + { + "epoch": 15.034628378378379, + "grad_norm": 0.3246876299381256, + "learning_rate": 2.740133333333333e-05, + "loss": 0.0196, + "step": 17801 + }, + { + "epoch": 15.035472972972974, + "grad_norm": 0.17559856176376343, + "learning_rate": 2.7401e-05, + "loss": 0.0091, + "step": 17802 + }, + { + "epoch": 15.036317567567568, + "grad_norm": 0.6271378993988037, + "learning_rate": 2.7400666666666667e-05, + "loss": 0.0082, + "step": 17803 + }, + { + "epoch": 15.037162162162161, + "grad_norm": 0.26438772678375244, + "learning_rate": 2.7400333333333333e-05, + "loss": 0.0121, + "step": 17804 + }, + { + "epoch": 15.038006756756756, + "grad_norm": 0.5367305874824524, + "learning_rate": 2.7400000000000002e-05, + "loss": 0.0156, + "step": 17805 + }, + { + "epoch": 15.03885135135135, + "grad_norm": 0.7372673153877258, + "learning_rate": 2.7399666666666668e-05, + "loss": 0.012, + "step": 17806 + }, + { + "epoch": 15.039695945945946, + "grad_norm": 0.29792672395706177, + "learning_rate": 2.7399333333333334e-05, + "loss": 0.0144, + "step": 17807 + }, + { + "epoch": 15.04054054054054, + "grad_norm": 0.41072067618370056, + "learning_rate": 2.7399e-05, + "loss": 0.01, + "step": 17808 + }, + { + "epoch": 15.041385135135135, + "grad_norm": 1.0822685956954956, + "learning_rate": 2.739866666666667e-05, + "loss": 0.0853, + "step": 17809 + }, + { + "epoch": 15.04222972972973, + "grad_norm": 0.3221876919269562, + "learning_rate": 2.739833333333333e-05, + "loss": 0.0201, + "step": 17810 + }, + { + "epoch": 15.043074324324325, + "grad_norm": 0.5168948769569397, + "learning_rate": 2.7398e-05, + "loss": 0.1404, + "step": 17811 + }, + { + "epoch": 15.04391891891892, + "grad_norm": 0.4330027997493744, + "learning_rate": 2.739766666666667e-05, + "loss": 0.106, + "step": 17812 + }, + { + "epoch": 15.044763513513514, + "grad_norm": 0.5735679268836975, + "learning_rate": 2.7397333333333333e-05, + "loss": 0.0482, + "step": 17813 + }, + { + "epoch": 15.045608108108109, + "grad_norm": 0.22369323670864105, + "learning_rate": 2.7397000000000002e-05, + "loss": 0.0219, + "step": 17814 + }, + { + "epoch": 15.046452702702704, + "grad_norm": 0.24221563339233398, + "learning_rate": 2.7396666666666668e-05, + "loss": 0.0192, + "step": 17815 + }, + { + "epoch": 15.047297297297296, + "grad_norm": 0.31863462924957275, + "learning_rate": 2.7396333333333334e-05, + "loss": 0.024, + "step": 17816 + }, + { + "epoch": 15.048141891891891, + "grad_norm": 0.16595731675624847, + "learning_rate": 2.7396e-05, + "loss": 0.0157, + "step": 17817 + }, + { + "epoch": 15.048986486486486, + "grad_norm": 0.248616024851799, + "learning_rate": 2.739566666666667e-05, + "loss": 0.0129, + "step": 17818 + }, + { + "epoch": 15.04983108108108, + "grad_norm": 0.37562257051467896, + "learning_rate": 2.7395333333333332e-05, + "loss": 0.0204, + "step": 17819 + }, + { + "epoch": 15.050675675675675, + "grad_norm": 0.2661958336830139, + "learning_rate": 2.7395e-05, + "loss": 0.0204, + "step": 17820 + }, + { + "epoch": 15.05152027027027, + "grad_norm": 0.16078482568264008, + "learning_rate": 2.739466666666667e-05, + "loss": 0.0151, + "step": 17821 + }, + { + "epoch": 15.052364864864865, + "grad_norm": 0.22930246591567993, + "learning_rate": 2.7394333333333333e-05, + "loss": 0.0113, + "step": 17822 + }, + { + "epoch": 15.05320945945946, + "grad_norm": 0.4380252957344055, + "learning_rate": 2.7394000000000002e-05, + "loss": 0.0178, + "step": 17823 + }, + { + "epoch": 15.054054054054054, + "grad_norm": 0.19941411912441254, + "learning_rate": 2.7393666666666665e-05, + "loss": 0.0197, + "step": 17824 + }, + { + "epoch": 15.05489864864865, + "grad_norm": 0.15048326551914215, + "learning_rate": 2.7393333333333334e-05, + "loss": 0.0095, + "step": 17825 + }, + { + "epoch": 15.055743243243244, + "grad_norm": 0.24978230893611908, + "learning_rate": 2.7393e-05, + "loss": 0.0225, + "step": 17826 + }, + { + "epoch": 15.056587837837839, + "grad_norm": 0.2672854959964752, + "learning_rate": 2.7392666666666666e-05, + "loss": 0.0124, + "step": 17827 + }, + { + "epoch": 15.057432432432432, + "grad_norm": 0.18090564012527466, + "learning_rate": 2.7392333333333332e-05, + "loss": 0.0149, + "step": 17828 + }, + { + "epoch": 15.058277027027026, + "grad_norm": 0.3704013228416443, + "learning_rate": 2.7392e-05, + "loss": 0.0257, + "step": 17829 + }, + { + "epoch": 15.059121621621621, + "grad_norm": 0.2924617826938629, + "learning_rate": 2.7391666666666667e-05, + "loss": 0.014, + "step": 17830 + }, + { + "epoch": 15.059966216216216, + "grad_norm": 0.25719907879829407, + "learning_rate": 2.7391333333333333e-05, + "loss": 0.0109, + "step": 17831 + }, + { + "epoch": 15.06081081081081, + "grad_norm": 0.34644079208374023, + "learning_rate": 2.7391000000000003e-05, + "loss": 0.0183, + "step": 17832 + }, + { + "epoch": 15.061655405405405, + "grad_norm": 0.20059962570667267, + "learning_rate": 2.7390666666666665e-05, + "loss": 0.0083, + "step": 17833 + }, + { + "epoch": 15.0625, + "grad_norm": 0.3518586754798889, + "learning_rate": 2.7390333333333334e-05, + "loss": 0.0198, + "step": 17834 + }, + { + "epoch": 15.063344594594595, + "grad_norm": 0.17779262363910675, + "learning_rate": 2.739e-05, + "loss": 0.0183, + "step": 17835 + }, + { + "epoch": 15.06418918918919, + "grad_norm": 0.2991625666618347, + "learning_rate": 2.7389666666666666e-05, + "loss": 0.0201, + "step": 17836 + }, + { + "epoch": 15.065033783783784, + "grad_norm": 0.2607681453227997, + "learning_rate": 2.7389333333333336e-05, + "loss": 0.0137, + "step": 17837 + }, + { + "epoch": 15.065878378378379, + "grad_norm": 0.44577333331108093, + "learning_rate": 2.7389e-05, + "loss": 0.0133, + "step": 17838 + }, + { + "epoch": 15.066722972972974, + "grad_norm": 0.5051764249801636, + "learning_rate": 2.7388666666666667e-05, + "loss": 0.0193, + "step": 17839 + }, + { + "epoch": 15.067567567567568, + "grad_norm": 0.23460949957370758, + "learning_rate": 2.7388333333333333e-05, + "loss": 0.0136, + "step": 17840 + }, + { + "epoch": 15.068412162162161, + "grad_norm": 0.2044050693511963, + "learning_rate": 2.7388000000000003e-05, + "loss": 0.0133, + "step": 17841 + }, + { + "epoch": 15.069256756756756, + "grad_norm": 0.5256674885749817, + "learning_rate": 2.7387666666666665e-05, + "loss": 0.0187, + "step": 17842 + }, + { + "epoch": 15.07010135135135, + "grad_norm": 0.25067973136901855, + "learning_rate": 2.7387333333333335e-05, + "loss": 0.0125, + "step": 17843 + }, + { + "epoch": 15.070945945945946, + "grad_norm": 0.128666952252388, + "learning_rate": 2.7387e-05, + "loss": 0.0083, + "step": 17844 + }, + { + "epoch": 15.07179054054054, + "grad_norm": 0.21883659064769745, + "learning_rate": 2.7386666666666666e-05, + "loss": 0.0086, + "step": 17845 + }, + { + "epoch": 15.072635135135135, + "grad_norm": 0.16294431686401367, + "learning_rate": 2.7386333333333336e-05, + "loss": 0.0064, + "step": 17846 + }, + { + "epoch": 15.07347972972973, + "grad_norm": 0.253785640001297, + "learning_rate": 2.7386000000000002e-05, + "loss": 0.0186, + "step": 17847 + }, + { + "epoch": 15.074324324324325, + "grad_norm": 0.22649720311164856, + "learning_rate": 2.7385666666666668e-05, + "loss": 0.0111, + "step": 17848 + }, + { + "epoch": 15.07516891891892, + "grad_norm": 0.26405343413352966, + "learning_rate": 2.7385333333333334e-05, + "loss": 0.0137, + "step": 17849 + }, + { + "epoch": 15.076013513513514, + "grad_norm": 0.17064374685287476, + "learning_rate": 2.7385000000000003e-05, + "loss": 0.0063, + "step": 17850 + }, + { + "epoch": 15.076858108108109, + "grad_norm": 0.1792953461408615, + "learning_rate": 2.7384666666666665e-05, + "loss": 0.01, + "step": 17851 + }, + { + "epoch": 15.077702702702704, + "grad_norm": 0.2844046354293823, + "learning_rate": 2.7384333333333335e-05, + "loss": 0.0063, + "step": 17852 + }, + { + "epoch": 15.078547297297296, + "grad_norm": 1.2834088802337646, + "learning_rate": 2.7383999999999997e-05, + "loss": 0.0193, + "step": 17853 + }, + { + "epoch": 15.079391891891891, + "grad_norm": 0.4065328538417816, + "learning_rate": 2.7383666666666667e-05, + "loss": 0.01, + "step": 17854 + }, + { + "epoch": 15.080236486486486, + "grad_norm": 0.3119105398654938, + "learning_rate": 2.7383333333333336e-05, + "loss": 0.012, + "step": 17855 + }, + { + "epoch": 15.08108108108108, + "grad_norm": 0.4629499912261963, + "learning_rate": 2.7383e-05, + "loss": 0.0124, + "step": 17856 + }, + { + "epoch": 15.081925675675675, + "grad_norm": 0.3042718768119812, + "learning_rate": 2.7382666666666668e-05, + "loss": 0.0154, + "step": 17857 + }, + { + "epoch": 15.08277027027027, + "grad_norm": 0.5756263732910156, + "learning_rate": 2.7382333333333334e-05, + "loss": 0.0102, + "step": 17858 + }, + { + "epoch": 15.083614864864865, + "grad_norm": 1.7138687372207642, + "learning_rate": 2.7382e-05, + "loss": 0.027, + "step": 17859 + }, + { + "epoch": 15.08445945945946, + "grad_norm": 0.4515911340713501, + "learning_rate": 2.7381666666666666e-05, + "loss": 0.0294, + "step": 17860 + }, + { + "epoch": 15.085304054054054, + "grad_norm": 0.557171642780304, + "learning_rate": 2.7381333333333335e-05, + "loss": 0.1541, + "step": 17861 + }, + { + "epoch": 15.08614864864865, + "grad_norm": 0.6626456379890442, + "learning_rate": 2.7381e-05, + "loss": 0.0858, + "step": 17862 + }, + { + "epoch": 15.086993243243244, + "grad_norm": 0.5498499870300293, + "learning_rate": 2.7380666666666667e-05, + "loss": 0.0588, + "step": 17863 + }, + { + "epoch": 15.087837837837839, + "grad_norm": 0.29978853464126587, + "learning_rate": 2.7380333333333336e-05, + "loss": 0.0299, + "step": 17864 + }, + { + "epoch": 15.088682432432432, + "grad_norm": 0.2471051812171936, + "learning_rate": 2.738e-05, + "loss": 0.0153, + "step": 17865 + }, + { + "epoch": 15.089527027027026, + "grad_norm": 0.3246035873889923, + "learning_rate": 2.7379666666666668e-05, + "loss": 0.0173, + "step": 17866 + }, + { + "epoch": 15.090371621621621, + "grad_norm": 0.45935705304145813, + "learning_rate": 2.7379333333333334e-05, + "loss": 0.0154, + "step": 17867 + }, + { + "epoch": 15.091216216216216, + "grad_norm": 0.197805255651474, + "learning_rate": 2.7379e-05, + "loss": 0.0086, + "step": 17868 + }, + { + "epoch": 15.09206081081081, + "grad_norm": 0.2953302264213562, + "learning_rate": 2.7378666666666666e-05, + "loss": 0.0304, + "step": 17869 + }, + { + "epoch": 15.092905405405405, + "grad_norm": 0.230815589427948, + "learning_rate": 2.7378333333333335e-05, + "loss": 0.0216, + "step": 17870 + }, + { + "epoch": 15.09375, + "grad_norm": 0.43817228078842163, + "learning_rate": 2.7378e-05, + "loss": 0.0201, + "step": 17871 + }, + { + "epoch": 15.094594594594595, + "grad_norm": 0.19492942094802856, + "learning_rate": 2.7377666666666667e-05, + "loss": 0.0162, + "step": 17872 + }, + { + "epoch": 15.09543918918919, + "grad_norm": 0.2752986550331116, + "learning_rate": 2.7377333333333336e-05, + "loss": 0.0115, + "step": 17873 + }, + { + "epoch": 15.096283783783784, + "grad_norm": 0.28321385383605957, + "learning_rate": 2.7377e-05, + "loss": 0.0113, + "step": 17874 + }, + { + "epoch": 15.097128378378379, + "grad_norm": 0.34797409176826477, + "learning_rate": 2.7376666666666668e-05, + "loss": 0.0265, + "step": 17875 + }, + { + "epoch": 15.097972972972974, + "grad_norm": 0.1460452824831009, + "learning_rate": 2.7376333333333334e-05, + "loss": 0.0126, + "step": 17876 + }, + { + "epoch": 15.098817567567568, + "grad_norm": 0.4767667055130005, + "learning_rate": 2.7376e-05, + "loss": 0.021, + "step": 17877 + }, + { + "epoch": 15.099662162162161, + "grad_norm": 0.23207077383995056, + "learning_rate": 2.7375666666666666e-05, + "loss": 0.0119, + "step": 17878 + }, + { + "epoch": 15.100506756756756, + "grad_norm": 0.187340646982193, + "learning_rate": 2.7375333333333335e-05, + "loss": 0.013, + "step": 17879 + }, + { + "epoch": 15.10135135135135, + "grad_norm": 0.2060144543647766, + "learning_rate": 2.7375e-05, + "loss": 0.0156, + "step": 17880 + }, + { + "epoch": 15.102195945945946, + "grad_norm": 0.19954657554626465, + "learning_rate": 2.7374666666666667e-05, + "loss": 0.0148, + "step": 17881 + }, + { + "epoch": 15.10304054054054, + "grad_norm": 0.19631366431713104, + "learning_rate": 2.7374333333333337e-05, + "loss": 0.0116, + "step": 17882 + }, + { + "epoch": 15.103885135135135, + "grad_norm": 0.31576478481292725, + "learning_rate": 2.7374e-05, + "loss": 0.0258, + "step": 17883 + }, + { + "epoch": 15.10472972972973, + "grad_norm": 0.21412576735019684, + "learning_rate": 2.737366666666667e-05, + "loss": 0.0176, + "step": 17884 + }, + { + "epoch": 15.105574324324325, + "grad_norm": 0.2978406548500061, + "learning_rate": 2.737333333333333e-05, + "loss": 0.0262, + "step": 17885 + }, + { + "epoch": 15.10641891891892, + "grad_norm": 0.49513736367225647, + "learning_rate": 2.7373e-05, + "loss": 0.0223, + "step": 17886 + }, + { + "epoch": 15.107263513513514, + "grad_norm": 0.438024640083313, + "learning_rate": 2.737266666666667e-05, + "loss": 0.01, + "step": 17887 + }, + { + "epoch": 15.108108108108109, + "grad_norm": 0.25288188457489014, + "learning_rate": 2.7372333333333332e-05, + "loss": 0.014, + "step": 17888 + }, + { + "epoch": 15.108952702702704, + "grad_norm": 0.1908874660730362, + "learning_rate": 2.7372e-05, + "loss": 0.0116, + "step": 17889 + }, + { + "epoch": 15.109797297297296, + "grad_norm": 0.3601333200931549, + "learning_rate": 2.7371666666666667e-05, + "loss": 0.0143, + "step": 17890 + }, + { + "epoch": 15.110641891891891, + "grad_norm": 0.5587716102600098, + "learning_rate": 2.7371333333333333e-05, + "loss": 0.0339, + "step": 17891 + }, + { + "epoch": 15.111486486486486, + "grad_norm": 0.46506235003471375, + "learning_rate": 2.7371e-05, + "loss": 0.0237, + "step": 17892 + }, + { + "epoch": 15.11233108108108, + "grad_norm": 0.5142996311187744, + "learning_rate": 2.737066666666667e-05, + "loss": 0.0239, + "step": 17893 + }, + { + "epoch": 15.113175675675675, + "grad_norm": 0.3320270776748657, + "learning_rate": 2.737033333333333e-05, + "loss": 0.0193, + "step": 17894 + }, + { + "epoch": 15.11402027027027, + "grad_norm": 0.1940671056509018, + "learning_rate": 2.737e-05, + "loss": 0.0101, + "step": 17895 + }, + { + "epoch": 15.114864864864865, + "grad_norm": 0.27859723567962646, + "learning_rate": 2.736966666666667e-05, + "loss": 0.0132, + "step": 17896 + }, + { + "epoch": 15.11570945945946, + "grad_norm": 0.22287088632583618, + "learning_rate": 2.7369333333333332e-05, + "loss": 0.0207, + "step": 17897 + }, + { + "epoch": 15.116554054054054, + "grad_norm": 0.22517576813697815, + "learning_rate": 2.7369000000000002e-05, + "loss": 0.0147, + "step": 17898 + }, + { + "epoch": 15.11739864864865, + "grad_norm": 0.18707115948200226, + "learning_rate": 2.7368666666666668e-05, + "loss": 0.0152, + "step": 17899 + }, + { + "epoch": 15.118243243243244, + "grad_norm": 0.2764994204044342, + "learning_rate": 2.7368333333333334e-05, + "loss": 0.0169, + "step": 17900 + }, + { + "epoch": 15.119087837837839, + "grad_norm": 0.20916599035263062, + "learning_rate": 2.7368e-05, + "loss": 0.0146, + "step": 17901 + }, + { + "epoch": 15.119932432432432, + "grad_norm": 0.22287094593048096, + "learning_rate": 2.736766666666667e-05, + "loss": 0.0087, + "step": 17902 + }, + { + "epoch": 15.120777027027026, + "grad_norm": 0.3334982693195343, + "learning_rate": 2.736733333333333e-05, + "loss": 0.0181, + "step": 17903 + }, + { + "epoch": 15.121621621621621, + "grad_norm": 0.17745226621627808, + "learning_rate": 2.7367e-05, + "loss": 0.01, + "step": 17904 + }, + { + "epoch": 15.122466216216216, + "grad_norm": 0.17874053120613098, + "learning_rate": 2.736666666666667e-05, + "loss": 0.0073, + "step": 17905 + }, + { + "epoch": 15.12331081081081, + "grad_norm": 0.20301346480846405, + "learning_rate": 2.7366333333333333e-05, + "loss": 0.0089, + "step": 17906 + }, + { + "epoch": 15.124155405405405, + "grad_norm": 0.4859938323497772, + "learning_rate": 2.7366000000000002e-05, + "loss": 0.0158, + "step": 17907 + }, + { + "epoch": 15.125, + "grad_norm": 0.37216049432754517, + "learning_rate": 2.7365666666666668e-05, + "loss": 0.0185, + "step": 17908 + }, + { + "epoch": 15.125844594594595, + "grad_norm": 0.41287627816200256, + "learning_rate": 2.7365333333333334e-05, + "loss": 0.0131, + "step": 17909 + }, + { + "epoch": 15.12668918918919, + "grad_norm": 0.371700644493103, + "learning_rate": 2.7365e-05, + "loss": 0.0101, + "step": 17910 + }, + { + "epoch": 15.127533783783784, + "grad_norm": 0.5538496971130371, + "learning_rate": 2.736466666666667e-05, + "loss": 0.1617, + "step": 17911 + }, + { + "epoch": 15.128378378378379, + "grad_norm": 0.42266231775283813, + "learning_rate": 2.7364333333333335e-05, + "loss": 0.0743, + "step": 17912 + }, + { + "epoch": 15.129222972972974, + "grad_norm": 0.3603571951389313, + "learning_rate": 2.7364e-05, + "loss": 0.0612, + "step": 17913 + }, + { + "epoch": 15.130067567567568, + "grad_norm": 0.2904580533504486, + "learning_rate": 2.7363666666666667e-05, + "loss": 0.021, + "step": 17914 + }, + { + "epoch": 15.130912162162161, + "grad_norm": 0.2537823021411896, + "learning_rate": 2.7363333333333333e-05, + "loss": 0.0224, + "step": 17915 + }, + { + "epoch": 15.131756756756756, + "grad_norm": 0.41148045659065247, + "learning_rate": 2.7363000000000002e-05, + "loss": 0.0195, + "step": 17916 + }, + { + "epoch": 15.13260135135135, + "grad_norm": 0.34218892455101013, + "learning_rate": 2.7362666666666665e-05, + "loss": 0.0159, + "step": 17917 + }, + { + "epoch": 15.133445945945946, + "grad_norm": 0.15405705571174622, + "learning_rate": 2.7362333333333334e-05, + "loss": 0.0174, + "step": 17918 + }, + { + "epoch": 15.13429054054054, + "grad_norm": 0.24384364485740662, + "learning_rate": 2.7362e-05, + "loss": 0.025, + "step": 17919 + }, + { + "epoch": 15.135135135135135, + "grad_norm": 0.4384803771972656, + "learning_rate": 2.7361666666666666e-05, + "loss": 0.022, + "step": 17920 + }, + { + "epoch": 15.13597972972973, + "grad_norm": 0.211355060338974, + "learning_rate": 2.7361333333333335e-05, + "loss": 0.0143, + "step": 17921 + }, + { + "epoch": 15.136824324324325, + "grad_norm": 0.29673150181770325, + "learning_rate": 2.7361e-05, + "loss": 0.0255, + "step": 17922 + }, + { + "epoch": 15.13766891891892, + "grad_norm": 0.1848924458026886, + "learning_rate": 2.7360666666666667e-05, + "loss": 0.0099, + "step": 17923 + }, + { + "epoch": 15.138513513513514, + "grad_norm": 0.17067047953605652, + "learning_rate": 2.7360333333333333e-05, + "loss": 0.0177, + "step": 17924 + }, + { + "epoch": 15.139358108108109, + "grad_norm": 0.22936032712459564, + "learning_rate": 2.7360000000000002e-05, + "loss": 0.0085, + "step": 17925 + }, + { + "epoch": 15.140202702702704, + "grad_norm": 0.3937418460845947, + "learning_rate": 2.7359666666666665e-05, + "loss": 0.0242, + "step": 17926 + }, + { + "epoch": 15.141047297297296, + "grad_norm": 0.2056439220905304, + "learning_rate": 2.7359333333333334e-05, + "loss": 0.0186, + "step": 17927 + }, + { + "epoch": 15.141891891891891, + "grad_norm": 0.22797517478466034, + "learning_rate": 2.7359e-05, + "loss": 0.0134, + "step": 17928 + }, + { + "epoch": 15.142736486486486, + "grad_norm": 0.1745963990688324, + "learning_rate": 2.7358666666666666e-05, + "loss": 0.008, + "step": 17929 + }, + { + "epoch": 15.14358108108108, + "grad_norm": 0.21645978093147278, + "learning_rate": 2.7358333333333335e-05, + "loss": 0.0155, + "step": 17930 + }, + { + "epoch": 15.144425675675675, + "grad_norm": 0.19989944994449615, + "learning_rate": 2.7358e-05, + "loss": 0.0108, + "step": 17931 + }, + { + "epoch": 15.14527027027027, + "grad_norm": 0.1926981508731842, + "learning_rate": 2.7357666666666667e-05, + "loss": 0.0167, + "step": 17932 + }, + { + "epoch": 15.146114864864865, + "grad_norm": 0.20132455229759216, + "learning_rate": 2.7357333333333333e-05, + "loss": 0.0123, + "step": 17933 + }, + { + "epoch": 15.14695945945946, + "grad_norm": 0.18613150715827942, + "learning_rate": 2.7357000000000003e-05, + "loss": 0.0128, + "step": 17934 + }, + { + "epoch": 15.147804054054054, + "grad_norm": 0.6037470102310181, + "learning_rate": 2.7356666666666665e-05, + "loss": 0.0138, + "step": 17935 + }, + { + "epoch": 15.14864864864865, + "grad_norm": 0.16777366399765015, + "learning_rate": 2.7356333333333334e-05, + "loss": 0.0136, + "step": 17936 + }, + { + "epoch": 15.149493243243244, + "grad_norm": 0.1524130403995514, + "learning_rate": 2.7356000000000004e-05, + "loss": 0.0076, + "step": 17937 + }, + { + "epoch": 15.150337837837839, + "grad_norm": 0.27257898449897766, + "learning_rate": 2.7355666666666666e-05, + "loss": 0.0173, + "step": 17938 + }, + { + "epoch": 15.151182432432432, + "grad_norm": 0.17007775604724884, + "learning_rate": 2.7355333333333336e-05, + "loss": 0.0115, + "step": 17939 + }, + { + "epoch": 15.152027027027026, + "grad_norm": 0.5597246289253235, + "learning_rate": 2.7355e-05, + "loss": 0.0221, + "step": 17940 + }, + { + "epoch": 15.152871621621621, + "grad_norm": 0.24142958223819733, + "learning_rate": 2.7354666666666667e-05, + "loss": 0.0096, + "step": 17941 + }, + { + "epoch": 15.153716216216216, + "grad_norm": 0.3330017626285553, + "learning_rate": 2.7354333333333333e-05, + "loss": 0.0168, + "step": 17942 + }, + { + "epoch": 15.15456081081081, + "grad_norm": 0.325987309217453, + "learning_rate": 2.7354000000000003e-05, + "loss": 0.0127, + "step": 17943 + }, + { + "epoch": 15.155405405405405, + "grad_norm": 0.31468915939331055, + "learning_rate": 2.7353666666666665e-05, + "loss": 0.0134, + "step": 17944 + }, + { + "epoch": 15.15625, + "grad_norm": 0.24547742307186127, + "learning_rate": 2.7353333333333335e-05, + "loss": 0.0096, + "step": 17945 + }, + { + "epoch": 15.157094594594595, + "grad_norm": 0.29970622062683105, + "learning_rate": 2.7353e-05, + "loss": 0.0111, + "step": 17946 + }, + { + "epoch": 15.15793918918919, + "grad_norm": 0.10665251314640045, + "learning_rate": 2.7352666666666666e-05, + "loss": 0.0061, + "step": 17947 + }, + { + "epoch": 15.158783783783784, + "grad_norm": 0.3694668412208557, + "learning_rate": 2.7352333333333336e-05, + "loss": 0.0589, + "step": 17948 + }, + { + "epoch": 15.159628378378379, + "grad_norm": 0.4641677737236023, + "learning_rate": 2.7352e-05, + "loss": 0.0205, + "step": 17949 + }, + { + "epoch": 15.160472972972974, + "grad_norm": 0.2787037193775177, + "learning_rate": 2.7351666666666668e-05, + "loss": 0.0113, + "step": 17950 + }, + { + "epoch": 15.161317567567568, + "grad_norm": 0.3360656201839447, + "learning_rate": 2.7351333333333334e-05, + "loss": 0.0164, + "step": 17951 + }, + { + "epoch": 15.162162162162161, + "grad_norm": 0.37733879685401917, + "learning_rate": 2.7351e-05, + "loss": 0.0161, + "step": 17952 + }, + { + "epoch": 15.163006756756756, + "grad_norm": 0.2823868989944458, + "learning_rate": 2.7350666666666665e-05, + "loss": 0.0119, + "step": 17953 + }, + { + "epoch": 15.16385135135135, + "grad_norm": 0.2068454474210739, + "learning_rate": 2.7350333333333335e-05, + "loss": 0.0123, + "step": 17954 + }, + { + "epoch": 15.164695945945946, + "grad_norm": 0.26467642188072205, + "learning_rate": 2.735e-05, + "loss": 0.0111, + "step": 17955 + }, + { + "epoch": 15.16554054054054, + "grad_norm": 0.2233419567346573, + "learning_rate": 2.7349666666666667e-05, + "loss": 0.0201, + "step": 17956 + }, + { + "epoch": 15.166385135135135, + "grad_norm": 0.7016283273696899, + "learning_rate": 2.7349333333333336e-05, + "loss": 0.0194, + "step": 17957 + }, + { + "epoch": 15.16722972972973, + "grad_norm": 0.8734582662582397, + "learning_rate": 2.7349e-05, + "loss": 0.0146, + "step": 17958 + }, + { + "epoch": 15.168074324324325, + "grad_norm": 0.4130729138851166, + "learning_rate": 2.7348666666666668e-05, + "loss": 0.0128, + "step": 17959 + }, + { + "epoch": 15.16891891891892, + "grad_norm": 0.6738098859786987, + "learning_rate": 2.7348333333333334e-05, + "loss": 0.0268, + "step": 17960 + }, + { + "epoch": 15.169763513513514, + "grad_norm": 0.506114661693573, + "learning_rate": 2.7348e-05, + "loss": 0.1485, + "step": 17961 + }, + { + "epoch": 15.170608108108109, + "grad_norm": 0.561430037021637, + "learning_rate": 2.734766666666667e-05, + "loss": 0.0895, + "step": 17962 + }, + { + "epoch": 15.171452702702704, + "grad_norm": 0.361662894487381, + "learning_rate": 2.7347333333333335e-05, + "loss": 0.0607, + "step": 17963 + }, + { + "epoch": 15.172297297297296, + "grad_norm": 0.2944692373275757, + "learning_rate": 2.7347e-05, + "loss": 0.026, + "step": 17964 + }, + { + "epoch": 15.173141891891891, + "grad_norm": 0.3623046875, + "learning_rate": 2.7346666666666667e-05, + "loss": 0.014, + "step": 17965 + }, + { + "epoch": 15.173986486486486, + "grad_norm": 0.21617428958415985, + "learning_rate": 2.7346333333333336e-05, + "loss": 0.0169, + "step": 17966 + }, + { + "epoch": 15.17483108108108, + "grad_norm": 0.17889651656150818, + "learning_rate": 2.7346e-05, + "loss": 0.0161, + "step": 17967 + }, + { + "epoch": 15.175675675675675, + "grad_norm": 0.24654808640480042, + "learning_rate": 2.7345666666666668e-05, + "loss": 0.0169, + "step": 17968 + }, + { + "epoch": 15.17652027027027, + "grad_norm": 0.4556775987148285, + "learning_rate": 2.7345333333333334e-05, + "loss": 0.024, + "step": 17969 + }, + { + "epoch": 15.177364864864865, + "grad_norm": 0.25740447640419006, + "learning_rate": 2.7345e-05, + "loss": 0.0218, + "step": 17970 + }, + { + "epoch": 15.17820945945946, + "grad_norm": 0.2302960604429245, + "learning_rate": 2.734466666666667e-05, + "loss": 0.0176, + "step": 17971 + }, + { + "epoch": 15.179054054054054, + "grad_norm": 0.13188496232032776, + "learning_rate": 2.7344333333333335e-05, + "loss": 0.0084, + "step": 17972 + }, + { + "epoch": 15.17989864864865, + "grad_norm": 0.3154216408729553, + "learning_rate": 2.7344e-05, + "loss": 0.0222, + "step": 17973 + }, + { + "epoch": 15.180743243243244, + "grad_norm": 0.2062019258737564, + "learning_rate": 2.7343666666666667e-05, + "loss": 0.0189, + "step": 17974 + }, + { + "epoch": 15.181587837837839, + "grad_norm": 0.5420456528663635, + "learning_rate": 2.7343333333333333e-05, + "loss": 0.0127, + "step": 17975 + }, + { + "epoch": 15.182432432432432, + "grad_norm": 0.35083064436912537, + "learning_rate": 2.7343e-05, + "loss": 0.0131, + "step": 17976 + }, + { + "epoch": 15.183277027027026, + "grad_norm": 0.42440465092658997, + "learning_rate": 2.7342666666666668e-05, + "loss": 0.0248, + "step": 17977 + }, + { + "epoch": 15.184121621621621, + "grad_norm": 0.17084020376205444, + "learning_rate": 2.7342333333333334e-05, + "loss": 0.012, + "step": 17978 + }, + { + "epoch": 15.184966216216216, + "grad_norm": 0.1872795969247818, + "learning_rate": 2.7342e-05, + "loss": 0.0172, + "step": 17979 + }, + { + "epoch": 15.18581081081081, + "grad_norm": 0.5783357620239258, + "learning_rate": 2.734166666666667e-05, + "loss": 0.0155, + "step": 17980 + }, + { + "epoch": 15.186655405405405, + "grad_norm": 0.3878561556339264, + "learning_rate": 2.7341333333333332e-05, + "loss": 0.0225, + "step": 17981 + }, + { + "epoch": 15.1875, + "grad_norm": 0.17163485288619995, + "learning_rate": 2.7341e-05, + "loss": 0.0148, + "step": 17982 + }, + { + "epoch": 15.188344594594595, + "grad_norm": 0.3971611261367798, + "learning_rate": 2.7340666666666667e-05, + "loss": 0.0177, + "step": 17983 + }, + { + "epoch": 15.18918918918919, + "grad_norm": 0.324217289686203, + "learning_rate": 2.7340333333333333e-05, + "loss": 0.0126, + "step": 17984 + }, + { + "epoch": 15.190033783783784, + "grad_norm": 0.5552520751953125, + "learning_rate": 2.734e-05, + "loss": 0.0103, + "step": 17985 + }, + { + "epoch": 15.190878378378379, + "grad_norm": 0.2872346043586731, + "learning_rate": 2.733966666666667e-05, + "loss": 0.0176, + "step": 17986 + }, + { + "epoch": 15.191722972972974, + "grad_norm": 0.2498611956834793, + "learning_rate": 2.7339333333333334e-05, + "loss": 0.0162, + "step": 17987 + }, + { + "epoch": 15.192567567567568, + "grad_norm": 0.2570425868034363, + "learning_rate": 2.7339e-05, + "loss": 0.011, + "step": 17988 + }, + { + "epoch": 15.193412162162161, + "grad_norm": 0.5945459604263306, + "learning_rate": 2.733866666666667e-05, + "loss": 0.0142, + "step": 17989 + }, + { + "epoch": 15.194256756756756, + "grad_norm": 0.20843090116977692, + "learning_rate": 2.7338333333333332e-05, + "loss": 0.0149, + "step": 17990 + }, + { + "epoch": 15.19510135135135, + "grad_norm": 0.12806685268878937, + "learning_rate": 2.7338e-05, + "loss": 0.008, + "step": 17991 + }, + { + "epoch": 15.195945945945946, + "grad_norm": 0.25281310081481934, + "learning_rate": 2.7337666666666667e-05, + "loss": 0.0256, + "step": 17992 + }, + { + "epoch": 15.19679054054054, + "grad_norm": 0.2366315722465515, + "learning_rate": 2.7337333333333333e-05, + "loss": 0.0132, + "step": 17993 + }, + { + "epoch": 15.197635135135135, + "grad_norm": 0.22954225540161133, + "learning_rate": 2.7337e-05, + "loss": 0.0135, + "step": 17994 + }, + { + "epoch": 15.19847972972973, + "grad_norm": 0.178816556930542, + "learning_rate": 2.733666666666667e-05, + "loss": 0.013, + "step": 17995 + }, + { + "epoch": 15.199324324324325, + "grad_norm": 0.18306119740009308, + "learning_rate": 2.7336333333333335e-05, + "loss": 0.0135, + "step": 17996 + }, + { + "epoch": 15.20016891891892, + "grad_norm": 0.33379143476486206, + "learning_rate": 2.7336e-05, + "loss": 0.0181, + "step": 17997 + }, + { + "epoch": 15.201013513513514, + "grad_norm": 0.28154850006103516, + "learning_rate": 2.733566666666667e-05, + "loss": 0.0111, + "step": 17998 + }, + { + "epoch": 15.201858108108109, + "grad_norm": 0.36878034472465515, + "learning_rate": 2.7335333333333332e-05, + "loss": 0.0221, + "step": 17999 + }, + { + "epoch": 15.202702702702704, + "grad_norm": 0.274894118309021, + "learning_rate": 2.7335000000000002e-05, + "loss": 0.0079, + "step": 18000 + }, + { + "epoch": 15.202702702702704, + "eval_cer": 0.08898883623805316, + "eval_loss": 0.297380268573761, + "eval_runtime": 8.5975, + "eval_samples_per_second": 11.748, + "eval_steps_per_second": 0.465, + "eval_wer": 0.2843915343915344, + "step": 18000 + }, + { + "epoch": 15.203547297297296, + "grad_norm": 0.6884171962738037, + "learning_rate": 2.7334666666666668e-05, + "loss": 0.0207, + "step": 18001 + }, + { + "epoch": 15.204391891891891, + "grad_norm": 0.45236337184906006, + "learning_rate": 2.7334333333333334e-05, + "loss": 0.0166, + "step": 18002 + }, + { + "epoch": 15.205236486486486, + "grad_norm": 0.4233097732067108, + "learning_rate": 2.7334e-05, + "loss": 0.0113, + "step": 18003 + }, + { + "epoch": 15.20608108108108, + "grad_norm": 0.6514233946800232, + "learning_rate": 2.7333666666666665e-05, + "loss": 0.0075, + "step": 18004 + }, + { + "epoch": 15.206925675675675, + "grad_norm": 0.46647870540618896, + "learning_rate": 2.7333333333333335e-05, + "loss": 0.0263, + "step": 18005 + }, + { + "epoch": 15.20777027027027, + "grad_norm": 0.5643413066864014, + "learning_rate": 2.7333e-05, + "loss": 0.0167, + "step": 18006 + }, + { + "epoch": 15.208614864864865, + "grad_norm": 0.20291204750537872, + "learning_rate": 2.7332666666666667e-05, + "loss": 0.0127, + "step": 18007 + }, + { + "epoch": 15.20945945945946, + "grad_norm": 0.9980204701423645, + "learning_rate": 2.7332333333333333e-05, + "loss": 0.0189, + "step": 18008 + }, + { + "epoch": 15.210304054054054, + "grad_norm": 0.6378279328346252, + "learning_rate": 2.7332000000000002e-05, + "loss": 0.0181, + "step": 18009 + }, + { + "epoch": 15.21114864864865, + "grad_norm": Infinity, + "learning_rate": 2.7332000000000002e-05, + "loss": 0.032, + "step": 18010 + }, + { + "epoch": 15.211993243243244, + "grad_norm": 0.5557295680046082, + "learning_rate": 2.7331666666666664e-05, + "loss": 0.1477, + "step": 18011 + }, + { + "epoch": 15.212837837837839, + "grad_norm": 0.8927812576293945, + "learning_rate": 2.7331333333333334e-05, + "loss": 0.1085, + "step": 18012 + }, + { + "epoch": 15.213682432432432, + "grad_norm": 0.43545806407928467, + "learning_rate": 2.7331000000000003e-05, + "loss": 0.0675, + "step": 18013 + }, + { + "epoch": 15.214527027027026, + "grad_norm": 0.34141433238983154, + "learning_rate": 2.7330666666666666e-05, + "loss": 0.0706, + "step": 18014 + }, + { + "epoch": 15.215371621621621, + "grad_norm": 0.22476395964622498, + "learning_rate": 2.7330333333333335e-05, + "loss": 0.0243, + "step": 18015 + }, + { + "epoch": 15.216216216216216, + "grad_norm": 0.3321186304092407, + "learning_rate": 2.733e-05, + "loss": 0.0182, + "step": 18016 + }, + { + "epoch": 15.21706081081081, + "grad_norm": 0.17994526028633118, + "learning_rate": 2.7329666666666667e-05, + "loss": 0.0161, + "step": 18017 + }, + { + "epoch": 15.217905405405405, + "grad_norm": 0.4083399474620819, + "learning_rate": 2.7329333333333333e-05, + "loss": 0.0157, + "step": 18018 + }, + { + "epoch": 15.21875, + "grad_norm": 0.4028252959251404, + "learning_rate": 2.7329000000000002e-05, + "loss": 0.0083, + "step": 18019 + }, + { + "epoch": 15.219594594594595, + "grad_norm": 0.313871830701828, + "learning_rate": 2.7328666666666665e-05, + "loss": 0.0191, + "step": 18020 + }, + { + "epoch": 15.22043918918919, + "grad_norm": 0.24096208810806274, + "learning_rate": 2.7328333333333334e-05, + "loss": 0.0164, + "step": 18021 + }, + { + "epoch": 15.221283783783784, + "grad_norm": 0.13362915813922882, + "learning_rate": 2.7328000000000003e-05, + "loss": 0.0141, + "step": 18022 + }, + { + "epoch": 15.222128378378379, + "grad_norm": 0.527019739151001, + "learning_rate": 2.7327666666666666e-05, + "loss": 0.0231, + "step": 18023 + }, + { + "epoch": 15.222972972972974, + "grad_norm": 0.21414807438850403, + "learning_rate": 2.7327333333333335e-05, + "loss": 0.0152, + "step": 18024 + }, + { + "epoch": 15.223817567567568, + "grad_norm": 0.2880793511867523, + "learning_rate": 2.7327e-05, + "loss": 0.0107, + "step": 18025 + }, + { + "epoch": 15.224662162162161, + "grad_norm": 0.13065050542354584, + "learning_rate": 2.7326666666666667e-05, + "loss": 0.0081, + "step": 18026 + }, + { + "epoch": 15.225506756756756, + "grad_norm": 0.31522125005722046, + "learning_rate": 2.7326333333333333e-05, + "loss": 0.0216, + "step": 18027 + }, + { + "epoch": 15.22635135135135, + "grad_norm": 0.3303487300872803, + "learning_rate": 2.7326000000000002e-05, + "loss": 0.0219, + "step": 18028 + }, + { + "epoch": 15.227195945945946, + "grad_norm": 0.2528036832809448, + "learning_rate": 2.7325666666666668e-05, + "loss": 0.0108, + "step": 18029 + }, + { + "epoch": 15.22804054054054, + "grad_norm": 0.2161412537097931, + "learning_rate": 2.7325333333333334e-05, + "loss": 0.0176, + "step": 18030 + }, + { + "epoch": 15.228885135135135, + "grad_norm": 0.35133227705955505, + "learning_rate": 2.7325000000000004e-05, + "loss": 0.0218, + "step": 18031 + }, + { + "epoch": 15.22972972972973, + "grad_norm": 0.16542959213256836, + "learning_rate": 2.7324666666666666e-05, + "loss": 0.0092, + "step": 18032 + }, + { + "epoch": 15.230574324324325, + "grad_norm": 0.3449667692184448, + "learning_rate": 2.7324333333333335e-05, + "loss": 0.0155, + "step": 18033 + }, + { + "epoch": 15.23141891891892, + "grad_norm": 0.18022876977920532, + "learning_rate": 2.7324e-05, + "loss": 0.0151, + "step": 18034 + }, + { + "epoch": 15.232263513513514, + "grad_norm": 0.3123176395893097, + "learning_rate": 2.7323666666666667e-05, + "loss": 0.019, + "step": 18035 + }, + { + "epoch": 15.233108108108109, + "grad_norm": 0.23571038246154785, + "learning_rate": 2.7323333333333333e-05, + "loss": 0.0133, + "step": 18036 + }, + { + "epoch": 15.233952702702704, + "grad_norm": 0.3529551923274994, + "learning_rate": 2.7323e-05, + "loss": 0.0262, + "step": 18037 + }, + { + "epoch": 15.234797297297296, + "grad_norm": 0.19583848118782043, + "learning_rate": 2.732266666666667e-05, + "loss": 0.0153, + "step": 18038 + }, + { + "epoch": 15.235641891891891, + "grad_norm": 0.23663070797920227, + "learning_rate": 2.7322333333333334e-05, + "loss": 0.0118, + "step": 18039 + }, + { + "epoch": 15.236486486486486, + "grad_norm": 0.2644903063774109, + "learning_rate": 2.7322e-05, + "loss": 0.0123, + "step": 18040 + }, + { + "epoch": 15.23733108108108, + "grad_norm": 0.16185057163238525, + "learning_rate": 2.7321666666666666e-05, + "loss": 0.0137, + "step": 18041 + }, + { + "epoch": 15.238175675675675, + "grad_norm": 0.3715205788612366, + "learning_rate": 2.7321333333333336e-05, + "loss": 0.0142, + "step": 18042 + }, + { + "epoch": 15.23902027027027, + "grad_norm": 0.16248869895935059, + "learning_rate": 2.7320999999999998e-05, + "loss": 0.0131, + "step": 18043 + }, + { + "epoch": 15.239864864864865, + "grad_norm": 0.28569504618644714, + "learning_rate": 2.7320666666666667e-05, + "loss": 0.017, + "step": 18044 + }, + { + "epoch": 15.24070945945946, + "grad_norm": 0.4315585792064667, + "learning_rate": 2.7320333333333333e-05, + "loss": 0.0217, + "step": 18045 + }, + { + "epoch": 15.241554054054054, + "grad_norm": 0.35730820894241333, + "learning_rate": 2.732e-05, + "loss": 0.0235, + "step": 18046 + }, + { + "epoch": 15.24239864864865, + "grad_norm": 0.46389514207839966, + "learning_rate": 2.731966666666667e-05, + "loss": 0.0274, + "step": 18047 + }, + { + "epoch": 15.243243243243244, + "grad_norm": 0.37407228350639343, + "learning_rate": 2.7319333333333335e-05, + "loss": 0.0108, + "step": 18048 + }, + { + "epoch": 15.244087837837839, + "grad_norm": 0.45212313532829285, + "learning_rate": 2.7319e-05, + "loss": 0.0156, + "step": 18049 + }, + { + "epoch": 15.244932432432432, + "grad_norm": 0.461635559797287, + "learning_rate": 2.7318666666666666e-05, + "loss": 0.0136, + "step": 18050 + }, + { + "epoch": 15.245777027027026, + "grad_norm": 0.24401965737342834, + "learning_rate": 2.7318333333333336e-05, + "loss": 0.0129, + "step": 18051 + }, + { + "epoch": 15.246621621621621, + "grad_norm": 0.36235666275024414, + "learning_rate": 2.7318e-05, + "loss": 0.0115, + "step": 18052 + }, + { + "epoch": 15.247466216216216, + "grad_norm": 0.4597361385822296, + "learning_rate": 2.7317666666666668e-05, + "loss": 0.0231, + "step": 18053 + }, + { + "epoch": 15.24831081081081, + "grad_norm": 0.31811365485191345, + "learning_rate": 2.7317333333333334e-05, + "loss": 0.0212, + "step": 18054 + }, + { + "epoch": 15.249155405405405, + "grad_norm": 0.44167929887771606, + "learning_rate": 2.7317e-05, + "loss": 0.0098, + "step": 18055 + }, + { + "epoch": 15.25, + "grad_norm": 0.24184057116508484, + "learning_rate": 2.731666666666667e-05, + "loss": 0.0168, + "step": 18056 + }, + { + "epoch": 15.250844594594595, + "grad_norm": 0.45723429322242737, + "learning_rate": 2.7316333333333335e-05, + "loss": 0.0179, + "step": 18057 + }, + { + "epoch": 15.25168918918919, + "grad_norm": 0.4305016100406647, + "learning_rate": 2.7316e-05, + "loss": 0.0141, + "step": 18058 + }, + { + "epoch": 15.252533783783784, + "grad_norm": 0.560950517654419, + "learning_rate": 2.7315666666666667e-05, + "loss": 0.0262, + "step": 18059 + }, + { + "epoch": 15.253378378378379, + "grad_norm": 0.7123531699180603, + "learning_rate": 2.7315333333333336e-05, + "loss": 0.0222, + "step": 18060 + }, + { + "epoch": 15.254222972972974, + "grad_norm": 0.5025020837783813, + "learning_rate": 2.7315e-05, + "loss": 0.1398, + "step": 18061 + }, + { + "epoch": 15.255067567567568, + "grad_norm": 0.3675677478313446, + "learning_rate": 2.7314666666666668e-05, + "loss": 0.0707, + "step": 18062 + }, + { + "epoch": 15.255912162162161, + "grad_norm": 0.5161731839179993, + "learning_rate": 2.7314333333333337e-05, + "loss": 0.0293, + "step": 18063 + }, + { + "epoch": 15.256756756756756, + "grad_norm": 0.3793841004371643, + "learning_rate": 2.7314e-05, + "loss": 0.0178, + "step": 18064 + }, + { + "epoch": 15.25760135135135, + "grad_norm": 0.29865071177482605, + "learning_rate": 2.731366666666667e-05, + "loss": 0.0257, + "step": 18065 + }, + { + "epoch": 15.258445945945946, + "grad_norm": 0.3236338794231415, + "learning_rate": 2.731333333333333e-05, + "loss": 0.0231, + "step": 18066 + }, + { + "epoch": 15.25929054054054, + "grad_norm": 0.3206884264945984, + "learning_rate": 2.7313e-05, + "loss": 0.0229, + "step": 18067 + }, + { + "epoch": 15.260135135135135, + "grad_norm": 0.269334614276886, + "learning_rate": 2.7312666666666667e-05, + "loss": 0.0162, + "step": 18068 + }, + { + "epoch": 15.26097972972973, + "grad_norm": 0.22553153336048126, + "learning_rate": 2.7312333333333333e-05, + "loss": 0.0188, + "step": 18069 + }, + { + "epoch": 15.261824324324325, + "grad_norm": 0.2777046263217926, + "learning_rate": 2.7312e-05, + "loss": 0.0171, + "step": 18070 + }, + { + "epoch": 15.26266891891892, + "grad_norm": 0.31364706158638, + "learning_rate": 2.7311666666666668e-05, + "loss": 0.0197, + "step": 18071 + }, + { + "epoch": 15.263513513513514, + "grad_norm": 0.16358643770217896, + "learning_rate": 2.7311333333333334e-05, + "loss": 0.0109, + "step": 18072 + }, + { + "epoch": 15.264358108108109, + "grad_norm": 0.21767258644104004, + "learning_rate": 2.7311e-05, + "loss": 0.0153, + "step": 18073 + }, + { + "epoch": 15.265202702702704, + "grad_norm": 0.30326706171035767, + "learning_rate": 2.731066666666667e-05, + "loss": 0.018, + "step": 18074 + }, + { + "epoch": 15.266047297297296, + "grad_norm": 0.2761289179325104, + "learning_rate": 2.7310333333333332e-05, + "loss": 0.0116, + "step": 18075 + }, + { + "epoch": 15.266891891891891, + "grad_norm": 0.32228994369506836, + "learning_rate": 2.731e-05, + "loss": 0.0158, + "step": 18076 + }, + { + "epoch": 15.267736486486486, + "grad_norm": 0.3274790048599243, + "learning_rate": 2.7309666666666667e-05, + "loss": 0.0197, + "step": 18077 + }, + { + "epoch": 15.26858108108108, + "grad_norm": 0.22883330285549164, + "learning_rate": 2.7309333333333333e-05, + "loss": 0.0132, + "step": 18078 + }, + { + "epoch": 15.269425675675675, + "grad_norm": 0.16066063940525055, + "learning_rate": 2.7309000000000002e-05, + "loss": 0.0124, + "step": 18079 + }, + { + "epoch": 15.27027027027027, + "grad_norm": 0.1992311030626297, + "learning_rate": 2.7308666666666668e-05, + "loss": 0.0121, + "step": 18080 + }, + { + "epoch": 15.271114864864865, + "grad_norm": 0.17975643277168274, + "learning_rate": 2.7308333333333334e-05, + "loss": 0.0115, + "step": 18081 + }, + { + "epoch": 15.27195945945946, + "grad_norm": 0.3044675588607788, + "learning_rate": 2.7308e-05, + "loss": 0.0112, + "step": 18082 + }, + { + "epoch": 15.272804054054054, + "grad_norm": 0.2529187798500061, + "learning_rate": 2.730766666666667e-05, + "loss": 0.0166, + "step": 18083 + }, + { + "epoch": 15.27364864864865, + "grad_norm": 0.17626024782657623, + "learning_rate": 2.7307333333333332e-05, + "loss": 0.0116, + "step": 18084 + }, + { + "epoch": 15.274493243243244, + "grad_norm": 0.21223124861717224, + "learning_rate": 2.7307e-05, + "loss": 0.013, + "step": 18085 + }, + { + "epoch": 15.275337837837839, + "grad_norm": 0.30752238631248474, + "learning_rate": 2.7306666666666667e-05, + "loss": 0.017, + "step": 18086 + }, + { + "epoch": 15.276182432432432, + "grad_norm": 0.23284029960632324, + "learning_rate": 2.7306333333333333e-05, + "loss": 0.0108, + "step": 18087 + }, + { + "epoch": 15.277027027027026, + "grad_norm": 0.3871188461780548, + "learning_rate": 2.7306000000000002e-05, + "loss": 0.0147, + "step": 18088 + }, + { + "epoch": 15.277871621621621, + "grad_norm": 0.2201136201620102, + "learning_rate": 2.730566666666667e-05, + "loss": 0.0104, + "step": 18089 + }, + { + "epoch": 15.278716216216216, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.7305333333333334e-05, + "loss": 0.0078, + "step": 18090 + }, + { + "epoch": 15.27956081081081, + "grad_norm": 0.2685273587703705, + "learning_rate": 2.7305e-05, + "loss": 0.0181, + "step": 18091 + }, + { + "epoch": 15.280405405405405, + "grad_norm": 0.23905067145824432, + "learning_rate": 2.730466666666667e-05, + "loss": 0.0124, + "step": 18092 + }, + { + "epoch": 15.28125, + "grad_norm": 0.3577647805213928, + "learning_rate": 2.7304333333333332e-05, + "loss": 0.0223, + "step": 18093 + }, + { + "epoch": 15.282094594594595, + "grad_norm": 0.28120824694633484, + "learning_rate": 2.7304e-05, + "loss": 0.0179, + "step": 18094 + }, + { + "epoch": 15.28293918918919, + "grad_norm": 0.4008473753929138, + "learning_rate": 2.7303666666666667e-05, + "loss": 0.0673, + "step": 18095 + }, + { + "epoch": 15.283783783783784, + "grad_norm": 0.29272469878196716, + "learning_rate": 2.7303333333333333e-05, + "loss": 0.0137, + "step": 18096 + }, + { + "epoch": 15.284628378378379, + "grad_norm": 0.4393337368965149, + "learning_rate": 2.7303000000000003e-05, + "loss": 0.0078, + "step": 18097 + }, + { + "epoch": 15.285472972972974, + "grad_norm": 0.33568379282951355, + "learning_rate": 2.7302666666666665e-05, + "loss": 0.0112, + "step": 18098 + }, + { + "epoch": 15.286317567567568, + "grad_norm": 0.25462356209754944, + "learning_rate": 2.7302333333333335e-05, + "loss": 0.0102, + "step": 18099 + }, + { + "epoch": 15.287162162162161, + "grad_norm": 0.22550411522388458, + "learning_rate": 2.7302e-05, + "loss": 0.0158, + "step": 18100 + }, + { + "epoch": 15.288006756756756, + "grad_norm": 0.43074271082878113, + "learning_rate": 2.7301666666666666e-05, + "loss": 0.0119, + "step": 18101 + }, + { + "epoch": 15.28885135135135, + "grad_norm": 0.21467554569244385, + "learning_rate": 2.7301333333333332e-05, + "loss": 0.012, + "step": 18102 + }, + { + "epoch": 15.289695945945946, + "grad_norm": 0.2623595595359802, + "learning_rate": 2.7301000000000002e-05, + "loss": 0.015, + "step": 18103 + }, + { + "epoch": 15.29054054054054, + "grad_norm": 0.7048852443695068, + "learning_rate": 2.7300666666666668e-05, + "loss": 0.0165, + "step": 18104 + }, + { + "epoch": 15.291385135135135, + "grad_norm": 0.43037140369415283, + "learning_rate": 2.7300333333333334e-05, + "loss": 0.0259, + "step": 18105 + }, + { + "epoch": 15.29222972972973, + "grad_norm": 0.48702800273895264, + "learning_rate": 2.7300000000000003e-05, + "loss": 0.0264, + "step": 18106 + }, + { + "epoch": 15.293074324324325, + "grad_norm": 0.8652085661888123, + "learning_rate": 2.7299666666666665e-05, + "loss": 0.0228, + "step": 18107 + }, + { + "epoch": 15.29391891891892, + "grad_norm": 0.21200202405452728, + "learning_rate": 2.7299333333333335e-05, + "loss": 0.0079, + "step": 18108 + }, + { + "epoch": 15.294763513513514, + "grad_norm": 0.4933154284954071, + "learning_rate": 2.7299e-05, + "loss": 0.0168, + "step": 18109 + }, + { + "epoch": 15.295608108108109, + "grad_norm": 1.0807663202285767, + "learning_rate": 2.7298666666666667e-05, + "loss": 0.0277, + "step": 18110 + }, + { + "epoch": 15.296452702702704, + "grad_norm": 0.517859160900116, + "learning_rate": 2.7298333333333333e-05, + "loss": 0.1434, + "step": 18111 + }, + { + "epoch": 15.297297297297296, + "grad_norm": 0.4425228536128998, + "learning_rate": 2.7298000000000002e-05, + "loss": 0.1021, + "step": 18112 + }, + { + "epoch": 15.298141891891891, + "grad_norm": 0.4050738215446472, + "learning_rate": 2.7297666666666668e-05, + "loss": 0.0595, + "step": 18113 + }, + { + "epoch": 15.298986486486486, + "grad_norm": 0.22147856652736664, + "learning_rate": 2.7297333333333334e-05, + "loss": 0.0295, + "step": 18114 + }, + { + "epoch": 15.29983108108108, + "grad_norm": 0.23629705607891083, + "learning_rate": 2.7297000000000003e-05, + "loss": 0.0191, + "step": 18115 + }, + { + "epoch": 15.300675675675675, + "grad_norm": 0.20761924982070923, + "learning_rate": 2.7296666666666666e-05, + "loss": 0.0187, + "step": 18116 + }, + { + "epoch": 15.30152027027027, + "grad_norm": 0.35340723395347595, + "learning_rate": 2.7296333333333335e-05, + "loss": 0.0255, + "step": 18117 + }, + { + "epoch": 15.302364864864865, + "grad_norm": 0.27313369512557983, + "learning_rate": 2.7296e-05, + "loss": 0.0221, + "step": 18118 + }, + { + "epoch": 15.30320945945946, + "grad_norm": 0.30548301339149475, + "learning_rate": 2.7295666666666667e-05, + "loss": 0.029, + "step": 18119 + }, + { + "epoch": 15.304054054054054, + "grad_norm": 0.20436853170394897, + "learning_rate": 2.7295333333333333e-05, + "loss": 0.0146, + "step": 18120 + }, + { + "epoch": 15.30489864864865, + "grad_norm": 0.16933098435401917, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0131, + "step": 18121 + }, + { + "epoch": 15.305743243243244, + "grad_norm": 0.21088926494121552, + "learning_rate": 2.7294666666666668e-05, + "loss": 0.0158, + "step": 18122 + }, + { + "epoch": 15.306587837837839, + "grad_norm": 0.190530464053154, + "learning_rate": 2.7294333333333334e-05, + "loss": 0.0152, + "step": 18123 + }, + { + "epoch": 15.307432432432432, + "grad_norm": 0.2261914610862732, + "learning_rate": 2.7294000000000003e-05, + "loss": 0.0189, + "step": 18124 + }, + { + "epoch": 15.308277027027026, + "grad_norm": 0.28845447301864624, + "learning_rate": 2.7293666666666666e-05, + "loss": 0.0174, + "step": 18125 + }, + { + "epoch": 15.309121621621621, + "grad_norm": 0.2143895924091339, + "learning_rate": 2.7293333333333335e-05, + "loss": 0.0095, + "step": 18126 + }, + { + "epoch": 15.309966216216216, + "grad_norm": 0.25440075993537903, + "learning_rate": 2.7292999999999998e-05, + "loss": 0.0121, + "step": 18127 + }, + { + "epoch": 15.31081081081081, + "grad_norm": 0.475199818611145, + "learning_rate": 2.7292666666666667e-05, + "loss": 0.0275, + "step": 18128 + }, + { + "epoch": 15.311655405405405, + "grad_norm": 0.30652719736099243, + "learning_rate": 2.7292333333333336e-05, + "loss": 0.009, + "step": 18129 + }, + { + "epoch": 15.3125, + "grad_norm": 0.23043271899223328, + "learning_rate": 2.7292e-05, + "loss": 0.0209, + "step": 18130 + }, + { + "epoch": 15.313344594594595, + "grad_norm": 0.3730531930923462, + "learning_rate": 2.7291666666666668e-05, + "loss": 0.0206, + "step": 18131 + }, + { + "epoch": 15.31418918918919, + "grad_norm": 0.1940453201532364, + "learning_rate": 2.7291333333333334e-05, + "loss": 0.0121, + "step": 18132 + }, + { + "epoch": 15.315033783783784, + "grad_norm": 0.1979050487279892, + "learning_rate": 2.7291e-05, + "loss": 0.006, + "step": 18133 + }, + { + "epoch": 15.315878378378379, + "grad_norm": 0.4511745572090149, + "learning_rate": 2.7290666666666666e-05, + "loss": 0.0134, + "step": 18134 + }, + { + "epoch": 15.316722972972974, + "grad_norm": 0.16231025755405426, + "learning_rate": 2.7290333333333335e-05, + "loss": 0.0066, + "step": 18135 + }, + { + "epoch": 15.317567567567568, + "grad_norm": 0.3858218789100647, + "learning_rate": 2.7289999999999998e-05, + "loss": 0.0204, + "step": 18136 + }, + { + "epoch": 15.318412162162161, + "grad_norm": 0.18805305659770966, + "learning_rate": 2.7289666666666667e-05, + "loss": 0.0159, + "step": 18137 + }, + { + "epoch": 15.319256756756756, + "grad_norm": 0.29254671931266785, + "learning_rate": 2.7289333333333337e-05, + "loss": 0.0142, + "step": 18138 + }, + { + "epoch": 15.32010135135135, + "grad_norm": 0.27514737844467163, + "learning_rate": 2.7289e-05, + "loss": 0.0123, + "step": 18139 + }, + { + "epoch": 15.320945945945946, + "grad_norm": 0.4019746780395508, + "learning_rate": 2.728866666666667e-05, + "loss": 0.0228, + "step": 18140 + }, + { + "epoch": 15.32179054054054, + "grad_norm": 0.20910108089447021, + "learning_rate": 2.7288333333333334e-05, + "loss": 0.0104, + "step": 18141 + }, + { + "epoch": 15.322635135135135, + "grad_norm": 0.47775495052337646, + "learning_rate": 2.7288e-05, + "loss": 0.0193, + "step": 18142 + }, + { + "epoch": 15.32347972972973, + "grad_norm": 0.21171313524246216, + "learning_rate": 2.7287666666666666e-05, + "loss": 0.0244, + "step": 18143 + }, + { + "epoch": 15.324324324324325, + "grad_norm": 0.1851014345884323, + "learning_rate": 2.7287333333333336e-05, + "loss": 0.0094, + "step": 18144 + }, + { + "epoch": 15.32516891891892, + "grad_norm": 0.237221822142601, + "learning_rate": 2.7286999999999998e-05, + "loss": 0.0143, + "step": 18145 + }, + { + "epoch": 15.326013513513514, + "grad_norm": 0.22040848433971405, + "learning_rate": 2.7286666666666667e-05, + "loss": 0.0149, + "step": 18146 + }, + { + "epoch": 15.326858108108109, + "grad_norm": 0.27407798171043396, + "learning_rate": 2.7286333333333337e-05, + "loss": 0.02, + "step": 18147 + }, + { + "epoch": 15.327702702702704, + "grad_norm": 0.23007410764694214, + "learning_rate": 2.7286e-05, + "loss": 0.0175, + "step": 18148 + }, + { + "epoch": 15.328547297297296, + "grad_norm": 0.2538798153400421, + "learning_rate": 2.728566666666667e-05, + "loss": 0.0109, + "step": 18149 + }, + { + "epoch": 15.329391891891891, + "grad_norm": 0.23991242051124573, + "learning_rate": 2.7285333333333335e-05, + "loss": 0.011, + "step": 18150 + }, + { + "epoch": 15.330236486486486, + "grad_norm": 0.2587308883666992, + "learning_rate": 2.7285e-05, + "loss": 0.0156, + "step": 18151 + }, + { + "epoch": 15.33108108108108, + "grad_norm": 0.1705309897661209, + "learning_rate": 2.7284666666666666e-05, + "loss": 0.0095, + "step": 18152 + }, + { + "epoch": 15.331925675675675, + "grad_norm": 0.30370935797691345, + "learning_rate": 2.7284333333333336e-05, + "loss": 0.0197, + "step": 18153 + }, + { + "epoch": 15.33277027027027, + "grad_norm": 0.42020168900489807, + "learning_rate": 2.7284e-05, + "loss": 0.0164, + "step": 18154 + }, + { + "epoch": 15.333614864864865, + "grad_norm": 0.21133655309677124, + "learning_rate": 2.7283666666666668e-05, + "loss": 0.0101, + "step": 18155 + }, + { + "epoch": 15.33445945945946, + "grad_norm": 0.19568075239658356, + "learning_rate": 2.7283333333333337e-05, + "loss": 0.0079, + "step": 18156 + }, + { + "epoch": 15.335304054054054, + "grad_norm": 0.2273695021867752, + "learning_rate": 2.7283e-05, + "loss": 0.0132, + "step": 18157 + }, + { + "epoch": 15.33614864864865, + "grad_norm": 0.4070001542568207, + "learning_rate": 2.728266666666667e-05, + "loss": 0.0118, + "step": 18158 + }, + { + "epoch": 15.336993243243244, + "grad_norm": 0.22409509122371674, + "learning_rate": 2.728233333333333e-05, + "loss": 0.0129, + "step": 18159 + }, + { + "epoch": 15.337837837837839, + "grad_norm": 0.6407039761543274, + "learning_rate": 2.7282e-05, + "loss": 0.0288, + "step": 18160 + }, + { + "epoch": 15.338682432432432, + "grad_norm": 0.5562635660171509, + "learning_rate": 2.7281666666666667e-05, + "loss": 0.1582, + "step": 18161 + }, + { + "epoch": 15.339527027027026, + "grad_norm": 0.28805845975875854, + "learning_rate": 2.7281333333333333e-05, + "loss": 0.0779, + "step": 18162 + }, + { + "epoch": 15.340371621621621, + "grad_norm": 0.3597899079322815, + "learning_rate": 2.7281000000000002e-05, + "loss": 0.0477, + "step": 18163 + }, + { + "epoch": 15.341216216216216, + "grad_norm": 0.32380327582359314, + "learning_rate": 2.7280666666666668e-05, + "loss": 0.0393, + "step": 18164 + }, + { + "epoch": 15.34206081081081, + "grad_norm": 0.46556851267814636, + "learning_rate": 2.7280333333333334e-05, + "loss": 0.0299, + "step": 18165 + }, + { + "epoch": 15.342905405405405, + "grad_norm": 0.44254133105278015, + "learning_rate": 2.728e-05, + "loss": 0.0239, + "step": 18166 + }, + { + "epoch": 15.34375, + "grad_norm": 0.14098747074604034, + "learning_rate": 2.727966666666667e-05, + "loss": 0.0067, + "step": 18167 + }, + { + "epoch": 15.344594594594595, + "grad_norm": 0.24688515067100525, + "learning_rate": 2.727933333333333e-05, + "loss": 0.0159, + "step": 18168 + }, + { + "epoch": 15.34543918918919, + "grad_norm": 0.21849723160266876, + "learning_rate": 2.7279e-05, + "loss": 0.0129, + "step": 18169 + }, + { + "epoch": 15.346283783783784, + "grad_norm": 0.5169563293457031, + "learning_rate": 2.7278666666666667e-05, + "loss": 0.017, + "step": 18170 + }, + { + "epoch": 15.347128378378379, + "grad_norm": 0.2775770127773285, + "learning_rate": 2.7278333333333333e-05, + "loss": 0.0217, + "step": 18171 + }, + { + "epoch": 15.347972972972974, + "grad_norm": 0.394681453704834, + "learning_rate": 2.7278000000000002e-05, + "loss": 0.0161, + "step": 18172 + }, + { + "epoch": 15.348817567567568, + "grad_norm": 0.2284432202577591, + "learning_rate": 2.7277666666666668e-05, + "loss": 0.0129, + "step": 18173 + }, + { + "epoch": 15.349662162162161, + "grad_norm": 0.3541112542152405, + "learning_rate": 2.7277333333333334e-05, + "loss": 0.0169, + "step": 18174 + }, + { + "epoch": 15.350506756756756, + "grad_norm": 0.2723517119884491, + "learning_rate": 2.7277e-05, + "loss": 0.0149, + "step": 18175 + }, + { + "epoch": 15.35135135135135, + "grad_norm": 0.4624632000923157, + "learning_rate": 2.727666666666667e-05, + "loss": 0.0217, + "step": 18176 + }, + { + "epoch": 15.352195945945946, + "grad_norm": 0.24644659459590912, + "learning_rate": 2.7276333333333332e-05, + "loss": 0.0169, + "step": 18177 + }, + { + "epoch": 15.35304054054054, + "grad_norm": 0.16635344922542572, + "learning_rate": 2.7276e-05, + "loss": 0.0101, + "step": 18178 + }, + { + "epoch": 15.353885135135135, + "grad_norm": 0.24038229882717133, + "learning_rate": 2.727566666666667e-05, + "loss": 0.0129, + "step": 18179 + }, + { + "epoch": 15.35472972972973, + "grad_norm": 0.6718043684959412, + "learning_rate": 2.7275333333333333e-05, + "loss": 0.0187, + "step": 18180 + }, + { + "epoch": 15.355574324324325, + "grad_norm": 0.32913219928741455, + "learning_rate": 2.7275000000000002e-05, + "loss": 0.0103, + "step": 18181 + }, + { + "epoch": 15.35641891891892, + "grad_norm": 0.46454814076423645, + "learning_rate": 2.7274666666666668e-05, + "loss": 0.0159, + "step": 18182 + }, + { + "epoch": 15.357263513513514, + "grad_norm": 0.1424124538898468, + "learning_rate": 2.7274333333333334e-05, + "loss": 0.0096, + "step": 18183 + }, + { + "epoch": 15.358108108108109, + "grad_norm": 0.32734981179237366, + "learning_rate": 2.7274e-05, + "loss": 0.0112, + "step": 18184 + }, + { + "epoch": 15.358952702702704, + "grad_norm": 0.3541088104248047, + "learning_rate": 2.727366666666667e-05, + "loss": 0.0138, + "step": 18185 + }, + { + "epoch": 15.359797297297296, + "grad_norm": 0.30191555619239807, + "learning_rate": 2.7273333333333332e-05, + "loss": 0.0247, + "step": 18186 + }, + { + "epoch": 15.360641891891891, + "grad_norm": 0.2799540162086487, + "learning_rate": 2.7273e-05, + "loss": 0.0165, + "step": 18187 + }, + { + "epoch": 15.361486486486486, + "grad_norm": 0.3149529993534088, + "learning_rate": 2.7272666666666667e-05, + "loss": 0.0168, + "step": 18188 + }, + { + "epoch": 15.36233108108108, + "grad_norm": 0.791868269443512, + "learning_rate": 2.7272333333333333e-05, + "loss": 0.0142, + "step": 18189 + }, + { + "epoch": 15.363175675675675, + "grad_norm": 0.23110046982765198, + "learning_rate": 2.7272000000000002e-05, + "loss": 0.0094, + "step": 18190 + }, + { + "epoch": 15.36402027027027, + "grad_norm": 0.27223169803619385, + "learning_rate": 2.7271666666666665e-05, + "loss": 0.0189, + "step": 18191 + }, + { + "epoch": 15.364864864864865, + "grad_norm": 0.146432563662529, + "learning_rate": 2.7271333333333334e-05, + "loss": 0.0104, + "step": 18192 + }, + { + "epoch": 15.36570945945946, + "grad_norm": 0.2913115620613098, + "learning_rate": 2.7271e-05, + "loss": 0.0161, + "step": 18193 + }, + { + "epoch": 15.366554054054054, + "grad_norm": 0.17614227533340454, + "learning_rate": 2.7270666666666666e-05, + "loss": 0.0111, + "step": 18194 + }, + { + "epoch": 15.36739864864865, + "grad_norm": 0.45663389563560486, + "learning_rate": 2.7270333333333332e-05, + "loss": 0.0113, + "step": 18195 + }, + { + "epoch": 15.368243243243244, + "grad_norm": 0.18032196164131165, + "learning_rate": 2.727e-05, + "loss": 0.0098, + "step": 18196 + }, + { + "epoch": 15.369087837837839, + "grad_norm": 0.3186451494693756, + "learning_rate": 2.7269666666666667e-05, + "loss": 0.0191, + "step": 18197 + }, + { + "epoch": 15.369932432432432, + "grad_norm": 0.34658846259117126, + "learning_rate": 2.7269333333333333e-05, + "loss": 0.0118, + "step": 18198 + }, + { + "epoch": 15.370777027027026, + "grad_norm": 0.31376394629478455, + "learning_rate": 2.7269000000000003e-05, + "loss": 0.0204, + "step": 18199 + }, + { + "epoch": 15.371621621621621, + "grad_norm": 0.28807976841926575, + "learning_rate": 2.7268666666666665e-05, + "loss": 0.016, + "step": 18200 + }, + { + "epoch": 15.372466216216216, + "grad_norm": 0.17428526282310486, + "learning_rate": 2.7268333333333335e-05, + "loss": 0.011, + "step": 18201 + }, + { + "epoch": 15.37331081081081, + "grad_norm": 0.6147226095199585, + "learning_rate": 2.7268e-05, + "loss": 0.015, + "step": 18202 + }, + { + "epoch": 15.374155405405405, + "grad_norm": 0.2913954257965088, + "learning_rate": 2.7267666666666666e-05, + "loss": 0.0181, + "step": 18203 + }, + { + "epoch": 15.375, + "grad_norm": 0.33630016446113586, + "learning_rate": 2.7267333333333336e-05, + "loss": 0.0172, + "step": 18204 + }, + { + "epoch": 15.375844594594595, + "grad_norm": 0.26981672644615173, + "learning_rate": 2.7267e-05, + "loss": 0.0185, + "step": 18205 + }, + { + "epoch": 15.37668918918919, + "grad_norm": 0.4033236801624298, + "learning_rate": 2.7266666666666668e-05, + "loss": 0.0126, + "step": 18206 + }, + { + "epoch": 15.377533783783784, + "grad_norm": 0.43923163414001465, + "learning_rate": 2.7266333333333334e-05, + "loss": 0.0162, + "step": 18207 + }, + { + "epoch": 15.378378378378379, + "grad_norm": 0.5771351456642151, + "learning_rate": 2.7266000000000003e-05, + "loss": 0.0152, + "step": 18208 + }, + { + "epoch": 15.379222972972974, + "grad_norm": 0.5772134065628052, + "learning_rate": 2.7265666666666665e-05, + "loss": 0.0297, + "step": 18209 + }, + { + "epoch": 15.380067567567568, + "grad_norm": 0.4545937478542328, + "learning_rate": 2.7265333333333335e-05, + "loss": 0.0102, + "step": 18210 + }, + { + "epoch": 15.380912162162161, + "grad_norm": 0.4172285199165344, + "learning_rate": 2.7265e-05, + "loss": 0.1345, + "step": 18211 + }, + { + "epoch": 15.381756756756756, + "grad_norm": 0.49393391609191895, + "learning_rate": 2.7264666666666667e-05, + "loss": 0.0825, + "step": 18212 + }, + { + "epoch": 15.38260135135135, + "grad_norm": 0.4057717025279999, + "learning_rate": 2.7264333333333336e-05, + "loss": 0.0607, + "step": 18213 + }, + { + "epoch": 15.383445945945946, + "grad_norm": 0.31476619839668274, + "learning_rate": 2.7264000000000002e-05, + "loss": 0.0247, + "step": 18214 + }, + { + "epoch": 15.38429054054054, + "grad_norm": 0.20496322214603424, + "learning_rate": 2.7263666666666668e-05, + "loss": 0.0169, + "step": 18215 + }, + { + "epoch": 15.385135135135135, + "grad_norm": 0.34170597791671753, + "learning_rate": 2.7263333333333334e-05, + "loss": 0.0208, + "step": 18216 + }, + { + "epoch": 15.38597972972973, + "grad_norm": 1.352895736694336, + "learning_rate": 2.7263e-05, + "loss": 0.0204, + "step": 18217 + }, + { + "epoch": 15.386824324324325, + "grad_norm": 0.18277104198932648, + "learning_rate": 2.7262666666666666e-05, + "loss": 0.016, + "step": 18218 + }, + { + "epoch": 15.38766891891892, + "grad_norm": 0.28226950764656067, + "learning_rate": 2.7262333333333335e-05, + "loss": 0.0199, + "step": 18219 + }, + { + "epoch": 15.388513513513514, + "grad_norm": 0.4412308931350708, + "learning_rate": 2.7261999999999997e-05, + "loss": 0.0272, + "step": 18220 + }, + { + "epoch": 15.389358108108109, + "grad_norm": 0.15789277851581573, + "learning_rate": 2.7261666666666667e-05, + "loss": 0.0084, + "step": 18221 + }, + { + "epoch": 15.390202702702704, + "grad_norm": 0.27613216638565063, + "learning_rate": 2.7261333333333336e-05, + "loss": 0.0176, + "step": 18222 + }, + { + "epoch": 15.391047297297296, + "grad_norm": 0.27034878730773926, + "learning_rate": 2.7261e-05, + "loss": 0.0197, + "step": 18223 + }, + { + "epoch": 15.391891891891891, + "grad_norm": 0.18364688754081726, + "learning_rate": 2.7260666666666668e-05, + "loss": 0.0145, + "step": 18224 + }, + { + "epoch": 15.392736486486486, + "grad_norm": 0.6616547703742981, + "learning_rate": 2.7260333333333334e-05, + "loss": 0.02, + "step": 18225 + }, + { + "epoch": 15.39358108108108, + "grad_norm": 0.5953474044799805, + "learning_rate": 2.726e-05, + "loss": 0.0226, + "step": 18226 + }, + { + "epoch": 15.394425675675675, + "grad_norm": 0.17555752396583557, + "learning_rate": 2.7259666666666666e-05, + "loss": 0.0144, + "step": 18227 + }, + { + "epoch": 15.39527027027027, + "grad_norm": 0.21275027096271515, + "learning_rate": 2.7259333333333335e-05, + "loss": 0.013, + "step": 18228 + }, + { + "epoch": 15.396114864864865, + "grad_norm": 0.18618468940258026, + "learning_rate": 2.7259e-05, + "loss": 0.0108, + "step": 18229 + }, + { + "epoch": 15.39695945945946, + "grad_norm": 0.17703182995319366, + "learning_rate": 2.7258666666666667e-05, + "loss": 0.0132, + "step": 18230 + }, + { + "epoch": 15.397804054054054, + "grad_norm": 0.19928333163261414, + "learning_rate": 2.7258333333333336e-05, + "loss": 0.018, + "step": 18231 + }, + { + "epoch": 15.39864864864865, + "grad_norm": 0.20501494407653809, + "learning_rate": 2.7258e-05, + "loss": 0.0179, + "step": 18232 + }, + { + "epoch": 15.399493243243244, + "grad_norm": 0.3188529312610626, + "learning_rate": 2.7257666666666668e-05, + "loss": 0.0199, + "step": 18233 + }, + { + "epoch": 15.400337837837839, + "grad_norm": 0.18112292885780334, + "learning_rate": 2.7257333333333334e-05, + "loss": 0.0129, + "step": 18234 + }, + { + "epoch": 15.401182432432432, + "grad_norm": 0.294336199760437, + "learning_rate": 2.7257e-05, + "loss": 0.014, + "step": 18235 + }, + { + "epoch": 15.402027027027026, + "grad_norm": 0.40510839223861694, + "learning_rate": 2.7256666666666666e-05, + "loss": 0.0197, + "step": 18236 + }, + { + "epoch": 15.402871621621621, + "grad_norm": 0.22977231442928314, + "learning_rate": 2.7256333333333335e-05, + "loss": 0.0194, + "step": 18237 + }, + { + "epoch": 15.403716216216216, + "grad_norm": 0.13460876047611237, + "learning_rate": 2.7256e-05, + "loss": 0.0068, + "step": 18238 + }, + { + "epoch": 15.40456081081081, + "grad_norm": 0.7694187164306641, + "learning_rate": 2.7255666666666667e-05, + "loss": 0.0136, + "step": 18239 + }, + { + "epoch": 15.405405405405405, + "grad_norm": 0.40295201539993286, + "learning_rate": 2.7255333333333337e-05, + "loss": 0.0169, + "step": 18240 + }, + { + "epoch": 15.40625, + "grad_norm": 0.26557543873786926, + "learning_rate": 2.7255e-05, + "loss": 0.0157, + "step": 18241 + }, + { + "epoch": 15.407094594594595, + "grad_norm": 0.20019736886024475, + "learning_rate": 2.725466666666667e-05, + "loss": 0.0081, + "step": 18242 + }, + { + "epoch": 15.40793918918919, + "grad_norm": 0.20870839059352875, + "learning_rate": 2.7254333333333334e-05, + "loss": 0.0115, + "step": 18243 + }, + { + "epoch": 15.408783783783784, + "grad_norm": 0.601452112197876, + "learning_rate": 2.7254e-05, + "loss": 0.0129, + "step": 18244 + }, + { + "epoch": 15.409628378378379, + "grad_norm": 0.29493653774261475, + "learning_rate": 2.7253666666666666e-05, + "loss": 0.0176, + "step": 18245 + }, + { + "epoch": 15.410472972972974, + "grad_norm": 0.393322229385376, + "learning_rate": 2.7253333333333336e-05, + "loss": 0.0149, + "step": 18246 + }, + { + "epoch": 15.411317567567568, + "grad_norm": 0.27638524770736694, + "learning_rate": 2.7253e-05, + "loss": 0.0192, + "step": 18247 + }, + { + "epoch": 15.412162162162161, + "grad_norm": 0.5546443462371826, + "learning_rate": 2.7252666666666667e-05, + "loss": 0.0188, + "step": 18248 + }, + { + "epoch": 15.413006756756756, + "grad_norm": 0.583680272102356, + "learning_rate": 2.7252333333333333e-05, + "loss": 0.0276, + "step": 18249 + }, + { + "epoch": 15.41385135135135, + "grad_norm": 0.41100919246673584, + "learning_rate": 2.7252e-05, + "loss": 0.0163, + "step": 18250 + }, + { + "epoch": 15.414695945945946, + "grad_norm": 0.37214770913124084, + "learning_rate": 2.725166666666667e-05, + "loss": 0.0188, + "step": 18251 + }, + { + "epoch": 15.41554054054054, + "grad_norm": 0.2680487036705017, + "learning_rate": 2.725133333333333e-05, + "loss": 0.0156, + "step": 18252 + }, + { + "epoch": 15.416385135135135, + "grad_norm": 0.2705312967300415, + "learning_rate": 2.7251e-05, + "loss": 0.0098, + "step": 18253 + }, + { + "epoch": 15.41722972972973, + "grad_norm": 0.371869295835495, + "learning_rate": 2.725066666666667e-05, + "loss": 0.0538, + "step": 18254 + }, + { + "epoch": 15.418074324324325, + "grad_norm": 0.20775990188121796, + "learning_rate": 2.7250333333333332e-05, + "loss": 0.0127, + "step": 18255 + }, + { + "epoch": 15.41891891891892, + "grad_norm": 0.2909325063228607, + "learning_rate": 2.725e-05, + "loss": 0.0147, + "step": 18256 + }, + { + "epoch": 15.419763513513514, + "grad_norm": 0.7069351673126221, + "learning_rate": 2.7249666666666668e-05, + "loss": 0.0176, + "step": 18257 + }, + { + "epoch": 15.420608108108109, + "grad_norm": 0.3319748640060425, + "learning_rate": 2.7249333333333334e-05, + "loss": 0.0143, + "step": 18258 + }, + { + "epoch": 15.421452702702704, + "grad_norm": 0.279380202293396, + "learning_rate": 2.7249e-05, + "loss": 0.0181, + "step": 18259 + }, + { + "epoch": 15.422297297297296, + "grad_norm": 0.4535934627056122, + "learning_rate": 2.724866666666667e-05, + "loss": 0.027, + "step": 18260 + }, + { + "epoch": 15.423141891891891, + "grad_norm": 0.4201138913631439, + "learning_rate": 2.724833333333333e-05, + "loss": 0.1377, + "step": 18261 + }, + { + "epoch": 15.423986486486486, + "grad_norm": 0.6190393567085266, + "learning_rate": 2.7248e-05, + "loss": 0.0952, + "step": 18262 + }, + { + "epoch": 15.42483108108108, + "grad_norm": 0.40885263681411743, + "learning_rate": 2.724766666666667e-05, + "loss": 0.0411, + "step": 18263 + }, + { + "epoch": 15.425675675675675, + "grad_norm": 0.29977846145629883, + "learning_rate": 2.7247333333333333e-05, + "loss": 0.0235, + "step": 18264 + }, + { + "epoch": 15.42652027027027, + "grad_norm": 0.2059001475572586, + "learning_rate": 2.7247000000000002e-05, + "loss": 0.0159, + "step": 18265 + }, + { + "epoch": 15.427364864864865, + "grad_norm": 0.5702077746391296, + "learning_rate": 2.7246666666666668e-05, + "loss": 0.0161, + "step": 18266 + }, + { + "epoch": 15.42820945945946, + "grad_norm": 0.33066847920417786, + "learning_rate": 2.7246333333333334e-05, + "loss": 0.0227, + "step": 18267 + }, + { + "epoch": 15.429054054054054, + "grad_norm": 0.5580555200576782, + "learning_rate": 2.7246e-05, + "loss": 0.0202, + "step": 18268 + }, + { + "epoch": 15.42989864864865, + "grad_norm": 0.2595452666282654, + "learning_rate": 2.724566666666667e-05, + "loss": 0.0119, + "step": 18269 + }, + { + "epoch": 15.430743243243244, + "grad_norm": 0.2268189787864685, + "learning_rate": 2.724533333333333e-05, + "loss": 0.0124, + "step": 18270 + }, + { + "epoch": 15.431587837837839, + "grad_norm": 0.3197750449180603, + "learning_rate": 2.7245e-05, + "loss": 0.0224, + "step": 18271 + }, + { + "epoch": 15.432432432432432, + "grad_norm": 0.19056126475334167, + "learning_rate": 2.724466666666667e-05, + "loss": 0.0061, + "step": 18272 + }, + { + "epoch": 15.433277027027026, + "grad_norm": 0.30589109659194946, + "learning_rate": 2.7244333333333333e-05, + "loss": 0.0241, + "step": 18273 + }, + { + "epoch": 15.434121621621621, + "grad_norm": 0.4860737919807434, + "learning_rate": 2.7244000000000002e-05, + "loss": 0.0311, + "step": 18274 + }, + { + "epoch": 15.434966216216216, + "grad_norm": 0.2312534600496292, + "learning_rate": 2.7243666666666668e-05, + "loss": 0.0141, + "step": 18275 + }, + { + "epoch": 15.43581081081081, + "grad_norm": 0.2753034830093384, + "learning_rate": 2.7243333333333334e-05, + "loss": 0.0105, + "step": 18276 + }, + { + "epoch": 15.436655405405405, + "grad_norm": 0.16449005901813507, + "learning_rate": 2.7243e-05, + "loss": 0.0128, + "step": 18277 + }, + { + "epoch": 15.4375, + "grad_norm": 0.3878950774669647, + "learning_rate": 2.7242666666666666e-05, + "loss": 0.01, + "step": 18278 + }, + { + "epoch": 15.438344594594595, + "grad_norm": 0.17168429493904114, + "learning_rate": 2.7242333333333335e-05, + "loss": 0.0121, + "step": 18279 + }, + { + "epoch": 15.43918918918919, + "grad_norm": 0.33676448464393616, + "learning_rate": 2.7242e-05, + "loss": 0.0222, + "step": 18280 + }, + { + "epoch": 15.440033783783784, + "grad_norm": 0.1596338301897049, + "learning_rate": 2.7241666666666667e-05, + "loss": 0.0104, + "step": 18281 + }, + { + "epoch": 15.440878378378379, + "grad_norm": 0.20567022264003754, + "learning_rate": 2.7241333333333333e-05, + "loss": 0.0125, + "step": 18282 + }, + { + "epoch": 15.441722972972974, + "grad_norm": 0.49194100499153137, + "learning_rate": 2.7241000000000002e-05, + "loss": 0.0141, + "step": 18283 + }, + { + "epoch": 15.442567567567568, + "grad_norm": 0.18296411633491516, + "learning_rate": 2.7240666666666665e-05, + "loss": 0.0128, + "step": 18284 + }, + { + "epoch": 15.443412162162161, + "grad_norm": 0.22873634099960327, + "learning_rate": 2.7240333333333334e-05, + "loss": 0.0192, + "step": 18285 + }, + { + "epoch": 15.444256756756756, + "grad_norm": 0.4811018407344818, + "learning_rate": 2.724e-05, + "loss": 0.0183, + "step": 18286 + }, + { + "epoch": 15.44510135135135, + "grad_norm": 1.043718934059143, + "learning_rate": 2.7239666666666666e-05, + "loss": 0.0544, + "step": 18287 + }, + { + "epoch": 15.445945945945946, + "grad_norm": 0.14807848632335663, + "learning_rate": 2.7239333333333335e-05, + "loss": 0.0102, + "step": 18288 + }, + { + "epoch": 15.44679054054054, + "grad_norm": 0.25974130630493164, + "learning_rate": 2.7239e-05, + "loss": 0.0154, + "step": 18289 + }, + { + "epoch": 15.447635135135135, + "grad_norm": 0.2658577859401703, + "learning_rate": 2.7238666666666667e-05, + "loss": 0.0102, + "step": 18290 + }, + { + "epoch": 15.44847972972973, + "grad_norm": 0.5247237086296082, + "learning_rate": 2.7238333333333333e-05, + "loss": 0.0185, + "step": 18291 + }, + { + "epoch": 15.449324324324325, + "grad_norm": 0.19945402443408966, + "learning_rate": 2.7238000000000002e-05, + "loss": 0.0198, + "step": 18292 + }, + { + "epoch": 15.45016891891892, + "grad_norm": 0.13627532124519348, + "learning_rate": 2.7237666666666665e-05, + "loss": 0.007, + "step": 18293 + }, + { + "epoch": 15.451013513513514, + "grad_norm": 0.2098246067762375, + "learning_rate": 2.7237333333333334e-05, + "loss": 0.0123, + "step": 18294 + }, + { + "epoch": 15.451858108108109, + "grad_norm": 0.26012295484542847, + "learning_rate": 2.7237e-05, + "loss": 0.0217, + "step": 18295 + }, + { + "epoch": 15.452702702702704, + "grad_norm": 0.18601368367671967, + "learning_rate": 2.7236666666666666e-05, + "loss": 0.0108, + "step": 18296 + }, + { + "epoch": 15.453547297297296, + "grad_norm": 0.2417616844177246, + "learning_rate": 2.7236333333333336e-05, + "loss": 0.0168, + "step": 18297 + }, + { + "epoch": 15.454391891891891, + "grad_norm": 0.22080105543136597, + "learning_rate": 2.7236e-05, + "loss": 0.0114, + "step": 18298 + }, + { + "epoch": 15.455236486486486, + "grad_norm": 0.2053535282611847, + "learning_rate": 2.7235666666666667e-05, + "loss": 0.0112, + "step": 18299 + }, + { + "epoch": 15.45608108108108, + "grad_norm": 0.41501808166503906, + "learning_rate": 2.7235333333333333e-05, + "loss": 0.0148, + "step": 18300 + }, + { + "epoch": 15.456925675675675, + "grad_norm": 0.2595584988594055, + "learning_rate": 2.7235000000000003e-05, + "loss": 0.0114, + "step": 18301 + }, + { + "epoch": 15.45777027027027, + "grad_norm": 0.26253512501716614, + "learning_rate": 2.7234666666666665e-05, + "loss": 0.0085, + "step": 18302 + }, + { + "epoch": 15.458614864864865, + "grad_norm": 0.20268511772155762, + "learning_rate": 2.7234333333333335e-05, + "loss": 0.0108, + "step": 18303 + }, + { + "epoch": 15.45945945945946, + "grad_norm": 0.29282236099243164, + "learning_rate": 2.7234000000000004e-05, + "loss": 0.0119, + "step": 18304 + }, + { + "epoch": 15.460304054054054, + "grad_norm": 0.26683682203292847, + "learning_rate": 2.7233666666666666e-05, + "loss": 0.0186, + "step": 18305 + }, + { + "epoch": 15.46114864864865, + "grad_norm": 0.3679380714893341, + "learning_rate": 2.7233333333333336e-05, + "loss": 0.0205, + "step": 18306 + }, + { + "epoch": 15.461993243243244, + "grad_norm": 0.21486127376556396, + "learning_rate": 2.7233e-05, + "loss": 0.0099, + "step": 18307 + }, + { + "epoch": 15.462837837837839, + "grad_norm": 0.3010319471359253, + "learning_rate": 2.7232666666666668e-05, + "loss": 0.0196, + "step": 18308 + }, + { + "epoch": 15.463682432432432, + "grad_norm": 0.8516964316368103, + "learning_rate": 2.7232333333333334e-05, + "loss": 0.0248, + "step": 18309 + }, + { + "epoch": 15.464527027027026, + "grad_norm": 0.2654324769973755, + "learning_rate": 2.7232e-05, + "loss": 0.016, + "step": 18310 + }, + { + "epoch": 15.465371621621621, + "grad_norm": 0.8283497095108032, + "learning_rate": 2.7231666666666665e-05, + "loss": 0.1931, + "step": 18311 + }, + { + "epoch": 15.466216216216216, + "grad_norm": 0.5220261812210083, + "learning_rate": 2.7231333333333335e-05, + "loss": 0.0929, + "step": 18312 + }, + { + "epoch": 15.46706081081081, + "grad_norm": 0.3343309164047241, + "learning_rate": 2.7231e-05, + "loss": 0.0538, + "step": 18313 + }, + { + "epoch": 15.467905405405405, + "grad_norm": 0.2611255347728729, + "learning_rate": 2.7230666666666667e-05, + "loss": 0.0279, + "step": 18314 + }, + { + "epoch": 15.46875, + "grad_norm": 0.49671024084091187, + "learning_rate": 2.7230333333333336e-05, + "loss": 0.031, + "step": 18315 + }, + { + "epoch": 15.469594594594595, + "grad_norm": 0.19079187512397766, + "learning_rate": 2.723e-05, + "loss": 0.0202, + "step": 18316 + }, + { + "epoch": 15.47043918918919, + "grad_norm": 0.3242945671081543, + "learning_rate": 2.7229666666666668e-05, + "loss": 0.0279, + "step": 18317 + }, + { + "epoch": 15.471283783783784, + "grad_norm": 0.19917809963226318, + "learning_rate": 2.7229333333333334e-05, + "loss": 0.0237, + "step": 18318 + }, + { + "epoch": 15.472128378378379, + "grad_norm": 0.13986635208129883, + "learning_rate": 2.7229e-05, + "loss": 0.0093, + "step": 18319 + }, + { + "epoch": 15.472972972972974, + "grad_norm": 1.6597347259521484, + "learning_rate": 2.7228666666666666e-05, + "loss": 0.0176, + "step": 18320 + }, + { + "epoch": 15.473817567567568, + "grad_norm": 0.13557177782058716, + "learning_rate": 2.7228333333333335e-05, + "loss": 0.0135, + "step": 18321 + }, + { + "epoch": 15.474662162162161, + "grad_norm": 0.25107496976852417, + "learning_rate": 2.7228e-05, + "loss": 0.0157, + "step": 18322 + }, + { + "epoch": 15.475506756756756, + "grad_norm": 0.18076936900615692, + "learning_rate": 2.7227666666666667e-05, + "loss": 0.0114, + "step": 18323 + }, + { + "epoch": 15.47635135135135, + "grad_norm": 0.2163534313440323, + "learning_rate": 2.7227333333333336e-05, + "loss": 0.0221, + "step": 18324 + }, + { + "epoch": 15.477195945945946, + "grad_norm": 0.36313748359680176, + "learning_rate": 2.7227e-05, + "loss": 0.0129, + "step": 18325 + }, + { + "epoch": 15.47804054054054, + "grad_norm": 0.22429540753364563, + "learning_rate": 2.7226666666666668e-05, + "loss": 0.0121, + "step": 18326 + }, + { + "epoch": 15.478885135135135, + "grad_norm": 0.45253801345825195, + "learning_rate": 2.7226333333333334e-05, + "loss": 0.0134, + "step": 18327 + }, + { + "epoch": 15.47972972972973, + "grad_norm": 0.8378182649612427, + "learning_rate": 2.7226e-05, + "loss": 0.0124, + "step": 18328 + }, + { + "epoch": 15.480574324324325, + "grad_norm": 0.1255923956632614, + "learning_rate": 2.722566666666667e-05, + "loss": 0.0147, + "step": 18329 + }, + { + "epoch": 15.48141891891892, + "grad_norm": 0.33260810375213623, + "learning_rate": 2.7225333333333335e-05, + "loss": 0.0139, + "step": 18330 + }, + { + "epoch": 15.482263513513514, + "grad_norm": 0.2367350161075592, + "learning_rate": 2.7225e-05, + "loss": 0.0218, + "step": 18331 + }, + { + "epoch": 15.483108108108109, + "grad_norm": 0.296567440032959, + "learning_rate": 2.7224666666666667e-05, + "loss": 0.0338, + "step": 18332 + }, + { + "epoch": 15.483952702702704, + "grad_norm": 0.11538872867822647, + "learning_rate": 2.7224333333333336e-05, + "loss": 0.0075, + "step": 18333 + }, + { + "epoch": 15.484797297297296, + "grad_norm": 0.5133497714996338, + "learning_rate": 2.7224e-05, + "loss": 0.0191, + "step": 18334 + }, + { + "epoch": 15.485641891891891, + "grad_norm": 0.28440988063812256, + "learning_rate": 2.7223666666666668e-05, + "loss": 0.0195, + "step": 18335 + }, + { + "epoch": 15.486486486486486, + "grad_norm": 0.18671874701976776, + "learning_rate": 2.7223333333333334e-05, + "loss": 0.0094, + "step": 18336 + }, + { + "epoch": 15.48733108108108, + "grad_norm": 0.22658586502075195, + "learning_rate": 2.7223e-05, + "loss": 0.0093, + "step": 18337 + }, + { + "epoch": 15.488175675675675, + "grad_norm": 0.3034018874168396, + "learning_rate": 2.722266666666667e-05, + "loss": 0.0144, + "step": 18338 + }, + { + "epoch": 15.48902027027027, + "grad_norm": 0.2161872684955597, + "learning_rate": 2.7222333333333332e-05, + "loss": 0.0175, + "step": 18339 + }, + { + "epoch": 15.489864864864865, + "grad_norm": 0.33482038974761963, + "learning_rate": 2.7222e-05, + "loss": 0.0136, + "step": 18340 + }, + { + "epoch": 15.49070945945946, + "grad_norm": 0.39744889736175537, + "learning_rate": 2.7221666666666667e-05, + "loss": 0.0227, + "step": 18341 + }, + { + "epoch": 15.491554054054054, + "grad_norm": 0.24081110954284668, + "learning_rate": 2.7221333333333333e-05, + "loss": 0.019, + "step": 18342 + }, + { + "epoch": 15.49239864864865, + "grad_norm": 0.27403661608695984, + "learning_rate": 2.7221e-05, + "loss": 0.0183, + "step": 18343 + }, + { + "epoch": 15.493243243243244, + "grad_norm": 0.23477286100387573, + "learning_rate": 2.722066666666667e-05, + "loss": 0.0136, + "step": 18344 + }, + { + "epoch": 15.494087837837839, + "grad_norm": 0.20450928807258606, + "learning_rate": 2.722033333333333e-05, + "loss": 0.0104, + "step": 18345 + }, + { + "epoch": 15.494932432432432, + "grad_norm": 0.3823623061180115, + "learning_rate": 2.722e-05, + "loss": 0.0164, + "step": 18346 + }, + { + "epoch": 15.495777027027026, + "grad_norm": 0.2394438534975052, + "learning_rate": 2.721966666666667e-05, + "loss": 0.0113, + "step": 18347 + }, + { + "epoch": 15.496621621621621, + "grad_norm": 0.3741815388202667, + "learning_rate": 2.7219333333333332e-05, + "loss": 0.0277, + "step": 18348 + }, + { + "epoch": 15.497466216216216, + "grad_norm": 0.19929328560829163, + "learning_rate": 2.7219e-05, + "loss": 0.0109, + "step": 18349 + }, + { + "epoch": 15.49831081081081, + "grad_norm": 0.324985146522522, + "learning_rate": 2.7218666666666667e-05, + "loss": 0.0162, + "step": 18350 + }, + { + "epoch": 15.499155405405405, + "grad_norm": 0.24661307036876678, + "learning_rate": 2.7218333333333333e-05, + "loss": 0.0093, + "step": 18351 + }, + { + "epoch": 15.5, + "grad_norm": 0.2184455245733261, + "learning_rate": 2.7218e-05, + "loss": 0.0163, + "step": 18352 + }, + { + "epoch": 15.500844594594595, + "grad_norm": 0.2869340777397156, + "learning_rate": 2.721766666666667e-05, + "loss": 0.0154, + "step": 18353 + }, + { + "epoch": 15.50168918918919, + "grad_norm": 0.42777976393699646, + "learning_rate": 2.7217333333333335e-05, + "loss": 0.011, + "step": 18354 + }, + { + "epoch": 15.502533783783784, + "grad_norm": 0.719268262386322, + "learning_rate": 2.7217e-05, + "loss": 0.0238, + "step": 18355 + }, + { + "epoch": 15.503378378378379, + "grad_norm": 0.19346410036087036, + "learning_rate": 2.721666666666667e-05, + "loss": 0.0085, + "step": 18356 + }, + { + "epoch": 15.504222972972974, + "grad_norm": 0.3438552916049957, + "learning_rate": 2.7216333333333332e-05, + "loss": 0.0187, + "step": 18357 + }, + { + "epoch": 15.505067567567568, + "grad_norm": 0.2998259365558624, + "learning_rate": 2.7216e-05, + "loss": 0.0159, + "step": 18358 + }, + { + "epoch": 15.505912162162161, + "grad_norm": 1.0488535165786743, + "learning_rate": 2.7215666666666668e-05, + "loss": 0.0315, + "step": 18359 + }, + { + "epoch": 15.506756756756756, + "grad_norm": 0.6391551494598389, + "learning_rate": 2.7215333333333334e-05, + "loss": 0.0179, + "step": 18360 + }, + { + "epoch": 15.50760135135135, + "grad_norm": 0.6698426604270935, + "learning_rate": 2.7215e-05, + "loss": 0.1533, + "step": 18361 + }, + { + "epoch": 15.508445945945946, + "grad_norm": 0.5837449431419373, + "learning_rate": 2.721466666666667e-05, + "loss": 0.1041, + "step": 18362 + }, + { + "epoch": 15.50929054054054, + "grad_norm": 0.39170628786087036, + "learning_rate": 2.7214333333333335e-05, + "loss": 0.0438, + "step": 18363 + }, + { + "epoch": 15.510135135135135, + "grad_norm": 0.3778437077999115, + "learning_rate": 2.7214e-05, + "loss": 0.0183, + "step": 18364 + }, + { + "epoch": 15.51097972972973, + "grad_norm": 0.36065873503685, + "learning_rate": 2.721366666666667e-05, + "loss": 0.0265, + "step": 18365 + }, + { + "epoch": 15.511824324324325, + "grad_norm": 0.19485405087471008, + "learning_rate": 2.7213333333333333e-05, + "loss": 0.013, + "step": 18366 + }, + { + "epoch": 15.51266891891892, + "grad_norm": 0.2039327174425125, + "learning_rate": 2.7213000000000002e-05, + "loss": 0.0166, + "step": 18367 + }, + { + "epoch": 15.513513513513514, + "grad_norm": 0.32892441749572754, + "learning_rate": 2.7212666666666664e-05, + "loss": 0.0322, + "step": 18368 + }, + { + "epoch": 15.514358108108109, + "grad_norm": 0.2748868465423584, + "learning_rate": 2.7212333333333334e-05, + "loss": 0.0152, + "step": 18369 + }, + { + "epoch": 15.515202702702704, + "grad_norm": 0.22539052367210388, + "learning_rate": 2.7212e-05, + "loss": 0.0144, + "step": 18370 + }, + { + "epoch": 15.516047297297296, + "grad_norm": 0.30046606063842773, + "learning_rate": 2.7211666666666666e-05, + "loss": 0.0173, + "step": 18371 + }, + { + "epoch": 15.516891891891891, + "grad_norm": 0.30404800176620483, + "learning_rate": 2.7211333333333335e-05, + "loss": 0.0154, + "step": 18372 + }, + { + "epoch": 15.517736486486486, + "grad_norm": 0.411846399307251, + "learning_rate": 2.7211e-05, + "loss": 0.0192, + "step": 18373 + }, + { + "epoch": 15.51858108108108, + "grad_norm": 0.45027366280555725, + "learning_rate": 2.7210666666666667e-05, + "loss": 0.0133, + "step": 18374 + }, + { + "epoch": 15.519425675675675, + "grad_norm": 0.25917935371398926, + "learning_rate": 2.7210333333333333e-05, + "loss": 0.0128, + "step": 18375 + }, + { + "epoch": 15.52027027027027, + "grad_norm": 0.28777655959129333, + "learning_rate": 2.7210000000000002e-05, + "loss": 0.0254, + "step": 18376 + }, + { + "epoch": 15.521114864864865, + "grad_norm": 0.3552461564540863, + "learning_rate": 2.7209666666666665e-05, + "loss": 0.016, + "step": 18377 + }, + { + "epoch": 15.52195945945946, + "grad_norm": 0.4919074475765228, + "learning_rate": 2.7209333333333334e-05, + "loss": 0.0236, + "step": 18378 + }, + { + "epoch": 15.522804054054054, + "grad_norm": 0.31461694836616516, + "learning_rate": 2.7209000000000003e-05, + "loss": 0.0178, + "step": 18379 + }, + { + "epoch": 15.52364864864865, + "grad_norm": 0.3742353618144989, + "learning_rate": 2.7208666666666666e-05, + "loss": 0.0287, + "step": 18380 + }, + { + "epoch": 15.524493243243244, + "grad_norm": 0.2994607388973236, + "learning_rate": 2.7208333333333335e-05, + "loss": 0.0249, + "step": 18381 + }, + { + "epoch": 15.525337837837839, + "grad_norm": 0.17579005658626556, + "learning_rate": 2.7208e-05, + "loss": 0.0088, + "step": 18382 + }, + { + "epoch": 15.526182432432432, + "grad_norm": 0.526495099067688, + "learning_rate": 2.7207666666666667e-05, + "loss": 0.0209, + "step": 18383 + }, + { + "epoch": 15.527027027027026, + "grad_norm": 0.4650704860687256, + "learning_rate": 2.7207333333333333e-05, + "loss": 0.0264, + "step": 18384 + }, + { + "epoch": 15.527871621621621, + "grad_norm": 0.25520625710487366, + "learning_rate": 2.7207000000000002e-05, + "loss": 0.0158, + "step": 18385 + }, + { + "epoch": 15.528716216216216, + "grad_norm": 0.2179252654314041, + "learning_rate": 2.7206666666666665e-05, + "loss": 0.0095, + "step": 18386 + }, + { + "epoch": 15.52956081081081, + "grad_norm": 0.23888055980205536, + "learning_rate": 2.7206333333333334e-05, + "loss": 0.0192, + "step": 18387 + }, + { + "epoch": 15.530405405405405, + "grad_norm": 0.34659233689308167, + "learning_rate": 2.7206000000000003e-05, + "loss": 0.021, + "step": 18388 + }, + { + "epoch": 15.53125, + "grad_norm": 0.17351676523685455, + "learning_rate": 2.7205666666666666e-05, + "loss": 0.0106, + "step": 18389 + }, + { + "epoch": 15.532094594594595, + "grad_norm": 0.3869589567184448, + "learning_rate": 2.7205333333333335e-05, + "loss": 0.0164, + "step": 18390 + }, + { + "epoch": 15.53293918918919, + "grad_norm": 0.26255807280540466, + "learning_rate": 2.7205e-05, + "loss": 0.0149, + "step": 18391 + }, + { + "epoch": 15.533783783783784, + "grad_norm": 0.18108442425727844, + "learning_rate": 2.7204666666666667e-05, + "loss": 0.0084, + "step": 18392 + }, + { + "epoch": 15.534628378378379, + "grad_norm": 0.27791091799736023, + "learning_rate": 2.7204333333333333e-05, + "loss": 0.0206, + "step": 18393 + }, + { + "epoch": 15.535472972972974, + "grad_norm": 0.28631019592285156, + "learning_rate": 2.7204000000000002e-05, + "loss": 0.0124, + "step": 18394 + }, + { + "epoch": 15.536317567567568, + "grad_norm": 0.8955848217010498, + "learning_rate": 2.7203666666666665e-05, + "loss": 0.0263, + "step": 18395 + }, + { + "epoch": 15.537162162162161, + "grad_norm": 0.3129208981990814, + "learning_rate": 2.7203333333333334e-05, + "loss": 0.0159, + "step": 18396 + }, + { + "epoch": 15.538006756756756, + "grad_norm": 0.24704685807228088, + "learning_rate": 2.7203000000000004e-05, + "loss": 0.0125, + "step": 18397 + }, + { + "epoch": 15.53885135135135, + "grad_norm": 0.6058564186096191, + "learning_rate": 2.7202666666666666e-05, + "loss": 0.0196, + "step": 18398 + }, + { + "epoch": 15.539695945945946, + "grad_norm": 0.4813462793827057, + "learning_rate": 2.7202333333333336e-05, + "loss": 0.027, + "step": 18399 + }, + { + "epoch": 15.54054054054054, + "grad_norm": 0.3176117241382599, + "learning_rate": 2.7201999999999998e-05, + "loss": 0.0121, + "step": 18400 + }, + { + "epoch": 15.541385135135135, + "grad_norm": 0.38672325015068054, + "learning_rate": 2.7201666666666667e-05, + "loss": 0.0228, + "step": 18401 + }, + { + "epoch": 15.54222972972973, + "grad_norm": 0.5505911707878113, + "learning_rate": 2.7201333333333333e-05, + "loss": 0.0261, + "step": 18402 + }, + { + "epoch": 15.543074324324325, + "grad_norm": 0.7597894072532654, + "learning_rate": 2.7201e-05, + "loss": 0.0324, + "step": 18403 + }, + { + "epoch": 15.54391891891892, + "grad_norm": 0.2976904809474945, + "learning_rate": 2.720066666666667e-05, + "loss": 0.0211, + "step": 18404 + }, + { + "epoch": 15.544763513513514, + "grad_norm": 1.288330316543579, + "learning_rate": 2.7200333333333335e-05, + "loss": 0.03, + "step": 18405 + }, + { + "epoch": 15.545608108108109, + "grad_norm": 0.2518208622932434, + "learning_rate": 2.72e-05, + "loss": 0.0102, + "step": 18406 + }, + { + "epoch": 15.546452702702704, + "grad_norm": 0.47485804557800293, + "learning_rate": 2.7199666666666666e-05, + "loss": 0.0289, + "step": 18407 + }, + { + "epoch": 15.547297297297296, + "grad_norm": 0.3253189027309418, + "learning_rate": 2.7199333333333336e-05, + "loss": 0.017, + "step": 18408 + }, + { + "epoch": 15.548141891891891, + "grad_norm": 1.2124497890472412, + "learning_rate": 2.7198999999999998e-05, + "loss": 0.0351, + "step": 18409 + }, + { + "epoch": 15.548986486486486, + "grad_norm": 0.7023546695709229, + "learning_rate": 2.7198666666666668e-05, + "loss": 0.0352, + "step": 18410 + }, + { + "epoch": 15.54983108108108, + "grad_norm": 0.4847975969314575, + "learning_rate": 2.7198333333333334e-05, + "loss": 0.1468, + "step": 18411 + }, + { + "epoch": 15.550675675675675, + "grad_norm": 0.42876315116882324, + "learning_rate": 2.7198e-05, + "loss": 0.11, + "step": 18412 + }, + { + "epoch": 15.55152027027027, + "grad_norm": 0.4527089595794678, + "learning_rate": 2.719766666666667e-05, + "loss": 0.0559, + "step": 18413 + }, + { + "epoch": 15.552364864864865, + "grad_norm": 0.21133333444595337, + "learning_rate": 2.7197333333333335e-05, + "loss": 0.031, + "step": 18414 + }, + { + "epoch": 15.55320945945946, + "grad_norm": 0.2521105408668518, + "learning_rate": 2.7197e-05, + "loss": 0.028, + "step": 18415 + }, + { + "epoch": 15.554054054054054, + "grad_norm": 0.18757179379463196, + "learning_rate": 2.7196666666666667e-05, + "loss": 0.0213, + "step": 18416 + }, + { + "epoch": 15.55489864864865, + "grad_norm": 0.5505829453468323, + "learning_rate": 2.7196333333333336e-05, + "loss": 0.0154, + "step": 18417 + }, + { + "epoch": 15.555743243243244, + "grad_norm": 0.737625002861023, + "learning_rate": 2.7196e-05, + "loss": 0.0195, + "step": 18418 + }, + { + "epoch": 15.556587837837839, + "grad_norm": 0.19218730926513672, + "learning_rate": 2.7195666666666668e-05, + "loss": 0.0174, + "step": 18419 + }, + { + "epoch": 15.557432432432432, + "grad_norm": 0.18233583867549896, + "learning_rate": 2.7195333333333334e-05, + "loss": 0.0097, + "step": 18420 + }, + { + "epoch": 15.558277027027026, + "grad_norm": 0.21361364424228668, + "learning_rate": 2.7195e-05, + "loss": 0.0179, + "step": 18421 + }, + { + "epoch": 15.559121621621621, + "grad_norm": 0.2034163773059845, + "learning_rate": 2.719466666666667e-05, + "loss": 0.0199, + "step": 18422 + }, + { + "epoch": 15.559966216216216, + "grad_norm": 0.323891818523407, + "learning_rate": 2.7194333333333335e-05, + "loss": 0.0264, + "step": 18423 + }, + { + "epoch": 15.56081081081081, + "grad_norm": 0.43480098247528076, + "learning_rate": 2.7194e-05, + "loss": 0.023, + "step": 18424 + }, + { + "epoch": 15.561655405405405, + "grad_norm": 0.21111947298049927, + "learning_rate": 2.7193666666666667e-05, + "loss": 0.0151, + "step": 18425 + }, + { + "epoch": 15.5625, + "grad_norm": 0.42332708835601807, + "learning_rate": 2.7193333333333336e-05, + "loss": 0.0195, + "step": 18426 + }, + { + "epoch": 15.563344594594595, + "grad_norm": 0.19921675324440002, + "learning_rate": 2.7193e-05, + "loss": 0.0169, + "step": 18427 + }, + { + "epoch": 15.56418918918919, + "grad_norm": 0.9044076800346375, + "learning_rate": 2.7192666666666668e-05, + "loss": 0.0168, + "step": 18428 + }, + { + "epoch": 15.565033783783784, + "grad_norm": 0.21130158007144928, + "learning_rate": 2.7192333333333334e-05, + "loss": 0.0166, + "step": 18429 + }, + { + "epoch": 15.565878378378379, + "grad_norm": 0.2021932452917099, + "learning_rate": 2.7192e-05, + "loss": 0.0098, + "step": 18430 + }, + { + "epoch": 15.566722972972974, + "grad_norm": 0.25654298067092896, + "learning_rate": 2.719166666666667e-05, + "loss": 0.0113, + "step": 18431 + }, + { + "epoch": 15.567567567567568, + "grad_norm": 0.23271706700325012, + "learning_rate": 2.7191333333333332e-05, + "loss": 0.0147, + "step": 18432 + }, + { + "epoch": 15.568412162162161, + "grad_norm": 0.308329313993454, + "learning_rate": 2.7191e-05, + "loss": 0.0155, + "step": 18433 + }, + { + "epoch": 15.569256756756756, + "grad_norm": 0.19864365458488464, + "learning_rate": 2.7190666666666667e-05, + "loss": 0.0101, + "step": 18434 + }, + { + "epoch": 15.57010135135135, + "grad_norm": 0.18430761992931366, + "learning_rate": 2.7190333333333333e-05, + "loss": 0.0114, + "step": 18435 + }, + { + "epoch": 15.570945945945946, + "grad_norm": 0.27035555243492126, + "learning_rate": 2.719e-05, + "loss": 0.0137, + "step": 18436 + }, + { + "epoch": 15.57179054054054, + "grad_norm": 0.2982066869735718, + "learning_rate": 2.7189666666666668e-05, + "loss": 0.0174, + "step": 18437 + }, + { + "epoch": 15.572635135135135, + "grad_norm": 0.31930461525917053, + "learning_rate": 2.7189333333333334e-05, + "loss": 0.0128, + "step": 18438 + }, + { + "epoch": 15.57347972972973, + "grad_norm": 0.23239071667194366, + "learning_rate": 2.7189e-05, + "loss": 0.0093, + "step": 18439 + }, + { + "epoch": 15.574324324324325, + "grad_norm": 0.5040379166603088, + "learning_rate": 2.718866666666667e-05, + "loss": 0.0251, + "step": 18440 + }, + { + "epoch": 15.57516891891892, + "grad_norm": 1.4565709829330444, + "learning_rate": 2.7188333333333332e-05, + "loss": 0.049, + "step": 18441 + }, + { + "epoch": 15.576013513513514, + "grad_norm": 0.28211158514022827, + "learning_rate": 2.7188e-05, + "loss": 0.0153, + "step": 18442 + }, + { + "epoch": 15.576858108108109, + "grad_norm": 0.23948538303375244, + "learning_rate": 2.7187666666666667e-05, + "loss": 0.0133, + "step": 18443 + }, + { + "epoch": 15.577702702702704, + "grad_norm": 0.8586085438728333, + "learning_rate": 2.7187333333333333e-05, + "loss": 0.0349, + "step": 18444 + }, + { + "epoch": 15.578547297297296, + "grad_norm": 0.3598977327346802, + "learning_rate": 2.7187e-05, + "loss": 0.0163, + "step": 18445 + }, + { + "epoch": 15.579391891891891, + "grad_norm": 0.4046485424041748, + "learning_rate": 2.718666666666667e-05, + "loss": 0.0168, + "step": 18446 + }, + { + "epoch": 15.580236486486486, + "grad_norm": 0.24871641397476196, + "learning_rate": 2.7186333333333334e-05, + "loss": 0.0074, + "step": 18447 + }, + { + "epoch": 15.58108108108108, + "grad_norm": 0.3746741712093353, + "learning_rate": 2.7186e-05, + "loss": 0.0223, + "step": 18448 + }, + { + "epoch": 15.581925675675675, + "grad_norm": 0.42060476541519165, + "learning_rate": 2.718566666666667e-05, + "loss": 0.0132, + "step": 18449 + }, + { + "epoch": 15.58277027027027, + "grad_norm": 0.23647195100784302, + "learning_rate": 2.7185333333333332e-05, + "loss": 0.0144, + "step": 18450 + }, + { + "epoch": 15.583614864864865, + "grad_norm": 0.25558269023895264, + "learning_rate": 2.7185e-05, + "loss": 0.024, + "step": 18451 + }, + { + "epoch": 15.58445945945946, + "grad_norm": 0.3145117163658142, + "learning_rate": 2.7184666666666667e-05, + "loss": 0.016, + "step": 18452 + }, + { + "epoch": 15.585304054054054, + "grad_norm": 0.2943001091480255, + "learning_rate": 2.7184333333333333e-05, + "loss": 0.019, + "step": 18453 + }, + { + "epoch": 15.58614864864865, + "grad_norm": 0.5769764184951782, + "learning_rate": 2.7184000000000003e-05, + "loss": 0.0191, + "step": 18454 + }, + { + "epoch": 15.586993243243244, + "grad_norm": 0.2634657025337219, + "learning_rate": 2.718366666666667e-05, + "loss": 0.0164, + "step": 18455 + }, + { + "epoch": 15.587837837837839, + "grad_norm": 0.7216510772705078, + "learning_rate": 2.7183333333333335e-05, + "loss": 0.0132, + "step": 18456 + }, + { + "epoch": 15.588682432432432, + "grad_norm": 0.28744107484817505, + "learning_rate": 2.7183e-05, + "loss": 0.0124, + "step": 18457 + }, + { + "epoch": 15.589527027027026, + "grad_norm": 0.3285321295261383, + "learning_rate": 2.718266666666667e-05, + "loss": 0.0162, + "step": 18458 + }, + { + "epoch": 15.590371621621621, + "grad_norm": 0.2661290466785431, + "learning_rate": 2.7182333333333332e-05, + "loss": 0.0173, + "step": 18459 + }, + { + "epoch": 15.591216216216216, + "grad_norm": 0.5409630537033081, + "learning_rate": 2.7182e-05, + "loss": 0.0204, + "step": 18460 + }, + { + "epoch": 15.59206081081081, + "grad_norm": 0.6293913125991821, + "learning_rate": 2.7181666666666664e-05, + "loss": 0.1648, + "step": 18461 + }, + { + "epoch": 15.592905405405405, + "grad_norm": 0.3169848918914795, + "learning_rate": 2.7181333333333334e-05, + "loss": 0.0725, + "step": 18462 + }, + { + "epoch": 15.59375, + "grad_norm": 0.31282690167427063, + "learning_rate": 2.7181000000000003e-05, + "loss": 0.0472, + "step": 18463 + }, + { + "epoch": 15.594594594594595, + "grad_norm": 0.3442772626876831, + "learning_rate": 2.7180666666666665e-05, + "loss": 0.0343, + "step": 18464 + }, + { + "epoch": 15.59543918918919, + "grad_norm": 0.33892229199409485, + "learning_rate": 2.7180333333333335e-05, + "loss": 0.0317, + "step": 18465 + }, + { + "epoch": 15.596283783783784, + "grad_norm": 0.2320188581943512, + "learning_rate": 2.718e-05, + "loss": 0.0304, + "step": 18466 + }, + { + "epoch": 15.597128378378379, + "grad_norm": 0.2910996973514557, + "learning_rate": 2.7179666666666667e-05, + "loss": 0.0221, + "step": 18467 + }, + { + "epoch": 15.597972972972974, + "grad_norm": 0.35841014981269836, + "learning_rate": 2.7179333333333333e-05, + "loss": 0.0194, + "step": 18468 + }, + { + "epoch": 15.598817567567568, + "grad_norm": 0.2470480501651764, + "learning_rate": 2.7179000000000002e-05, + "loss": 0.0147, + "step": 18469 + }, + { + "epoch": 15.599662162162161, + "grad_norm": 0.14695125818252563, + "learning_rate": 2.7178666666666668e-05, + "loss": 0.0142, + "step": 18470 + }, + { + "epoch": 15.600506756756756, + "grad_norm": 0.33743804693222046, + "learning_rate": 2.7178333333333334e-05, + "loss": 0.0218, + "step": 18471 + }, + { + "epoch": 15.60135135135135, + "grad_norm": 0.7250595092773438, + "learning_rate": 2.7178000000000003e-05, + "loss": 0.0114, + "step": 18472 + }, + { + "epoch": 15.602195945945946, + "grad_norm": 0.18708369135856628, + "learning_rate": 2.7177666666666666e-05, + "loss": 0.0149, + "step": 18473 + }, + { + "epoch": 15.60304054054054, + "grad_norm": 0.20806559920310974, + "learning_rate": 2.7177333333333335e-05, + "loss": 0.0158, + "step": 18474 + }, + { + "epoch": 15.603885135135135, + "grad_norm": 0.3951265513896942, + "learning_rate": 2.7177e-05, + "loss": 0.0203, + "step": 18475 + }, + { + "epoch": 15.60472972972973, + "grad_norm": 0.18808308243751526, + "learning_rate": 2.7176666666666667e-05, + "loss": 0.0118, + "step": 18476 + }, + { + "epoch": 15.605574324324325, + "grad_norm": 0.2517838180065155, + "learning_rate": 2.7176333333333333e-05, + "loss": 0.021, + "step": 18477 + }, + { + "epoch": 15.60641891891892, + "grad_norm": 0.18638913333415985, + "learning_rate": 2.7176000000000002e-05, + "loss": 0.0093, + "step": 18478 + }, + { + "epoch": 15.607263513513514, + "grad_norm": 0.256876677274704, + "learning_rate": 2.7175666666666668e-05, + "loss": 0.0276, + "step": 18479 + }, + { + "epoch": 15.608108108108109, + "grad_norm": 0.2232247143983841, + "learning_rate": 2.7175333333333334e-05, + "loss": 0.015, + "step": 18480 + }, + { + "epoch": 15.608952702702704, + "grad_norm": 0.2973925769329071, + "learning_rate": 2.7175000000000003e-05, + "loss": 0.0157, + "step": 18481 + }, + { + "epoch": 15.609797297297296, + "grad_norm": 0.21110694110393524, + "learning_rate": 2.7174666666666666e-05, + "loss": 0.0132, + "step": 18482 + }, + { + "epoch": 15.610641891891891, + "grad_norm": 0.21169088780879974, + "learning_rate": 2.7174333333333335e-05, + "loss": 0.0192, + "step": 18483 + }, + { + "epoch": 15.611486486486486, + "grad_norm": 0.3002253472805023, + "learning_rate": 2.7174e-05, + "loss": 0.0565, + "step": 18484 + }, + { + "epoch": 15.61233108108108, + "grad_norm": 0.18633602559566498, + "learning_rate": 2.7173666666666667e-05, + "loss": 0.0082, + "step": 18485 + }, + { + "epoch": 15.613175675675675, + "grad_norm": 0.25016796588897705, + "learning_rate": 2.7173333333333333e-05, + "loss": 0.016, + "step": 18486 + }, + { + "epoch": 15.61402027027027, + "grad_norm": 0.4243975281715393, + "learning_rate": 2.7173000000000002e-05, + "loss": 0.0247, + "step": 18487 + }, + { + "epoch": 15.614864864864865, + "grad_norm": 0.2087721824645996, + "learning_rate": 2.7172666666666668e-05, + "loss": 0.0142, + "step": 18488 + }, + { + "epoch": 15.61570945945946, + "grad_norm": 1.0771822929382324, + "learning_rate": 2.7172333333333334e-05, + "loss": 0.0195, + "step": 18489 + }, + { + "epoch": 15.616554054054054, + "grad_norm": 0.4194962680339813, + "learning_rate": 2.7172e-05, + "loss": 0.0161, + "step": 18490 + }, + { + "epoch": 15.61739864864865, + "grad_norm": 0.11989229917526245, + "learning_rate": 2.7171666666666666e-05, + "loss": 0.0076, + "step": 18491 + }, + { + "epoch": 15.618243243243244, + "grad_norm": 0.31048914790153503, + "learning_rate": 2.7171333333333335e-05, + "loss": 0.0128, + "step": 18492 + }, + { + "epoch": 15.619087837837839, + "grad_norm": 0.20574980974197388, + "learning_rate": 2.7170999999999998e-05, + "loss": 0.0118, + "step": 18493 + }, + { + "epoch": 15.619932432432432, + "grad_norm": 0.11667665839195251, + "learning_rate": 2.7170666666666667e-05, + "loss": 0.007, + "step": 18494 + }, + { + "epoch": 15.620777027027026, + "grad_norm": 0.32408496737480164, + "learning_rate": 2.7170333333333333e-05, + "loss": 0.0192, + "step": 18495 + }, + { + "epoch": 15.621621621621621, + "grad_norm": 1.7512731552124023, + "learning_rate": 2.717e-05, + "loss": 0.0277, + "step": 18496 + }, + { + "epoch": 15.622466216216216, + "grad_norm": 0.43431127071380615, + "learning_rate": 2.716966666666667e-05, + "loss": 0.0201, + "step": 18497 + }, + { + "epoch": 15.62331081081081, + "grad_norm": 0.30795061588287354, + "learning_rate": 2.7169333333333334e-05, + "loss": 0.0165, + "step": 18498 + }, + { + "epoch": 15.624155405405405, + "grad_norm": 0.5241383910179138, + "learning_rate": 2.7169e-05, + "loss": 0.0243, + "step": 18499 + }, + { + "epoch": 15.625, + "grad_norm": 0.2798970341682434, + "learning_rate": 2.7168666666666666e-05, + "loss": 0.0178, + "step": 18500 + }, + { + "epoch": 15.625844594594595, + "grad_norm": 0.3254292607307434, + "learning_rate": 2.7168333333333336e-05, + "loss": 0.0116, + "step": 18501 + }, + { + "epoch": 15.62668918918919, + "grad_norm": 0.332354336977005, + "learning_rate": 2.7167999999999998e-05, + "loss": 0.0119, + "step": 18502 + }, + { + "epoch": 15.627533783783784, + "grad_norm": 0.3609098196029663, + "learning_rate": 2.7167666666666667e-05, + "loss": 0.0086, + "step": 18503 + }, + { + "epoch": 15.628378378378379, + "grad_norm": 0.3170948922634125, + "learning_rate": 2.7167333333333337e-05, + "loss": 0.0177, + "step": 18504 + }, + { + "epoch": 15.629222972972974, + "grad_norm": 0.36143508553504944, + "learning_rate": 2.7167e-05, + "loss": 0.013, + "step": 18505 + }, + { + "epoch": 15.630067567567568, + "grad_norm": 0.3219981789588928, + "learning_rate": 2.716666666666667e-05, + "loss": 0.0243, + "step": 18506 + }, + { + "epoch": 15.630912162162161, + "grad_norm": 0.6101748943328857, + "learning_rate": 2.7166333333333335e-05, + "loss": 0.0228, + "step": 18507 + }, + { + "epoch": 15.631756756756756, + "grad_norm": 0.5234553217887878, + "learning_rate": 2.7166e-05, + "loss": 0.0268, + "step": 18508 + }, + { + "epoch": 15.63260135135135, + "grad_norm": 0.21003931760787964, + "learning_rate": 2.7165666666666666e-05, + "loss": 0.0132, + "step": 18509 + }, + { + "epoch": 15.633445945945946, + "grad_norm": 0.7365099191665649, + "learning_rate": 2.7165333333333336e-05, + "loss": 0.015, + "step": 18510 + }, + { + "epoch": 15.63429054054054, + "grad_norm": 0.46885156631469727, + "learning_rate": 2.7164999999999998e-05, + "loss": 0.1468, + "step": 18511 + }, + { + "epoch": 15.635135135135135, + "grad_norm": 0.6175428628921509, + "learning_rate": 2.7164666666666668e-05, + "loss": 0.1123, + "step": 18512 + }, + { + "epoch": 15.63597972972973, + "grad_norm": 0.3334709107875824, + "learning_rate": 2.7164333333333337e-05, + "loss": 0.0436, + "step": 18513 + }, + { + "epoch": 15.636824324324325, + "grad_norm": 0.23931661248207092, + "learning_rate": 2.7164e-05, + "loss": 0.0279, + "step": 18514 + }, + { + "epoch": 15.63766891891892, + "grad_norm": 0.24414020776748657, + "learning_rate": 2.716366666666667e-05, + "loss": 0.0219, + "step": 18515 + }, + { + "epoch": 15.638513513513514, + "grad_norm": 0.46607154607772827, + "learning_rate": 2.7163333333333335e-05, + "loss": 0.0353, + "step": 18516 + }, + { + "epoch": 15.639358108108109, + "grad_norm": 0.1472661793231964, + "learning_rate": 2.7163e-05, + "loss": 0.0095, + "step": 18517 + }, + { + "epoch": 15.640202702702704, + "grad_norm": 0.22560542821884155, + "learning_rate": 2.7162666666666667e-05, + "loss": 0.02, + "step": 18518 + }, + { + "epoch": 15.641047297297296, + "grad_norm": 0.2801091969013214, + "learning_rate": 2.7162333333333336e-05, + "loss": 0.0282, + "step": 18519 + }, + { + "epoch": 15.641891891891891, + "grad_norm": 0.21151170134544373, + "learning_rate": 2.7162000000000002e-05, + "loss": 0.0123, + "step": 18520 + }, + { + "epoch": 15.642736486486486, + "grad_norm": 0.1473148763179779, + "learning_rate": 2.7161666666666668e-05, + "loss": 0.0099, + "step": 18521 + }, + { + "epoch": 15.64358108108108, + "grad_norm": 0.27852147817611694, + "learning_rate": 2.7161333333333334e-05, + "loss": 0.03, + "step": 18522 + }, + { + "epoch": 15.644425675675675, + "grad_norm": 0.23591287434101105, + "learning_rate": 2.7161e-05, + "loss": 0.0234, + "step": 18523 + }, + { + "epoch": 15.64527027027027, + "grad_norm": 0.21101056039333344, + "learning_rate": 2.716066666666667e-05, + "loss": 0.0159, + "step": 18524 + }, + { + "epoch": 15.646114864864865, + "grad_norm": 1.0095971822738647, + "learning_rate": 2.716033333333333e-05, + "loss": 0.0218, + "step": 18525 + }, + { + "epoch": 15.64695945945946, + "grad_norm": 0.1467965841293335, + "learning_rate": 2.716e-05, + "loss": 0.013, + "step": 18526 + }, + { + "epoch": 15.647804054054054, + "grad_norm": 0.31199875473976135, + "learning_rate": 2.7159666666666667e-05, + "loss": 0.0162, + "step": 18527 + }, + { + "epoch": 15.64864864864865, + "grad_norm": 0.3044928014278412, + "learning_rate": 2.7159333333333333e-05, + "loss": 0.0208, + "step": 18528 + }, + { + "epoch": 15.649493243243244, + "grad_norm": 0.2880210280418396, + "learning_rate": 2.7159000000000002e-05, + "loss": 0.0204, + "step": 18529 + }, + { + "epoch": 15.650337837837839, + "grad_norm": 0.20491783320903778, + "learning_rate": 2.7158666666666668e-05, + "loss": 0.0152, + "step": 18530 + }, + { + "epoch": 15.651182432432432, + "grad_norm": 0.1730993390083313, + "learning_rate": 2.7158333333333334e-05, + "loss": 0.0135, + "step": 18531 + }, + { + "epoch": 15.652027027027026, + "grad_norm": 0.1859225630760193, + "learning_rate": 2.7158e-05, + "loss": 0.0078, + "step": 18532 + }, + { + "epoch": 15.652871621621621, + "grad_norm": 0.15741027891635895, + "learning_rate": 2.715766666666667e-05, + "loss": 0.0103, + "step": 18533 + }, + { + "epoch": 15.653716216216216, + "grad_norm": 0.2803872227668762, + "learning_rate": 2.7157333333333332e-05, + "loss": 0.0154, + "step": 18534 + }, + { + "epoch": 15.65456081081081, + "grad_norm": 0.257946252822876, + "learning_rate": 2.7157e-05, + "loss": 0.0224, + "step": 18535 + }, + { + "epoch": 15.655405405405405, + "grad_norm": 0.3648444414138794, + "learning_rate": 2.7156666666666667e-05, + "loss": 0.018, + "step": 18536 + }, + { + "epoch": 15.65625, + "grad_norm": 0.2710329592227936, + "learning_rate": 2.7156333333333333e-05, + "loss": 0.0189, + "step": 18537 + }, + { + "epoch": 15.657094594594595, + "grad_norm": 0.2598293721675873, + "learning_rate": 2.7156000000000002e-05, + "loss": 0.0156, + "step": 18538 + }, + { + "epoch": 15.65793918918919, + "grad_norm": 0.3662302792072296, + "learning_rate": 2.7155666666666668e-05, + "loss": 0.0185, + "step": 18539 + }, + { + "epoch": 15.658783783783784, + "grad_norm": 0.2487301528453827, + "learning_rate": 2.7155333333333334e-05, + "loss": 0.0167, + "step": 18540 + }, + { + "epoch": 15.659628378378379, + "grad_norm": 0.2906597852706909, + "learning_rate": 2.7155e-05, + "loss": 0.0251, + "step": 18541 + }, + { + "epoch": 15.660472972972974, + "grad_norm": 0.2136707305908203, + "learning_rate": 2.715466666666667e-05, + "loss": 0.0149, + "step": 18542 + }, + { + "epoch": 15.661317567567568, + "grad_norm": 0.3459275960922241, + "learning_rate": 2.7154333333333332e-05, + "loss": 0.014, + "step": 18543 + }, + { + "epoch": 15.662162162162161, + "grad_norm": 0.32171908020973206, + "learning_rate": 2.7154e-05, + "loss": 0.0124, + "step": 18544 + }, + { + "epoch": 15.663006756756756, + "grad_norm": 0.9886259436607361, + "learning_rate": 2.7153666666666667e-05, + "loss": 0.0146, + "step": 18545 + }, + { + "epoch": 15.66385135135135, + "grad_norm": 0.17517463862895966, + "learning_rate": 2.7153333333333333e-05, + "loss": 0.0131, + "step": 18546 + }, + { + "epoch": 15.664695945945946, + "grad_norm": 0.21029581129550934, + "learning_rate": 2.7153000000000002e-05, + "loss": 0.0135, + "step": 18547 + }, + { + "epoch": 15.66554054054054, + "grad_norm": 0.21862119436264038, + "learning_rate": 2.715266666666667e-05, + "loss": 0.0112, + "step": 18548 + }, + { + "epoch": 15.666385135135135, + "grad_norm": 0.1302947700023651, + "learning_rate": 2.7152333333333334e-05, + "loss": 0.0072, + "step": 18549 + }, + { + "epoch": 15.66722972972973, + "grad_norm": 0.23352205753326416, + "learning_rate": 2.7152e-05, + "loss": 0.0131, + "step": 18550 + }, + { + "epoch": 15.668074324324325, + "grad_norm": 0.5075459480285645, + "learning_rate": 2.7151666666666666e-05, + "loss": 0.0191, + "step": 18551 + }, + { + "epoch": 15.66891891891892, + "grad_norm": 0.1326143741607666, + "learning_rate": 2.7151333333333332e-05, + "loss": 0.0104, + "step": 18552 + }, + { + "epoch": 15.669763513513514, + "grad_norm": 0.27368682622909546, + "learning_rate": 2.7151e-05, + "loss": 0.0225, + "step": 18553 + }, + { + "epoch": 15.670608108108109, + "grad_norm": 0.3044329583644867, + "learning_rate": 2.7150666666666667e-05, + "loss": 0.0192, + "step": 18554 + }, + { + "epoch": 15.671452702702704, + "grad_norm": 0.18793760240077972, + "learning_rate": 2.7150333333333333e-05, + "loss": 0.014, + "step": 18555 + }, + { + "epoch": 15.672297297297296, + "grad_norm": 0.2896483242511749, + "learning_rate": 2.7150000000000003e-05, + "loss": 0.0186, + "step": 18556 + }, + { + "epoch": 15.673141891891891, + "grad_norm": 0.291373074054718, + "learning_rate": 2.7149666666666665e-05, + "loss": 0.02, + "step": 18557 + }, + { + "epoch": 15.673986486486486, + "grad_norm": 0.29518115520477295, + "learning_rate": 2.7149333333333335e-05, + "loss": 0.0193, + "step": 18558 + }, + { + "epoch": 15.67483108108108, + "grad_norm": 0.7488310933113098, + "learning_rate": 2.7149e-05, + "loss": 0.0318, + "step": 18559 + }, + { + "epoch": 15.675675675675675, + "grad_norm": 1.15260910987854, + "learning_rate": 2.7148666666666666e-05, + "loss": 0.0356, + "step": 18560 + }, + { + "epoch": 15.67652027027027, + "grad_norm": 0.5878828167915344, + "learning_rate": 2.7148333333333332e-05, + "loss": 0.1671, + "step": 18561 + }, + { + "epoch": 15.677364864864865, + "grad_norm": 0.7740805745124817, + "learning_rate": 2.7148e-05, + "loss": 0.115, + "step": 18562 + }, + { + "epoch": 15.67820945945946, + "grad_norm": 0.4073130190372467, + "learning_rate": 2.7147666666666668e-05, + "loss": 0.0628, + "step": 18563 + }, + { + "epoch": 15.679054054054054, + "grad_norm": 0.25874873995780945, + "learning_rate": 2.7147333333333334e-05, + "loss": 0.0273, + "step": 18564 + }, + { + "epoch": 15.67989864864865, + "grad_norm": 0.2809062600135803, + "learning_rate": 2.7147000000000003e-05, + "loss": 0.0218, + "step": 18565 + }, + { + "epoch": 15.680743243243244, + "grad_norm": 0.2677428722381592, + "learning_rate": 2.7146666666666665e-05, + "loss": 0.0271, + "step": 18566 + }, + { + "epoch": 15.681587837837839, + "grad_norm": 0.5477888584136963, + "learning_rate": 2.7146333333333335e-05, + "loss": 0.0451, + "step": 18567 + }, + { + "epoch": 15.682432432432432, + "grad_norm": 0.44919711351394653, + "learning_rate": 2.7146e-05, + "loss": 0.0255, + "step": 18568 + }, + { + "epoch": 15.683277027027026, + "grad_norm": 0.22981247305870056, + "learning_rate": 2.7145666666666667e-05, + "loss": 0.0179, + "step": 18569 + }, + { + "epoch": 15.684121621621621, + "grad_norm": 0.22879353165626526, + "learning_rate": 2.7145333333333336e-05, + "loss": 0.0164, + "step": 18570 + }, + { + "epoch": 15.684966216216216, + "grad_norm": 0.19787649810314178, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0109, + "step": 18571 + }, + { + "epoch": 15.68581081081081, + "grad_norm": 0.1759536862373352, + "learning_rate": 2.7144666666666668e-05, + "loss": 0.0126, + "step": 18572 + }, + { + "epoch": 15.686655405405405, + "grad_norm": 0.21286867558956146, + "learning_rate": 2.7144333333333334e-05, + "loss": 0.016, + "step": 18573 + }, + { + "epoch": 15.6875, + "grad_norm": 0.3384721577167511, + "learning_rate": 2.7144000000000003e-05, + "loss": 0.0334, + "step": 18574 + }, + { + "epoch": 15.688344594594595, + "grad_norm": 0.2346629947423935, + "learning_rate": 2.7143666666666666e-05, + "loss": 0.0167, + "step": 18575 + }, + { + "epoch": 15.68918918918919, + "grad_norm": 0.3288659453392029, + "learning_rate": 2.7143333333333335e-05, + "loss": 0.0292, + "step": 18576 + }, + { + "epoch": 15.690033783783784, + "grad_norm": 0.1775103658437729, + "learning_rate": 2.7143e-05, + "loss": 0.0121, + "step": 18577 + }, + { + "epoch": 15.690878378378379, + "grad_norm": 0.655321478843689, + "learning_rate": 2.7142666666666667e-05, + "loss": 0.0307, + "step": 18578 + }, + { + "epoch": 15.691722972972974, + "grad_norm": 0.39282023906707764, + "learning_rate": 2.7142333333333336e-05, + "loss": 0.0213, + "step": 18579 + }, + { + "epoch": 15.692567567567568, + "grad_norm": 0.5063250660896301, + "learning_rate": 2.7142e-05, + "loss": 0.026, + "step": 18580 + }, + { + "epoch": 15.693412162162161, + "grad_norm": 0.23945622146129608, + "learning_rate": 2.7141666666666668e-05, + "loss": 0.0261, + "step": 18581 + }, + { + "epoch": 15.694256756756756, + "grad_norm": 0.20191411674022675, + "learning_rate": 2.7141333333333334e-05, + "loss": 0.0156, + "step": 18582 + }, + { + "epoch": 15.69510135135135, + "grad_norm": 0.16052131354808807, + "learning_rate": 2.7141e-05, + "loss": 0.0129, + "step": 18583 + }, + { + "epoch": 15.695945945945946, + "grad_norm": 0.0995149165391922, + "learning_rate": 2.7140666666666666e-05, + "loss": 0.0105, + "step": 18584 + }, + { + "epoch": 15.69679054054054, + "grad_norm": 0.22068163752555847, + "learning_rate": 2.7140333333333335e-05, + "loss": 0.0104, + "step": 18585 + }, + { + "epoch": 15.697635135135135, + "grad_norm": 0.13028188049793243, + "learning_rate": 2.7139999999999998e-05, + "loss": 0.0127, + "step": 18586 + }, + { + "epoch": 15.69847972972973, + "grad_norm": 0.315193772315979, + "learning_rate": 2.7139666666666667e-05, + "loss": 0.0192, + "step": 18587 + }, + { + "epoch": 15.699324324324325, + "grad_norm": 0.29484277963638306, + "learning_rate": 2.7139333333333336e-05, + "loss": 0.0114, + "step": 18588 + }, + { + "epoch": 15.70016891891892, + "grad_norm": 0.270673930644989, + "learning_rate": 2.7139e-05, + "loss": 0.0141, + "step": 18589 + }, + { + "epoch": 15.701013513513514, + "grad_norm": 0.1454775631427765, + "learning_rate": 2.7138666666666668e-05, + "loss": 0.0103, + "step": 18590 + }, + { + "epoch": 15.701858108108109, + "grad_norm": 0.4085286557674408, + "learning_rate": 2.7138333333333334e-05, + "loss": 0.0109, + "step": 18591 + }, + { + "epoch": 15.702702702702704, + "grad_norm": 0.2733286917209625, + "learning_rate": 2.7138e-05, + "loss": 0.0223, + "step": 18592 + }, + { + "epoch": 15.703547297297296, + "grad_norm": 0.27358296513557434, + "learning_rate": 2.7137666666666666e-05, + "loss": 0.0111, + "step": 18593 + }, + { + "epoch": 15.704391891891891, + "grad_norm": 0.24647177755832672, + "learning_rate": 2.7137333333333335e-05, + "loss": 0.0099, + "step": 18594 + }, + { + "epoch": 15.705236486486486, + "grad_norm": 0.21549929678440094, + "learning_rate": 2.7137e-05, + "loss": 0.0173, + "step": 18595 + }, + { + "epoch": 15.70608108108108, + "grad_norm": 0.27461645007133484, + "learning_rate": 2.7136666666666667e-05, + "loss": 0.011, + "step": 18596 + }, + { + "epoch": 15.706925675675675, + "grad_norm": 0.1540439873933792, + "learning_rate": 2.7136333333333336e-05, + "loss": 0.0082, + "step": 18597 + }, + { + "epoch": 15.70777027027027, + "grad_norm": 0.2996853291988373, + "learning_rate": 2.7136e-05, + "loss": 0.0163, + "step": 18598 + }, + { + "epoch": 15.708614864864865, + "grad_norm": 0.2647879421710968, + "learning_rate": 2.713566666666667e-05, + "loss": 0.0174, + "step": 18599 + }, + { + "epoch": 15.70945945945946, + "grad_norm": 0.29372406005859375, + "learning_rate": 2.7135333333333334e-05, + "loss": 0.0132, + "step": 18600 + }, + { + "epoch": 15.710304054054054, + "grad_norm": 1.3935765027999878, + "learning_rate": 2.7135e-05, + "loss": 0.0131, + "step": 18601 + }, + { + "epoch": 15.71114864864865, + "grad_norm": 0.16782818734645844, + "learning_rate": 2.7134666666666666e-05, + "loss": 0.0082, + "step": 18602 + }, + { + "epoch": 15.711993243243244, + "grad_norm": 0.5896353125572205, + "learning_rate": 2.7134333333333335e-05, + "loss": 0.0156, + "step": 18603 + }, + { + "epoch": 15.712837837837839, + "grad_norm": 0.27601659297943115, + "learning_rate": 2.7134e-05, + "loss": 0.0112, + "step": 18604 + }, + { + "epoch": 15.713682432432432, + "grad_norm": 0.172328382730484, + "learning_rate": 2.7133666666666667e-05, + "loss": 0.0118, + "step": 18605 + }, + { + "epoch": 15.714527027027026, + "grad_norm": 0.7520842552185059, + "learning_rate": 2.7133333333333337e-05, + "loss": 0.0216, + "step": 18606 + }, + { + "epoch": 15.715371621621621, + "grad_norm": 0.5293852090835571, + "learning_rate": 2.7133e-05, + "loss": 0.0181, + "step": 18607 + }, + { + "epoch": 15.716216216216216, + "grad_norm": 0.9344848394393921, + "learning_rate": 2.713266666666667e-05, + "loss": 0.0328, + "step": 18608 + }, + { + "epoch": 15.71706081081081, + "grad_norm": 3.4174439907073975, + "learning_rate": 2.7132333333333335e-05, + "loss": 0.0352, + "step": 18609 + }, + { + "epoch": 15.717905405405405, + "grad_norm": 0.4734511375427246, + "learning_rate": 2.7132e-05, + "loss": 0.0263, + "step": 18610 + }, + { + "epoch": 15.71875, + "grad_norm": 0.5346171259880066, + "learning_rate": 2.7131666666666666e-05, + "loss": 0.1445, + "step": 18611 + }, + { + "epoch": 15.719594594594595, + "grad_norm": 0.5339146256446838, + "learning_rate": 2.7131333333333332e-05, + "loss": 0.0945, + "step": 18612 + }, + { + "epoch": 15.72043918918919, + "grad_norm": 0.3791370093822479, + "learning_rate": 2.7131e-05, + "loss": 0.0467, + "step": 18613 + }, + { + "epoch": 15.721283783783784, + "grad_norm": 0.5084345936775208, + "learning_rate": 2.7130666666666668e-05, + "loss": 0.0403, + "step": 18614 + }, + { + "epoch": 15.722128378378379, + "grad_norm": 0.26210638880729675, + "learning_rate": 2.7130333333333334e-05, + "loss": 0.0166, + "step": 18615 + }, + { + "epoch": 15.722972972972974, + "grad_norm": 0.20969295501708984, + "learning_rate": 2.713e-05, + "loss": 0.0239, + "step": 18616 + }, + { + "epoch": 15.723817567567568, + "grad_norm": 0.2855505049228668, + "learning_rate": 2.712966666666667e-05, + "loss": 0.025, + "step": 18617 + }, + { + "epoch": 15.724662162162161, + "grad_norm": 0.4415738582611084, + "learning_rate": 2.712933333333333e-05, + "loss": 0.0238, + "step": 18618 + }, + { + "epoch": 15.725506756756756, + "grad_norm": 0.26481789350509644, + "learning_rate": 2.7129e-05, + "loss": 0.0188, + "step": 18619 + }, + { + "epoch": 15.72635135135135, + "grad_norm": 0.21263590455055237, + "learning_rate": 2.712866666666667e-05, + "loss": 0.0125, + "step": 18620 + }, + { + "epoch": 15.727195945945946, + "grad_norm": 0.4039628207683563, + "learning_rate": 2.7128333333333333e-05, + "loss": 0.0209, + "step": 18621 + }, + { + "epoch": 15.72804054054054, + "grad_norm": 0.33292651176452637, + "learning_rate": 2.7128000000000002e-05, + "loss": 0.0191, + "step": 18622 + }, + { + "epoch": 15.728885135135135, + "grad_norm": 0.24187727272510529, + "learning_rate": 2.7127666666666668e-05, + "loss": 0.0164, + "step": 18623 + }, + { + "epoch": 15.72972972972973, + "grad_norm": 0.4855310618877411, + "learning_rate": 2.7127333333333334e-05, + "loss": 0.0342, + "step": 18624 + }, + { + "epoch": 15.730574324324325, + "grad_norm": 0.21794643998146057, + "learning_rate": 2.7127e-05, + "loss": 0.0124, + "step": 18625 + }, + { + "epoch": 15.73141891891892, + "grad_norm": 0.2471878081560135, + "learning_rate": 2.712666666666667e-05, + "loss": 0.017, + "step": 18626 + }, + { + "epoch": 15.732263513513514, + "grad_norm": 0.17637306451797485, + "learning_rate": 2.712633333333333e-05, + "loss": 0.0118, + "step": 18627 + }, + { + "epoch": 15.733108108108109, + "grad_norm": 0.5215626955032349, + "learning_rate": 2.7126e-05, + "loss": 0.0175, + "step": 18628 + }, + { + "epoch": 15.733952702702704, + "grad_norm": 0.24027617275714874, + "learning_rate": 2.712566666666667e-05, + "loss": 0.0127, + "step": 18629 + }, + { + "epoch": 15.734797297297296, + "grad_norm": 0.27033576369285583, + "learning_rate": 2.7125333333333333e-05, + "loss": 0.0272, + "step": 18630 + }, + { + "epoch": 15.735641891891891, + "grad_norm": 0.17373578250408173, + "learning_rate": 2.7125000000000002e-05, + "loss": 0.0141, + "step": 18631 + }, + { + "epoch": 15.736486486486486, + "grad_norm": 0.4426910877227783, + "learning_rate": 2.7124666666666668e-05, + "loss": 0.0131, + "step": 18632 + }, + { + "epoch": 15.73733108108108, + "grad_norm": 0.2598005533218384, + "learning_rate": 2.7124333333333334e-05, + "loss": 0.0194, + "step": 18633 + }, + { + "epoch": 15.738175675675675, + "grad_norm": 0.29625776410102844, + "learning_rate": 2.7124e-05, + "loss": 0.0176, + "step": 18634 + }, + { + "epoch": 15.73902027027027, + "grad_norm": 0.20171590149402618, + "learning_rate": 2.712366666666667e-05, + "loss": 0.0135, + "step": 18635 + }, + { + "epoch": 15.739864864864865, + "grad_norm": 0.21213780343532562, + "learning_rate": 2.7123333333333332e-05, + "loss": 0.0114, + "step": 18636 + }, + { + "epoch": 15.74070945945946, + "grad_norm": 0.21102063357830048, + "learning_rate": 2.7123e-05, + "loss": 0.0163, + "step": 18637 + }, + { + "epoch": 15.741554054054054, + "grad_norm": 0.214180588722229, + "learning_rate": 2.712266666666667e-05, + "loss": 0.014, + "step": 18638 + }, + { + "epoch": 15.74239864864865, + "grad_norm": 0.27420783042907715, + "learning_rate": 2.7122333333333333e-05, + "loss": 0.0122, + "step": 18639 + }, + { + "epoch": 15.743243243243244, + "grad_norm": 0.3106343746185303, + "learning_rate": 2.7122000000000002e-05, + "loss": 0.0139, + "step": 18640 + }, + { + "epoch": 15.744087837837839, + "grad_norm": 0.16904741525650024, + "learning_rate": 2.7121666666666665e-05, + "loss": 0.0145, + "step": 18641 + }, + { + "epoch": 15.744932432432432, + "grad_norm": 0.2750586271286011, + "learning_rate": 2.7121333333333334e-05, + "loss": 0.018, + "step": 18642 + }, + { + "epoch": 15.745777027027026, + "grad_norm": 0.2763146162033081, + "learning_rate": 2.7121e-05, + "loss": 0.0122, + "step": 18643 + }, + { + "epoch": 15.746621621621621, + "grad_norm": 0.4591847062110901, + "learning_rate": 2.7120666666666666e-05, + "loss": 0.0165, + "step": 18644 + }, + { + "epoch": 15.747466216216216, + "grad_norm": 0.9607972502708435, + "learning_rate": 2.7120333333333335e-05, + "loss": 0.0243, + "step": 18645 + }, + { + "epoch": 15.74831081081081, + "grad_norm": 0.7416853308677673, + "learning_rate": 2.712e-05, + "loss": 0.0069, + "step": 18646 + }, + { + "epoch": 15.749155405405405, + "grad_norm": 0.11929646134376526, + "learning_rate": 2.7119666666666667e-05, + "loss": 0.0098, + "step": 18647 + }, + { + "epoch": 15.75, + "grad_norm": 0.46932023763656616, + "learning_rate": 2.7119333333333333e-05, + "loss": 0.0181, + "step": 18648 + }, + { + "epoch": 15.750844594594595, + "grad_norm": 0.25556156039237976, + "learning_rate": 2.7119000000000002e-05, + "loss": 0.0099, + "step": 18649 + }, + { + "epoch": 15.75168918918919, + "grad_norm": 0.29844537377357483, + "learning_rate": 2.7118666666666665e-05, + "loss": 0.0139, + "step": 18650 + }, + { + "epoch": 15.752533783783784, + "grad_norm": 0.19022372364997864, + "learning_rate": 2.7118333333333334e-05, + "loss": 0.014, + "step": 18651 + }, + { + "epoch": 15.753378378378379, + "grad_norm": 0.22540390491485596, + "learning_rate": 2.7118e-05, + "loss": 0.0103, + "step": 18652 + }, + { + "epoch": 15.754222972972974, + "grad_norm": 0.28912532329559326, + "learning_rate": 2.7117666666666666e-05, + "loss": 0.0153, + "step": 18653 + }, + { + "epoch": 15.755067567567568, + "grad_norm": 0.30728521943092346, + "learning_rate": 2.7117333333333335e-05, + "loss": 0.0176, + "step": 18654 + }, + { + "epoch": 15.755912162162161, + "grad_norm": 0.24886958301067352, + "learning_rate": 2.7117e-05, + "loss": 0.0117, + "step": 18655 + }, + { + "epoch": 15.756756756756756, + "grad_norm": 0.18022754788398743, + "learning_rate": 2.7116666666666667e-05, + "loss": 0.0138, + "step": 18656 + }, + { + "epoch": 15.75760135135135, + "grad_norm": 0.36925244331359863, + "learning_rate": 2.7116333333333333e-05, + "loss": 0.0197, + "step": 18657 + }, + { + "epoch": 15.758445945945946, + "grad_norm": 0.4658627510070801, + "learning_rate": 2.7116000000000003e-05, + "loss": 0.0281, + "step": 18658 + }, + { + "epoch": 15.75929054054054, + "grad_norm": 0.6006466746330261, + "learning_rate": 2.7115666666666665e-05, + "loss": 0.0325, + "step": 18659 + }, + { + "epoch": 15.760135135135135, + "grad_norm": 0.33421972393989563, + "learning_rate": 2.7115333333333334e-05, + "loss": 0.019, + "step": 18660 + }, + { + "epoch": 15.76097972972973, + "grad_norm": 0.4413016736507416, + "learning_rate": 2.7115e-05, + "loss": 0.1356, + "step": 18661 + }, + { + "epoch": 15.761824324324325, + "grad_norm": 0.5056946277618408, + "learning_rate": 2.7114666666666666e-05, + "loss": 0.1037, + "step": 18662 + }, + { + "epoch": 15.76266891891892, + "grad_norm": 0.23522379994392395, + "learning_rate": 2.7114333333333336e-05, + "loss": 0.0342, + "step": 18663 + }, + { + "epoch": 15.763513513513514, + "grad_norm": 0.25994837284088135, + "learning_rate": 2.7114e-05, + "loss": 0.0322, + "step": 18664 + }, + { + "epoch": 15.764358108108109, + "grad_norm": 0.3501163423061371, + "learning_rate": 2.7113666666666668e-05, + "loss": 0.0275, + "step": 18665 + }, + { + "epoch": 15.765202702702704, + "grad_norm": 0.22949761152267456, + "learning_rate": 2.7113333333333333e-05, + "loss": 0.0195, + "step": 18666 + }, + { + "epoch": 15.766047297297296, + "grad_norm": 0.20538318157196045, + "learning_rate": 2.7113000000000003e-05, + "loss": 0.0224, + "step": 18667 + }, + { + "epoch": 15.766891891891891, + "grad_norm": 1.3401113748550415, + "learning_rate": 2.7112666666666665e-05, + "loss": 0.0188, + "step": 18668 + }, + { + "epoch": 15.767736486486486, + "grad_norm": 0.3512064218521118, + "learning_rate": 2.7112333333333335e-05, + "loss": 0.0126, + "step": 18669 + }, + { + "epoch": 15.76858108108108, + "grad_norm": 0.18248651921749115, + "learning_rate": 2.7112000000000004e-05, + "loss": 0.0151, + "step": 18670 + }, + { + "epoch": 15.769425675675675, + "grad_norm": 0.14357687532901764, + "learning_rate": 2.7111666666666667e-05, + "loss": 0.0132, + "step": 18671 + }, + { + "epoch": 15.77027027027027, + "grad_norm": 0.17029854655265808, + "learning_rate": 2.7111333333333336e-05, + "loss": 0.012, + "step": 18672 + }, + { + "epoch": 15.771114864864865, + "grad_norm": 0.23034170269966125, + "learning_rate": 2.7111e-05, + "loss": 0.0216, + "step": 18673 + }, + { + "epoch": 15.77195945945946, + "grad_norm": 0.4725148677825928, + "learning_rate": 2.7110666666666668e-05, + "loss": 0.0152, + "step": 18674 + }, + { + "epoch": 15.772804054054054, + "grad_norm": 0.3600834310054779, + "learning_rate": 2.7110333333333334e-05, + "loss": 0.0294, + "step": 18675 + }, + { + "epoch": 15.77364864864865, + "grad_norm": 0.17081834375858307, + "learning_rate": 2.711e-05, + "loss": 0.0105, + "step": 18676 + }, + { + "epoch": 15.774493243243244, + "grad_norm": 0.18434084951877594, + "learning_rate": 2.7109666666666666e-05, + "loss": 0.0198, + "step": 18677 + }, + { + "epoch": 15.775337837837839, + "grad_norm": 0.20651914179325104, + "learning_rate": 2.7109333333333335e-05, + "loss": 0.0176, + "step": 18678 + }, + { + "epoch": 15.776182432432432, + "grad_norm": 0.23305897414684296, + "learning_rate": 2.7109e-05, + "loss": 0.0177, + "step": 18679 + }, + { + "epoch": 15.777027027027026, + "grad_norm": 0.580467939376831, + "learning_rate": 2.7108666666666667e-05, + "loss": 0.0578, + "step": 18680 + }, + { + "epoch": 15.777871621621621, + "grad_norm": 0.2632788419723511, + "learning_rate": 2.7108333333333336e-05, + "loss": 0.0153, + "step": 18681 + }, + { + "epoch": 15.778716216216216, + "grad_norm": 0.3093242645263672, + "learning_rate": 2.7108e-05, + "loss": 0.0187, + "step": 18682 + }, + { + "epoch": 15.77956081081081, + "grad_norm": 0.3007364571094513, + "learning_rate": 2.7107666666666668e-05, + "loss": 0.01, + "step": 18683 + }, + { + "epoch": 15.780405405405405, + "grad_norm": 0.29001691937446594, + "learning_rate": 2.7107333333333334e-05, + "loss": 0.0151, + "step": 18684 + }, + { + "epoch": 15.78125, + "grad_norm": 0.33839496970176697, + "learning_rate": 2.7107e-05, + "loss": 0.0104, + "step": 18685 + }, + { + "epoch": 15.782094594594595, + "grad_norm": 0.27081719040870667, + "learning_rate": 2.7106666666666666e-05, + "loss": 0.0164, + "step": 18686 + }, + { + "epoch": 15.78293918918919, + "grad_norm": 0.1676793098449707, + "learning_rate": 2.7106333333333335e-05, + "loss": 0.0112, + "step": 18687 + }, + { + "epoch": 15.783783783783784, + "grad_norm": 0.17746050655841827, + "learning_rate": 2.7106e-05, + "loss": 0.0105, + "step": 18688 + }, + { + "epoch": 15.784628378378379, + "grad_norm": 0.28407108783721924, + "learning_rate": 2.7105666666666667e-05, + "loss": 0.0175, + "step": 18689 + }, + { + "epoch": 15.785472972972974, + "grad_norm": 0.20660458505153656, + "learning_rate": 2.7105333333333336e-05, + "loss": 0.0143, + "step": 18690 + }, + { + "epoch": 15.786317567567568, + "grad_norm": 0.5446896553039551, + "learning_rate": 2.7105e-05, + "loss": 0.0156, + "step": 18691 + }, + { + "epoch": 15.787162162162161, + "grad_norm": 0.1969040483236313, + "learning_rate": 2.7104666666666668e-05, + "loss": 0.0158, + "step": 18692 + }, + { + "epoch": 15.788006756756756, + "grad_norm": 0.29429882764816284, + "learning_rate": 2.7104333333333334e-05, + "loss": 0.0132, + "step": 18693 + }, + { + "epoch": 15.78885135135135, + "grad_norm": 0.21226724982261658, + "learning_rate": 2.7104e-05, + "loss": 0.0096, + "step": 18694 + }, + { + "epoch": 15.789695945945946, + "grad_norm": 0.15921856462955475, + "learning_rate": 2.710366666666667e-05, + "loss": 0.011, + "step": 18695 + }, + { + "epoch": 15.79054054054054, + "grad_norm": 0.6061167120933533, + "learning_rate": 2.7103333333333335e-05, + "loss": 0.0284, + "step": 18696 + }, + { + "epoch": 15.791385135135135, + "grad_norm": 0.16545146703720093, + "learning_rate": 2.7103e-05, + "loss": 0.0137, + "step": 18697 + }, + { + "epoch": 15.79222972972973, + "grad_norm": 0.24558842182159424, + "learning_rate": 2.7102666666666667e-05, + "loss": 0.0156, + "step": 18698 + }, + { + "epoch": 15.793074324324325, + "grad_norm": 0.1871388852596283, + "learning_rate": 2.7102333333333336e-05, + "loss": 0.0097, + "step": 18699 + }, + { + "epoch": 15.79391891891892, + "grad_norm": 0.1840459406375885, + "learning_rate": 2.7102e-05, + "loss": 0.0084, + "step": 18700 + }, + { + "epoch": 15.794763513513514, + "grad_norm": 0.348924458026886, + "learning_rate": 2.710166666666667e-05, + "loss": 0.0278, + "step": 18701 + }, + { + "epoch": 15.795608108108109, + "grad_norm": 0.24565649032592773, + "learning_rate": 2.710133333333333e-05, + "loss": 0.0129, + "step": 18702 + }, + { + "epoch": 15.796452702702704, + "grad_norm": 1.0431499481201172, + "learning_rate": 2.7101e-05, + "loss": 0.0163, + "step": 18703 + }, + { + "epoch": 15.797297297297296, + "grad_norm": 0.37507569789886475, + "learning_rate": 2.710066666666667e-05, + "loss": 0.0103, + "step": 18704 + }, + { + "epoch": 15.798141891891891, + "grad_norm": 0.41606053709983826, + "learning_rate": 2.7100333333333332e-05, + "loss": 0.0138, + "step": 18705 + }, + { + "epoch": 15.798986486486486, + "grad_norm": 0.28012412786483765, + "learning_rate": 2.71e-05, + "loss": 0.0092, + "step": 18706 + }, + { + "epoch": 15.79983108108108, + "grad_norm": 0.3164566159248352, + "learning_rate": 2.7099666666666667e-05, + "loss": 0.013, + "step": 18707 + }, + { + "epoch": 15.800675675675675, + "grad_norm": 0.5695411562919617, + "learning_rate": 2.7099333333333333e-05, + "loss": 0.012, + "step": 18708 + }, + { + "epoch": 15.80152027027027, + "grad_norm": 0.7085590958595276, + "learning_rate": 2.7099e-05, + "loss": 0.0222, + "step": 18709 + }, + { + "epoch": 15.802364864864865, + "grad_norm": 1.9140058755874634, + "learning_rate": 2.709866666666667e-05, + "loss": 0.0452, + "step": 18710 + }, + { + "epoch": 15.80320945945946, + "grad_norm": 0.6922261118888855, + "learning_rate": 2.709833333333333e-05, + "loss": 0.1532, + "step": 18711 + }, + { + "epoch": 15.804054054054054, + "grad_norm": 0.4634554386138916, + "learning_rate": 2.7098e-05, + "loss": 0.0819, + "step": 18712 + }, + { + "epoch": 15.80489864864865, + "grad_norm": 0.39096057415008545, + "learning_rate": 2.709766666666667e-05, + "loss": 0.0422, + "step": 18713 + }, + { + "epoch": 15.805743243243244, + "grad_norm": 0.33308079838752747, + "learning_rate": 2.7097333333333332e-05, + "loss": 0.054, + "step": 18714 + }, + { + "epoch": 15.806587837837839, + "grad_norm": 0.5493724942207336, + "learning_rate": 2.7097e-05, + "loss": 0.0255, + "step": 18715 + }, + { + "epoch": 15.807432432432432, + "grad_norm": 0.21385660767555237, + "learning_rate": 2.7096666666666668e-05, + "loss": 0.0226, + "step": 18716 + }, + { + "epoch": 15.808277027027026, + "grad_norm": 0.46845489740371704, + "learning_rate": 2.7096333333333333e-05, + "loss": 0.024, + "step": 18717 + }, + { + "epoch": 15.809121621621621, + "grad_norm": 0.34053224325180054, + "learning_rate": 2.7096e-05, + "loss": 0.0208, + "step": 18718 + }, + { + "epoch": 15.809966216216216, + "grad_norm": 0.3668852150440216, + "learning_rate": 2.709566666666667e-05, + "loss": 0.0257, + "step": 18719 + }, + { + "epoch": 15.81081081081081, + "grad_norm": 0.13919681310653687, + "learning_rate": 2.7095333333333335e-05, + "loss": 0.0091, + "step": 18720 + }, + { + "epoch": 15.811655405405405, + "grad_norm": 0.3000916838645935, + "learning_rate": 2.7095e-05, + "loss": 0.0174, + "step": 18721 + }, + { + "epoch": 15.8125, + "grad_norm": 0.20283928513526917, + "learning_rate": 2.709466666666667e-05, + "loss": 0.0168, + "step": 18722 + }, + { + "epoch": 15.813344594594595, + "grad_norm": 0.17297010123729706, + "learning_rate": 2.7094333333333332e-05, + "loss": 0.0105, + "step": 18723 + }, + { + "epoch": 15.81418918918919, + "grad_norm": 0.19875282049179077, + "learning_rate": 2.7094000000000002e-05, + "loss": 0.0096, + "step": 18724 + }, + { + "epoch": 15.815033783783784, + "grad_norm": 0.27260175347328186, + "learning_rate": 2.7093666666666668e-05, + "loss": 0.0155, + "step": 18725 + }, + { + "epoch": 15.815878378378379, + "grad_norm": 0.25195837020874023, + "learning_rate": 2.7093333333333334e-05, + "loss": 0.0213, + "step": 18726 + }, + { + "epoch": 15.816722972972974, + "grad_norm": 0.37729185819625854, + "learning_rate": 2.7093e-05, + "loss": 0.0211, + "step": 18727 + }, + { + "epoch": 15.817567567567568, + "grad_norm": 0.22893033921718597, + "learning_rate": 2.709266666666667e-05, + "loss": 0.0143, + "step": 18728 + }, + { + "epoch": 15.818412162162161, + "grad_norm": 0.2192000299692154, + "learning_rate": 2.7092333333333335e-05, + "loss": 0.0141, + "step": 18729 + }, + { + "epoch": 15.819256756756756, + "grad_norm": 0.3119589388370514, + "learning_rate": 2.7092e-05, + "loss": 0.018, + "step": 18730 + }, + { + "epoch": 15.82010135135135, + "grad_norm": 0.21879792213439941, + "learning_rate": 2.7091666666666667e-05, + "loss": 0.0159, + "step": 18731 + }, + { + "epoch": 15.820945945945946, + "grad_norm": 0.5260905027389526, + "learning_rate": 2.7091333333333333e-05, + "loss": 0.0241, + "step": 18732 + }, + { + "epoch": 15.82179054054054, + "grad_norm": 0.2474813014268875, + "learning_rate": 2.7091000000000002e-05, + "loss": 0.0137, + "step": 18733 + }, + { + "epoch": 15.822635135135135, + "grad_norm": 0.16739992797374725, + "learning_rate": 2.7090666666666665e-05, + "loss": 0.0114, + "step": 18734 + }, + { + "epoch": 15.82347972972973, + "grad_norm": 0.1696195900440216, + "learning_rate": 2.7090333333333334e-05, + "loss": 0.0137, + "step": 18735 + }, + { + "epoch": 15.824324324324325, + "grad_norm": 0.7392054796218872, + "learning_rate": 2.709e-05, + "loss": 0.026, + "step": 18736 + }, + { + "epoch": 15.82516891891892, + "grad_norm": 0.15343378484249115, + "learning_rate": 2.7089666666666666e-05, + "loss": 0.0152, + "step": 18737 + }, + { + "epoch": 15.826013513513514, + "grad_norm": 0.30221325159072876, + "learning_rate": 2.7089333333333335e-05, + "loss": 0.0143, + "step": 18738 + }, + { + "epoch": 15.826858108108109, + "grad_norm": 0.23585093021392822, + "learning_rate": 2.7089e-05, + "loss": 0.0128, + "step": 18739 + }, + { + "epoch": 15.827702702702704, + "grad_norm": 0.32336103916168213, + "learning_rate": 2.7088666666666667e-05, + "loss": 0.0232, + "step": 18740 + }, + { + "epoch": 15.828547297297296, + "grad_norm": 0.1467314511537552, + "learning_rate": 2.7088333333333333e-05, + "loss": 0.0118, + "step": 18741 + }, + { + "epoch": 15.829391891891891, + "grad_norm": 0.578624963760376, + "learning_rate": 2.7088000000000002e-05, + "loss": 0.0292, + "step": 18742 + }, + { + "epoch": 15.830236486486486, + "grad_norm": 0.16609948873519897, + "learning_rate": 2.7087666666666665e-05, + "loss": 0.0126, + "step": 18743 + }, + { + "epoch": 15.83108108108108, + "grad_norm": 0.34367144107818604, + "learning_rate": 2.7087333333333334e-05, + "loss": 0.0167, + "step": 18744 + }, + { + "epoch": 15.831925675675675, + "grad_norm": 0.15720365941524506, + "learning_rate": 2.7087000000000003e-05, + "loss": 0.0167, + "step": 18745 + }, + { + "epoch": 15.83277027027027, + "grad_norm": 0.20935609936714172, + "learning_rate": 2.7086666666666666e-05, + "loss": 0.0153, + "step": 18746 + }, + { + "epoch": 15.833614864864865, + "grad_norm": 0.695347011089325, + "learning_rate": 2.7086333333333335e-05, + "loss": 0.0121, + "step": 18747 + }, + { + "epoch": 15.83445945945946, + "grad_norm": 1.3406293392181396, + "learning_rate": 2.7086e-05, + "loss": 0.0162, + "step": 18748 + }, + { + "epoch": 15.835304054054054, + "grad_norm": 0.3939487040042877, + "learning_rate": 2.7085666666666667e-05, + "loss": 0.0158, + "step": 18749 + }, + { + "epoch": 15.83614864864865, + "grad_norm": 0.39404284954071045, + "learning_rate": 2.7085333333333333e-05, + "loss": 0.0151, + "step": 18750 + }, + { + "epoch": 15.836993243243244, + "grad_norm": 0.5256537199020386, + "learning_rate": 2.7085000000000002e-05, + "loss": 0.0196, + "step": 18751 + }, + { + "epoch": 15.837837837837839, + "grad_norm": 0.4569038152694702, + "learning_rate": 2.7084666666666665e-05, + "loss": 0.0156, + "step": 18752 + }, + { + "epoch": 15.838682432432432, + "grad_norm": 0.368308424949646, + "learning_rate": 2.7084333333333334e-05, + "loss": 0.0203, + "step": 18753 + }, + { + "epoch": 15.839527027027026, + "grad_norm": 0.4142323434352875, + "learning_rate": 2.7084000000000004e-05, + "loss": 0.0277, + "step": 18754 + }, + { + "epoch": 15.840371621621621, + "grad_norm": 0.3151550889015198, + "learning_rate": 2.7083666666666666e-05, + "loss": 0.014, + "step": 18755 + }, + { + "epoch": 15.841216216216216, + "grad_norm": 0.275553435087204, + "learning_rate": 2.7083333333333335e-05, + "loss": 0.0176, + "step": 18756 + }, + { + "epoch": 15.84206081081081, + "grad_norm": 1.3722316026687622, + "learning_rate": 2.7083e-05, + "loss": 0.0156, + "step": 18757 + }, + { + "epoch": 15.842905405405405, + "grad_norm": 0.16342507302761078, + "learning_rate": 2.7082666666666667e-05, + "loss": 0.0077, + "step": 18758 + }, + { + "epoch": 15.84375, + "grad_norm": 0.24285389482975006, + "learning_rate": 2.7082333333333333e-05, + "loss": 0.0101, + "step": 18759 + }, + { + "epoch": 15.844594594594595, + "grad_norm": 0.9218359589576721, + "learning_rate": 2.7082000000000003e-05, + "loss": 0.0464, + "step": 18760 + }, + { + "epoch": 15.84543918918919, + "grad_norm": 0.49483001232147217, + "learning_rate": 2.7081666666666665e-05, + "loss": 0.1549, + "step": 18761 + }, + { + "epoch": 15.846283783783784, + "grad_norm": 0.8384677767753601, + "learning_rate": 2.7081333333333334e-05, + "loss": 0.1333, + "step": 18762 + }, + { + "epoch": 15.847128378378379, + "grad_norm": 0.3385794758796692, + "learning_rate": 2.7081e-05, + "loss": 0.0563, + "step": 18763 + }, + { + "epoch": 15.847972972972974, + "grad_norm": 0.4150884449481964, + "learning_rate": 2.7080666666666666e-05, + "loss": 0.0284, + "step": 18764 + }, + { + "epoch": 15.848817567567568, + "grad_norm": 0.28614628314971924, + "learning_rate": 2.7080333333333336e-05, + "loss": 0.0281, + "step": 18765 + }, + { + "epoch": 15.849662162162161, + "grad_norm": 0.20743659138679504, + "learning_rate": 2.7079999999999998e-05, + "loss": 0.0109, + "step": 18766 + }, + { + "epoch": 15.850506756756756, + "grad_norm": 0.37036582827568054, + "learning_rate": 2.7079666666666668e-05, + "loss": 0.014, + "step": 18767 + }, + { + "epoch": 15.85135135135135, + "grad_norm": 0.30271831154823303, + "learning_rate": 2.7079333333333333e-05, + "loss": 0.0169, + "step": 18768 + }, + { + "epoch": 15.852195945945946, + "grad_norm": 0.3071022927761078, + "learning_rate": 2.7079e-05, + "loss": 0.0277, + "step": 18769 + }, + { + "epoch": 15.85304054054054, + "grad_norm": 0.3329978287220001, + "learning_rate": 2.707866666666667e-05, + "loss": 0.0148, + "step": 18770 + }, + { + "epoch": 15.853885135135135, + "grad_norm": 0.2239314764738083, + "learning_rate": 2.7078333333333335e-05, + "loss": 0.0136, + "step": 18771 + }, + { + "epoch": 15.85472972972973, + "grad_norm": 0.314155250787735, + "learning_rate": 2.7078e-05, + "loss": 0.0188, + "step": 18772 + }, + { + "epoch": 15.855574324324325, + "grad_norm": 0.300162672996521, + "learning_rate": 2.7077666666666667e-05, + "loss": 0.0147, + "step": 18773 + }, + { + "epoch": 15.85641891891892, + "grad_norm": 0.22201554477214813, + "learning_rate": 2.7077333333333336e-05, + "loss": 0.0209, + "step": 18774 + }, + { + "epoch": 15.857263513513514, + "grad_norm": 0.24284541606903076, + "learning_rate": 2.7077e-05, + "loss": 0.0148, + "step": 18775 + }, + { + "epoch": 15.858108108108109, + "grad_norm": 0.2492896169424057, + "learning_rate": 2.7076666666666668e-05, + "loss": 0.0144, + "step": 18776 + }, + { + "epoch": 15.858952702702704, + "grad_norm": 0.5438525080680847, + "learning_rate": 2.7076333333333334e-05, + "loss": 0.0311, + "step": 18777 + }, + { + "epoch": 15.859797297297296, + "grad_norm": 0.12860828638076782, + "learning_rate": 2.7076e-05, + "loss": 0.0075, + "step": 18778 + }, + { + "epoch": 15.860641891891891, + "grad_norm": 0.17737817764282227, + "learning_rate": 2.707566666666667e-05, + "loss": 0.0165, + "step": 18779 + }, + { + "epoch": 15.861486486486486, + "grad_norm": 0.14427590370178223, + "learning_rate": 2.7075333333333335e-05, + "loss": 0.0118, + "step": 18780 + }, + { + "epoch": 15.86233108108108, + "grad_norm": 0.12606418132781982, + "learning_rate": 2.7075e-05, + "loss": 0.0089, + "step": 18781 + }, + { + "epoch": 15.863175675675675, + "grad_norm": 0.16112020611763, + "learning_rate": 2.7074666666666667e-05, + "loss": 0.0108, + "step": 18782 + }, + { + "epoch": 15.86402027027027, + "grad_norm": 0.2942800521850586, + "learning_rate": 2.7074333333333336e-05, + "loss": 0.015, + "step": 18783 + }, + { + "epoch": 15.864864864864865, + "grad_norm": 0.14953093230724335, + "learning_rate": 2.7074e-05, + "loss": 0.0106, + "step": 18784 + }, + { + "epoch": 15.86570945945946, + "grad_norm": 0.6907792091369629, + "learning_rate": 2.7073666666666668e-05, + "loss": 0.0204, + "step": 18785 + }, + { + "epoch": 15.866554054054054, + "grad_norm": 0.5582908391952515, + "learning_rate": 2.7073333333333334e-05, + "loss": 0.027, + "step": 18786 + }, + { + "epoch": 15.86739864864865, + "grad_norm": 0.23508158326148987, + "learning_rate": 2.7073e-05, + "loss": 0.0149, + "step": 18787 + }, + { + "epoch": 15.868243243243244, + "grad_norm": 0.18185435235500336, + "learning_rate": 2.707266666666667e-05, + "loss": 0.0084, + "step": 18788 + }, + { + "epoch": 15.869087837837839, + "grad_norm": 0.40880030393600464, + "learning_rate": 2.7072333333333335e-05, + "loss": 0.0159, + "step": 18789 + }, + { + "epoch": 15.869932432432432, + "grad_norm": 0.23396874964237213, + "learning_rate": 2.7072e-05, + "loss": 0.0202, + "step": 18790 + }, + { + "epoch": 15.870777027027026, + "grad_norm": 0.213005930185318, + "learning_rate": 2.7071666666666667e-05, + "loss": 0.0207, + "step": 18791 + }, + { + "epoch": 15.871621621621621, + "grad_norm": 0.3195456266403198, + "learning_rate": 2.7071333333333333e-05, + "loss": 0.0128, + "step": 18792 + }, + { + "epoch": 15.872466216216216, + "grad_norm": 0.330335795879364, + "learning_rate": 2.7071e-05, + "loss": 0.0216, + "step": 18793 + }, + { + "epoch": 15.87331081081081, + "grad_norm": 0.36698487401008606, + "learning_rate": 2.7070666666666668e-05, + "loss": 0.0623, + "step": 18794 + }, + { + "epoch": 15.874155405405405, + "grad_norm": 0.27115464210510254, + "learning_rate": 2.7070333333333334e-05, + "loss": 0.0266, + "step": 18795 + }, + { + "epoch": 15.875, + "grad_norm": 0.17226135730743408, + "learning_rate": 2.707e-05, + "loss": 0.0116, + "step": 18796 + }, + { + "epoch": 15.875844594594595, + "grad_norm": 0.3834768831729889, + "learning_rate": 2.706966666666667e-05, + "loss": 0.0131, + "step": 18797 + }, + { + "epoch": 15.87668918918919, + "grad_norm": 0.28075888752937317, + "learning_rate": 2.7069333333333332e-05, + "loss": 0.0194, + "step": 18798 + }, + { + "epoch": 15.877533783783784, + "grad_norm": 0.37093544006347656, + "learning_rate": 2.7069e-05, + "loss": 0.0142, + "step": 18799 + }, + { + "epoch": 15.878378378378379, + "grad_norm": 0.44840097427368164, + "learning_rate": 2.7068666666666667e-05, + "loss": 0.0236, + "step": 18800 + }, + { + "epoch": 15.879222972972974, + "grad_norm": 0.22083839774131775, + "learning_rate": 2.7068333333333333e-05, + "loss": 0.0132, + "step": 18801 + }, + { + "epoch": 15.880067567567568, + "grad_norm": 0.2727466821670532, + "learning_rate": 2.7068e-05, + "loss": 0.0126, + "step": 18802 + }, + { + "epoch": 15.880912162162161, + "grad_norm": 0.16644899547100067, + "learning_rate": 2.706766666666667e-05, + "loss": 0.0144, + "step": 18803 + }, + { + "epoch": 15.881756756756756, + "grad_norm": 0.46181347966194153, + "learning_rate": 2.7067333333333334e-05, + "loss": 0.0236, + "step": 18804 + }, + { + "epoch": 15.88260135135135, + "grad_norm": 0.40460190176963806, + "learning_rate": 2.7067e-05, + "loss": 0.0391, + "step": 18805 + }, + { + "epoch": 15.883445945945946, + "grad_norm": 0.22998186945915222, + "learning_rate": 2.706666666666667e-05, + "loss": 0.0131, + "step": 18806 + }, + { + "epoch": 15.88429054054054, + "grad_norm": 0.5097019672393799, + "learning_rate": 2.7066333333333332e-05, + "loss": 0.0168, + "step": 18807 + }, + { + "epoch": 15.885135135135135, + "grad_norm": 0.1833622306585312, + "learning_rate": 2.7066e-05, + "loss": 0.0144, + "step": 18808 + }, + { + "epoch": 15.88597972972973, + "grad_norm": 0.2673334777355194, + "learning_rate": 2.7065666666666667e-05, + "loss": 0.0197, + "step": 18809 + }, + { + "epoch": 15.886824324324325, + "grad_norm": 0.3773963451385498, + "learning_rate": 2.7065333333333333e-05, + "loss": 0.0691, + "step": 18810 + }, + { + "epoch": 15.88766891891892, + "grad_norm": 0.6612576842308044, + "learning_rate": 2.7065e-05, + "loss": 0.1954, + "step": 18811 + }, + { + "epoch": 15.888513513513514, + "grad_norm": 0.4621056914329529, + "learning_rate": 2.706466666666667e-05, + "loss": 0.0852, + "step": 18812 + }, + { + "epoch": 15.889358108108109, + "grad_norm": 0.48038795590400696, + "learning_rate": 2.7064333333333334e-05, + "loss": 0.0482, + "step": 18813 + }, + { + "epoch": 15.890202702702704, + "grad_norm": 0.3736734390258789, + "learning_rate": 2.7064e-05, + "loss": 0.0404, + "step": 18814 + }, + { + "epoch": 15.891047297297296, + "grad_norm": 0.4475392699241638, + "learning_rate": 2.706366666666667e-05, + "loss": 0.02, + "step": 18815 + }, + { + "epoch": 15.891891891891891, + "grad_norm": 0.20399700105190277, + "learning_rate": 2.7063333333333332e-05, + "loss": 0.0222, + "step": 18816 + }, + { + "epoch": 15.892736486486486, + "grad_norm": 0.24301202595233917, + "learning_rate": 2.7063e-05, + "loss": 0.0175, + "step": 18817 + }, + { + "epoch": 15.89358108108108, + "grad_norm": 0.3093893826007843, + "learning_rate": 2.7062666666666668e-05, + "loss": 0.0173, + "step": 18818 + }, + { + "epoch": 15.894425675675675, + "grad_norm": 0.3126246929168701, + "learning_rate": 2.7062333333333333e-05, + "loss": 0.0134, + "step": 18819 + }, + { + "epoch": 15.89527027027027, + "grad_norm": 0.16510546207427979, + "learning_rate": 2.7062000000000003e-05, + "loss": 0.0119, + "step": 18820 + }, + { + "epoch": 15.896114864864865, + "grad_norm": 0.1699393093585968, + "learning_rate": 2.706166666666667e-05, + "loss": 0.009, + "step": 18821 + }, + { + "epoch": 15.89695945945946, + "grad_norm": 0.4145408868789673, + "learning_rate": 2.7061333333333335e-05, + "loss": 0.0224, + "step": 18822 + }, + { + "epoch": 15.897804054054054, + "grad_norm": 0.20672264695167542, + "learning_rate": 2.7061e-05, + "loss": 0.013, + "step": 18823 + }, + { + "epoch": 15.89864864864865, + "grad_norm": 0.4432198405265808, + "learning_rate": 2.7060666666666667e-05, + "loss": 0.0329, + "step": 18824 + }, + { + "epoch": 15.899493243243244, + "grad_norm": 0.19188496470451355, + "learning_rate": 2.7060333333333332e-05, + "loss": 0.0191, + "step": 18825 + }, + { + "epoch": 15.900337837837839, + "grad_norm": 0.4188947081565857, + "learning_rate": 2.7060000000000002e-05, + "loss": 0.0197, + "step": 18826 + }, + { + "epoch": 15.901182432432432, + "grad_norm": 0.9434117674827576, + "learning_rate": 2.7059666666666664e-05, + "loss": 0.0136, + "step": 18827 + }, + { + "epoch": 15.902027027027026, + "grad_norm": 0.2614683508872986, + "learning_rate": 2.7059333333333334e-05, + "loss": 0.0116, + "step": 18828 + }, + { + "epoch": 15.902871621621621, + "grad_norm": 0.1765451282262802, + "learning_rate": 2.7059000000000003e-05, + "loss": 0.0104, + "step": 18829 + }, + { + "epoch": 15.903716216216216, + "grad_norm": 0.1282753050327301, + "learning_rate": 2.7058666666666666e-05, + "loss": 0.0072, + "step": 18830 + }, + { + "epoch": 15.90456081081081, + "grad_norm": 0.13382567465305328, + "learning_rate": 2.7058333333333335e-05, + "loss": 0.0127, + "step": 18831 + }, + { + "epoch": 15.905405405405405, + "grad_norm": 0.3103150427341461, + "learning_rate": 2.7058e-05, + "loss": 0.0152, + "step": 18832 + }, + { + "epoch": 15.90625, + "grad_norm": 0.21313530206680298, + "learning_rate": 2.7057666666666667e-05, + "loss": 0.0118, + "step": 18833 + }, + { + "epoch": 15.907094594594595, + "grad_norm": 0.21333695948123932, + "learning_rate": 2.7057333333333333e-05, + "loss": 0.0135, + "step": 18834 + }, + { + "epoch": 15.90793918918919, + "grad_norm": 0.13778306543827057, + "learning_rate": 2.7057000000000002e-05, + "loss": 0.0132, + "step": 18835 + }, + { + "epoch": 15.908783783783784, + "grad_norm": 0.19674652814865112, + "learning_rate": 2.7056666666666665e-05, + "loss": 0.0078, + "step": 18836 + }, + { + "epoch": 15.909628378378379, + "grad_norm": 0.3084276616573334, + "learning_rate": 2.7056333333333334e-05, + "loss": 0.0159, + "step": 18837 + }, + { + "epoch": 15.910472972972974, + "grad_norm": 0.30669933557510376, + "learning_rate": 2.7056000000000003e-05, + "loss": 0.021, + "step": 18838 + }, + { + "epoch": 15.911317567567568, + "grad_norm": 0.19681695103645325, + "learning_rate": 2.7055666666666666e-05, + "loss": 0.0112, + "step": 18839 + }, + { + "epoch": 15.912162162162161, + "grad_norm": 0.11599000543355942, + "learning_rate": 2.7055333333333335e-05, + "loss": 0.0085, + "step": 18840 + }, + { + "epoch": 15.913006756756756, + "grad_norm": 0.18450240790843964, + "learning_rate": 2.7055e-05, + "loss": 0.0059, + "step": 18841 + }, + { + "epoch": 15.91385135135135, + "grad_norm": 0.34262165427207947, + "learning_rate": 2.7054666666666667e-05, + "loss": 0.018, + "step": 18842 + }, + { + "epoch": 15.914695945945946, + "grad_norm": 0.31010901927948, + "learning_rate": 2.7054333333333333e-05, + "loss": 0.0109, + "step": 18843 + }, + { + "epoch": 15.91554054054054, + "grad_norm": 0.15586665272712708, + "learning_rate": 2.7054000000000002e-05, + "loss": 0.0082, + "step": 18844 + }, + { + "epoch": 15.916385135135135, + "grad_norm": 0.35201504826545715, + "learning_rate": 2.7053666666666668e-05, + "loss": 0.0131, + "step": 18845 + }, + { + "epoch": 15.91722972972973, + "grad_norm": 0.26627302169799805, + "learning_rate": 2.7053333333333334e-05, + "loss": 0.0171, + "step": 18846 + }, + { + "epoch": 15.918074324324325, + "grad_norm": 0.25373387336730957, + "learning_rate": 2.7053000000000003e-05, + "loss": 0.0109, + "step": 18847 + }, + { + "epoch": 15.91891891891892, + "grad_norm": 0.2545987665653229, + "learning_rate": 2.7052666666666666e-05, + "loss": 0.0148, + "step": 18848 + }, + { + "epoch": 15.919763513513514, + "grad_norm": 0.6553526520729065, + "learning_rate": 2.7052333333333335e-05, + "loss": 0.0181, + "step": 18849 + }, + { + "epoch": 15.920608108108109, + "grad_norm": 0.3194291889667511, + "learning_rate": 2.7052e-05, + "loss": 0.0079, + "step": 18850 + }, + { + "epoch": 15.921452702702704, + "grad_norm": 0.9239679574966431, + "learning_rate": 2.7051666666666667e-05, + "loss": 0.0177, + "step": 18851 + }, + { + "epoch": 15.922297297297296, + "grad_norm": 0.412945955991745, + "learning_rate": 2.7051333333333333e-05, + "loss": 0.0123, + "step": 18852 + }, + { + "epoch": 15.923141891891891, + "grad_norm": 0.576506495475769, + "learning_rate": 2.7051e-05, + "loss": 0.0291, + "step": 18853 + }, + { + "epoch": 15.923986486486486, + "grad_norm": 0.17186634242534637, + "learning_rate": 2.705066666666667e-05, + "loss": 0.0118, + "step": 18854 + }, + { + "epoch": 15.92483108108108, + "grad_norm": 0.22594767808914185, + "learning_rate": 2.7050333333333334e-05, + "loss": 0.0091, + "step": 18855 + }, + { + "epoch": 15.925675675675675, + "grad_norm": 0.28844860196113586, + "learning_rate": 2.705e-05, + "loss": 0.0103, + "step": 18856 + }, + { + "epoch": 15.92652027027027, + "grad_norm": 0.3155047595500946, + "learning_rate": 2.7049666666666666e-05, + "loss": 0.0159, + "step": 18857 + }, + { + "epoch": 15.927364864864865, + "grad_norm": 0.5784269571304321, + "learning_rate": 2.7049333333333335e-05, + "loss": 0.0162, + "step": 18858 + }, + { + "epoch": 15.92820945945946, + "grad_norm": 0.39389315247535706, + "learning_rate": 2.7048999999999998e-05, + "loss": 0.0153, + "step": 18859 + }, + { + "epoch": 15.929054054054054, + "grad_norm": 0.5635330080986023, + "learning_rate": 2.7048666666666667e-05, + "loss": 0.0222, + "step": 18860 + }, + { + "epoch": 15.92989864864865, + "grad_norm": 0.6537546515464783, + "learning_rate": 2.7048333333333333e-05, + "loss": 0.1834, + "step": 18861 + }, + { + "epoch": 15.930743243243244, + "grad_norm": 0.516494870185852, + "learning_rate": 2.7048e-05, + "loss": 0.1076, + "step": 18862 + }, + { + "epoch": 15.931587837837839, + "grad_norm": 0.3557286560535431, + "learning_rate": 2.704766666666667e-05, + "loss": 0.0268, + "step": 18863 + }, + { + "epoch": 15.932432432432432, + "grad_norm": 0.1849779486656189, + "learning_rate": 2.7047333333333334e-05, + "loss": 0.0265, + "step": 18864 + }, + { + "epoch": 15.933277027027026, + "grad_norm": 0.30054613947868347, + "learning_rate": 2.7047e-05, + "loss": 0.0334, + "step": 18865 + }, + { + "epoch": 15.934121621621621, + "grad_norm": 0.16799534857273102, + "learning_rate": 2.7046666666666666e-05, + "loss": 0.0181, + "step": 18866 + }, + { + "epoch": 15.934966216216216, + "grad_norm": 0.29504716396331787, + "learning_rate": 2.7046333333333336e-05, + "loss": 0.0231, + "step": 18867 + }, + { + "epoch": 15.93581081081081, + "grad_norm": 0.18099656701087952, + "learning_rate": 2.7045999999999998e-05, + "loss": 0.0189, + "step": 18868 + }, + { + "epoch": 15.936655405405405, + "grad_norm": 0.24607239663600922, + "learning_rate": 2.7045666666666668e-05, + "loss": 0.0217, + "step": 18869 + }, + { + "epoch": 15.9375, + "grad_norm": 0.24153929948806763, + "learning_rate": 2.7045333333333337e-05, + "loss": 0.0197, + "step": 18870 + }, + { + "epoch": 15.938344594594595, + "grad_norm": 0.1850704699754715, + "learning_rate": 2.7045e-05, + "loss": 0.0117, + "step": 18871 + }, + { + "epoch": 15.93918918918919, + "grad_norm": 0.2396479696035385, + "learning_rate": 2.704466666666667e-05, + "loss": 0.0173, + "step": 18872 + }, + { + "epoch": 15.940033783783784, + "grad_norm": 0.4435042440891266, + "learning_rate": 2.7044333333333335e-05, + "loss": 0.0237, + "step": 18873 + }, + { + "epoch": 15.940878378378379, + "grad_norm": 0.24509479105472565, + "learning_rate": 2.7044e-05, + "loss": 0.0225, + "step": 18874 + }, + { + "epoch": 15.941722972972974, + "grad_norm": 0.29623880982398987, + "learning_rate": 2.7043666666666667e-05, + "loss": 0.0207, + "step": 18875 + }, + { + "epoch": 15.942567567567568, + "grad_norm": 0.2278040498495102, + "learning_rate": 2.7043333333333336e-05, + "loss": 0.01, + "step": 18876 + }, + { + "epoch": 15.943412162162161, + "grad_norm": 0.25297045707702637, + "learning_rate": 2.7043e-05, + "loss": 0.0238, + "step": 18877 + }, + { + "epoch": 15.944256756756756, + "grad_norm": 0.2226988822221756, + "learning_rate": 2.7042666666666668e-05, + "loss": 0.012, + "step": 18878 + }, + { + "epoch": 15.94510135135135, + "grad_norm": 0.3274124562740326, + "learning_rate": 2.7042333333333337e-05, + "loss": 0.0245, + "step": 18879 + }, + { + "epoch": 15.945945945945946, + "grad_norm": 0.2993306815624237, + "learning_rate": 2.7042e-05, + "loss": 0.0204, + "step": 18880 + }, + { + "epoch": 15.94679054054054, + "grad_norm": 0.44122231006622314, + "learning_rate": 2.704166666666667e-05, + "loss": 0.0286, + "step": 18881 + }, + { + "epoch": 15.947635135135135, + "grad_norm": 0.2630763351917267, + "learning_rate": 2.7041333333333335e-05, + "loss": 0.0137, + "step": 18882 + }, + { + "epoch": 15.94847972972973, + "grad_norm": 0.16958993673324585, + "learning_rate": 2.7041e-05, + "loss": 0.0144, + "step": 18883 + }, + { + "epoch": 15.949324324324325, + "grad_norm": 0.19210058450698853, + "learning_rate": 2.7040666666666667e-05, + "loss": 0.0134, + "step": 18884 + }, + { + "epoch": 15.95016891891892, + "grad_norm": 0.583527147769928, + "learning_rate": 2.7040333333333333e-05, + "loss": 0.0173, + "step": 18885 + }, + { + "epoch": 15.951013513513514, + "grad_norm": 0.22587142884731293, + "learning_rate": 2.704e-05, + "loss": 0.0146, + "step": 18886 + }, + { + "epoch": 15.951858108108109, + "grad_norm": 0.1940801441669464, + "learning_rate": 2.7039666666666668e-05, + "loss": 0.009, + "step": 18887 + }, + { + "epoch": 15.952702702702704, + "grad_norm": 0.25088274478912354, + "learning_rate": 2.7039333333333334e-05, + "loss": 0.019, + "step": 18888 + }, + { + "epoch": 15.953547297297296, + "grad_norm": 0.13529253005981445, + "learning_rate": 2.7039e-05, + "loss": 0.0085, + "step": 18889 + }, + { + "epoch": 15.954391891891891, + "grad_norm": 0.2649870812892914, + "learning_rate": 2.703866666666667e-05, + "loss": 0.0132, + "step": 18890 + }, + { + "epoch": 15.955236486486486, + "grad_norm": 0.12326785176992416, + "learning_rate": 2.703833333333333e-05, + "loss": 0.0071, + "step": 18891 + }, + { + "epoch": 15.95608108108108, + "grad_norm": 0.25371798872947693, + "learning_rate": 2.7038e-05, + "loss": 0.0158, + "step": 18892 + }, + { + "epoch": 15.956925675675675, + "grad_norm": 0.23793622851371765, + "learning_rate": 2.7037666666666667e-05, + "loss": 0.0123, + "step": 18893 + }, + { + "epoch": 15.95777027027027, + "grad_norm": 0.20132878422737122, + "learning_rate": 2.7037333333333333e-05, + "loss": 0.0077, + "step": 18894 + }, + { + "epoch": 15.958614864864865, + "grad_norm": 0.20020291209220886, + "learning_rate": 2.7037000000000002e-05, + "loss": 0.0114, + "step": 18895 + }, + { + "epoch": 15.95945945945946, + "grad_norm": 0.12031635642051697, + "learning_rate": 2.7036666666666668e-05, + "loss": 0.0116, + "step": 18896 + }, + { + "epoch": 15.960304054054054, + "grad_norm": 0.35592520236968994, + "learning_rate": 2.7036333333333334e-05, + "loss": 0.0107, + "step": 18897 + }, + { + "epoch": 15.96114864864865, + "grad_norm": 0.1693996787071228, + "learning_rate": 2.7036e-05, + "loss": 0.0111, + "step": 18898 + }, + { + "epoch": 15.961993243243244, + "grad_norm": 0.24791668355464935, + "learning_rate": 2.703566666666667e-05, + "loss": 0.013, + "step": 18899 + }, + { + "epoch": 15.962837837837839, + "grad_norm": 0.3076660931110382, + "learning_rate": 2.7035333333333332e-05, + "loss": 0.0163, + "step": 18900 + }, + { + "epoch": 15.963682432432432, + "grad_norm": 0.5759793519973755, + "learning_rate": 2.7035e-05, + "loss": 0.0174, + "step": 18901 + }, + { + "epoch": 15.964527027027026, + "grad_norm": 0.7135964632034302, + "learning_rate": 2.7034666666666667e-05, + "loss": 0.0187, + "step": 18902 + }, + { + "epoch": 15.965371621621621, + "grad_norm": 0.21387101709842682, + "learning_rate": 2.7034333333333333e-05, + "loss": 0.0173, + "step": 18903 + }, + { + "epoch": 15.966216216216216, + "grad_norm": 0.10202141851186752, + "learning_rate": 2.7034000000000002e-05, + "loss": 0.0048, + "step": 18904 + }, + { + "epoch": 15.96706081081081, + "grad_norm": 0.34674879908561707, + "learning_rate": 2.703366666666667e-05, + "loss": 0.0143, + "step": 18905 + }, + { + "epoch": 15.967905405405405, + "grad_norm": 0.25236597657203674, + "learning_rate": 2.7033333333333334e-05, + "loss": 0.0181, + "step": 18906 + }, + { + "epoch": 15.96875, + "grad_norm": 0.28578832745552063, + "learning_rate": 2.7033e-05, + "loss": 0.0164, + "step": 18907 + }, + { + "epoch": 15.969594594594595, + "grad_norm": 0.7483678460121155, + "learning_rate": 2.703266666666667e-05, + "loss": 0.0249, + "step": 18908 + }, + { + "epoch": 15.97043918918919, + "grad_norm": 0.42669951915740967, + "learning_rate": 2.7032333333333332e-05, + "loss": 0.0217, + "step": 18909 + }, + { + "epoch": 15.971283783783784, + "grad_norm": 0.4086074233055115, + "learning_rate": 2.7032e-05, + "loss": 0.0304, + "step": 18910 + }, + { + "epoch": 15.972128378378379, + "grad_norm": 0.6443332433700562, + "learning_rate": 2.7031666666666667e-05, + "loss": 0.1828, + "step": 18911 + }, + { + "epoch": 15.972972972972974, + "grad_norm": 0.3256075978279114, + "learning_rate": 2.7031333333333333e-05, + "loss": 0.058, + "step": 18912 + }, + { + "epoch": 15.973817567567568, + "grad_norm": 0.4985482692718506, + "learning_rate": 2.7031000000000003e-05, + "loss": 0.0335, + "step": 18913 + }, + { + "epoch": 15.974662162162161, + "grad_norm": 0.5144703388214111, + "learning_rate": 2.7030666666666665e-05, + "loss": 0.0173, + "step": 18914 + }, + { + "epoch": 15.975506756756756, + "grad_norm": 0.22378501296043396, + "learning_rate": 2.7030333333333334e-05, + "loss": 0.0172, + "step": 18915 + }, + { + "epoch": 15.97635135135135, + "grad_norm": 0.23188939690589905, + "learning_rate": 2.703e-05, + "loss": 0.0234, + "step": 18916 + }, + { + "epoch": 15.977195945945946, + "grad_norm": 0.2432371973991394, + "learning_rate": 2.7029666666666666e-05, + "loss": 0.0291, + "step": 18917 + }, + { + "epoch": 15.97804054054054, + "grad_norm": 0.22301085293293, + "learning_rate": 2.7029333333333332e-05, + "loss": 0.0263, + "step": 18918 + }, + { + "epoch": 15.978885135135135, + "grad_norm": 0.17018528282642365, + "learning_rate": 2.7029e-05, + "loss": 0.0122, + "step": 18919 + }, + { + "epoch": 15.97972972972973, + "grad_norm": 0.15237727761268616, + "learning_rate": 2.7028666666666668e-05, + "loss": 0.0113, + "step": 18920 + }, + { + "epoch": 15.980574324324325, + "grad_norm": 0.18256962299346924, + "learning_rate": 2.7028333333333333e-05, + "loss": 0.0145, + "step": 18921 + }, + { + "epoch": 15.98141891891892, + "grad_norm": 0.23990577459335327, + "learning_rate": 2.7028000000000003e-05, + "loss": 0.0183, + "step": 18922 + }, + { + "epoch": 15.982263513513514, + "grad_norm": 0.2512304484844208, + "learning_rate": 2.7027666666666665e-05, + "loss": 0.0096, + "step": 18923 + }, + { + "epoch": 15.983108108108109, + "grad_norm": 0.24380725622177124, + "learning_rate": 2.7027333333333335e-05, + "loss": 0.0098, + "step": 18924 + }, + { + "epoch": 15.983952702702704, + "grad_norm": 0.3394589424133301, + "learning_rate": 2.7027e-05, + "loss": 0.0176, + "step": 18925 + }, + { + "epoch": 15.984797297297296, + "grad_norm": 0.6180864572525024, + "learning_rate": 2.7026666666666667e-05, + "loss": 0.0198, + "step": 18926 + }, + { + "epoch": 15.985641891891891, + "grad_norm": 0.9660511016845703, + "learning_rate": 2.7026333333333332e-05, + "loss": 0.0159, + "step": 18927 + }, + { + "epoch": 15.986486486486486, + "grad_norm": 0.29548195004463196, + "learning_rate": 2.7026000000000002e-05, + "loss": 0.0146, + "step": 18928 + }, + { + "epoch": 15.98733108108108, + "grad_norm": 0.4281396269798279, + "learning_rate": 2.7025666666666668e-05, + "loss": 0.0245, + "step": 18929 + }, + { + "epoch": 15.988175675675675, + "grad_norm": 0.2845509350299835, + "learning_rate": 2.7025333333333334e-05, + "loss": 0.0105, + "step": 18930 + }, + { + "epoch": 15.98902027027027, + "grad_norm": 0.18011987209320068, + "learning_rate": 2.7025000000000003e-05, + "loss": 0.014, + "step": 18931 + }, + { + "epoch": 15.989864864864865, + "grad_norm": 0.19641205668449402, + "learning_rate": 2.7024666666666666e-05, + "loss": 0.0092, + "step": 18932 + }, + { + "epoch": 15.99070945945946, + "grad_norm": 0.42221829295158386, + "learning_rate": 2.7024333333333335e-05, + "loss": 0.0162, + "step": 18933 + }, + { + "epoch": 15.991554054054054, + "grad_norm": 0.7741987705230713, + "learning_rate": 2.7024e-05, + "loss": 0.0423, + "step": 18934 + }, + { + "epoch": 15.99239864864865, + "grad_norm": 0.43355610966682434, + "learning_rate": 2.7023666666666667e-05, + "loss": 0.0154, + "step": 18935 + }, + { + "epoch": 15.993243243243244, + "grad_norm": 0.21689914166927338, + "learning_rate": 2.7023333333333333e-05, + "loss": 0.0234, + "step": 18936 + }, + { + "epoch": 15.994087837837839, + "grad_norm": 0.5040437579154968, + "learning_rate": 2.7023000000000002e-05, + "loss": 0.0182, + "step": 18937 + }, + { + "epoch": 15.994932432432432, + "grad_norm": 0.4167857766151428, + "learning_rate": 2.7022666666666668e-05, + "loss": 0.0133, + "step": 18938 + }, + { + "epoch": 15.995777027027026, + "grad_norm": 0.737220048904419, + "learning_rate": 2.7022333333333334e-05, + "loss": 0.0199, + "step": 18939 + }, + { + "epoch": 15.996621621621621, + "grad_norm": 0.3822496831417084, + "learning_rate": 2.7022000000000003e-05, + "loss": 0.0154, + "step": 18940 + }, + { + "epoch": 15.997466216216216, + "grad_norm": 0.8749369382858276, + "learning_rate": 2.7021666666666666e-05, + "loss": 0.034, + "step": 18941 + }, + { + "epoch": 15.99831081081081, + "grad_norm": 0.47848764061927795, + "learning_rate": 2.7021333333333335e-05, + "loss": 0.0169, + "step": 18942 + }, + { + "epoch": 15.999155405405405, + "grad_norm": 0.21957193315029144, + "learning_rate": 2.7020999999999998e-05, + "loss": 0.011, + "step": 18943 + }, + { + "epoch": 16.0, + "grad_norm": 0.4119029939174652, + "learning_rate": 2.7020666666666667e-05, + "loss": 0.0221, + "step": 18944 + }, + { + "epoch": 16.000844594594593, + "grad_norm": 0.6321733593940735, + "learning_rate": 2.7020333333333336e-05, + "loss": 0.1186, + "step": 18945 + }, + { + "epoch": 16.00168918918919, + "grad_norm": 0.5644442439079285, + "learning_rate": 2.702e-05, + "loss": 0.0832, + "step": 18946 + }, + { + "epoch": 16.002533783783782, + "grad_norm": 0.2862599790096283, + "learning_rate": 2.7019666666666668e-05, + "loss": 0.0433, + "step": 18947 + }, + { + "epoch": 16.00337837837838, + "grad_norm": 0.19299070537090302, + "learning_rate": 2.7019333333333334e-05, + "loss": 0.0211, + "step": 18948 + }, + { + "epoch": 16.004222972972972, + "grad_norm": 0.37322503328323364, + "learning_rate": 2.7019e-05, + "loss": 0.0232, + "step": 18949 + }, + { + "epoch": 16.00506756756757, + "grad_norm": 0.2164585292339325, + "learning_rate": 2.7018666666666666e-05, + "loss": 0.0099, + "step": 18950 + }, + { + "epoch": 16.00591216216216, + "grad_norm": 0.341135710477829, + "learning_rate": 2.7018333333333335e-05, + "loss": 0.0103, + "step": 18951 + }, + { + "epoch": 16.006756756756758, + "grad_norm": 0.34524697065353394, + "learning_rate": 2.7017999999999998e-05, + "loss": 0.0138, + "step": 18952 + }, + { + "epoch": 16.00760135135135, + "grad_norm": 0.4143409729003906, + "learning_rate": 2.7017666666666667e-05, + "loss": 0.0213, + "step": 18953 + }, + { + "epoch": 16.008445945945947, + "grad_norm": 0.4751981496810913, + "learning_rate": 2.7017333333333336e-05, + "loss": 0.0174, + "step": 18954 + }, + { + "epoch": 16.00929054054054, + "grad_norm": 0.17956262826919556, + "learning_rate": 2.7017e-05, + "loss": 0.0094, + "step": 18955 + }, + { + "epoch": 16.010135135135137, + "grad_norm": 0.2720603048801422, + "learning_rate": 2.701666666666667e-05, + "loss": 0.0188, + "step": 18956 + }, + { + "epoch": 16.01097972972973, + "grad_norm": 0.5537415146827698, + "learning_rate": 2.7016333333333334e-05, + "loss": 0.0107, + "step": 18957 + }, + { + "epoch": 16.011824324324323, + "grad_norm": 0.15573784708976746, + "learning_rate": 2.7016e-05, + "loss": 0.0086, + "step": 18958 + }, + { + "epoch": 16.01266891891892, + "grad_norm": 0.24943673610687256, + "learning_rate": 2.7015666666666666e-05, + "loss": 0.0093, + "step": 18959 + }, + { + "epoch": 16.013513513513512, + "grad_norm": 0.3397156894207001, + "learning_rate": 2.7015333333333335e-05, + "loss": 0.0093, + "step": 18960 + }, + { + "epoch": 16.01435810810811, + "grad_norm": 0.5707467198371887, + "learning_rate": 2.7015e-05, + "loss": 0.0191, + "step": 18961 + }, + { + "epoch": 16.0152027027027, + "grad_norm": 0.19566673040390015, + "learning_rate": 2.7014666666666667e-05, + "loss": 0.0085, + "step": 18962 + }, + { + "epoch": 16.0160472972973, + "grad_norm": 0.20858021080493927, + "learning_rate": 2.7014333333333337e-05, + "loss": 0.0096, + "step": 18963 + }, + { + "epoch": 16.01689189189189, + "grad_norm": 0.38161835074424744, + "learning_rate": 2.7014e-05, + "loss": 0.0157, + "step": 18964 + }, + { + "epoch": 16.017736486486488, + "grad_norm": 0.28362664580345154, + "learning_rate": 2.701366666666667e-05, + "loss": 0.0173, + "step": 18965 + }, + { + "epoch": 16.01858108108108, + "grad_norm": 0.3455907702445984, + "learning_rate": 2.7013333333333334e-05, + "loss": 0.0205, + "step": 18966 + }, + { + "epoch": 16.019425675675677, + "grad_norm": 0.3430241048336029, + "learning_rate": 2.7013e-05, + "loss": 0.0157, + "step": 18967 + }, + { + "epoch": 16.02027027027027, + "grad_norm": 0.2674371004104614, + "learning_rate": 2.7012666666666666e-05, + "loss": 0.0095, + "step": 18968 + }, + { + "epoch": 16.021114864864863, + "grad_norm": 0.47311505675315857, + "learning_rate": 2.7012333333333336e-05, + "loss": 0.0317, + "step": 18969 + }, + { + "epoch": 16.02195945945946, + "grad_norm": 0.17933370172977448, + "learning_rate": 2.7012e-05, + "loss": 0.0085, + "step": 18970 + }, + { + "epoch": 16.022804054054053, + "grad_norm": 0.45508649945259094, + "learning_rate": 2.7011666666666668e-05, + "loss": 0.0171, + "step": 18971 + }, + { + "epoch": 16.02364864864865, + "grad_norm": 0.31161993741989136, + "learning_rate": 2.7011333333333337e-05, + "loss": 0.0095, + "step": 18972 + }, + { + "epoch": 16.024493243243242, + "grad_norm": 0.2645605206489563, + "learning_rate": 2.7011e-05, + "loss": 0.0117, + "step": 18973 + }, + { + "epoch": 16.02533783783784, + "grad_norm": 0.10327792167663574, + "learning_rate": 2.701066666666667e-05, + "loss": 0.0055, + "step": 18974 + }, + { + "epoch": 16.02618243243243, + "grad_norm": 0.1905670017004013, + "learning_rate": 2.701033333333333e-05, + "loss": 0.0115, + "step": 18975 + }, + { + "epoch": 16.027027027027028, + "grad_norm": 0.2701975107192993, + "learning_rate": 2.701e-05, + "loss": 0.0181, + "step": 18976 + }, + { + "epoch": 16.02787162162162, + "grad_norm": 0.3207516074180603, + "learning_rate": 2.7009666666666667e-05, + "loss": 0.0128, + "step": 18977 + }, + { + "epoch": 16.028716216216218, + "grad_norm": 0.168283149600029, + "learning_rate": 2.7009333333333332e-05, + "loss": 0.0098, + "step": 18978 + }, + { + "epoch": 16.02956081081081, + "grad_norm": 0.15174083411693573, + "learning_rate": 2.7009000000000002e-05, + "loss": 0.0119, + "step": 18979 + }, + { + "epoch": 16.030405405405407, + "grad_norm": 0.31633925437927246, + "learning_rate": 2.7008666666666668e-05, + "loss": 0.0112, + "step": 18980 + }, + { + "epoch": 16.03125, + "grad_norm": 0.5581297278404236, + "learning_rate": 2.7008333333333334e-05, + "loss": 0.0183, + "step": 18981 + }, + { + "epoch": 16.032094594594593, + "grad_norm": 0.18757742643356323, + "learning_rate": 2.7008e-05, + "loss": 0.0155, + "step": 18982 + }, + { + "epoch": 16.03293918918919, + "grad_norm": 0.14947406947612762, + "learning_rate": 2.700766666666667e-05, + "loss": 0.007, + "step": 18983 + }, + { + "epoch": 16.033783783783782, + "grad_norm": 1.0703396797180176, + "learning_rate": 2.700733333333333e-05, + "loss": 0.0177, + "step": 18984 + }, + { + "epoch": 16.03462837837838, + "grad_norm": 0.2638567388057709, + "learning_rate": 2.7007e-05, + "loss": 0.0104, + "step": 18985 + }, + { + "epoch": 16.035472972972972, + "grad_norm": 0.6872798800468445, + "learning_rate": 2.7006666666666667e-05, + "loss": 0.0245, + "step": 18986 + }, + { + "epoch": 16.03631756756757, + "grad_norm": 0.5201773047447205, + "learning_rate": 2.7006333333333333e-05, + "loss": 0.0156, + "step": 18987 + }, + { + "epoch": 16.03716216216216, + "grad_norm": 0.20256367325782776, + "learning_rate": 2.7006000000000002e-05, + "loss": 0.0104, + "step": 18988 + }, + { + "epoch": 16.038006756756758, + "grad_norm": 0.35367655754089355, + "learning_rate": 2.7005666666666668e-05, + "loss": 0.0184, + "step": 18989 + }, + { + "epoch": 16.03885135135135, + "grad_norm": 0.8182762861251831, + "learning_rate": 2.7005333333333334e-05, + "loss": 0.0202, + "step": 18990 + }, + { + "epoch": 16.039695945945947, + "grad_norm": 0.4495563805103302, + "learning_rate": 2.7005e-05, + "loss": 0.0186, + "step": 18991 + }, + { + "epoch": 16.04054054054054, + "grad_norm": 1.5800673961639404, + "learning_rate": 2.700466666666667e-05, + "loss": 0.0144, + "step": 18992 + }, + { + "epoch": 16.041385135135137, + "grad_norm": 0.8964859247207642, + "learning_rate": 2.700433333333333e-05, + "loss": 0.0219, + "step": 18993 + }, + { + "epoch": 16.04222972972973, + "grad_norm": 0.35774317383766174, + "learning_rate": 2.7004e-05, + "loss": 0.0144, + "step": 18994 + }, + { + "epoch": 16.043074324324323, + "grad_norm": 0.9202538728713989, + "learning_rate": 2.700366666666667e-05, + "loss": 0.1664, + "step": 18995 + }, + { + "epoch": 16.04391891891892, + "grad_norm": 0.4201194643974304, + "learning_rate": 2.7003333333333333e-05, + "loss": 0.0777, + "step": 18996 + }, + { + "epoch": 16.044763513513512, + "grad_norm": 0.38394537568092346, + "learning_rate": 2.7003000000000002e-05, + "loss": 0.0388, + "step": 18997 + }, + { + "epoch": 16.04560810810811, + "grad_norm": 0.3647363781929016, + "learning_rate": 2.7002666666666668e-05, + "loss": 0.0406, + "step": 18998 + }, + { + "epoch": 16.0464527027027, + "grad_norm": 0.27066001296043396, + "learning_rate": 2.7002333333333334e-05, + "loss": 0.0208, + "step": 18999 + }, + { + "epoch": 16.0472972972973, + "grad_norm": 0.40812283754348755, + "learning_rate": 2.7002e-05, + "loss": 0.0352, + "step": 19000 + }, + { + "epoch": 16.0472972972973, + "eval_cer": 0.09300457794554655, + "eval_loss": 0.25963857769966125, + "eval_runtime": 8.6426, + "eval_samples_per_second": 11.686, + "eval_steps_per_second": 0.463, + "eval_wer": 0.2958553791887125, + "step": 19000 + }, + { + "epoch": 16.04814189189189, + "grad_norm": 0.23464694619178772, + "learning_rate": 2.700166666666667e-05, + "loss": 0.0156, + "step": 19001 + }, + { + "epoch": 16.048986486486488, + "grad_norm": 0.2345747947692871, + "learning_rate": 2.7001333333333332e-05, + "loss": 0.0125, + "step": 19002 + }, + { + "epoch": 16.04983108108108, + "grad_norm": 0.39202919602394104, + "learning_rate": 2.7001e-05, + "loss": 0.0189, + "step": 19003 + }, + { + "epoch": 16.050675675675677, + "grad_norm": 0.1496170461177826, + "learning_rate": 2.7000666666666667e-05, + "loss": 0.0137, + "step": 19004 + }, + { + "epoch": 16.05152027027027, + "grad_norm": 0.28966251015663147, + "learning_rate": 2.7000333333333333e-05, + "loss": 0.0168, + "step": 19005 + }, + { + "epoch": 16.052364864864863, + "grad_norm": 0.2715795636177063, + "learning_rate": 2.7000000000000002e-05, + "loss": 0.0223, + "step": 19006 + }, + { + "epoch": 16.05320945945946, + "grad_norm": 0.21958030760288239, + "learning_rate": 2.6999666666666665e-05, + "loss": 0.014, + "step": 19007 + }, + { + "epoch": 16.054054054054053, + "grad_norm": 0.2935420572757721, + "learning_rate": 2.6999333333333334e-05, + "loss": 0.0145, + "step": 19008 + }, + { + "epoch": 16.05489864864865, + "grad_norm": 0.19486260414123535, + "learning_rate": 2.6999e-05, + "loss": 0.0116, + "step": 19009 + }, + { + "epoch": 16.055743243243242, + "grad_norm": 0.16414806246757507, + "learning_rate": 2.6998666666666666e-05, + "loss": 0.0164, + "step": 19010 + }, + { + "epoch": 16.05658783783784, + "grad_norm": 0.15592654049396515, + "learning_rate": 2.6998333333333335e-05, + "loss": 0.0118, + "step": 19011 + }, + { + "epoch": 16.05743243243243, + "grad_norm": 0.21121399104595184, + "learning_rate": 2.6998e-05, + "loss": 0.0129, + "step": 19012 + }, + { + "epoch": 16.058277027027028, + "grad_norm": 0.19802705943584442, + "learning_rate": 2.6997666666666667e-05, + "loss": 0.0104, + "step": 19013 + }, + { + "epoch": 16.05912162162162, + "grad_norm": 0.14865070581436157, + "learning_rate": 2.6997333333333333e-05, + "loss": 0.0077, + "step": 19014 + }, + { + "epoch": 16.059966216216218, + "grad_norm": 0.19859330356121063, + "learning_rate": 2.6997000000000003e-05, + "loss": 0.013, + "step": 19015 + }, + { + "epoch": 16.06081081081081, + "grad_norm": 0.13254739344120026, + "learning_rate": 2.6996666666666665e-05, + "loss": 0.0068, + "step": 19016 + }, + { + "epoch": 16.061655405405407, + "grad_norm": 0.6553070545196533, + "learning_rate": 2.6996333333333334e-05, + "loss": 0.0177, + "step": 19017 + }, + { + "epoch": 16.0625, + "grad_norm": 0.4011949300765991, + "learning_rate": 2.6996e-05, + "loss": 0.0187, + "step": 19018 + }, + { + "epoch": 16.063344594594593, + "grad_norm": 0.3609631657600403, + "learning_rate": 2.6995666666666666e-05, + "loss": 0.0115, + "step": 19019 + }, + { + "epoch": 16.06418918918919, + "grad_norm": 0.32779502868652344, + "learning_rate": 2.6995333333333336e-05, + "loss": 0.01, + "step": 19020 + }, + { + "epoch": 16.065033783783782, + "grad_norm": 0.3398602306842804, + "learning_rate": 2.6995e-05, + "loss": 0.0172, + "step": 19021 + }, + { + "epoch": 16.06587837837838, + "grad_norm": 0.2640920877456665, + "learning_rate": 2.6994666666666668e-05, + "loss": 0.0155, + "step": 19022 + }, + { + "epoch": 16.066722972972972, + "grad_norm": 0.23968163132667542, + "learning_rate": 2.6994333333333333e-05, + "loss": 0.0105, + "step": 19023 + }, + { + "epoch": 16.06756756756757, + "grad_norm": 0.6163328289985657, + "learning_rate": 2.6994000000000003e-05, + "loss": 0.0285, + "step": 19024 + }, + { + "epoch": 16.06841216216216, + "grad_norm": 0.15507137775421143, + "learning_rate": 2.6993666666666665e-05, + "loss": 0.0084, + "step": 19025 + }, + { + "epoch": 16.069256756756758, + "grad_norm": 0.34600165486335754, + "learning_rate": 2.6993333333333335e-05, + "loss": 0.0109, + "step": 19026 + }, + { + "epoch": 16.07010135135135, + "grad_norm": 0.26656684279441833, + "learning_rate": 2.6993e-05, + "loss": 0.0127, + "step": 19027 + }, + { + "epoch": 16.070945945945947, + "grad_norm": 0.34257495403289795, + "learning_rate": 2.6992666666666667e-05, + "loss": 0.0098, + "step": 19028 + }, + { + "epoch": 16.07179054054054, + "grad_norm": 0.21748508512973785, + "learning_rate": 2.6992333333333336e-05, + "loss": 0.017, + "step": 19029 + }, + { + "epoch": 16.072635135135137, + "grad_norm": 0.2160845845937729, + "learning_rate": 2.6992000000000002e-05, + "loss": 0.0154, + "step": 19030 + }, + { + "epoch": 16.07347972972973, + "grad_norm": 0.3357551693916321, + "learning_rate": 2.6991666666666668e-05, + "loss": 0.0157, + "step": 19031 + }, + { + "epoch": 16.074324324324323, + "grad_norm": 0.5616337060928345, + "learning_rate": 2.6991333333333334e-05, + "loss": 0.0167, + "step": 19032 + }, + { + "epoch": 16.07516891891892, + "grad_norm": 0.25123703479766846, + "learning_rate": 2.6991000000000003e-05, + "loss": 0.0156, + "step": 19033 + }, + { + "epoch": 16.076013513513512, + "grad_norm": 0.14878472685813904, + "learning_rate": 2.6990666666666666e-05, + "loss": 0.0079, + "step": 19034 + }, + { + "epoch": 16.07685810810811, + "grad_norm": 0.5795447826385498, + "learning_rate": 2.6990333333333335e-05, + "loss": 0.0174, + "step": 19035 + }, + { + "epoch": 16.0777027027027, + "grad_norm": 0.2678561508655548, + "learning_rate": 2.699e-05, + "loss": 0.0064, + "step": 19036 + }, + { + "epoch": 16.0785472972973, + "grad_norm": 0.3419835865497589, + "learning_rate": 2.6989666666666667e-05, + "loss": 0.0111, + "step": 19037 + }, + { + "epoch": 16.07939189189189, + "grad_norm": 0.22936256229877472, + "learning_rate": 2.6989333333333336e-05, + "loss": 0.0089, + "step": 19038 + }, + { + "epoch": 16.080236486486488, + "grad_norm": 0.2093910425901413, + "learning_rate": 2.6989e-05, + "loss": 0.0142, + "step": 19039 + }, + { + "epoch": 16.08108108108108, + "grad_norm": 0.3602679669857025, + "learning_rate": 2.6988666666666668e-05, + "loss": 0.009, + "step": 19040 + }, + { + "epoch": 16.081925675675677, + "grad_norm": 1.0083303451538086, + "learning_rate": 2.6988333333333334e-05, + "loss": 0.0192, + "step": 19041 + }, + { + "epoch": 16.08277027027027, + "grad_norm": 0.5215113759040833, + "learning_rate": 2.6988e-05, + "loss": 0.0142, + "step": 19042 + }, + { + "epoch": 16.083614864864863, + "grad_norm": 0.5954417586326599, + "learning_rate": 2.6987666666666666e-05, + "loss": 0.0355, + "step": 19043 + }, + { + "epoch": 16.08445945945946, + "grad_norm": 0.7231013178825378, + "learning_rate": 2.6987333333333335e-05, + "loss": 0.0254, + "step": 19044 + }, + { + "epoch": 16.085304054054053, + "grad_norm": 1.2272833585739136, + "learning_rate": 2.6987e-05, + "loss": 0.1933, + "step": 19045 + }, + { + "epoch": 16.08614864864865, + "grad_norm": 0.4711672365665436, + "learning_rate": 2.6986666666666667e-05, + "loss": 0.1066, + "step": 19046 + }, + { + "epoch": 16.086993243243242, + "grad_norm": 0.5765762329101562, + "learning_rate": 2.6986333333333336e-05, + "loss": 0.0614, + "step": 19047 + }, + { + "epoch": 16.08783783783784, + "grad_norm": 0.28749796748161316, + "learning_rate": 2.6986e-05, + "loss": 0.0206, + "step": 19048 + }, + { + "epoch": 16.08868243243243, + "grad_norm": 0.20439353585243225, + "learning_rate": 2.6985666666666668e-05, + "loss": 0.0223, + "step": 19049 + }, + { + "epoch": 16.089527027027028, + "grad_norm": 0.21609415113925934, + "learning_rate": 2.6985333333333334e-05, + "loss": 0.0175, + "step": 19050 + }, + { + "epoch": 16.09037162162162, + "grad_norm": 0.35154327750205994, + "learning_rate": 2.6985e-05, + "loss": 0.0172, + "step": 19051 + }, + { + "epoch": 16.091216216216218, + "grad_norm": 0.44154199957847595, + "learning_rate": 2.6984666666666666e-05, + "loss": 0.0138, + "step": 19052 + }, + { + "epoch": 16.09206081081081, + "grad_norm": 0.16727213561534882, + "learning_rate": 2.6984333333333335e-05, + "loss": 0.0156, + "step": 19053 + }, + { + "epoch": 16.092905405405407, + "grad_norm": 0.26246291399002075, + "learning_rate": 2.6984e-05, + "loss": 0.0125, + "step": 19054 + }, + { + "epoch": 16.09375, + "grad_norm": 0.5175496339797974, + "learning_rate": 2.6983666666666667e-05, + "loss": 0.0185, + "step": 19055 + }, + { + "epoch": 16.094594594594593, + "grad_norm": 0.29622572660446167, + "learning_rate": 2.6983333333333336e-05, + "loss": 0.019, + "step": 19056 + }, + { + "epoch": 16.09543918918919, + "grad_norm": 0.1160205751657486, + "learning_rate": 2.6983e-05, + "loss": 0.0103, + "step": 19057 + }, + { + "epoch": 16.096283783783782, + "grad_norm": 0.22810788452625275, + "learning_rate": 2.698266666666667e-05, + "loss": 0.0136, + "step": 19058 + }, + { + "epoch": 16.09712837837838, + "grad_norm": 0.1956898719072342, + "learning_rate": 2.6982333333333334e-05, + "loss": 0.0174, + "step": 19059 + }, + { + "epoch": 16.097972972972972, + "grad_norm": 0.18883618712425232, + "learning_rate": 2.6982e-05, + "loss": 0.0154, + "step": 19060 + }, + { + "epoch": 16.09881756756757, + "grad_norm": 0.7749740481376648, + "learning_rate": 2.698166666666667e-05, + "loss": 0.0161, + "step": 19061 + }, + { + "epoch": 16.09966216216216, + "grad_norm": 0.4287874102592468, + "learning_rate": 2.6981333333333335e-05, + "loss": 0.0241, + "step": 19062 + }, + { + "epoch": 16.100506756756758, + "grad_norm": 0.5161810517311096, + "learning_rate": 2.6981e-05, + "loss": 0.014, + "step": 19063 + }, + { + "epoch": 16.10135135135135, + "grad_norm": 0.2639031708240509, + "learning_rate": 2.6980666666666667e-05, + "loss": 0.0089, + "step": 19064 + }, + { + "epoch": 16.102195945945947, + "grad_norm": 0.23968330025672913, + "learning_rate": 2.6980333333333333e-05, + "loss": 0.021, + "step": 19065 + }, + { + "epoch": 16.10304054054054, + "grad_norm": 0.21387390792369843, + "learning_rate": 2.698e-05, + "loss": 0.0106, + "step": 19066 + }, + { + "epoch": 16.103885135135137, + "grad_norm": 0.19448862969875336, + "learning_rate": 2.697966666666667e-05, + "loss": 0.0073, + "step": 19067 + }, + { + "epoch": 16.10472972972973, + "grad_norm": 0.2063804715871811, + "learning_rate": 2.697933333333333e-05, + "loss": 0.0091, + "step": 19068 + }, + { + "epoch": 16.105574324324323, + "grad_norm": 0.2641753554344177, + "learning_rate": 2.6979e-05, + "loss": 0.0149, + "step": 19069 + }, + { + "epoch": 16.10641891891892, + "grad_norm": 1.3552238941192627, + "learning_rate": 2.697866666666667e-05, + "loss": 0.0167, + "step": 19070 + }, + { + "epoch": 16.107263513513512, + "grad_norm": 0.30816152691841125, + "learning_rate": 2.6978333333333332e-05, + "loss": 0.0119, + "step": 19071 + }, + { + "epoch": 16.10810810810811, + "grad_norm": 0.9123364686965942, + "learning_rate": 2.6978e-05, + "loss": 0.0131, + "step": 19072 + }, + { + "epoch": 16.1089527027027, + "grad_norm": 0.31297311186790466, + "learning_rate": 2.6977666666666668e-05, + "loss": 0.0172, + "step": 19073 + }, + { + "epoch": 16.1097972972973, + "grad_norm": 0.2565706670284271, + "learning_rate": 2.6977333333333333e-05, + "loss": 0.0064, + "step": 19074 + }, + { + "epoch": 16.11064189189189, + "grad_norm": 0.20419050753116608, + "learning_rate": 2.6977e-05, + "loss": 0.0126, + "step": 19075 + }, + { + "epoch": 16.111486486486488, + "grad_norm": 0.2533123791217804, + "learning_rate": 2.697666666666667e-05, + "loss": 0.0157, + "step": 19076 + }, + { + "epoch": 16.11233108108108, + "grad_norm": 0.19083242118358612, + "learning_rate": 2.697633333333333e-05, + "loss": 0.0121, + "step": 19077 + }, + { + "epoch": 16.113175675675677, + "grad_norm": 0.2834113538265228, + "learning_rate": 2.6976e-05, + "loss": 0.0136, + "step": 19078 + }, + { + "epoch": 16.11402027027027, + "grad_norm": 0.22238636016845703, + "learning_rate": 2.697566666666667e-05, + "loss": 0.0159, + "step": 19079 + }, + { + "epoch": 16.114864864864863, + "grad_norm": 0.19893598556518555, + "learning_rate": 2.6975333333333332e-05, + "loss": 0.0141, + "step": 19080 + }, + { + "epoch": 16.11570945945946, + "grad_norm": 0.13433071970939636, + "learning_rate": 2.6975000000000002e-05, + "loss": 0.0082, + "step": 19081 + }, + { + "epoch": 16.116554054054053, + "grad_norm": 0.13843873143196106, + "learning_rate": 2.6974666666666668e-05, + "loss": 0.0083, + "step": 19082 + }, + { + "epoch": 16.11739864864865, + "grad_norm": 0.3320448696613312, + "learning_rate": 2.6974333333333334e-05, + "loss": 0.0166, + "step": 19083 + }, + { + "epoch": 16.118243243243242, + "grad_norm": 0.24719582498073578, + "learning_rate": 2.6974e-05, + "loss": 0.0187, + "step": 19084 + }, + { + "epoch": 16.11908783783784, + "grad_norm": 0.2035215198993683, + "learning_rate": 2.697366666666667e-05, + "loss": 0.0141, + "step": 19085 + }, + { + "epoch": 16.11993243243243, + "grad_norm": 0.1913788914680481, + "learning_rate": 2.6973333333333335e-05, + "loss": 0.0106, + "step": 19086 + }, + { + "epoch": 16.120777027027028, + "grad_norm": 0.9191031455993652, + "learning_rate": 2.6973e-05, + "loss": 0.0403, + "step": 19087 + }, + { + "epoch": 16.12162162162162, + "grad_norm": 0.19811229407787323, + "learning_rate": 2.697266666666667e-05, + "loss": 0.0114, + "step": 19088 + }, + { + "epoch": 16.122466216216218, + "grad_norm": 0.22891733050346375, + "learning_rate": 2.6972333333333333e-05, + "loss": 0.0136, + "step": 19089 + }, + { + "epoch": 16.12331081081081, + "grad_norm": 0.5489579439163208, + "learning_rate": 2.6972000000000002e-05, + "loss": 0.0104, + "step": 19090 + }, + { + "epoch": 16.124155405405407, + "grad_norm": 0.5671960115432739, + "learning_rate": 2.6971666666666668e-05, + "loss": 0.0182, + "step": 19091 + }, + { + "epoch": 16.125, + "grad_norm": 0.4319444000720978, + "learning_rate": 2.6971333333333334e-05, + "loss": 0.0091, + "step": 19092 + }, + { + "epoch": 16.125844594594593, + "grad_norm": 0.226178377866745, + "learning_rate": 2.6971e-05, + "loss": 0.0127, + "step": 19093 + }, + { + "epoch": 16.12668918918919, + "grad_norm": 0.6423830986022949, + "learning_rate": 2.6970666666666666e-05, + "loss": 0.0244, + "step": 19094 + }, + { + "epoch": 16.127533783783782, + "grad_norm": 1.538854718208313, + "learning_rate": 2.6970333333333335e-05, + "loss": 0.2441, + "step": 19095 + }, + { + "epoch": 16.12837837837838, + "grad_norm": 0.39314472675323486, + "learning_rate": 2.697e-05, + "loss": 0.0858, + "step": 19096 + }, + { + "epoch": 16.129222972972972, + "grad_norm": 0.32734763622283936, + "learning_rate": 2.6969666666666667e-05, + "loss": 0.0322, + "step": 19097 + }, + { + "epoch": 16.13006756756757, + "grad_norm": 0.32168853282928467, + "learning_rate": 2.6969333333333333e-05, + "loss": 0.025, + "step": 19098 + }, + { + "epoch": 16.13091216216216, + "grad_norm": 0.30248528718948364, + "learning_rate": 2.6969000000000002e-05, + "loss": 0.0196, + "step": 19099 + }, + { + "epoch": 16.131756756756758, + "grad_norm": 0.19757427275180817, + "learning_rate": 2.6968666666666665e-05, + "loss": 0.0194, + "step": 19100 + }, + { + "epoch": 16.13260135135135, + "grad_norm": 0.2510763704776764, + "learning_rate": 2.6968333333333334e-05, + "loss": 0.0224, + "step": 19101 + }, + { + "epoch": 16.133445945945947, + "grad_norm": 0.21594616770744324, + "learning_rate": 2.6968e-05, + "loss": 0.0127, + "step": 19102 + }, + { + "epoch": 16.13429054054054, + "grad_norm": 0.4470423460006714, + "learning_rate": 2.6967666666666666e-05, + "loss": 0.0141, + "step": 19103 + }, + { + "epoch": 16.135135135135137, + "grad_norm": 0.23797763884067535, + "learning_rate": 2.6967333333333335e-05, + "loss": 0.0209, + "step": 19104 + }, + { + "epoch": 16.13597972972973, + "grad_norm": 0.18252140283584595, + "learning_rate": 2.6967e-05, + "loss": 0.0139, + "step": 19105 + }, + { + "epoch": 16.136824324324323, + "grad_norm": 0.29270726442337036, + "learning_rate": 2.6966666666666667e-05, + "loss": 0.0174, + "step": 19106 + }, + { + "epoch": 16.13766891891892, + "grad_norm": 0.27521857619285583, + "learning_rate": 2.6966333333333333e-05, + "loss": 0.0168, + "step": 19107 + }, + { + "epoch": 16.138513513513512, + "grad_norm": 0.34565702080726624, + "learning_rate": 2.6966000000000002e-05, + "loss": 0.019, + "step": 19108 + }, + { + "epoch": 16.13935810810811, + "grad_norm": 0.167056143283844, + "learning_rate": 2.6965666666666665e-05, + "loss": 0.0078, + "step": 19109 + }, + { + "epoch": 16.1402027027027, + "grad_norm": 0.27978676557540894, + "learning_rate": 2.6965333333333334e-05, + "loss": 0.0223, + "step": 19110 + }, + { + "epoch": 16.1410472972973, + "grad_norm": 0.15572193264961243, + "learning_rate": 2.6965000000000004e-05, + "loss": 0.0121, + "step": 19111 + }, + { + "epoch": 16.14189189189189, + "grad_norm": 0.41491562128067017, + "learning_rate": 2.6964666666666666e-05, + "loss": 0.0183, + "step": 19112 + }, + { + "epoch": 16.142736486486488, + "grad_norm": 0.2381228506565094, + "learning_rate": 2.6964333333333335e-05, + "loss": 0.0137, + "step": 19113 + }, + { + "epoch": 16.14358108108108, + "grad_norm": 0.34003567695617676, + "learning_rate": 2.6964e-05, + "loss": 0.0221, + "step": 19114 + }, + { + "epoch": 16.144425675675677, + "grad_norm": 0.2835814356803894, + "learning_rate": 2.6963666666666667e-05, + "loss": 0.0127, + "step": 19115 + }, + { + "epoch": 16.14527027027027, + "grad_norm": 0.30856555700302124, + "learning_rate": 2.6963333333333333e-05, + "loss": 0.0126, + "step": 19116 + }, + { + "epoch": 16.146114864864863, + "grad_norm": 0.2676299810409546, + "learning_rate": 2.6963000000000003e-05, + "loss": 0.0215, + "step": 19117 + }, + { + "epoch": 16.14695945945946, + "grad_norm": 0.2430361658334732, + "learning_rate": 2.6962666666666665e-05, + "loss": 0.0138, + "step": 19118 + }, + { + "epoch": 16.147804054054053, + "grad_norm": 0.19018132984638214, + "learning_rate": 2.6962333333333334e-05, + "loss": 0.0111, + "step": 19119 + }, + { + "epoch": 16.14864864864865, + "grad_norm": 0.2531176507472992, + "learning_rate": 2.6962000000000004e-05, + "loss": 0.0191, + "step": 19120 + }, + { + "epoch": 16.149493243243242, + "grad_norm": 0.6513605713844299, + "learning_rate": 2.6961666666666666e-05, + "loss": 0.0353, + "step": 19121 + }, + { + "epoch": 16.15033783783784, + "grad_norm": 0.10806529968976974, + "learning_rate": 2.6961333333333336e-05, + "loss": 0.0046, + "step": 19122 + }, + { + "epoch": 16.15118243243243, + "grad_norm": 0.23200882971286774, + "learning_rate": 2.6961e-05, + "loss": 0.0139, + "step": 19123 + }, + { + "epoch": 16.152027027027028, + "grad_norm": 0.37022730708122253, + "learning_rate": 2.6960666666666668e-05, + "loss": 0.0131, + "step": 19124 + }, + { + "epoch": 16.15287162162162, + "grad_norm": 0.16901758313179016, + "learning_rate": 2.6960333333333333e-05, + "loss": 0.008, + "step": 19125 + }, + { + "epoch": 16.153716216216218, + "grad_norm": 0.22188180685043335, + "learning_rate": 2.696e-05, + "loss": 0.0121, + "step": 19126 + }, + { + "epoch": 16.15456081081081, + "grad_norm": 0.42149943113327026, + "learning_rate": 2.6959666666666665e-05, + "loss": 0.0175, + "step": 19127 + }, + { + "epoch": 16.155405405405407, + "grad_norm": 0.3336746394634247, + "learning_rate": 2.6959333333333335e-05, + "loss": 0.0115, + "step": 19128 + }, + { + "epoch": 16.15625, + "grad_norm": 0.2294071912765503, + "learning_rate": 2.6959e-05, + "loss": 0.0068, + "step": 19129 + }, + { + "epoch": 16.157094594594593, + "grad_norm": 0.22233176231384277, + "learning_rate": 2.6958666666666667e-05, + "loss": 0.009, + "step": 19130 + }, + { + "epoch": 16.15793918918919, + "grad_norm": 0.26424553990364075, + "learning_rate": 2.6958333333333336e-05, + "loss": 0.0113, + "step": 19131 + }, + { + "epoch": 16.158783783783782, + "grad_norm": 0.4493183493614197, + "learning_rate": 2.6958e-05, + "loss": 0.01, + "step": 19132 + }, + { + "epoch": 16.15962837837838, + "grad_norm": 0.376857191324234, + "learning_rate": 2.6957666666666668e-05, + "loss": 0.0172, + "step": 19133 + }, + { + "epoch": 16.160472972972972, + "grad_norm": 0.3899010121822357, + "learning_rate": 2.6957333333333334e-05, + "loss": 0.0191, + "step": 19134 + }, + { + "epoch": 16.16131756756757, + "grad_norm": 0.5653181076049805, + "learning_rate": 2.6957e-05, + "loss": 0.0198, + "step": 19135 + }, + { + "epoch": 16.16216216216216, + "grad_norm": 0.2555639147758484, + "learning_rate": 2.695666666666667e-05, + "loss": 0.0128, + "step": 19136 + }, + { + "epoch": 16.163006756756758, + "grad_norm": 0.3586733937263489, + "learning_rate": 2.6956333333333335e-05, + "loss": 0.0123, + "step": 19137 + }, + { + "epoch": 16.16385135135135, + "grad_norm": 0.4492076635360718, + "learning_rate": 2.6956e-05, + "loss": 0.013, + "step": 19138 + }, + { + "epoch": 16.164695945945947, + "grad_norm": 0.30334073305130005, + "learning_rate": 2.6955666666666667e-05, + "loss": 0.0123, + "step": 19139 + }, + { + "epoch": 16.16554054054054, + "grad_norm": 0.477026104927063, + "learning_rate": 2.6955333333333336e-05, + "loss": 0.0104, + "step": 19140 + }, + { + "epoch": 16.166385135135137, + "grad_norm": 0.31719788908958435, + "learning_rate": 2.6955e-05, + "loss": 0.0203, + "step": 19141 + }, + { + "epoch": 16.16722972972973, + "grad_norm": 0.6018878221511841, + "learning_rate": 2.6954666666666668e-05, + "loss": 0.0401, + "step": 19142 + }, + { + "epoch": 16.168074324324323, + "grad_norm": 0.7246050238609314, + "learning_rate": 2.6954333333333334e-05, + "loss": 0.029, + "step": 19143 + }, + { + "epoch": 16.16891891891892, + "grad_norm": 0.28777241706848145, + "learning_rate": 2.6954e-05, + "loss": 0.0137, + "step": 19144 + }, + { + "epoch": 16.169763513513512, + "grad_norm": 3.186394214630127, + "learning_rate": 2.695366666666667e-05, + "loss": 0.2496, + "step": 19145 + }, + { + "epoch": 16.17060810810811, + "grad_norm": 0.5446413159370422, + "learning_rate": 2.6953333333333335e-05, + "loss": 0.0812, + "step": 19146 + }, + { + "epoch": 16.1714527027027, + "grad_norm": 0.5015683770179749, + "learning_rate": 2.6953e-05, + "loss": 0.0317, + "step": 19147 + }, + { + "epoch": 16.1722972972973, + "grad_norm": 0.28486886620521545, + "learning_rate": 2.6952666666666667e-05, + "loss": 0.019, + "step": 19148 + }, + { + "epoch": 16.17314189189189, + "grad_norm": 0.370370090007782, + "learning_rate": 2.6952333333333336e-05, + "loss": 0.0246, + "step": 19149 + }, + { + "epoch": 16.173986486486488, + "grad_norm": 0.25164729356765747, + "learning_rate": 2.6952e-05, + "loss": 0.0265, + "step": 19150 + }, + { + "epoch": 16.17483108108108, + "grad_norm": 0.20523248612880707, + "learning_rate": 2.6951666666666668e-05, + "loss": 0.0247, + "step": 19151 + }, + { + "epoch": 16.175675675675677, + "grad_norm": 0.1589275300502777, + "learning_rate": 2.6951333333333334e-05, + "loss": 0.0077, + "step": 19152 + }, + { + "epoch": 16.17652027027027, + "grad_norm": 0.3085521459579468, + "learning_rate": 2.6951e-05, + "loss": 0.0203, + "step": 19153 + }, + { + "epoch": 16.177364864864863, + "grad_norm": 0.297552227973938, + "learning_rate": 2.695066666666667e-05, + "loss": 0.0154, + "step": 19154 + }, + { + "epoch": 16.17820945945946, + "grad_norm": 0.12961605191230774, + "learning_rate": 2.6950333333333332e-05, + "loss": 0.0077, + "step": 19155 + }, + { + "epoch": 16.179054054054053, + "grad_norm": 0.19077792763710022, + "learning_rate": 2.695e-05, + "loss": 0.009, + "step": 19156 + }, + { + "epoch": 16.17989864864865, + "grad_norm": 0.19936823844909668, + "learning_rate": 2.6949666666666667e-05, + "loss": 0.0127, + "step": 19157 + }, + { + "epoch": 16.180743243243242, + "grad_norm": 0.691940188407898, + "learning_rate": 2.6949333333333333e-05, + "loss": 0.0152, + "step": 19158 + }, + { + "epoch": 16.18158783783784, + "grad_norm": 0.26773345470428467, + "learning_rate": 2.6949e-05, + "loss": 0.0177, + "step": 19159 + }, + { + "epoch": 16.18243243243243, + "grad_norm": 0.2133043259382248, + "learning_rate": 2.6948666666666668e-05, + "loss": 0.0096, + "step": 19160 + }, + { + "epoch": 16.183277027027028, + "grad_norm": 0.4262356758117676, + "learning_rate": 2.6948333333333334e-05, + "loss": 0.0214, + "step": 19161 + }, + { + "epoch": 16.18412162162162, + "grad_norm": 0.19365458190441132, + "learning_rate": 2.6948e-05, + "loss": 0.0107, + "step": 19162 + }, + { + "epoch": 16.184966216216218, + "grad_norm": 0.20634201169013977, + "learning_rate": 2.694766666666667e-05, + "loss": 0.0144, + "step": 19163 + }, + { + "epoch": 16.18581081081081, + "grad_norm": 0.17155417799949646, + "learning_rate": 2.6947333333333332e-05, + "loss": 0.0162, + "step": 19164 + }, + { + "epoch": 16.186655405405407, + "grad_norm": 0.1993480622768402, + "learning_rate": 2.6947e-05, + "loss": 0.011, + "step": 19165 + }, + { + "epoch": 16.1875, + "grad_norm": 0.2602609395980835, + "learning_rate": 2.6946666666666667e-05, + "loss": 0.0119, + "step": 19166 + }, + { + "epoch": 16.188344594594593, + "grad_norm": 0.24432651698589325, + "learning_rate": 2.6946333333333333e-05, + "loss": 0.0106, + "step": 19167 + }, + { + "epoch": 16.18918918918919, + "grad_norm": 0.5220144391059875, + "learning_rate": 2.6946e-05, + "loss": 0.0105, + "step": 19168 + }, + { + "epoch": 16.190033783783782, + "grad_norm": 0.31930309534072876, + "learning_rate": 2.694566666666667e-05, + "loss": 0.0231, + "step": 19169 + }, + { + "epoch": 16.19087837837838, + "grad_norm": 0.22465524077415466, + "learning_rate": 2.6945333333333334e-05, + "loss": 0.0133, + "step": 19170 + }, + { + "epoch": 16.191722972972972, + "grad_norm": 0.37013760209083557, + "learning_rate": 2.6945e-05, + "loss": 0.0201, + "step": 19171 + }, + { + "epoch": 16.19256756756757, + "grad_norm": 0.34544241428375244, + "learning_rate": 2.694466666666667e-05, + "loss": 0.0171, + "step": 19172 + }, + { + "epoch": 16.19341216216216, + "grad_norm": 0.1023310199379921, + "learning_rate": 2.6944333333333332e-05, + "loss": 0.0058, + "step": 19173 + }, + { + "epoch": 16.194256756756758, + "grad_norm": 0.32897067070007324, + "learning_rate": 2.6944e-05, + "loss": 0.0153, + "step": 19174 + }, + { + "epoch": 16.19510135135135, + "grad_norm": 0.4281437397003174, + "learning_rate": 2.6943666666666667e-05, + "loss": 0.0181, + "step": 19175 + }, + { + "epoch": 16.195945945945947, + "grad_norm": 0.132788747549057, + "learning_rate": 2.6943333333333333e-05, + "loss": 0.0106, + "step": 19176 + }, + { + "epoch": 16.19679054054054, + "grad_norm": 0.34052932262420654, + "learning_rate": 2.6943e-05, + "loss": 0.0147, + "step": 19177 + }, + { + "epoch": 16.197635135135137, + "grad_norm": 0.20465949177742004, + "learning_rate": 2.694266666666667e-05, + "loss": 0.0088, + "step": 19178 + }, + { + "epoch": 16.19847972972973, + "grad_norm": 0.17185181379318237, + "learning_rate": 2.6942333333333335e-05, + "loss": 0.007, + "step": 19179 + }, + { + "epoch": 16.199324324324323, + "grad_norm": 0.08989804238080978, + "learning_rate": 2.6942e-05, + "loss": 0.0063, + "step": 19180 + }, + { + "epoch": 16.20016891891892, + "grad_norm": 0.36561715602874756, + "learning_rate": 2.694166666666667e-05, + "loss": 0.0145, + "step": 19181 + }, + { + "epoch": 16.201013513513512, + "grad_norm": 0.2557700276374817, + "learning_rate": 2.6941333333333332e-05, + "loss": 0.0207, + "step": 19182 + }, + { + "epoch": 16.20185810810811, + "grad_norm": 0.3398689329624176, + "learning_rate": 2.6941000000000002e-05, + "loss": 0.0161, + "step": 19183 + }, + { + "epoch": 16.2027027027027, + "grad_norm": 0.2641391158103943, + "learning_rate": 2.6940666666666668e-05, + "loss": 0.0121, + "step": 19184 + }, + { + "epoch": 16.2035472972973, + "grad_norm": 0.19183526933193207, + "learning_rate": 2.6940333333333334e-05, + "loss": 0.0112, + "step": 19185 + }, + { + "epoch": 16.20439189189189, + "grad_norm": 0.26184117794036865, + "learning_rate": 2.6940000000000003e-05, + "loss": 0.0092, + "step": 19186 + }, + { + "epoch": 16.205236486486488, + "grad_norm": 0.39102399349212646, + "learning_rate": 2.6939666666666666e-05, + "loss": 0.0186, + "step": 19187 + }, + { + "epoch": 16.20608108108108, + "grad_norm": 0.26112502813339233, + "learning_rate": 2.6939333333333335e-05, + "loss": 0.0119, + "step": 19188 + }, + { + "epoch": 16.206925675675677, + "grad_norm": 0.20279012620449066, + "learning_rate": 2.6939e-05, + "loss": 0.0131, + "step": 19189 + }, + { + "epoch": 16.20777027027027, + "grad_norm": 0.2067413181066513, + "learning_rate": 2.6938666666666667e-05, + "loss": 0.0174, + "step": 19190 + }, + { + "epoch": 16.208614864864863, + "grad_norm": 0.32755544781684875, + "learning_rate": 2.6938333333333333e-05, + "loss": 0.0137, + "step": 19191 + }, + { + "epoch": 16.20945945945946, + "grad_norm": 0.4138539433479309, + "learning_rate": 2.6938000000000002e-05, + "loss": 0.017, + "step": 19192 + }, + { + "epoch": 16.210304054054053, + "grad_norm": 0.5082986950874329, + "learning_rate": 2.6937666666666665e-05, + "loss": 0.0669, + "step": 19193 + }, + { + "epoch": 16.21114864864865, + "grad_norm": 0.3811717927455902, + "learning_rate": 2.6937333333333334e-05, + "loss": 0.0147, + "step": 19194 + }, + { + "epoch": 16.211993243243242, + "grad_norm": 1.0181859731674194, + "learning_rate": 2.6937000000000003e-05, + "loss": 0.1749, + "step": 19195 + }, + { + "epoch": 16.21283783783784, + "grad_norm": 0.4132888913154602, + "learning_rate": 2.6936666666666666e-05, + "loss": 0.0739, + "step": 19196 + }, + { + "epoch": 16.21368243243243, + "grad_norm": 0.26524433493614197, + "learning_rate": 2.6936333333333335e-05, + "loss": 0.0357, + "step": 19197 + }, + { + "epoch": 16.214527027027028, + "grad_norm": 0.26575979590415955, + "learning_rate": 2.6936e-05, + "loss": 0.0196, + "step": 19198 + }, + { + "epoch": 16.21537162162162, + "grad_norm": 0.2339753657579422, + "learning_rate": 2.6935666666666667e-05, + "loss": 0.0191, + "step": 19199 + }, + { + "epoch": 16.216216216216218, + "grad_norm": 0.24071872234344482, + "learning_rate": 2.6935333333333333e-05, + "loss": 0.0162, + "step": 19200 + }, + { + "epoch": 16.21706081081081, + "grad_norm": 0.1988564431667328, + "learning_rate": 2.6935000000000002e-05, + "loss": 0.0135, + "step": 19201 + }, + { + "epoch": 16.217905405405407, + "grad_norm": 0.36906540393829346, + "learning_rate": 2.6934666666666665e-05, + "loss": 0.027, + "step": 19202 + }, + { + "epoch": 16.21875, + "grad_norm": 0.2901764214038849, + "learning_rate": 2.6934333333333334e-05, + "loss": 0.0167, + "step": 19203 + }, + { + "epoch": 16.219594594594593, + "grad_norm": 0.18933576345443726, + "learning_rate": 2.6934000000000003e-05, + "loss": 0.0088, + "step": 19204 + }, + { + "epoch": 16.22043918918919, + "grad_norm": 0.1672268658876419, + "learning_rate": 2.6933666666666666e-05, + "loss": 0.0131, + "step": 19205 + }, + { + "epoch": 16.221283783783782, + "grad_norm": 0.2796870470046997, + "learning_rate": 2.6933333333333335e-05, + "loss": 0.0081, + "step": 19206 + }, + { + "epoch": 16.22212837837838, + "grad_norm": 0.16404040157794952, + "learning_rate": 2.6933e-05, + "loss": 0.0097, + "step": 19207 + }, + { + "epoch": 16.222972972972972, + "grad_norm": 0.3213350772857666, + "learning_rate": 2.6932666666666667e-05, + "loss": 0.0125, + "step": 19208 + }, + { + "epoch": 16.22381756756757, + "grad_norm": 0.25741979479789734, + "learning_rate": 2.6932333333333333e-05, + "loss": 0.0135, + "step": 19209 + }, + { + "epoch": 16.22466216216216, + "grad_norm": 0.2548295855522156, + "learning_rate": 2.6932000000000002e-05, + "loss": 0.0123, + "step": 19210 + }, + { + "epoch": 16.225506756756758, + "grad_norm": 0.17718642950057983, + "learning_rate": 2.6931666666666668e-05, + "loss": 0.0084, + "step": 19211 + }, + { + "epoch": 16.22635135135135, + "grad_norm": 0.32097622752189636, + "learning_rate": 2.6931333333333334e-05, + "loss": 0.0124, + "step": 19212 + }, + { + "epoch": 16.227195945945947, + "grad_norm": 0.2069326937198639, + "learning_rate": 2.6931000000000004e-05, + "loss": 0.0103, + "step": 19213 + }, + { + "epoch": 16.22804054054054, + "grad_norm": 0.2564847767353058, + "learning_rate": 2.6930666666666666e-05, + "loss": 0.0117, + "step": 19214 + }, + { + "epoch": 16.228885135135137, + "grad_norm": 0.1653338223695755, + "learning_rate": 2.6930333333333335e-05, + "loss": 0.0104, + "step": 19215 + }, + { + "epoch": 16.22972972972973, + "grad_norm": 0.28162506222724915, + "learning_rate": 2.6929999999999998e-05, + "loss": 0.0126, + "step": 19216 + }, + { + "epoch": 16.230574324324323, + "grad_norm": 0.3230029046535492, + "learning_rate": 2.6929666666666667e-05, + "loss": 0.0142, + "step": 19217 + }, + { + "epoch": 16.23141891891892, + "grad_norm": 0.20186719298362732, + "learning_rate": 2.6929333333333333e-05, + "loss": 0.0067, + "step": 19218 + }, + { + "epoch": 16.232263513513512, + "grad_norm": 0.22126170992851257, + "learning_rate": 2.6929e-05, + "loss": 0.0086, + "step": 19219 + }, + { + "epoch": 16.23310810810811, + "grad_norm": 0.24802033603191376, + "learning_rate": 2.692866666666667e-05, + "loss": 0.0093, + "step": 19220 + }, + { + "epoch": 16.2339527027027, + "grad_norm": 0.19494207203388214, + "learning_rate": 2.6928333333333334e-05, + "loss": 0.0066, + "step": 19221 + }, + { + "epoch": 16.2347972972973, + "grad_norm": 0.2367819994688034, + "learning_rate": 2.6928e-05, + "loss": 0.021, + "step": 19222 + }, + { + "epoch": 16.23564189189189, + "grad_norm": 0.24939361214637756, + "learning_rate": 2.6927666666666666e-05, + "loss": 0.0152, + "step": 19223 + }, + { + "epoch": 16.236486486486488, + "grad_norm": 0.12467286735773087, + "learning_rate": 2.6927333333333336e-05, + "loss": 0.0092, + "step": 19224 + }, + { + "epoch": 16.23733108108108, + "grad_norm": 0.24932211637496948, + "learning_rate": 2.6926999999999998e-05, + "loss": 0.0129, + "step": 19225 + }, + { + "epoch": 16.238175675675677, + "grad_norm": 0.6839396357536316, + "learning_rate": 2.6926666666666667e-05, + "loss": 0.0215, + "step": 19226 + }, + { + "epoch": 16.23902027027027, + "grad_norm": 0.10713373124599457, + "learning_rate": 2.6926333333333333e-05, + "loss": 0.0069, + "step": 19227 + }, + { + "epoch": 16.239864864864863, + "grad_norm": 0.3355658948421478, + "learning_rate": 2.6926e-05, + "loss": 0.0136, + "step": 19228 + }, + { + "epoch": 16.24070945945946, + "grad_norm": 0.1621003895998001, + "learning_rate": 2.692566666666667e-05, + "loss": 0.0082, + "step": 19229 + }, + { + "epoch": 16.241554054054053, + "grad_norm": 0.35414767265319824, + "learning_rate": 2.6925333333333335e-05, + "loss": 0.0189, + "step": 19230 + }, + { + "epoch": 16.24239864864865, + "grad_norm": 0.16585958003997803, + "learning_rate": 2.6925e-05, + "loss": 0.0082, + "step": 19231 + }, + { + "epoch": 16.243243243243242, + "grad_norm": 0.17285865545272827, + "learning_rate": 2.6924666666666666e-05, + "loss": 0.0119, + "step": 19232 + }, + { + "epoch": 16.24408783783784, + "grad_norm": 0.21189995110034943, + "learning_rate": 2.6924333333333336e-05, + "loss": 0.014, + "step": 19233 + }, + { + "epoch": 16.24493243243243, + "grad_norm": 0.1941458135843277, + "learning_rate": 2.6924e-05, + "loss": 0.01, + "step": 19234 + }, + { + "epoch": 16.245777027027028, + "grad_norm": 0.2597445547580719, + "learning_rate": 2.6923666666666668e-05, + "loss": 0.0136, + "step": 19235 + }, + { + "epoch": 16.24662162162162, + "grad_norm": 0.31574317812919617, + "learning_rate": 2.6923333333333337e-05, + "loss": 0.0121, + "step": 19236 + }, + { + "epoch": 16.247466216216218, + "grad_norm": 0.384294718503952, + "learning_rate": 2.6923e-05, + "loss": 0.0156, + "step": 19237 + }, + { + "epoch": 16.24831081081081, + "grad_norm": 0.3037026524543762, + "learning_rate": 2.692266666666667e-05, + "loss": 0.009, + "step": 19238 + }, + { + "epoch": 16.249155405405407, + "grad_norm": 0.25124451518058777, + "learning_rate": 2.6922333333333335e-05, + "loss": 0.0101, + "step": 19239 + }, + { + "epoch": 16.25, + "grad_norm": 0.47419416904449463, + "learning_rate": 2.6922e-05, + "loss": 0.0228, + "step": 19240 + }, + { + "epoch": 16.250844594594593, + "grad_norm": 0.7632764577865601, + "learning_rate": 2.6921666666666667e-05, + "loss": 0.014, + "step": 19241 + }, + { + "epoch": 16.25168918918919, + "grad_norm": 0.28942981362342834, + "learning_rate": 2.6921333333333336e-05, + "loss": 0.0124, + "step": 19242 + }, + { + "epoch": 16.252533783783782, + "grad_norm": 0.28502869606018066, + "learning_rate": 2.6921e-05, + "loss": 0.0114, + "step": 19243 + }, + { + "epoch": 16.25337837837838, + "grad_norm": 0.7388613224029541, + "learning_rate": 2.6920666666666668e-05, + "loss": 0.0177, + "step": 19244 + }, + { + "epoch": 16.254222972972972, + "grad_norm": 0.7239223122596741, + "learning_rate": 2.6920333333333334e-05, + "loss": 0.1375, + "step": 19245 + }, + { + "epoch": 16.25506756756757, + "grad_norm": 0.33885037899017334, + "learning_rate": 2.692e-05, + "loss": 0.0575, + "step": 19246 + }, + { + "epoch": 16.25591216216216, + "grad_norm": 0.2795621454715729, + "learning_rate": 2.691966666666667e-05, + "loss": 0.0322, + "step": 19247 + }, + { + "epoch": 16.256756756756758, + "grad_norm": 0.21968206763267517, + "learning_rate": 2.691933333333333e-05, + "loss": 0.0264, + "step": 19248 + }, + { + "epoch": 16.25760135135135, + "grad_norm": 0.2621563673019409, + "learning_rate": 2.6919e-05, + "loss": 0.0133, + "step": 19249 + }, + { + "epoch": 16.258445945945947, + "grad_norm": 0.23282714188098907, + "learning_rate": 2.6918666666666667e-05, + "loss": 0.0178, + "step": 19250 + }, + { + "epoch": 16.25929054054054, + "grad_norm": 0.4400377869606018, + "learning_rate": 2.6918333333333333e-05, + "loss": 0.02, + "step": 19251 + }, + { + "epoch": 16.260135135135137, + "grad_norm": 0.3522021770477295, + "learning_rate": 2.6918e-05, + "loss": 0.0237, + "step": 19252 + }, + { + "epoch": 16.26097972972973, + "grad_norm": 0.13204549252986908, + "learning_rate": 2.6917666666666668e-05, + "loss": 0.0084, + "step": 19253 + }, + { + "epoch": 16.261824324324323, + "grad_norm": 0.2614590525627136, + "learning_rate": 2.6917333333333334e-05, + "loss": 0.0215, + "step": 19254 + }, + { + "epoch": 16.26266891891892, + "grad_norm": 0.15032710134983063, + "learning_rate": 2.6917e-05, + "loss": 0.0098, + "step": 19255 + }, + { + "epoch": 16.263513513513512, + "grad_norm": 0.20476244390010834, + "learning_rate": 2.691666666666667e-05, + "loss": 0.0137, + "step": 19256 + }, + { + "epoch": 16.26435810810811, + "grad_norm": 0.24492593109607697, + "learning_rate": 2.6916333333333332e-05, + "loss": 0.0148, + "step": 19257 + }, + { + "epoch": 16.2652027027027, + "grad_norm": 0.6205217838287354, + "learning_rate": 2.6916e-05, + "loss": 0.0118, + "step": 19258 + }, + { + "epoch": 16.2660472972973, + "grad_norm": 0.30976808071136475, + "learning_rate": 2.6915666666666667e-05, + "loss": 0.0173, + "step": 19259 + }, + { + "epoch": 16.26689189189189, + "grad_norm": 0.1306973099708557, + "learning_rate": 2.6915333333333333e-05, + "loss": 0.0077, + "step": 19260 + }, + { + "epoch": 16.267736486486488, + "grad_norm": 0.27100372314453125, + "learning_rate": 2.6915000000000002e-05, + "loss": 0.0114, + "step": 19261 + }, + { + "epoch": 16.26858108108108, + "grad_norm": 0.5574339628219604, + "learning_rate": 2.6914666666666668e-05, + "loss": 0.0314, + "step": 19262 + }, + { + "epoch": 16.269425675675677, + "grad_norm": 0.21133536100387573, + "learning_rate": 2.6914333333333334e-05, + "loss": 0.0164, + "step": 19263 + }, + { + "epoch": 16.27027027027027, + "grad_norm": 0.24457591772079468, + "learning_rate": 2.6914e-05, + "loss": 0.019, + "step": 19264 + }, + { + "epoch": 16.271114864864863, + "grad_norm": 0.25413015484809875, + "learning_rate": 2.691366666666667e-05, + "loss": 0.0082, + "step": 19265 + }, + { + "epoch": 16.27195945945946, + "grad_norm": 0.32643869519233704, + "learning_rate": 2.6913333333333332e-05, + "loss": 0.0152, + "step": 19266 + }, + { + "epoch": 16.272804054054053, + "grad_norm": 0.19672857224941254, + "learning_rate": 2.6913e-05, + "loss": 0.0097, + "step": 19267 + }, + { + "epoch": 16.27364864864865, + "grad_norm": 0.23860901594161987, + "learning_rate": 2.6912666666666667e-05, + "loss": 0.0129, + "step": 19268 + }, + { + "epoch": 16.274493243243242, + "grad_norm": 0.2889743447303772, + "learning_rate": 2.6912333333333333e-05, + "loss": 0.0139, + "step": 19269 + }, + { + "epoch": 16.27533783783784, + "grad_norm": 0.25420913100242615, + "learning_rate": 2.6912000000000003e-05, + "loss": 0.019, + "step": 19270 + }, + { + "epoch": 16.27618243243243, + "grad_norm": 0.19389063119888306, + "learning_rate": 2.691166666666667e-05, + "loss": 0.0097, + "step": 19271 + }, + { + "epoch": 16.277027027027028, + "grad_norm": 0.15969951450824738, + "learning_rate": 2.6911333333333334e-05, + "loss": 0.0106, + "step": 19272 + }, + { + "epoch": 16.27787162162162, + "grad_norm": 0.6504871845245361, + "learning_rate": 2.6911e-05, + "loss": 0.011, + "step": 19273 + }, + { + "epoch": 16.278716216216218, + "grad_norm": 0.10924866795539856, + "learning_rate": 2.691066666666667e-05, + "loss": 0.0078, + "step": 19274 + }, + { + "epoch": 16.27956081081081, + "grad_norm": 0.18045161664485931, + "learning_rate": 2.6910333333333332e-05, + "loss": 0.0124, + "step": 19275 + }, + { + "epoch": 16.280405405405407, + "grad_norm": 0.24381133913993835, + "learning_rate": 2.691e-05, + "loss": 0.0161, + "step": 19276 + }, + { + "epoch": 16.28125, + "grad_norm": 0.24846595525741577, + "learning_rate": 2.6909666666666664e-05, + "loss": 0.0136, + "step": 19277 + }, + { + "epoch": 16.282094594594593, + "grad_norm": 0.5856099724769592, + "learning_rate": 2.6909333333333333e-05, + "loss": 0.0167, + "step": 19278 + }, + { + "epoch": 16.28293918918919, + "grad_norm": 0.19652500748634338, + "learning_rate": 2.6909000000000003e-05, + "loss": 0.0128, + "step": 19279 + }, + { + "epoch": 16.283783783783782, + "grad_norm": 0.6241533160209656, + "learning_rate": 2.6908666666666665e-05, + "loss": 0.0133, + "step": 19280 + }, + { + "epoch": 16.28462837837838, + "grad_norm": 0.2112320363521576, + "learning_rate": 2.6908333333333335e-05, + "loss": 0.0112, + "step": 19281 + }, + { + "epoch": 16.285472972972972, + "grad_norm": 0.21507865190505981, + "learning_rate": 2.6908e-05, + "loss": 0.0099, + "step": 19282 + }, + { + "epoch": 16.28631756756757, + "grad_norm": 0.2598866820335388, + "learning_rate": 2.6907666666666666e-05, + "loss": 0.013, + "step": 19283 + }, + { + "epoch": 16.28716216216216, + "grad_norm": 0.1667809784412384, + "learning_rate": 2.6907333333333332e-05, + "loss": 0.0154, + "step": 19284 + }, + { + "epoch": 16.288006756756758, + "grad_norm": 1.494378685951233, + "learning_rate": 2.6907000000000002e-05, + "loss": 0.0231, + "step": 19285 + }, + { + "epoch": 16.28885135135135, + "grad_norm": 0.6447586417198181, + "learning_rate": 2.6906666666666668e-05, + "loss": 0.0152, + "step": 19286 + }, + { + "epoch": 16.289695945945947, + "grad_norm": 0.2856926918029785, + "learning_rate": 2.6906333333333334e-05, + "loss": 0.0102, + "step": 19287 + }, + { + "epoch": 16.29054054054054, + "grad_norm": 0.24896077811717987, + "learning_rate": 2.6906000000000003e-05, + "loss": 0.0168, + "step": 19288 + }, + { + "epoch": 16.291385135135137, + "grad_norm": 0.33246949315071106, + "learning_rate": 2.6905666666666665e-05, + "loss": 0.0185, + "step": 19289 + }, + { + "epoch": 16.29222972972973, + "grad_norm": 0.2710575461387634, + "learning_rate": 2.6905333333333335e-05, + "loss": 0.0112, + "step": 19290 + }, + { + "epoch": 16.293074324324323, + "grad_norm": 0.5936405062675476, + "learning_rate": 2.6905e-05, + "loss": 0.016, + "step": 19291 + }, + { + "epoch": 16.29391891891892, + "grad_norm": 0.2835101783275604, + "learning_rate": 2.6904666666666667e-05, + "loss": 0.0075, + "step": 19292 + }, + { + "epoch": 16.294763513513512, + "grad_norm": 0.686079204082489, + "learning_rate": 2.6904333333333333e-05, + "loss": 0.0211, + "step": 19293 + }, + { + "epoch": 16.29560810810811, + "grad_norm": 0.3802734613418579, + "learning_rate": 2.6904000000000002e-05, + "loss": 0.0221, + "step": 19294 + }, + { + "epoch": 16.2964527027027, + "grad_norm": 0.7445623278617859, + "learning_rate": 2.6903666666666668e-05, + "loss": 0.1497, + "step": 19295 + }, + { + "epoch": 16.2972972972973, + "grad_norm": 0.45330801606178284, + "learning_rate": 2.6903333333333334e-05, + "loss": 0.1026, + "step": 19296 + }, + { + "epoch": 16.29814189189189, + "grad_norm": 0.3336142599582672, + "learning_rate": 2.6903000000000003e-05, + "loss": 0.0499, + "step": 19297 + }, + { + "epoch": 16.298986486486488, + "grad_norm": 0.4085596799850464, + "learning_rate": 2.6902666666666666e-05, + "loss": 0.0294, + "step": 19298 + }, + { + "epoch": 16.29983108108108, + "grad_norm": 0.27651262283325195, + "learning_rate": 2.6902333333333335e-05, + "loss": 0.0257, + "step": 19299 + }, + { + "epoch": 16.300675675675677, + "grad_norm": 0.2786552608013153, + "learning_rate": 2.6902e-05, + "loss": 0.0194, + "step": 19300 + }, + { + "epoch": 16.30152027027027, + "grad_norm": 0.2047901302576065, + "learning_rate": 2.6901666666666667e-05, + "loss": 0.0154, + "step": 19301 + }, + { + "epoch": 16.302364864864863, + "grad_norm": 0.4022664725780487, + "learning_rate": 2.6901333333333333e-05, + "loss": 0.0251, + "step": 19302 + }, + { + "epoch": 16.30320945945946, + "grad_norm": 0.3998076021671295, + "learning_rate": 2.6901000000000002e-05, + "loss": 0.0172, + "step": 19303 + }, + { + "epoch": 16.304054054054053, + "grad_norm": 0.20187510550022125, + "learning_rate": 2.6900666666666668e-05, + "loss": 0.012, + "step": 19304 + }, + { + "epoch": 16.30489864864865, + "grad_norm": 0.21852971613407135, + "learning_rate": 2.6900333333333334e-05, + "loss": 0.0118, + "step": 19305 + }, + { + "epoch": 16.305743243243242, + "grad_norm": 0.2071145474910736, + "learning_rate": 2.69e-05, + "loss": 0.011, + "step": 19306 + }, + { + "epoch": 16.30658783783784, + "grad_norm": 0.21817587316036224, + "learning_rate": 2.6899666666666666e-05, + "loss": 0.0112, + "step": 19307 + }, + { + "epoch": 16.30743243243243, + "grad_norm": 0.11432158201932907, + "learning_rate": 2.6899333333333335e-05, + "loss": 0.0104, + "step": 19308 + }, + { + "epoch": 16.308277027027028, + "grad_norm": 0.23655878007411957, + "learning_rate": 2.6898999999999998e-05, + "loss": 0.0157, + "step": 19309 + }, + { + "epoch": 16.30912162162162, + "grad_norm": 0.2921530604362488, + "learning_rate": 2.6898666666666667e-05, + "loss": 0.0225, + "step": 19310 + }, + { + "epoch": 16.309966216216218, + "grad_norm": 0.2628394067287445, + "learning_rate": 2.6898333333333336e-05, + "loss": 0.021, + "step": 19311 + }, + { + "epoch": 16.31081081081081, + "grad_norm": 0.22558185458183289, + "learning_rate": 2.6898e-05, + "loss": 0.0133, + "step": 19312 + }, + { + "epoch": 16.311655405405407, + "grad_norm": 0.21511277556419373, + "learning_rate": 2.6897666666666668e-05, + "loss": 0.007, + "step": 19313 + }, + { + "epoch": 16.3125, + "grad_norm": 0.10944570600986481, + "learning_rate": 2.6897333333333334e-05, + "loss": 0.0051, + "step": 19314 + }, + { + "epoch": 16.313344594594593, + "grad_norm": 0.25951194763183594, + "learning_rate": 2.6897e-05, + "loss": 0.0171, + "step": 19315 + }, + { + "epoch": 16.31418918918919, + "grad_norm": 0.2997587025165558, + "learning_rate": 2.6896666666666666e-05, + "loss": 0.0124, + "step": 19316 + }, + { + "epoch": 16.315033783783782, + "grad_norm": 0.14906078577041626, + "learning_rate": 2.6896333333333335e-05, + "loss": 0.0109, + "step": 19317 + }, + { + "epoch": 16.31587837837838, + "grad_norm": 0.22185136377811432, + "learning_rate": 2.6895999999999998e-05, + "loss": 0.0117, + "step": 19318 + }, + { + "epoch": 16.316722972972972, + "grad_norm": 0.1369655430316925, + "learning_rate": 2.6895666666666667e-05, + "loss": 0.0087, + "step": 19319 + }, + { + "epoch": 16.31756756756757, + "grad_norm": 0.6764651536941528, + "learning_rate": 2.6895333333333337e-05, + "loss": 0.0153, + "step": 19320 + }, + { + "epoch": 16.31841216216216, + "grad_norm": 0.2625580430030823, + "learning_rate": 2.6895e-05, + "loss": 0.0196, + "step": 19321 + }, + { + "epoch": 16.319256756756758, + "grad_norm": 0.1942031979560852, + "learning_rate": 2.689466666666667e-05, + "loss": 0.0093, + "step": 19322 + }, + { + "epoch": 16.32010135135135, + "grad_norm": 0.46661001443862915, + "learning_rate": 2.6894333333333334e-05, + "loss": 0.025, + "step": 19323 + }, + { + "epoch": 16.320945945945947, + "grad_norm": 0.4055994749069214, + "learning_rate": 2.6894e-05, + "loss": 0.0152, + "step": 19324 + }, + { + "epoch": 16.32179054054054, + "grad_norm": 0.25765830278396606, + "learning_rate": 2.6893666666666666e-05, + "loss": 0.0149, + "step": 19325 + }, + { + "epoch": 16.322635135135137, + "grad_norm": 0.16439469158649445, + "learning_rate": 2.6893333333333336e-05, + "loss": 0.012, + "step": 19326 + }, + { + "epoch": 16.32347972972973, + "grad_norm": 0.15747766196727753, + "learning_rate": 2.6892999999999998e-05, + "loss": 0.0087, + "step": 19327 + }, + { + "epoch": 16.324324324324323, + "grad_norm": 0.27452006936073303, + "learning_rate": 2.6892666666666667e-05, + "loss": 0.02, + "step": 19328 + }, + { + "epoch": 16.32516891891892, + "grad_norm": 0.27794042229652405, + "learning_rate": 2.6892333333333337e-05, + "loss": 0.012, + "step": 19329 + }, + { + "epoch": 16.326013513513512, + "grad_norm": 0.1780550330877304, + "learning_rate": 2.6892e-05, + "loss": 0.009, + "step": 19330 + }, + { + "epoch": 16.32685810810811, + "grad_norm": 0.26798391342163086, + "learning_rate": 2.689166666666667e-05, + "loss": 0.0122, + "step": 19331 + }, + { + "epoch": 16.3277027027027, + "grad_norm": 0.5226688981056213, + "learning_rate": 2.6891333333333335e-05, + "loss": 0.0617, + "step": 19332 + }, + { + "epoch": 16.3285472972973, + "grad_norm": 0.17452074587345123, + "learning_rate": 2.6891e-05, + "loss": 0.0123, + "step": 19333 + }, + { + "epoch": 16.32939189189189, + "grad_norm": 0.18145841360092163, + "learning_rate": 2.6890666666666666e-05, + "loss": 0.0103, + "step": 19334 + }, + { + "epoch": 16.330236486486488, + "grad_norm": 0.2296178936958313, + "learning_rate": 2.6890333333333336e-05, + "loss": 0.0101, + "step": 19335 + }, + { + "epoch": 16.33108108108108, + "grad_norm": 1.1827443838119507, + "learning_rate": 2.6890000000000002e-05, + "loss": 0.0099, + "step": 19336 + }, + { + "epoch": 16.331925675675677, + "grad_norm": 0.21177664399147034, + "learning_rate": 2.6889666666666668e-05, + "loss": 0.013, + "step": 19337 + }, + { + "epoch": 16.33277027027027, + "grad_norm": 0.3388921320438385, + "learning_rate": 2.6889333333333334e-05, + "loss": 0.0125, + "step": 19338 + }, + { + "epoch": 16.333614864864863, + "grad_norm": 0.6706394553184509, + "learning_rate": 2.6889e-05, + "loss": 0.0219, + "step": 19339 + }, + { + "epoch": 16.33445945945946, + "grad_norm": 0.184416264295578, + "learning_rate": 2.688866666666667e-05, + "loss": 0.0096, + "step": 19340 + }, + { + "epoch": 16.335304054054053, + "grad_norm": 0.1284676343202591, + "learning_rate": 2.688833333333333e-05, + "loss": 0.0063, + "step": 19341 + }, + { + "epoch": 16.33614864864865, + "grad_norm": 0.7445410490036011, + "learning_rate": 2.6888e-05, + "loss": 0.0255, + "step": 19342 + }, + { + "epoch": 16.336993243243242, + "grad_norm": 0.3275003433227539, + "learning_rate": 2.6887666666666667e-05, + "loss": 0.0155, + "step": 19343 + }, + { + "epoch": 16.33783783783784, + "grad_norm": 0.8680418729782104, + "learning_rate": 2.6887333333333333e-05, + "loss": 0.0317, + "step": 19344 + }, + { + "epoch": 16.33868243243243, + "grad_norm": 0.6917017698287964, + "learning_rate": 2.6887000000000002e-05, + "loss": 0.12, + "step": 19345 + }, + { + "epoch": 16.339527027027028, + "grad_norm": 0.3661770522594452, + "learning_rate": 2.6886666666666668e-05, + "loss": 0.0876, + "step": 19346 + }, + { + "epoch": 16.34037162162162, + "grad_norm": 0.29823699593544006, + "learning_rate": 2.6886333333333334e-05, + "loss": 0.037, + "step": 19347 + }, + { + "epoch": 16.341216216216218, + "grad_norm": 0.5278337597846985, + "learning_rate": 2.6886e-05, + "loss": 0.0269, + "step": 19348 + }, + { + "epoch": 16.34206081081081, + "grad_norm": 0.3284691274166107, + "learning_rate": 2.688566666666667e-05, + "loss": 0.0303, + "step": 19349 + }, + { + "epoch": 16.342905405405407, + "grad_norm": 0.1514424979686737, + "learning_rate": 2.688533333333333e-05, + "loss": 0.0131, + "step": 19350 + }, + { + "epoch": 16.34375, + "grad_norm": 0.3846130967140198, + "learning_rate": 2.6885e-05, + "loss": 0.0153, + "step": 19351 + }, + { + "epoch": 16.344594594594593, + "grad_norm": 0.21822555363178253, + "learning_rate": 2.6884666666666667e-05, + "loss": 0.014, + "step": 19352 + }, + { + "epoch": 16.34543918918919, + "grad_norm": 0.22350727021694183, + "learning_rate": 2.6884333333333333e-05, + "loss": 0.0116, + "step": 19353 + }, + { + "epoch": 16.346283783783782, + "grad_norm": 0.3573997914791107, + "learning_rate": 2.6884000000000002e-05, + "loss": 0.0218, + "step": 19354 + }, + { + "epoch": 16.34712837837838, + "grad_norm": 0.44302132725715637, + "learning_rate": 2.6883666666666668e-05, + "loss": 0.0121, + "step": 19355 + }, + { + "epoch": 16.347972972972972, + "grad_norm": 0.24883703887462616, + "learning_rate": 2.6883333333333334e-05, + "loss": 0.0126, + "step": 19356 + }, + { + "epoch": 16.34881756756757, + "grad_norm": 0.3662465810775757, + "learning_rate": 2.6883e-05, + "loss": 0.0282, + "step": 19357 + }, + { + "epoch": 16.34966216216216, + "grad_norm": 0.19458021223545074, + "learning_rate": 2.688266666666667e-05, + "loss": 0.0112, + "step": 19358 + }, + { + "epoch": 16.350506756756758, + "grad_norm": 0.9535660147666931, + "learning_rate": 2.6882333333333332e-05, + "loss": 0.0233, + "step": 19359 + }, + { + "epoch": 16.35135135135135, + "grad_norm": 0.2435533106327057, + "learning_rate": 2.6882e-05, + "loss": 0.0158, + "step": 19360 + }, + { + "epoch": 16.352195945945947, + "grad_norm": 0.2015494406223297, + "learning_rate": 2.688166666666667e-05, + "loss": 0.0104, + "step": 19361 + }, + { + "epoch": 16.35304054054054, + "grad_norm": 0.17094503343105316, + "learning_rate": 2.6881333333333333e-05, + "loss": 0.0121, + "step": 19362 + }, + { + "epoch": 16.353885135135137, + "grad_norm": 0.4073604941368103, + "learning_rate": 2.6881000000000002e-05, + "loss": 0.023, + "step": 19363 + }, + { + "epoch": 16.35472972972973, + "grad_norm": 0.1864444762468338, + "learning_rate": 2.6880666666666668e-05, + "loss": 0.0132, + "step": 19364 + }, + { + "epoch": 16.355574324324323, + "grad_norm": 0.3685154616832733, + "learning_rate": 2.6880333333333334e-05, + "loss": 0.0121, + "step": 19365 + }, + { + "epoch": 16.35641891891892, + "grad_norm": 0.2816293239593506, + "learning_rate": 2.688e-05, + "loss": 0.0167, + "step": 19366 + }, + { + "epoch": 16.357263513513512, + "grad_norm": 0.2487460970878601, + "learning_rate": 2.6879666666666666e-05, + "loss": 0.0151, + "step": 19367 + }, + { + "epoch": 16.35810810810811, + "grad_norm": 0.13037356734275818, + "learning_rate": 2.6879333333333332e-05, + "loss": 0.0083, + "step": 19368 + }, + { + "epoch": 16.3589527027027, + "grad_norm": 0.3092624247074127, + "learning_rate": 2.6879e-05, + "loss": 0.0138, + "step": 19369 + }, + { + "epoch": 16.3597972972973, + "grad_norm": 0.289609432220459, + "learning_rate": 2.6878666666666667e-05, + "loss": 0.0147, + "step": 19370 + }, + { + "epoch": 16.36064189189189, + "grad_norm": 0.2894192934036255, + "learning_rate": 2.6878333333333333e-05, + "loss": 0.0162, + "step": 19371 + }, + { + "epoch": 16.361486486486488, + "grad_norm": 0.2763635814189911, + "learning_rate": 2.6878000000000003e-05, + "loss": 0.0083, + "step": 19372 + }, + { + "epoch": 16.36233108108108, + "grad_norm": 0.3427894711494446, + "learning_rate": 2.6877666666666665e-05, + "loss": 0.0156, + "step": 19373 + }, + { + "epoch": 16.363175675675677, + "grad_norm": 0.2523401379585266, + "learning_rate": 2.6877333333333334e-05, + "loss": 0.0139, + "step": 19374 + }, + { + "epoch": 16.36402027027027, + "grad_norm": 0.22448332607746124, + "learning_rate": 2.6877e-05, + "loss": 0.0155, + "step": 19375 + }, + { + "epoch": 16.364864864864863, + "grad_norm": 0.3823239803314209, + "learning_rate": 2.6876666666666666e-05, + "loss": 0.0124, + "step": 19376 + }, + { + "epoch": 16.36570945945946, + "grad_norm": 0.2978352904319763, + "learning_rate": 2.6876333333333332e-05, + "loss": 0.0204, + "step": 19377 + }, + { + "epoch": 16.366554054054053, + "grad_norm": 0.18651632964611053, + "learning_rate": 2.6876e-05, + "loss": 0.0115, + "step": 19378 + }, + { + "epoch": 16.36739864864865, + "grad_norm": 0.2120492160320282, + "learning_rate": 2.6875666666666667e-05, + "loss": 0.0127, + "step": 19379 + }, + { + "epoch": 16.368243243243242, + "grad_norm": 0.362397164106369, + "learning_rate": 2.6875333333333333e-05, + "loss": 0.0237, + "step": 19380 + }, + { + "epoch": 16.36908783783784, + "grad_norm": 0.4016982614994049, + "learning_rate": 2.6875000000000003e-05, + "loss": 0.0193, + "step": 19381 + }, + { + "epoch": 16.36993243243243, + "grad_norm": 0.2595267593860626, + "learning_rate": 2.6874666666666665e-05, + "loss": 0.0157, + "step": 19382 + }, + { + "epoch": 16.370777027027028, + "grad_norm": 0.2556735873222351, + "learning_rate": 2.6874333333333335e-05, + "loss": 0.0111, + "step": 19383 + }, + { + "epoch": 16.37162162162162, + "grad_norm": 0.20107401907444, + "learning_rate": 2.6874e-05, + "loss": 0.0104, + "step": 19384 + }, + { + "epoch": 16.372466216216218, + "grad_norm": 1.1447324752807617, + "learning_rate": 2.6873666666666666e-05, + "loss": 0.0273, + "step": 19385 + }, + { + "epoch": 16.37331081081081, + "grad_norm": 0.14742253720760345, + "learning_rate": 2.6873333333333336e-05, + "loss": 0.0065, + "step": 19386 + }, + { + "epoch": 16.374155405405407, + "grad_norm": 0.33034011721611023, + "learning_rate": 2.6873000000000002e-05, + "loss": 0.0207, + "step": 19387 + }, + { + "epoch": 16.375, + "grad_norm": 0.12039942294359207, + "learning_rate": 2.6872666666666668e-05, + "loss": 0.0064, + "step": 19388 + }, + { + "epoch": 16.375844594594593, + "grad_norm": 0.3567414879798889, + "learning_rate": 2.6872333333333334e-05, + "loss": 0.0177, + "step": 19389 + }, + { + "epoch": 16.37668918918919, + "grad_norm": 0.7211792469024658, + "learning_rate": 2.6872000000000003e-05, + "loss": 0.0272, + "step": 19390 + }, + { + "epoch": 16.377533783783782, + "grad_norm": 0.6149898767471313, + "learning_rate": 2.6871666666666665e-05, + "loss": 0.0156, + "step": 19391 + }, + { + "epoch": 16.37837837837838, + "grad_norm": 0.4008631110191345, + "learning_rate": 2.6871333333333335e-05, + "loss": 0.0165, + "step": 19392 + }, + { + "epoch": 16.379222972972972, + "grad_norm": 0.38510018587112427, + "learning_rate": 2.6871e-05, + "loss": 0.016, + "step": 19393 + }, + { + "epoch": 16.38006756756757, + "grad_norm": 0.5925344228744507, + "learning_rate": 2.6870666666666667e-05, + "loss": 0.0186, + "step": 19394 + }, + { + "epoch": 16.38091216216216, + "grad_norm": 1.3718810081481934, + "learning_rate": 2.6870333333333336e-05, + "loss": 0.1731, + "step": 19395 + }, + { + "epoch": 16.381756756756758, + "grad_norm": 0.35612794756889343, + "learning_rate": 2.6870000000000002e-05, + "loss": 0.0902, + "step": 19396 + }, + { + "epoch": 16.38260135135135, + "grad_norm": 0.40778663754463196, + "learning_rate": 2.6869666666666668e-05, + "loss": 0.0544, + "step": 19397 + }, + { + "epoch": 16.383445945945947, + "grad_norm": 0.3175700902938843, + "learning_rate": 2.6869333333333334e-05, + "loss": 0.0261, + "step": 19398 + }, + { + "epoch": 16.38429054054054, + "grad_norm": 0.551663339138031, + "learning_rate": 2.6869e-05, + "loss": 0.0269, + "step": 19399 + }, + { + "epoch": 16.385135135135137, + "grad_norm": 0.13557858765125275, + "learning_rate": 2.6868666666666666e-05, + "loss": 0.0153, + "step": 19400 + }, + { + "epoch": 16.38597972972973, + "grad_norm": 0.44428566098213196, + "learning_rate": 2.6868333333333335e-05, + "loss": 0.0262, + "step": 19401 + }, + { + "epoch": 16.386824324324323, + "grad_norm": 0.27010995149612427, + "learning_rate": 2.6867999999999998e-05, + "loss": 0.0206, + "step": 19402 + }, + { + "epoch": 16.38766891891892, + "grad_norm": 0.23252062499523163, + "learning_rate": 2.6867666666666667e-05, + "loss": 0.0136, + "step": 19403 + }, + { + "epoch": 16.388513513513512, + "grad_norm": 0.290050208568573, + "learning_rate": 2.6867333333333336e-05, + "loss": 0.015, + "step": 19404 + }, + { + "epoch": 16.38935810810811, + "grad_norm": 0.3560445010662079, + "learning_rate": 2.6867e-05, + "loss": 0.0136, + "step": 19405 + }, + { + "epoch": 16.3902027027027, + "grad_norm": 0.20043835043907166, + "learning_rate": 2.6866666666666668e-05, + "loss": 0.0216, + "step": 19406 + }, + { + "epoch": 16.3910472972973, + "grad_norm": 0.19747889041900635, + "learning_rate": 2.6866333333333334e-05, + "loss": 0.0164, + "step": 19407 + }, + { + "epoch": 16.39189189189189, + "grad_norm": 0.2612135410308838, + "learning_rate": 2.6866e-05, + "loss": 0.0138, + "step": 19408 + }, + { + "epoch": 16.392736486486488, + "grad_norm": 0.16221117973327637, + "learning_rate": 2.6865666666666666e-05, + "loss": 0.013, + "step": 19409 + }, + { + "epoch": 16.39358108108108, + "grad_norm": 0.44986510276794434, + "learning_rate": 2.6865333333333335e-05, + "loss": 0.0297, + "step": 19410 + }, + { + "epoch": 16.394425675675677, + "grad_norm": 0.1769234985113144, + "learning_rate": 2.6865e-05, + "loss": 0.0094, + "step": 19411 + }, + { + "epoch": 16.39527027027027, + "grad_norm": 0.31973981857299805, + "learning_rate": 2.6864666666666667e-05, + "loss": 0.0139, + "step": 19412 + }, + { + "epoch": 16.396114864864863, + "grad_norm": 0.1815357804298401, + "learning_rate": 2.6864333333333336e-05, + "loss": 0.0164, + "step": 19413 + }, + { + "epoch": 16.39695945945946, + "grad_norm": 0.2655523419380188, + "learning_rate": 2.6864e-05, + "loss": 0.0316, + "step": 19414 + }, + { + "epoch": 16.397804054054053, + "grad_norm": 0.5130230784416199, + "learning_rate": 2.6863666666666668e-05, + "loss": 0.0196, + "step": 19415 + }, + { + "epoch": 16.39864864864865, + "grad_norm": 0.1353745311498642, + "learning_rate": 2.6863333333333334e-05, + "loss": 0.0099, + "step": 19416 + }, + { + "epoch": 16.399493243243242, + "grad_norm": 0.33461031317710876, + "learning_rate": 2.6863e-05, + "loss": 0.0257, + "step": 19417 + }, + { + "epoch": 16.40033783783784, + "grad_norm": 0.6363779306411743, + "learning_rate": 2.6862666666666666e-05, + "loss": 0.0139, + "step": 19418 + }, + { + "epoch": 16.40118243243243, + "grad_norm": 0.30573487281799316, + "learning_rate": 2.6862333333333335e-05, + "loss": 0.0078, + "step": 19419 + }, + { + "epoch": 16.402027027027028, + "grad_norm": 0.2128751426935196, + "learning_rate": 2.6862e-05, + "loss": 0.011, + "step": 19420 + }, + { + "epoch": 16.40287162162162, + "grad_norm": 0.20716534554958344, + "learning_rate": 2.6861666666666667e-05, + "loss": 0.0107, + "step": 19421 + }, + { + "epoch": 16.403716216216218, + "grad_norm": 0.38202282786369324, + "learning_rate": 2.6861333333333337e-05, + "loss": 0.0095, + "step": 19422 + }, + { + "epoch": 16.40456081081081, + "grad_norm": 0.27038100361824036, + "learning_rate": 2.6861e-05, + "loss": 0.0111, + "step": 19423 + }, + { + "epoch": 16.405405405405407, + "grad_norm": 0.18224157392978668, + "learning_rate": 2.686066666666667e-05, + "loss": 0.0118, + "step": 19424 + }, + { + "epoch": 16.40625, + "grad_norm": 0.3102482259273529, + "learning_rate": 2.6860333333333334e-05, + "loss": 0.0191, + "step": 19425 + }, + { + "epoch": 16.407094594594593, + "grad_norm": 0.30313146114349365, + "learning_rate": 2.686e-05, + "loss": 0.0179, + "step": 19426 + }, + { + "epoch": 16.40793918918919, + "grad_norm": 0.24074101448059082, + "learning_rate": 2.6859666666666666e-05, + "loss": 0.0139, + "step": 19427 + }, + { + "epoch": 16.408783783783782, + "grad_norm": 0.20879612863063812, + "learning_rate": 2.6859333333333332e-05, + "loss": 0.0117, + "step": 19428 + }, + { + "epoch": 16.40962837837838, + "grad_norm": 0.15210597217082977, + "learning_rate": 2.6859e-05, + "loss": 0.0101, + "step": 19429 + }, + { + "epoch": 16.410472972972972, + "grad_norm": 0.3560755252838135, + "learning_rate": 2.6858666666666667e-05, + "loss": 0.0145, + "step": 19430 + }, + { + "epoch": 16.41131756756757, + "grad_norm": 0.2388422191143036, + "learning_rate": 2.6858333333333333e-05, + "loss": 0.0088, + "step": 19431 + }, + { + "epoch": 16.41216216216216, + "grad_norm": 0.22182083129882812, + "learning_rate": 2.6858e-05, + "loss": 0.0106, + "step": 19432 + }, + { + "epoch": 16.413006756756758, + "grad_norm": 0.156329944729805, + "learning_rate": 2.685766666666667e-05, + "loss": 0.0082, + "step": 19433 + }, + { + "epoch": 16.41385135135135, + "grad_norm": 0.17974674701690674, + "learning_rate": 2.685733333333333e-05, + "loss": 0.0135, + "step": 19434 + }, + { + "epoch": 16.414695945945947, + "grad_norm": 0.16074688732624054, + "learning_rate": 2.6857e-05, + "loss": 0.0078, + "step": 19435 + }, + { + "epoch": 16.41554054054054, + "grad_norm": 0.49899059534072876, + "learning_rate": 2.685666666666667e-05, + "loss": 0.0155, + "step": 19436 + }, + { + "epoch": 16.416385135135137, + "grad_norm": 0.19626069068908691, + "learning_rate": 2.6856333333333332e-05, + "loss": 0.0085, + "step": 19437 + }, + { + "epoch": 16.41722972972973, + "grad_norm": 0.6178881525993347, + "learning_rate": 2.6856000000000002e-05, + "loss": 0.0152, + "step": 19438 + }, + { + "epoch": 16.418074324324323, + "grad_norm": 0.18406009674072266, + "learning_rate": 2.6855666666666668e-05, + "loss": 0.0071, + "step": 19439 + }, + { + "epoch": 16.41891891891892, + "grad_norm": 0.3383699655532837, + "learning_rate": 2.6855333333333334e-05, + "loss": 0.0145, + "step": 19440 + }, + { + "epoch": 16.419763513513512, + "grad_norm": 0.42394566535949707, + "learning_rate": 2.6855e-05, + "loss": 0.0129, + "step": 19441 + }, + { + "epoch": 16.42060810810811, + "grad_norm": 0.2749832570552826, + "learning_rate": 2.685466666666667e-05, + "loss": 0.0156, + "step": 19442 + }, + { + "epoch": 16.4214527027027, + "grad_norm": 0.3224293291568756, + "learning_rate": 2.685433333333333e-05, + "loss": 0.0135, + "step": 19443 + }, + { + "epoch": 16.4222972972973, + "grad_norm": 0.6228426694869995, + "learning_rate": 2.6854e-05, + "loss": 0.0677, + "step": 19444 + }, + { + "epoch": 16.42314189189189, + "grad_norm": 0.5216479897499084, + "learning_rate": 2.685366666666667e-05, + "loss": 0.1462, + "step": 19445 + }, + { + "epoch": 16.423986486486488, + "grad_norm": 0.5890079140663147, + "learning_rate": 2.6853333333333333e-05, + "loss": 0.106, + "step": 19446 + }, + { + "epoch": 16.42483108108108, + "grad_norm": 0.432174950838089, + "learning_rate": 2.6853000000000002e-05, + "loss": 0.0503, + "step": 19447 + }, + { + "epoch": 16.425675675675677, + "grad_norm": 0.4075501561164856, + "learning_rate": 2.6852666666666668e-05, + "loss": 0.039, + "step": 19448 + }, + { + "epoch": 16.42652027027027, + "grad_norm": 0.2902076840400696, + "learning_rate": 2.6852333333333334e-05, + "loss": 0.0226, + "step": 19449 + }, + { + "epoch": 16.427364864864863, + "grad_norm": 0.2291741669178009, + "learning_rate": 2.6852e-05, + "loss": 0.0125, + "step": 19450 + }, + { + "epoch": 16.42820945945946, + "grad_norm": 0.2601809501647949, + "learning_rate": 2.685166666666667e-05, + "loss": 0.0195, + "step": 19451 + }, + { + "epoch": 16.429054054054053, + "grad_norm": 0.28031691908836365, + "learning_rate": 2.685133333333333e-05, + "loss": 0.0115, + "step": 19452 + }, + { + "epoch": 16.42989864864865, + "grad_norm": 0.1462888866662979, + "learning_rate": 2.6851e-05, + "loss": 0.0158, + "step": 19453 + }, + { + "epoch": 16.430743243243242, + "grad_norm": 0.3153122663497925, + "learning_rate": 2.685066666666667e-05, + "loss": 0.0197, + "step": 19454 + }, + { + "epoch": 16.43158783783784, + "grad_norm": 0.19299812614917755, + "learning_rate": 2.6850333333333333e-05, + "loss": 0.0088, + "step": 19455 + }, + { + "epoch": 16.43243243243243, + "grad_norm": 0.18631960451602936, + "learning_rate": 2.6850000000000002e-05, + "loss": 0.0079, + "step": 19456 + }, + { + "epoch": 16.433277027027028, + "grad_norm": 0.3193402588367462, + "learning_rate": 2.6849666666666665e-05, + "loss": 0.0167, + "step": 19457 + }, + { + "epoch": 16.43412162162162, + "grad_norm": 0.3479108512401581, + "learning_rate": 2.6849333333333334e-05, + "loss": 0.0113, + "step": 19458 + }, + { + "epoch": 16.434966216216218, + "grad_norm": 0.14019691944122314, + "learning_rate": 2.6849e-05, + "loss": 0.0114, + "step": 19459 + }, + { + "epoch": 16.43581081081081, + "grad_norm": 0.2321406453847885, + "learning_rate": 2.6848666666666666e-05, + "loss": 0.0164, + "step": 19460 + }, + { + "epoch": 16.436655405405407, + "grad_norm": 0.19330409169197083, + "learning_rate": 2.6848333333333335e-05, + "loss": 0.0139, + "step": 19461 + }, + { + "epoch": 16.4375, + "grad_norm": 0.5667406916618347, + "learning_rate": 2.6848e-05, + "loss": 0.0138, + "step": 19462 + }, + { + "epoch": 16.438344594594593, + "grad_norm": 0.21876467764377594, + "learning_rate": 2.6847666666666667e-05, + "loss": 0.0073, + "step": 19463 + }, + { + "epoch": 16.43918918918919, + "grad_norm": 0.3047693371772766, + "learning_rate": 2.6847333333333333e-05, + "loss": 0.0208, + "step": 19464 + }, + { + "epoch": 16.440033783783782, + "grad_norm": 0.2001260370016098, + "learning_rate": 2.6847000000000002e-05, + "loss": 0.0135, + "step": 19465 + }, + { + "epoch": 16.44087837837838, + "grad_norm": 0.14469002187252045, + "learning_rate": 2.6846666666666665e-05, + "loss": 0.0107, + "step": 19466 + }, + { + "epoch": 16.441722972972972, + "grad_norm": 0.24994277954101562, + "learning_rate": 2.6846333333333334e-05, + "loss": 0.0125, + "step": 19467 + }, + { + "epoch": 16.44256756756757, + "grad_norm": 0.1620766967535019, + "learning_rate": 2.6846e-05, + "loss": 0.0151, + "step": 19468 + }, + { + "epoch": 16.44341216216216, + "grad_norm": 0.17276285588741302, + "learning_rate": 2.6845666666666666e-05, + "loss": 0.0103, + "step": 19469 + }, + { + "epoch": 16.444256756756758, + "grad_norm": 0.3307458758354187, + "learning_rate": 2.6845333333333335e-05, + "loss": 0.013, + "step": 19470 + }, + { + "epoch": 16.44510135135135, + "grad_norm": 0.2843451201915741, + "learning_rate": 2.6845e-05, + "loss": 0.0092, + "step": 19471 + }, + { + "epoch": 16.445945945945947, + "grad_norm": 0.21294857561588287, + "learning_rate": 2.6844666666666667e-05, + "loss": 0.0116, + "step": 19472 + }, + { + "epoch": 16.44679054054054, + "grad_norm": 0.15939147770404816, + "learning_rate": 2.6844333333333333e-05, + "loss": 0.0104, + "step": 19473 + }, + { + "epoch": 16.447635135135137, + "grad_norm": 0.39566606283187866, + "learning_rate": 2.6844000000000003e-05, + "loss": 0.0191, + "step": 19474 + }, + { + "epoch": 16.44847972972973, + "grad_norm": 0.22217868268489838, + "learning_rate": 2.6843666666666665e-05, + "loss": 0.0084, + "step": 19475 + }, + { + "epoch": 16.449324324324323, + "grad_norm": 0.19553375244140625, + "learning_rate": 2.6843333333333334e-05, + "loss": 0.009, + "step": 19476 + }, + { + "epoch": 16.45016891891892, + "grad_norm": 0.2335994839668274, + "learning_rate": 2.6843e-05, + "loss": 0.0135, + "step": 19477 + }, + { + "epoch": 16.451013513513512, + "grad_norm": 0.2992324233055115, + "learning_rate": 2.6842666666666666e-05, + "loss": 0.0144, + "step": 19478 + }, + { + "epoch": 16.45185810810811, + "grad_norm": 0.10854735225439072, + "learning_rate": 2.6842333333333336e-05, + "loss": 0.0035, + "step": 19479 + }, + { + "epoch": 16.4527027027027, + "grad_norm": 0.16480214893817902, + "learning_rate": 2.6842e-05, + "loss": 0.0061, + "step": 19480 + }, + { + "epoch": 16.4535472972973, + "grad_norm": 0.12790147960186005, + "learning_rate": 2.6841666666666667e-05, + "loss": 0.0057, + "step": 19481 + }, + { + "epoch": 16.45439189189189, + "grad_norm": 0.24843229353427887, + "learning_rate": 2.6841333333333333e-05, + "loss": 0.0155, + "step": 19482 + }, + { + "epoch": 16.455236486486488, + "grad_norm": 0.36121445894241333, + "learning_rate": 2.6841000000000003e-05, + "loss": 0.015, + "step": 19483 + }, + { + "epoch": 16.45608108108108, + "grad_norm": 0.2545869052410126, + "learning_rate": 2.6840666666666665e-05, + "loss": 0.0155, + "step": 19484 + }, + { + "epoch": 16.456925675675677, + "grad_norm": 0.2000930905342102, + "learning_rate": 2.6840333333333335e-05, + "loss": 0.0098, + "step": 19485 + }, + { + "epoch": 16.45777027027027, + "grad_norm": 0.2672353982925415, + "learning_rate": 2.6840000000000004e-05, + "loss": 0.0145, + "step": 19486 + }, + { + "epoch": 16.458614864864863, + "grad_norm": 0.5015769600868225, + "learning_rate": 2.6839666666666666e-05, + "loss": 0.0319, + "step": 19487 + }, + { + "epoch": 16.45945945945946, + "grad_norm": 0.3783192038536072, + "learning_rate": 2.6839333333333336e-05, + "loss": 0.0268, + "step": 19488 + }, + { + "epoch": 16.460304054054053, + "grad_norm": 0.19076408445835114, + "learning_rate": 2.6839e-05, + "loss": 0.0113, + "step": 19489 + }, + { + "epoch": 16.46114864864865, + "grad_norm": 0.9311555027961731, + "learning_rate": 2.6838666666666668e-05, + "loss": 0.0219, + "step": 19490 + }, + { + "epoch": 16.461993243243242, + "grad_norm": 0.11374102532863617, + "learning_rate": 2.6838333333333334e-05, + "loss": 0.0054, + "step": 19491 + }, + { + "epoch": 16.46283783783784, + "grad_norm": 0.26984134316444397, + "learning_rate": 2.6838e-05, + "loss": 0.0193, + "step": 19492 + }, + { + "epoch": 16.46368243243243, + "grad_norm": 0.7140309810638428, + "learning_rate": 2.6837666666666665e-05, + "loss": 0.0211, + "step": 19493 + }, + { + "epoch": 16.464527027027028, + "grad_norm": 1.8419605493545532, + "learning_rate": 2.6837333333333335e-05, + "loss": 0.04, + "step": 19494 + }, + { + "epoch": 16.46537162162162, + "grad_norm": 0.4229506850242615, + "learning_rate": 2.6837e-05, + "loss": 0.1278, + "step": 19495 + }, + { + "epoch": 16.466216216216218, + "grad_norm": 0.611250102519989, + "learning_rate": 2.6836666666666667e-05, + "loss": 0.0781, + "step": 19496 + }, + { + "epoch": 16.46706081081081, + "grad_norm": 0.2960658669471741, + "learning_rate": 2.6836333333333336e-05, + "loss": 0.0354, + "step": 19497 + }, + { + "epoch": 16.467905405405407, + "grad_norm": 0.21593812108039856, + "learning_rate": 2.6836e-05, + "loss": 0.0248, + "step": 19498 + }, + { + "epoch": 16.46875, + "grad_norm": 0.23343245685100555, + "learning_rate": 2.6835666666666668e-05, + "loss": 0.0111, + "step": 19499 + }, + { + "epoch": 16.469594594594593, + "grad_norm": 0.403019517660141, + "learning_rate": 2.6835333333333334e-05, + "loss": 0.0197, + "step": 19500 + }, + { + "epoch": 16.47043918918919, + "grad_norm": 0.30637043714523315, + "learning_rate": 2.6835e-05, + "loss": 0.0162, + "step": 19501 + }, + { + "epoch": 16.471283783783782, + "grad_norm": 0.18856677412986755, + "learning_rate": 2.6834666666666666e-05, + "loss": 0.0182, + "step": 19502 + }, + { + "epoch": 16.47212837837838, + "grad_norm": 0.17538577318191528, + "learning_rate": 2.6834333333333335e-05, + "loss": 0.0112, + "step": 19503 + }, + { + "epoch": 16.472972972972972, + "grad_norm": 0.16824087500572205, + "learning_rate": 2.6834e-05, + "loss": 0.0094, + "step": 19504 + }, + { + "epoch": 16.47381756756757, + "grad_norm": 0.2106093019247055, + "learning_rate": 2.6833666666666667e-05, + "loss": 0.0125, + "step": 19505 + }, + { + "epoch": 16.47466216216216, + "grad_norm": 0.2656978666782379, + "learning_rate": 2.6833333333333336e-05, + "loss": 0.0163, + "step": 19506 + }, + { + "epoch": 16.475506756756758, + "grad_norm": 0.2558043301105499, + "learning_rate": 2.6833e-05, + "loss": 0.0149, + "step": 19507 + }, + { + "epoch": 16.47635135135135, + "grad_norm": 0.25010669231414795, + "learning_rate": 2.6832666666666668e-05, + "loss": 0.0162, + "step": 19508 + }, + { + "epoch": 16.477195945945947, + "grad_norm": 0.19183295965194702, + "learning_rate": 2.6832333333333334e-05, + "loss": 0.0127, + "step": 19509 + }, + { + "epoch": 16.47804054054054, + "grad_norm": 0.20999376475811005, + "learning_rate": 2.6832e-05, + "loss": 0.021, + "step": 19510 + }, + { + "epoch": 16.478885135135137, + "grad_norm": 0.26383769512176514, + "learning_rate": 2.683166666666667e-05, + "loss": 0.0122, + "step": 19511 + }, + { + "epoch": 16.47972972972973, + "grad_norm": 0.2200082391500473, + "learning_rate": 2.6831333333333335e-05, + "loss": 0.0161, + "step": 19512 + }, + { + "epoch": 16.480574324324323, + "grad_norm": 0.2268911749124527, + "learning_rate": 2.6831e-05, + "loss": 0.0113, + "step": 19513 + }, + { + "epoch": 16.48141891891892, + "grad_norm": 0.40629154443740845, + "learning_rate": 2.6830666666666667e-05, + "loss": 0.0127, + "step": 19514 + }, + { + "epoch": 16.482263513513512, + "grad_norm": 0.148725226521492, + "learning_rate": 2.6830333333333336e-05, + "loss": 0.0137, + "step": 19515 + }, + { + "epoch": 16.48310810810811, + "grad_norm": 0.1704261302947998, + "learning_rate": 2.683e-05, + "loss": 0.0114, + "step": 19516 + }, + { + "epoch": 16.4839527027027, + "grad_norm": 0.3622487783432007, + "learning_rate": 2.6829666666666668e-05, + "loss": 0.019, + "step": 19517 + }, + { + "epoch": 16.4847972972973, + "grad_norm": 0.2945890724658966, + "learning_rate": 2.682933333333333e-05, + "loss": 0.0137, + "step": 19518 + }, + { + "epoch": 16.48564189189189, + "grad_norm": 0.339091420173645, + "learning_rate": 2.6829e-05, + "loss": 0.0219, + "step": 19519 + }, + { + "epoch": 16.486486486486488, + "grad_norm": 0.4131038188934326, + "learning_rate": 2.682866666666667e-05, + "loss": 0.0128, + "step": 19520 + }, + { + "epoch": 16.48733108108108, + "grad_norm": 0.3176792562007904, + "learning_rate": 2.6828333333333332e-05, + "loss": 0.0101, + "step": 19521 + }, + { + "epoch": 16.488175675675677, + "grad_norm": 0.2927132844924927, + "learning_rate": 2.6828e-05, + "loss": 0.0136, + "step": 19522 + }, + { + "epoch": 16.48902027027027, + "grad_norm": 0.8071916699409485, + "learning_rate": 2.6827666666666667e-05, + "loss": 0.0235, + "step": 19523 + }, + { + "epoch": 16.489864864864863, + "grad_norm": 0.20439837872982025, + "learning_rate": 2.6827333333333333e-05, + "loss": 0.0116, + "step": 19524 + }, + { + "epoch": 16.49070945945946, + "grad_norm": 0.20663069188594818, + "learning_rate": 2.6827e-05, + "loss": 0.0102, + "step": 19525 + }, + { + "epoch": 16.491554054054053, + "grad_norm": 0.21555103361606598, + "learning_rate": 2.682666666666667e-05, + "loss": 0.014, + "step": 19526 + }, + { + "epoch": 16.49239864864865, + "grad_norm": 0.23389685153961182, + "learning_rate": 2.6826333333333334e-05, + "loss": 0.0171, + "step": 19527 + }, + { + "epoch": 16.493243243243242, + "grad_norm": 0.27106133103370667, + "learning_rate": 2.6826e-05, + "loss": 0.0184, + "step": 19528 + }, + { + "epoch": 16.49408783783784, + "grad_norm": 0.1819290816783905, + "learning_rate": 2.682566666666667e-05, + "loss": 0.007, + "step": 19529 + }, + { + "epoch": 16.49493243243243, + "grad_norm": 0.4107491970062256, + "learning_rate": 2.6825333333333332e-05, + "loss": 0.0299, + "step": 19530 + }, + { + "epoch": 16.495777027027028, + "grad_norm": 0.37589308619499207, + "learning_rate": 2.6825e-05, + "loss": 0.0147, + "step": 19531 + }, + { + "epoch": 16.49662162162162, + "grad_norm": 0.27192801237106323, + "learning_rate": 2.6824666666666667e-05, + "loss": 0.0122, + "step": 19532 + }, + { + "epoch": 16.497466216216218, + "grad_norm": 0.5053355097770691, + "learning_rate": 2.6824333333333333e-05, + "loss": 0.0123, + "step": 19533 + }, + { + "epoch": 16.49831081081081, + "grad_norm": 0.30612286925315857, + "learning_rate": 2.6824e-05, + "loss": 0.0128, + "step": 19534 + }, + { + "epoch": 16.499155405405407, + "grad_norm": 0.20802877843379974, + "learning_rate": 2.682366666666667e-05, + "loss": 0.0069, + "step": 19535 + }, + { + "epoch": 16.5, + "grad_norm": 0.20747925341129303, + "learning_rate": 2.6823333333333335e-05, + "loss": 0.0059, + "step": 19536 + }, + { + "epoch": 16.500844594594593, + "grad_norm": 0.18318593502044678, + "learning_rate": 2.6823e-05, + "loss": 0.0093, + "step": 19537 + }, + { + "epoch": 16.50168918918919, + "grad_norm": 0.19324177503585815, + "learning_rate": 2.682266666666667e-05, + "loss": 0.0061, + "step": 19538 + }, + { + "epoch": 16.502533783783782, + "grad_norm": 0.37580159306526184, + "learning_rate": 2.6822333333333332e-05, + "loss": 0.0242, + "step": 19539 + }, + { + "epoch": 16.50337837837838, + "grad_norm": 0.2106221318244934, + "learning_rate": 2.6822000000000002e-05, + "loss": 0.0118, + "step": 19540 + }, + { + "epoch": 16.504222972972972, + "grad_norm": 0.19539117813110352, + "learning_rate": 2.6821666666666668e-05, + "loss": 0.0098, + "step": 19541 + }, + { + "epoch": 16.50506756756757, + "grad_norm": 0.3596411943435669, + "learning_rate": 2.6821333333333334e-05, + "loss": 0.0123, + "step": 19542 + }, + { + "epoch": 16.50591216216216, + "grad_norm": 1.1696304082870483, + "learning_rate": 2.6821e-05, + "loss": 0.0236, + "step": 19543 + }, + { + "epoch": 16.506756756756758, + "grad_norm": 0.4039745032787323, + "learning_rate": 2.682066666666667e-05, + "loss": 0.0165, + "step": 19544 + }, + { + "epoch": 16.50760135135135, + "grad_norm": 0.5758534669876099, + "learning_rate": 2.6820333333333335e-05, + "loss": 0.1223, + "step": 19545 + }, + { + "epoch": 16.508445945945947, + "grad_norm": 0.8750205636024475, + "learning_rate": 2.682e-05, + "loss": 0.0915, + "step": 19546 + }, + { + "epoch": 16.50929054054054, + "grad_norm": 0.2957712411880493, + "learning_rate": 2.681966666666667e-05, + "loss": 0.0291, + "step": 19547 + }, + { + "epoch": 16.510135135135137, + "grad_norm": 0.3357982039451599, + "learning_rate": 2.6819333333333333e-05, + "loss": 0.028, + "step": 19548 + }, + { + "epoch": 16.51097972972973, + "grad_norm": 0.1706397980451584, + "learning_rate": 2.6819000000000002e-05, + "loss": 0.0107, + "step": 19549 + }, + { + "epoch": 16.511824324324323, + "grad_norm": 0.17870639264583588, + "learning_rate": 2.6818666666666664e-05, + "loss": 0.014, + "step": 19550 + }, + { + "epoch": 16.51266891891892, + "grad_norm": 0.24931275844573975, + "learning_rate": 2.6818333333333334e-05, + "loss": 0.017, + "step": 19551 + }, + { + "epoch": 16.513513513513512, + "grad_norm": 0.2357298880815506, + "learning_rate": 2.6818e-05, + "loss": 0.0147, + "step": 19552 + }, + { + "epoch": 16.51435810810811, + "grad_norm": 0.24268551170825958, + "learning_rate": 2.6817666666666666e-05, + "loss": 0.015, + "step": 19553 + }, + { + "epoch": 16.5152027027027, + "grad_norm": 0.24600523710250854, + "learning_rate": 2.6817333333333335e-05, + "loss": 0.0185, + "step": 19554 + }, + { + "epoch": 16.5160472972973, + "grad_norm": 0.17310263216495514, + "learning_rate": 2.6817e-05, + "loss": 0.0128, + "step": 19555 + }, + { + "epoch": 16.51689189189189, + "grad_norm": 0.3009398281574249, + "learning_rate": 2.6816666666666667e-05, + "loss": 0.014, + "step": 19556 + }, + { + "epoch": 16.517736486486488, + "grad_norm": 0.4216502606868744, + "learning_rate": 2.6816333333333333e-05, + "loss": 0.0157, + "step": 19557 + }, + { + "epoch": 16.51858108108108, + "grad_norm": 0.14420577883720398, + "learning_rate": 2.6816000000000002e-05, + "loss": 0.0138, + "step": 19558 + }, + { + "epoch": 16.519425675675677, + "grad_norm": 0.3288867771625519, + "learning_rate": 2.6815666666666665e-05, + "loss": 0.0343, + "step": 19559 + }, + { + "epoch": 16.52027027027027, + "grad_norm": 0.18244855105876923, + "learning_rate": 2.6815333333333334e-05, + "loss": 0.0093, + "step": 19560 + }, + { + "epoch": 16.521114864864863, + "grad_norm": 0.23565281927585602, + "learning_rate": 2.6815000000000003e-05, + "loss": 0.0203, + "step": 19561 + }, + { + "epoch": 16.52195945945946, + "grad_norm": 0.1475265622138977, + "learning_rate": 2.6814666666666666e-05, + "loss": 0.0105, + "step": 19562 + }, + { + "epoch": 16.522804054054053, + "grad_norm": 0.23872174322605133, + "learning_rate": 2.6814333333333335e-05, + "loss": 0.0121, + "step": 19563 + }, + { + "epoch": 16.52364864864865, + "grad_norm": 0.10566062480211258, + "learning_rate": 2.6814e-05, + "loss": 0.009, + "step": 19564 + }, + { + "epoch": 16.524493243243242, + "grad_norm": 0.1742836982011795, + "learning_rate": 2.6813666666666667e-05, + "loss": 0.0116, + "step": 19565 + }, + { + "epoch": 16.52533783783784, + "grad_norm": 0.23231537640094757, + "learning_rate": 2.6813333333333333e-05, + "loss": 0.0128, + "step": 19566 + }, + { + "epoch": 16.52618243243243, + "grad_norm": 0.1098446324467659, + "learning_rate": 2.6813000000000002e-05, + "loss": 0.0077, + "step": 19567 + }, + { + "epoch": 16.527027027027028, + "grad_norm": 0.28634196519851685, + "learning_rate": 2.6812666666666665e-05, + "loss": 0.0129, + "step": 19568 + }, + { + "epoch": 16.52787162162162, + "grad_norm": 0.15250320732593536, + "learning_rate": 2.6812333333333334e-05, + "loss": 0.0086, + "step": 19569 + }, + { + "epoch": 16.528716216216218, + "grad_norm": 0.4923097491264343, + "learning_rate": 2.6812000000000004e-05, + "loss": 0.0185, + "step": 19570 + }, + { + "epoch": 16.52956081081081, + "grad_norm": 0.30758413672447205, + "learning_rate": 2.6811666666666666e-05, + "loss": 0.0218, + "step": 19571 + }, + { + "epoch": 16.530405405405407, + "grad_norm": 0.30511945486068726, + "learning_rate": 2.6811333333333335e-05, + "loss": 0.0133, + "step": 19572 + }, + { + "epoch": 16.53125, + "grad_norm": 0.2187240868806839, + "learning_rate": 2.6811e-05, + "loss": 0.0144, + "step": 19573 + }, + { + "epoch": 16.532094594594593, + "grad_norm": 0.20864365994930267, + "learning_rate": 2.6810666666666667e-05, + "loss": 0.009, + "step": 19574 + }, + { + "epoch": 16.53293918918919, + "grad_norm": 0.22541779279708862, + "learning_rate": 2.6810333333333333e-05, + "loss": 0.0134, + "step": 19575 + }, + { + "epoch": 16.533783783783782, + "grad_norm": 0.4317740201950073, + "learning_rate": 2.6810000000000003e-05, + "loss": 0.0197, + "step": 19576 + }, + { + "epoch": 16.53462837837838, + "grad_norm": 0.3059368133544922, + "learning_rate": 2.680966666666667e-05, + "loss": 0.018, + "step": 19577 + }, + { + "epoch": 16.535472972972972, + "grad_norm": 0.21003592014312744, + "learning_rate": 2.6809333333333334e-05, + "loss": 0.0141, + "step": 19578 + }, + { + "epoch": 16.53631756756757, + "grad_norm": 0.3089529871940613, + "learning_rate": 2.6809e-05, + "loss": 0.0173, + "step": 19579 + }, + { + "epoch": 16.53716216216216, + "grad_norm": 0.3342539072036743, + "learning_rate": 2.6808666666666666e-05, + "loss": 0.0118, + "step": 19580 + }, + { + "epoch": 16.538006756756758, + "grad_norm": 0.17753934860229492, + "learning_rate": 2.6808333333333336e-05, + "loss": 0.0132, + "step": 19581 + }, + { + "epoch": 16.53885135135135, + "grad_norm": 0.4431025981903076, + "learning_rate": 2.6807999999999998e-05, + "loss": 0.0276, + "step": 19582 + }, + { + "epoch": 16.539695945945947, + "grad_norm": 0.18693993985652924, + "learning_rate": 2.6807666666666667e-05, + "loss": 0.011, + "step": 19583 + }, + { + "epoch": 16.54054054054054, + "grad_norm": 0.38803020119667053, + "learning_rate": 2.6807333333333333e-05, + "loss": 0.0143, + "step": 19584 + }, + { + "epoch": 16.541385135135137, + "grad_norm": 0.4963170886039734, + "learning_rate": 2.6807e-05, + "loss": 0.0147, + "step": 19585 + }, + { + "epoch": 16.54222972972973, + "grad_norm": 0.4502437114715576, + "learning_rate": 2.680666666666667e-05, + "loss": 0.0189, + "step": 19586 + }, + { + "epoch": 16.543074324324323, + "grad_norm": 0.6886362433433533, + "learning_rate": 2.6806333333333335e-05, + "loss": 0.0225, + "step": 19587 + }, + { + "epoch": 16.54391891891892, + "grad_norm": 0.329865425825119, + "learning_rate": 2.6806e-05, + "loss": 0.0145, + "step": 19588 + }, + { + "epoch": 16.544763513513512, + "grad_norm": 0.13843417167663574, + "learning_rate": 2.6805666666666666e-05, + "loss": 0.0066, + "step": 19589 + }, + { + "epoch": 16.54560810810811, + "grad_norm": 0.11385305225849152, + "learning_rate": 2.6805333333333336e-05, + "loss": 0.005, + "step": 19590 + }, + { + "epoch": 16.5464527027027, + "grad_norm": 0.275094211101532, + "learning_rate": 2.6805e-05, + "loss": 0.0088, + "step": 19591 + }, + { + "epoch": 16.5472972972973, + "grad_norm": 0.2979552447795868, + "learning_rate": 2.6804666666666668e-05, + "loss": 0.0155, + "step": 19592 + }, + { + "epoch": 16.54814189189189, + "grad_norm": 0.8101019859313965, + "learning_rate": 2.6804333333333334e-05, + "loss": 0.0159, + "step": 19593 + }, + { + "epoch": 16.548986486486488, + "grad_norm": 0.3841153681278229, + "learning_rate": 2.6804e-05, + "loss": 0.0161, + "step": 19594 + }, + { + "epoch": 16.54983108108108, + "grad_norm": 0.6682618856430054, + "learning_rate": 2.680366666666667e-05, + "loss": 0.1776, + "step": 19595 + }, + { + "epoch": 16.550675675675677, + "grad_norm": 0.4883839786052704, + "learning_rate": 2.6803333333333335e-05, + "loss": 0.0698, + "step": 19596 + }, + { + "epoch": 16.55152027027027, + "grad_norm": 0.5826992988586426, + "learning_rate": 2.6803e-05, + "loss": 0.0599, + "step": 19597 + }, + { + "epoch": 16.552364864864863, + "grad_norm": 0.3052903115749359, + "learning_rate": 2.6802666666666667e-05, + "loss": 0.0403, + "step": 19598 + }, + { + "epoch": 16.55320945945946, + "grad_norm": 0.19211415946483612, + "learning_rate": 2.6802333333333336e-05, + "loss": 0.0137, + "step": 19599 + }, + { + "epoch": 16.554054054054053, + "grad_norm": 0.17433594167232513, + "learning_rate": 2.6802e-05, + "loss": 0.0131, + "step": 19600 + }, + { + "epoch": 16.55489864864865, + "grad_norm": 0.1589062213897705, + "learning_rate": 2.6801666666666668e-05, + "loss": 0.0119, + "step": 19601 + }, + { + "epoch": 16.555743243243242, + "grad_norm": 0.3414626717567444, + "learning_rate": 2.6801333333333334e-05, + "loss": 0.0207, + "step": 19602 + }, + { + "epoch": 16.55658783783784, + "grad_norm": 0.5336631536483765, + "learning_rate": 2.6801e-05, + "loss": 0.0159, + "step": 19603 + }, + { + "epoch": 16.55743243243243, + "grad_norm": 0.18324421346187592, + "learning_rate": 2.680066666666667e-05, + "loss": 0.0102, + "step": 19604 + }, + { + "epoch": 16.558277027027028, + "grad_norm": 0.3048180043697357, + "learning_rate": 2.6800333333333335e-05, + "loss": 0.0234, + "step": 19605 + }, + { + "epoch": 16.55912162162162, + "grad_norm": 0.1769506186246872, + "learning_rate": 2.68e-05, + "loss": 0.0105, + "step": 19606 + }, + { + "epoch": 16.559966216216218, + "grad_norm": 0.4218679666519165, + "learning_rate": 2.6799666666666667e-05, + "loss": 0.0167, + "step": 19607 + }, + { + "epoch": 16.56081081081081, + "grad_norm": 0.1738714724779129, + "learning_rate": 2.6799333333333333e-05, + "loss": 0.0124, + "step": 19608 + }, + { + "epoch": 16.561655405405407, + "grad_norm": 0.29298320412635803, + "learning_rate": 2.6799e-05, + "loss": 0.0187, + "step": 19609 + }, + { + "epoch": 16.5625, + "grad_norm": 0.2500016987323761, + "learning_rate": 2.6798666666666668e-05, + "loss": 0.0245, + "step": 19610 + }, + { + "epoch": 16.563344594594593, + "grad_norm": 0.2485433965921402, + "learning_rate": 2.6798333333333334e-05, + "loss": 0.0134, + "step": 19611 + }, + { + "epoch": 16.56418918918919, + "grad_norm": 0.07249834388494492, + "learning_rate": 2.6798e-05, + "loss": 0.0063, + "step": 19612 + }, + { + "epoch": 16.565033783783782, + "grad_norm": 0.21965885162353516, + "learning_rate": 2.679766666666667e-05, + "loss": 0.0104, + "step": 19613 + }, + { + "epoch": 16.56587837837838, + "grad_norm": 0.1827099621295929, + "learning_rate": 2.6797333333333332e-05, + "loss": 0.0083, + "step": 19614 + }, + { + "epoch": 16.566722972972972, + "grad_norm": 0.23010259866714478, + "learning_rate": 2.6797e-05, + "loss": 0.0171, + "step": 19615 + }, + { + "epoch": 16.56756756756757, + "grad_norm": 0.243382528424263, + "learning_rate": 2.6796666666666667e-05, + "loss": 0.0121, + "step": 19616 + }, + { + "epoch": 16.56841216216216, + "grad_norm": 0.16941605508327484, + "learning_rate": 2.6796333333333333e-05, + "loss": 0.0118, + "step": 19617 + }, + { + "epoch": 16.569256756756758, + "grad_norm": 0.17725007236003876, + "learning_rate": 2.6796e-05, + "loss": 0.0101, + "step": 19618 + }, + { + "epoch": 16.57010135135135, + "grad_norm": 0.25380000472068787, + "learning_rate": 2.6795666666666668e-05, + "loss": 0.0167, + "step": 19619 + }, + { + "epoch": 16.570945945945947, + "grad_norm": 0.2637317478656769, + "learning_rate": 2.6795333333333334e-05, + "loss": 0.0108, + "step": 19620 + }, + { + "epoch": 16.57179054054054, + "grad_norm": 0.41622045636177063, + "learning_rate": 2.6795e-05, + "loss": 0.0219, + "step": 19621 + }, + { + "epoch": 16.572635135135137, + "grad_norm": 0.2353658378124237, + "learning_rate": 2.679466666666667e-05, + "loss": 0.0164, + "step": 19622 + }, + { + "epoch": 16.57347972972973, + "grad_norm": 0.21036885678768158, + "learning_rate": 2.6794333333333332e-05, + "loss": 0.0147, + "step": 19623 + }, + { + "epoch": 16.574324324324323, + "grad_norm": 0.25686752796173096, + "learning_rate": 2.6794e-05, + "loss": 0.0098, + "step": 19624 + }, + { + "epoch": 16.57516891891892, + "grad_norm": 0.26871034502983093, + "learning_rate": 2.6793666666666667e-05, + "loss": 0.0112, + "step": 19625 + }, + { + "epoch": 16.576013513513512, + "grad_norm": 0.2671060264110565, + "learning_rate": 2.6793333333333333e-05, + "loss": 0.0129, + "step": 19626 + }, + { + "epoch": 16.57685810810811, + "grad_norm": 0.3310137689113617, + "learning_rate": 2.6793000000000002e-05, + "loss": 0.0144, + "step": 19627 + }, + { + "epoch": 16.5777027027027, + "grad_norm": 0.3106028139591217, + "learning_rate": 2.679266666666667e-05, + "loss": 0.012, + "step": 19628 + }, + { + "epoch": 16.5785472972973, + "grad_norm": 0.27781546115875244, + "learning_rate": 2.6792333333333334e-05, + "loss": 0.0136, + "step": 19629 + }, + { + "epoch": 16.57939189189189, + "grad_norm": 0.20655052363872528, + "learning_rate": 2.6792e-05, + "loss": 0.0176, + "step": 19630 + }, + { + "epoch": 16.580236486486488, + "grad_norm": 0.1281420886516571, + "learning_rate": 2.679166666666667e-05, + "loss": 0.0075, + "step": 19631 + }, + { + "epoch": 16.58108108108108, + "grad_norm": 0.5573215484619141, + "learning_rate": 2.6791333333333332e-05, + "loss": 0.008, + "step": 19632 + }, + { + "epoch": 16.581925675675677, + "grad_norm": 0.24396300315856934, + "learning_rate": 2.6791e-05, + "loss": 0.0123, + "step": 19633 + }, + { + "epoch": 16.58277027027027, + "grad_norm": 0.11969365924596786, + "learning_rate": 2.6790666666666667e-05, + "loss": 0.0083, + "step": 19634 + }, + { + "epoch": 16.583614864864863, + "grad_norm": 0.2692713439464569, + "learning_rate": 2.6790333333333333e-05, + "loss": 0.0124, + "step": 19635 + }, + { + "epoch": 16.58445945945946, + "grad_norm": 0.18368478119373322, + "learning_rate": 2.6790000000000003e-05, + "loss": 0.0087, + "step": 19636 + }, + { + "epoch": 16.585304054054053, + "grad_norm": 0.4186718463897705, + "learning_rate": 2.678966666666667e-05, + "loss": 0.0147, + "step": 19637 + }, + { + "epoch": 16.58614864864865, + "grad_norm": 0.34088265895843506, + "learning_rate": 2.6789333333333335e-05, + "loss": 0.0146, + "step": 19638 + }, + { + "epoch": 16.586993243243242, + "grad_norm": 0.26284265518188477, + "learning_rate": 2.6789e-05, + "loss": 0.0089, + "step": 19639 + }, + { + "epoch": 16.58783783783784, + "grad_norm": 0.28005361557006836, + "learning_rate": 2.6788666666666666e-05, + "loss": 0.0126, + "step": 19640 + }, + { + "epoch": 16.58868243243243, + "grad_norm": 0.4864574372768402, + "learning_rate": 2.6788333333333332e-05, + "loss": 0.0146, + "step": 19641 + }, + { + "epoch": 16.589527027027028, + "grad_norm": 0.11935680359601974, + "learning_rate": 2.6788000000000002e-05, + "loss": 0.0047, + "step": 19642 + }, + { + "epoch": 16.59037162162162, + "grad_norm": 0.3700384795665741, + "learning_rate": 2.6787666666666664e-05, + "loss": 0.0162, + "step": 19643 + }, + { + "epoch": 16.591216216216218, + "grad_norm": 0.4890389144420624, + "learning_rate": 2.6787333333333334e-05, + "loss": 0.0247, + "step": 19644 + }, + { + "epoch": 16.59206081081081, + "grad_norm": 0.6996914148330688, + "learning_rate": 2.6787000000000003e-05, + "loss": 0.1741, + "step": 19645 + }, + { + "epoch": 16.592905405405407, + "grad_norm": 0.4849957227706909, + "learning_rate": 2.6786666666666665e-05, + "loss": 0.0933, + "step": 19646 + }, + { + "epoch": 16.59375, + "grad_norm": 0.35311445593833923, + "learning_rate": 2.6786333333333335e-05, + "loss": 0.0545, + "step": 19647 + }, + { + "epoch": 16.594594594594593, + "grad_norm": 0.23567400872707367, + "learning_rate": 2.6786e-05, + "loss": 0.0185, + "step": 19648 + }, + { + "epoch": 16.59543918918919, + "grad_norm": 0.3102905750274658, + "learning_rate": 2.6785666666666667e-05, + "loss": 0.033, + "step": 19649 + }, + { + "epoch": 16.596283783783782, + "grad_norm": 0.25538334250450134, + "learning_rate": 2.6785333333333333e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 16.59712837837838, + "grad_norm": 0.19469939172267914, + "learning_rate": 2.6785000000000002e-05, + "loss": 0.0163, + "step": 19651 + }, + { + "epoch": 16.597972972972972, + "grad_norm": 0.19361411035060883, + "learning_rate": 2.6784666666666668e-05, + "loss": 0.0082, + "step": 19652 + }, + { + "epoch": 16.59881756756757, + "grad_norm": 0.1791425347328186, + "learning_rate": 2.6784333333333334e-05, + "loss": 0.0096, + "step": 19653 + }, + { + "epoch": 16.59966216216216, + "grad_norm": 0.25997960567474365, + "learning_rate": 2.6784000000000003e-05, + "loss": 0.0164, + "step": 19654 + }, + { + "epoch": 16.600506756756758, + "grad_norm": 0.1475990265607834, + "learning_rate": 2.6783666666666666e-05, + "loss": 0.0088, + "step": 19655 + }, + { + "epoch": 16.60135135135135, + "grad_norm": 0.19411127269268036, + "learning_rate": 2.6783333333333335e-05, + "loss": 0.0149, + "step": 19656 + }, + { + "epoch": 16.602195945945947, + "grad_norm": 0.17410144209861755, + "learning_rate": 2.6783e-05, + "loss": 0.0137, + "step": 19657 + }, + { + "epoch": 16.60304054054054, + "grad_norm": 0.45771524310112, + "learning_rate": 2.6782666666666667e-05, + "loss": 0.0236, + "step": 19658 + }, + { + "epoch": 16.603885135135137, + "grad_norm": 0.20712929964065552, + "learning_rate": 2.6782333333333333e-05, + "loss": 0.017, + "step": 19659 + }, + { + "epoch": 16.60472972972973, + "grad_norm": 0.2657155692577362, + "learning_rate": 2.6782000000000002e-05, + "loss": 0.0118, + "step": 19660 + }, + { + "epoch": 16.605574324324323, + "grad_norm": 0.20073214173316956, + "learning_rate": 2.6781666666666668e-05, + "loss": 0.0141, + "step": 19661 + }, + { + "epoch": 16.60641891891892, + "grad_norm": 0.16992273926734924, + "learning_rate": 2.6781333333333334e-05, + "loss": 0.0129, + "step": 19662 + }, + { + "epoch": 16.607263513513512, + "grad_norm": 0.25129586458206177, + "learning_rate": 2.6781000000000003e-05, + "loss": 0.0159, + "step": 19663 + }, + { + "epoch": 16.60810810810811, + "grad_norm": 0.17868909239768982, + "learning_rate": 2.6780666666666666e-05, + "loss": 0.0111, + "step": 19664 + }, + { + "epoch": 16.6089527027027, + "grad_norm": 0.5204295516014099, + "learning_rate": 2.6780333333333335e-05, + "loss": 0.0125, + "step": 19665 + }, + { + "epoch": 16.6097972972973, + "grad_norm": 0.9833913445472717, + "learning_rate": 2.678e-05, + "loss": 0.021, + "step": 19666 + }, + { + "epoch": 16.61064189189189, + "grad_norm": 0.23980167508125305, + "learning_rate": 2.6779666666666667e-05, + "loss": 0.0145, + "step": 19667 + }, + { + "epoch": 16.611486486486488, + "grad_norm": 0.24394552409648895, + "learning_rate": 2.6779333333333333e-05, + "loss": 0.0117, + "step": 19668 + }, + { + "epoch": 16.61233108108108, + "grad_norm": 0.1412329226732254, + "learning_rate": 2.6779e-05, + "loss": 0.007, + "step": 19669 + }, + { + "epoch": 16.613175675675677, + "grad_norm": 0.414661705493927, + "learning_rate": 2.6778666666666668e-05, + "loss": 0.0153, + "step": 19670 + }, + { + "epoch": 16.61402027027027, + "grad_norm": 0.21028244495391846, + "learning_rate": 2.6778333333333334e-05, + "loss": 0.0101, + "step": 19671 + }, + { + "epoch": 16.614864864864863, + "grad_norm": 0.20514293015003204, + "learning_rate": 2.6778e-05, + "loss": 0.0093, + "step": 19672 + }, + { + "epoch": 16.61570945945946, + "grad_norm": 0.178190216422081, + "learning_rate": 2.6777666666666666e-05, + "loss": 0.0148, + "step": 19673 + }, + { + "epoch": 16.616554054054053, + "grad_norm": 0.22508229315280914, + "learning_rate": 2.6777333333333335e-05, + "loss": 0.0126, + "step": 19674 + }, + { + "epoch": 16.61739864864865, + "grad_norm": 0.1304350048303604, + "learning_rate": 2.6776999999999998e-05, + "loss": 0.0083, + "step": 19675 + }, + { + "epoch": 16.618243243243242, + "grad_norm": 0.2766498029232025, + "learning_rate": 2.6776666666666667e-05, + "loss": 0.0171, + "step": 19676 + }, + { + "epoch": 16.61908783783784, + "grad_norm": 0.19217149913311005, + "learning_rate": 2.6776333333333337e-05, + "loss": 0.0099, + "step": 19677 + }, + { + "epoch": 16.61993243243243, + "grad_norm": 0.20488953590393066, + "learning_rate": 2.6776e-05, + "loss": 0.0167, + "step": 19678 + }, + { + "epoch": 16.620777027027028, + "grad_norm": 0.2972652018070221, + "learning_rate": 2.677566666666667e-05, + "loss": 0.0143, + "step": 19679 + }, + { + "epoch": 16.62162162162162, + "grad_norm": 0.8560664057731628, + "learning_rate": 2.6775333333333334e-05, + "loss": 0.0139, + "step": 19680 + }, + { + "epoch": 16.622466216216218, + "grad_norm": 0.16154709458351135, + "learning_rate": 2.6775e-05, + "loss": 0.0108, + "step": 19681 + }, + { + "epoch": 16.62331081081081, + "grad_norm": 2.388894557952881, + "learning_rate": 2.6774666666666666e-05, + "loss": 0.0667, + "step": 19682 + }, + { + "epoch": 16.624155405405407, + "grad_norm": 1.065908432006836, + "learning_rate": 2.6774333333333336e-05, + "loss": 0.0163, + "step": 19683 + }, + { + "epoch": 16.625, + "grad_norm": 0.19664804637432098, + "learning_rate": 2.6773999999999998e-05, + "loss": 0.0132, + "step": 19684 + }, + { + "epoch": 16.625844594594593, + "grad_norm": 0.3197903037071228, + "learning_rate": 2.6773666666666667e-05, + "loss": 0.0131, + "step": 19685 + }, + { + "epoch": 16.62668918918919, + "grad_norm": 0.6318233609199524, + "learning_rate": 2.6773333333333337e-05, + "loss": 0.0115, + "step": 19686 + }, + { + "epoch": 16.627533783783782, + "grad_norm": 0.6883628964424133, + "learning_rate": 2.6773e-05, + "loss": 0.0257, + "step": 19687 + }, + { + "epoch": 16.62837837837838, + "grad_norm": 0.3760039210319519, + "learning_rate": 2.677266666666667e-05, + "loss": 0.0112, + "step": 19688 + }, + { + "epoch": 16.629222972972972, + "grad_norm": 0.18542274832725525, + "learning_rate": 2.6772333333333335e-05, + "loss": 0.0106, + "step": 19689 + }, + { + "epoch": 16.63006756756757, + "grad_norm": 0.19158650934696198, + "learning_rate": 2.6772e-05, + "loss": 0.0139, + "step": 19690 + }, + { + "epoch": 16.63091216216216, + "grad_norm": 0.24013561010360718, + "learning_rate": 2.6771666666666666e-05, + "loss": 0.0098, + "step": 19691 + }, + { + "epoch": 16.631756756756758, + "grad_norm": 0.29934579133987427, + "learning_rate": 2.6771333333333336e-05, + "loss": 0.0069, + "step": 19692 + }, + { + "epoch": 16.63260135135135, + "grad_norm": 1.497539758682251, + "learning_rate": 2.6770999999999998e-05, + "loss": 0.0211, + "step": 19693 + }, + { + "epoch": 16.633445945945947, + "grad_norm": 1.2230948209762573, + "learning_rate": 2.6770666666666668e-05, + "loss": 0.0362, + "step": 19694 + }, + { + "epoch": 16.63429054054054, + "grad_norm": 0.6474587917327881, + "learning_rate": 2.6770333333333337e-05, + "loss": 0.1455, + "step": 19695 + }, + { + "epoch": 16.635135135135137, + "grad_norm": 0.4561708867549896, + "learning_rate": 2.677e-05, + "loss": 0.1048, + "step": 19696 + }, + { + "epoch": 16.63597972972973, + "grad_norm": 0.3983246982097626, + "learning_rate": 2.676966666666667e-05, + "loss": 0.0419, + "step": 19697 + }, + { + "epoch": 16.636824324324323, + "grad_norm": 0.3627442419528961, + "learning_rate": 2.6769333333333335e-05, + "loss": 0.0239, + "step": 19698 + }, + { + "epoch": 16.63766891891892, + "grad_norm": 0.26666101813316345, + "learning_rate": 2.6769e-05, + "loss": 0.0188, + "step": 19699 + }, + { + "epoch": 16.638513513513512, + "grad_norm": 0.3254402279853821, + "learning_rate": 2.6768666666666667e-05, + "loss": 0.0203, + "step": 19700 + }, + { + "epoch": 16.63935810810811, + "grad_norm": 0.2907671630382538, + "learning_rate": 2.6768333333333333e-05, + "loss": 0.0205, + "step": 19701 + }, + { + "epoch": 16.6402027027027, + "grad_norm": 0.25455448031425476, + "learning_rate": 2.6768000000000002e-05, + "loss": 0.0126, + "step": 19702 + }, + { + "epoch": 16.6410472972973, + "grad_norm": 0.2278524935245514, + "learning_rate": 2.6767666666666668e-05, + "loss": 0.0158, + "step": 19703 + }, + { + "epoch": 16.64189189189189, + "grad_norm": 0.5229688286781311, + "learning_rate": 2.6767333333333334e-05, + "loss": 0.0229, + "step": 19704 + }, + { + "epoch": 16.642736486486488, + "grad_norm": 0.2521594166755676, + "learning_rate": 2.6767e-05, + "loss": 0.0198, + "step": 19705 + }, + { + "epoch": 16.64358108108108, + "grad_norm": 0.3436254560947418, + "learning_rate": 2.676666666666667e-05, + "loss": 0.0127, + "step": 19706 + }, + { + "epoch": 16.644425675675677, + "grad_norm": 0.20491398870944977, + "learning_rate": 2.676633333333333e-05, + "loss": 0.01, + "step": 19707 + }, + { + "epoch": 16.64527027027027, + "grad_norm": 0.4816928207874298, + "learning_rate": 2.6766e-05, + "loss": 0.0265, + "step": 19708 + }, + { + "epoch": 16.646114864864863, + "grad_norm": 0.26033973693847656, + "learning_rate": 2.6765666666666667e-05, + "loss": 0.0196, + "step": 19709 + }, + { + "epoch": 16.64695945945946, + "grad_norm": 0.35802143812179565, + "learning_rate": 2.6765333333333333e-05, + "loss": 0.0124, + "step": 19710 + }, + { + "epoch": 16.647804054054053, + "grad_norm": 0.48838409781455994, + "learning_rate": 2.6765000000000002e-05, + "loss": 0.0145, + "step": 19711 + }, + { + "epoch": 16.64864864864865, + "grad_norm": 0.5270443558692932, + "learning_rate": 2.6764666666666668e-05, + "loss": 0.0141, + "step": 19712 + }, + { + "epoch": 16.649493243243242, + "grad_norm": 0.29766666889190674, + "learning_rate": 2.6764333333333334e-05, + "loss": 0.0141, + "step": 19713 + }, + { + "epoch": 16.65033783783784, + "grad_norm": 0.20740430057048798, + "learning_rate": 2.6764e-05, + "loss": 0.0177, + "step": 19714 + }, + { + "epoch": 16.65118243243243, + "grad_norm": 0.25915372371673584, + "learning_rate": 2.676366666666667e-05, + "loss": 0.0121, + "step": 19715 + }, + { + "epoch": 16.652027027027028, + "grad_norm": 0.21872971951961517, + "learning_rate": 2.6763333333333332e-05, + "loss": 0.0166, + "step": 19716 + }, + { + "epoch": 16.65287162162162, + "grad_norm": 0.17478974163532257, + "learning_rate": 2.6763e-05, + "loss": 0.0152, + "step": 19717 + }, + { + "epoch": 16.653716216216218, + "grad_norm": 0.2961994707584381, + "learning_rate": 2.6762666666666667e-05, + "loss": 0.0186, + "step": 19718 + }, + { + "epoch": 16.65456081081081, + "grad_norm": 0.5299079418182373, + "learning_rate": 2.6762333333333333e-05, + "loss": 0.0139, + "step": 19719 + }, + { + "epoch": 16.655405405405407, + "grad_norm": 0.5015304684638977, + "learning_rate": 2.6762000000000002e-05, + "loss": 0.0204, + "step": 19720 + }, + { + "epoch": 16.65625, + "grad_norm": 0.430972158908844, + "learning_rate": 2.6761666666666668e-05, + "loss": 0.0166, + "step": 19721 + }, + { + "epoch": 16.657094594594593, + "grad_norm": 0.33247146010398865, + "learning_rate": 2.6761333333333334e-05, + "loss": 0.0143, + "step": 19722 + }, + { + "epoch": 16.65793918918919, + "grad_norm": 0.2549715042114258, + "learning_rate": 2.6761e-05, + "loss": 0.0159, + "step": 19723 + }, + { + "epoch": 16.658783783783782, + "grad_norm": 0.15047821402549744, + "learning_rate": 2.676066666666667e-05, + "loss": 0.0079, + "step": 19724 + }, + { + "epoch": 16.65962837837838, + "grad_norm": 0.31141215562820435, + "learning_rate": 2.6760333333333332e-05, + "loss": 0.0092, + "step": 19725 + }, + { + "epoch": 16.660472972972972, + "grad_norm": 0.2088351547718048, + "learning_rate": 2.676e-05, + "loss": 0.008, + "step": 19726 + }, + { + "epoch": 16.66131756756757, + "grad_norm": 0.2023894488811493, + "learning_rate": 2.675966666666667e-05, + "loss": 0.0055, + "step": 19727 + }, + { + "epoch": 16.66216216216216, + "grad_norm": 0.17572084069252014, + "learning_rate": 2.6759333333333333e-05, + "loss": 0.0066, + "step": 19728 + }, + { + "epoch": 16.663006756756758, + "grad_norm": 0.2338770180940628, + "learning_rate": 2.6759000000000002e-05, + "loss": 0.0123, + "step": 19729 + }, + { + "epoch": 16.66385135135135, + "grad_norm": 0.26423943042755127, + "learning_rate": 2.6758666666666665e-05, + "loss": 0.0161, + "step": 19730 + }, + { + "epoch": 16.664695945945947, + "grad_norm": 0.32976987957954407, + "learning_rate": 2.6758333333333334e-05, + "loss": 0.0095, + "step": 19731 + }, + { + "epoch": 16.66554054054054, + "grad_norm": 0.21608763933181763, + "learning_rate": 2.6758e-05, + "loss": 0.0134, + "step": 19732 + }, + { + "epoch": 16.666385135135137, + "grad_norm": 1.3213192224502563, + "learning_rate": 2.6757666666666666e-05, + "loss": 0.0159, + "step": 19733 + }, + { + "epoch": 16.66722972972973, + "grad_norm": 0.3423996567726135, + "learning_rate": 2.6757333333333332e-05, + "loss": 0.0231, + "step": 19734 + }, + { + "epoch": 16.668074324324323, + "grad_norm": 0.45726269483566284, + "learning_rate": 2.6757e-05, + "loss": 0.0139, + "step": 19735 + }, + { + "epoch": 16.66891891891892, + "grad_norm": 1.679822325706482, + "learning_rate": 2.6756666666666667e-05, + "loss": 0.0115, + "step": 19736 + }, + { + "epoch": 16.669763513513512, + "grad_norm": 1.4289170503616333, + "learning_rate": 2.6756333333333333e-05, + "loss": 0.0241, + "step": 19737 + }, + { + "epoch": 16.67060810810811, + "grad_norm": 0.24985644221305847, + "learning_rate": 2.6756000000000003e-05, + "loss": 0.0143, + "step": 19738 + }, + { + "epoch": 16.6714527027027, + "grad_norm": 0.605256199836731, + "learning_rate": 2.6755666666666665e-05, + "loss": 0.059, + "step": 19739 + }, + { + "epoch": 16.6722972972973, + "grad_norm": 0.36632993817329407, + "learning_rate": 2.6755333333333335e-05, + "loss": 0.0232, + "step": 19740 + }, + { + "epoch": 16.67314189189189, + "grad_norm": 0.8169874548912048, + "learning_rate": 2.6755e-05, + "loss": 0.0235, + "step": 19741 + }, + { + "epoch": 16.673986486486488, + "grad_norm": 0.21101436018943787, + "learning_rate": 2.6754666666666666e-05, + "loss": 0.0123, + "step": 19742 + }, + { + "epoch": 16.67483108108108, + "grad_norm": 0.3730694353580475, + "learning_rate": 2.6754333333333332e-05, + "loss": 0.0218, + "step": 19743 + }, + { + "epoch": 16.675675675675677, + "grad_norm": 1.1550092697143555, + "learning_rate": 2.6754e-05, + "loss": 0.0351, + "step": 19744 + }, + { + "epoch": 16.67652027027027, + "grad_norm": 0.6485347151756287, + "learning_rate": 2.6753666666666668e-05, + "loss": 0.1592, + "step": 19745 + }, + { + "epoch": 16.677364864864863, + "grad_norm": 0.4143927991390228, + "learning_rate": 2.6753333333333334e-05, + "loss": 0.0981, + "step": 19746 + }, + { + "epoch": 16.67820945945946, + "grad_norm": 0.35555920004844666, + "learning_rate": 2.6753000000000003e-05, + "loss": 0.0724, + "step": 19747 + }, + { + "epoch": 16.679054054054053, + "grad_norm": 0.2624872326850891, + "learning_rate": 2.6752666666666665e-05, + "loss": 0.0423, + "step": 19748 + }, + { + "epoch": 16.67989864864865, + "grad_norm": 0.2647895812988281, + "learning_rate": 2.6752333333333335e-05, + "loss": 0.0239, + "step": 19749 + }, + { + "epoch": 16.680743243243242, + "grad_norm": 0.3505138158798218, + "learning_rate": 2.6752e-05, + "loss": 0.0344, + "step": 19750 + }, + { + "epoch": 16.68158783783784, + "grad_norm": 0.14824619889259338, + "learning_rate": 2.6751666666666667e-05, + "loss": 0.0123, + "step": 19751 + }, + { + "epoch": 16.68243243243243, + "grad_norm": 0.22241473197937012, + "learning_rate": 2.6751333333333336e-05, + "loss": 0.018, + "step": 19752 + }, + { + "epoch": 16.683277027027028, + "grad_norm": 0.25581657886505127, + "learning_rate": 2.6751000000000002e-05, + "loss": 0.0171, + "step": 19753 + }, + { + "epoch": 16.68412162162162, + "grad_norm": 0.18314576148986816, + "learning_rate": 2.6750666666666668e-05, + "loss": 0.0131, + "step": 19754 + }, + { + "epoch": 16.684966216216218, + "grad_norm": 0.4533959627151489, + "learning_rate": 2.6750333333333334e-05, + "loss": 0.0363, + "step": 19755 + }, + { + "epoch": 16.68581081081081, + "grad_norm": 0.22962123155593872, + "learning_rate": 2.6750000000000003e-05, + "loss": 0.0164, + "step": 19756 + }, + { + "epoch": 16.686655405405407, + "grad_norm": 0.1948186159133911, + "learning_rate": 2.6749666666666666e-05, + "loss": 0.021, + "step": 19757 + }, + { + "epoch": 16.6875, + "grad_norm": 0.23972763121128082, + "learning_rate": 2.6749333333333335e-05, + "loss": 0.0129, + "step": 19758 + }, + { + "epoch": 16.688344594594593, + "grad_norm": 0.2735106945037842, + "learning_rate": 2.6749e-05, + "loss": 0.0073, + "step": 19759 + }, + { + "epoch": 16.68918918918919, + "grad_norm": 0.16093960404396057, + "learning_rate": 2.6748666666666667e-05, + "loss": 0.0093, + "step": 19760 + }, + { + "epoch": 16.690033783783782, + "grad_norm": 0.24007615447044373, + "learning_rate": 2.6748333333333336e-05, + "loss": 0.0195, + "step": 19761 + }, + { + "epoch": 16.69087837837838, + "grad_norm": 0.28386691212654114, + "learning_rate": 2.6748e-05, + "loss": 0.016, + "step": 19762 + }, + { + "epoch": 16.691722972972972, + "grad_norm": 0.13939537107944489, + "learning_rate": 2.6747666666666668e-05, + "loss": 0.0107, + "step": 19763 + }, + { + "epoch": 16.69256756756757, + "grad_norm": 0.15558373928070068, + "learning_rate": 2.6747333333333334e-05, + "loss": 0.0103, + "step": 19764 + }, + { + "epoch": 16.69341216216216, + "grad_norm": 0.22777582705020905, + "learning_rate": 2.6747e-05, + "loss": 0.0145, + "step": 19765 + }, + { + "epoch": 16.694256756756758, + "grad_norm": 0.15719205141067505, + "learning_rate": 2.6746666666666666e-05, + "loss": 0.0125, + "step": 19766 + }, + { + "epoch": 16.69510135135135, + "grad_norm": 0.24734075367450714, + "learning_rate": 2.6746333333333335e-05, + "loss": 0.0511, + "step": 19767 + }, + { + "epoch": 16.695945945945947, + "grad_norm": 0.24458159506320953, + "learning_rate": 2.6745999999999998e-05, + "loss": 0.0216, + "step": 19768 + }, + { + "epoch": 16.69679054054054, + "grad_norm": 0.2290378212928772, + "learning_rate": 2.6745666666666667e-05, + "loss": 0.0108, + "step": 19769 + }, + { + "epoch": 16.697635135135137, + "grad_norm": 0.4530847370624542, + "learning_rate": 2.6745333333333336e-05, + "loss": 0.0241, + "step": 19770 + }, + { + "epoch": 16.69847972972973, + "grad_norm": 0.2336363047361374, + "learning_rate": 2.6745e-05, + "loss": 0.0093, + "step": 19771 + }, + { + "epoch": 16.699324324324323, + "grad_norm": 0.3865707516670227, + "learning_rate": 2.6744666666666668e-05, + "loss": 0.0155, + "step": 19772 + }, + { + "epoch": 16.70016891891892, + "grad_norm": 0.34270694851875305, + "learning_rate": 2.6744333333333334e-05, + "loss": 0.0165, + "step": 19773 + }, + { + "epoch": 16.701013513513512, + "grad_norm": 0.22067774832248688, + "learning_rate": 2.6744e-05, + "loss": 0.0126, + "step": 19774 + }, + { + "epoch": 16.70185810810811, + "grad_norm": 0.25973209738731384, + "learning_rate": 2.6743666666666666e-05, + "loss": 0.019, + "step": 19775 + }, + { + "epoch": 16.7027027027027, + "grad_norm": 0.3173901438713074, + "learning_rate": 2.6743333333333335e-05, + "loss": 0.0143, + "step": 19776 + }, + { + "epoch": 16.7035472972973, + "grad_norm": 0.2426389455795288, + "learning_rate": 2.6743e-05, + "loss": 0.0103, + "step": 19777 + }, + { + "epoch": 16.70439189189189, + "grad_norm": 0.2440708726644516, + "learning_rate": 2.6742666666666667e-05, + "loss": 0.0138, + "step": 19778 + }, + { + "epoch": 16.705236486486488, + "grad_norm": 0.5036393404006958, + "learning_rate": 2.6742333333333337e-05, + "loss": 0.0113, + "step": 19779 + }, + { + "epoch": 16.70608108108108, + "grad_norm": 0.16707319021224976, + "learning_rate": 2.6742e-05, + "loss": 0.0172, + "step": 19780 + }, + { + "epoch": 16.706925675675677, + "grad_norm": 0.15628743171691895, + "learning_rate": 2.674166666666667e-05, + "loss": 0.0135, + "step": 19781 + }, + { + "epoch": 16.70777027027027, + "grad_norm": 0.21712419390678406, + "learning_rate": 2.6741333333333334e-05, + "loss": 0.008, + "step": 19782 + }, + { + "epoch": 16.708614864864863, + "grad_norm": 0.49104076623916626, + "learning_rate": 2.6741e-05, + "loss": 0.0111, + "step": 19783 + }, + { + "epoch": 16.70945945945946, + "grad_norm": 0.3367590010166168, + "learning_rate": 2.6740666666666666e-05, + "loss": 0.0141, + "step": 19784 + }, + { + "epoch": 16.710304054054053, + "grad_norm": 0.25996506214141846, + "learning_rate": 2.6740333333333336e-05, + "loss": 0.0142, + "step": 19785 + }, + { + "epoch": 16.71114864864865, + "grad_norm": 0.6841756701469421, + "learning_rate": 2.674e-05, + "loss": 0.0094, + "step": 19786 + }, + { + "epoch": 16.711993243243242, + "grad_norm": 0.2212696224451065, + "learning_rate": 2.6739666666666667e-05, + "loss": 0.0136, + "step": 19787 + }, + { + "epoch": 16.71283783783784, + "grad_norm": 0.373617023229599, + "learning_rate": 2.6739333333333337e-05, + "loss": 0.0181, + "step": 19788 + }, + { + "epoch": 16.71368243243243, + "grad_norm": 0.6775513887405396, + "learning_rate": 2.6739e-05, + "loss": 0.0118, + "step": 19789 + }, + { + "epoch": 16.714527027027028, + "grad_norm": 0.864284873008728, + "learning_rate": 2.673866666666667e-05, + "loss": 0.0161, + "step": 19790 + }, + { + "epoch": 16.71537162162162, + "grad_norm": 0.35638928413391113, + "learning_rate": 2.673833333333333e-05, + "loss": 0.0084, + "step": 19791 + }, + { + "epoch": 16.716216216216218, + "grad_norm": 0.6962651014328003, + "learning_rate": 2.6738e-05, + "loss": 0.0245, + "step": 19792 + }, + { + "epoch": 16.71706081081081, + "grad_norm": 0.43913912773132324, + "learning_rate": 2.6737666666666666e-05, + "loss": 0.0182, + "step": 19793 + }, + { + "epoch": 16.717905405405407, + "grad_norm": 0.3581584095954895, + "learning_rate": 2.6737333333333332e-05, + "loss": 0.0151, + "step": 19794 + }, + { + "epoch": 16.71875, + "grad_norm": 0.45820820331573486, + "learning_rate": 2.6737e-05, + "loss": 0.1532, + "step": 19795 + }, + { + "epoch": 16.719594594594593, + "grad_norm": 0.5551906228065491, + "learning_rate": 2.6736666666666668e-05, + "loss": 0.1065, + "step": 19796 + }, + { + "epoch": 16.72043918918919, + "grad_norm": 0.3579467236995697, + "learning_rate": 2.6736333333333334e-05, + "loss": 0.0466, + "step": 19797 + }, + { + "epoch": 16.721283783783782, + "grad_norm": 0.3430355489253998, + "learning_rate": 2.6736e-05, + "loss": 0.0367, + "step": 19798 + }, + { + "epoch": 16.72212837837838, + "grad_norm": 0.39417126774787903, + "learning_rate": 2.673566666666667e-05, + "loss": 0.0219, + "step": 19799 + }, + { + "epoch": 16.722972972972972, + "grad_norm": 0.5524218082427979, + "learning_rate": 2.673533333333333e-05, + "loss": 0.0253, + "step": 19800 + }, + { + "epoch": 16.72381756756757, + "grad_norm": 0.17675311863422394, + "learning_rate": 2.6735e-05, + "loss": 0.0113, + "step": 19801 + }, + { + "epoch": 16.72466216216216, + "grad_norm": 0.4539445638656616, + "learning_rate": 2.673466666666667e-05, + "loss": 0.0265, + "step": 19802 + }, + { + "epoch": 16.725506756756758, + "grad_norm": 0.2767810523509979, + "learning_rate": 2.6734333333333333e-05, + "loss": 0.018, + "step": 19803 + }, + { + "epoch": 16.72635135135135, + "grad_norm": 0.21229559183120728, + "learning_rate": 2.6734000000000002e-05, + "loss": 0.0097, + "step": 19804 + }, + { + "epoch": 16.727195945945947, + "grad_norm": 0.3080562949180603, + "learning_rate": 2.6733666666666668e-05, + "loss": 0.0233, + "step": 19805 + }, + { + "epoch": 16.72804054054054, + "grad_norm": 0.3543163239955902, + "learning_rate": 2.6733333333333334e-05, + "loss": 0.0268, + "step": 19806 + }, + { + "epoch": 16.728885135135137, + "grad_norm": 0.1987989842891693, + "learning_rate": 2.6733e-05, + "loss": 0.0117, + "step": 19807 + }, + { + "epoch": 16.72972972972973, + "grad_norm": 0.24347665905952454, + "learning_rate": 2.673266666666667e-05, + "loss": 0.015, + "step": 19808 + }, + { + "epoch": 16.730574324324323, + "grad_norm": 0.21856525540351868, + "learning_rate": 2.673233333333333e-05, + "loss": 0.0156, + "step": 19809 + }, + { + "epoch": 16.73141891891892, + "grad_norm": 0.36064112186431885, + "learning_rate": 2.6732e-05, + "loss": 0.0136, + "step": 19810 + }, + { + "epoch": 16.732263513513512, + "grad_norm": 0.19558174908161163, + "learning_rate": 2.673166666666667e-05, + "loss": 0.0135, + "step": 19811 + }, + { + "epoch": 16.73310810810811, + "grad_norm": 0.3559859097003937, + "learning_rate": 2.6731333333333333e-05, + "loss": 0.0113, + "step": 19812 + }, + { + "epoch": 16.7339527027027, + "grad_norm": 0.42814210057258606, + "learning_rate": 2.6731000000000002e-05, + "loss": 0.0146, + "step": 19813 + }, + { + "epoch": 16.7347972972973, + "grad_norm": 0.1894991546869278, + "learning_rate": 2.6730666666666668e-05, + "loss": 0.013, + "step": 19814 + }, + { + "epoch": 16.73564189189189, + "grad_norm": 0.2340611219406128, + "learning_rate": 2.6730333333333334e-05, + "loss": 0.0191, + "step": 19815 + }, + { + "epoch": 16.736486486486488, + "grad_norm": 0.393179714679718, + "learning_rate": 2.673e-05, + "loss": 0.0188, + "step": 19816 + }, + { + "epoch": 16.73733108108108, + "grad_norm": 0.2578005790710449, + "learning_rate": 2.672966666666667e-05, + "loss": 0.0112, + "step": 19817 + }, + { + "epoch": 16.738175675675677, + "grad_norm": 0.41842976212501526, + "learning_rate": 2.6729333333333332e-05, + "loss": 0.0206, + "step": 19818 + }, + { + "epoch": 16.73902027027027, + "grad_norm": 0.27694594860076904, + "learning_rate": 2.6729e-05, + "loss": 0.0191, + "step": 19819 + }, + { + "epoch": 16.739864864864863, + "grad_norm": 0.49591782689094543, + "learning_rate": 2.6728666666666667e-05, + "loss": 0.0299, + "step": 19820 + }, + { + "epoch": 16.74070945945946, + "grad_norm": 0.4061344563961029, + "learning_rate": 2.6728333333333333e-05, + "loss": 0.0184, + "step": 19821 + }, + { + "epoch": 16.741554054054053, + "grad_norm": 0.3248167634010315, + "learning_rate": 2.6728000000000002e-05, + "loss": 0.0166, + "step": 19822 + }, + { + "epoch": 16.74239864864865, + "grad_norm": 0.3107462525367737, + "learning_rate": 2.6727666666666665e-05, + "loss": 0.0155, + "step": 19823 + }, + { + "epoch": 16.743243243243242, + "grad_norm": 0.44482946395874023, + "learning_rate": 2.6727333333333334e-05, + "loss": 0.0193, + "step": 19824 + }, + { + "epoch": 16.74408783783784, + "grad_norm": 0.36121320724487305, + "learning_rate": 2.6727e-05, + "loss": 0.0135, + "step": 19825 + }, + { + "epoch": 16.74493243243243, + "grad_norm": 0.28318896889686584, + "learning_rate": 2.6726666666666666e-05, + "loss": 0.0218, + "step": 19826 + }, + { + "epoch": 16.745777027027028, + "grad_norm": 0.2997457981109619, + "learning_rate": 2.6726333333333335e-05, + "loss": 0.0174, + "step": 19827 + }, + { + "epoch": 16.74662162162162, + "grad_norm": 0.394016295671463, + "learning_rate": 2.6726e-05, + "loss": 0.0631, + "step": 19828 + }, + { + "epoch": 16.747466216216218, + "grad_norm": 0.3715249300003052, + "learning_rate": 2.6725666666666667e-05, + "loss": 0.0141, + "step": 19829 + }, + { + "epoch": 16.74831081081081, + "grad_norm": 0.32825911045074463, + "learning_rate": 2.6725333333333333e-05, + "loss": 0.0168, + "step": 19830 + }, + { + "epoch": 16.749155405405407, + "grad_norm": 0.3008241653442383, + "learning_rate": 2.6725000000000002e-05, + "loss": 0.0145, + "step": 19831 + }, + { + "epoch": 16.75, + "grad_norm": 0.4571862518787384, + "learning_rate": 2.6724666666666665e-05, + "loss": 0.0234, + "step": 19832 + }, + { + "epoch": 16.750844594594593, + "grad_norm": 0.2735479772090912, + "learning_rate": 2.6724333333333334e-05, + "loss": 0.0146, + "step": 19833 + }, + { + "epoch": 16.75168918918919, + "grad_norm": 0.23305924236774445, + "learning_rate": 2.6724e-05, + "loss": 0.0128, + "step": 19834 + }, + { + "epoch": 16.752533783783782, + "grad_norm": 0.2730776369571686, + "learning_rate": 2.6723666666666666e-05, + "loss": 0.0154, + "step": 19835 + }, + { + "epoch": 16.75337837837838, + "grad_norm": 0.3416961431503296, + "learning_rate": 2.6723333333333336e-05, + "loss": 0.0139, + "step": 19836 + }, + { + "epoch": 16.754222972972972, + "grad_norm": 0.3455730378627777, + "learning_rate": 2.6723e-05, + "loss": 0.0099, + "step": 19837 + }, + { + "epoch": 16.75506756756757, + "grad_norm": 0.23445332050323486, + "learning_rate": 2.6722666666666667e-05, + "loss": 0.0123, + "step": 19838 + }, + { + "epoch": 16.75591216216216, + "grad_norm": 0.17545926570892334, + "learning_rate": 2.6722333333333333e-05, + "loss": 0.0075, + "step": 19839 + }, + { + "epoch": 16.756756756756758, + "grad_norm": 0.2429177165031433, + "learning_rate": 2.6722000000000003e-05, + "loss": 0.0119, + "step": 19840 + }, + { + "epoch": 16.75760135135135, + "grad_norm": 0.28588178753852844, + "learning_rate": 2.6721666666666665e-05, + "loss": 0.0106, + "step": 19841 + }, + { + "epoch": 16.758445945945947, + "grad_norm": 1.9578064680099487, + "learning_rate": 2.6721333333333335e-05, + "loss": 0.034, + "step": 19842 + }, + { + "epoch": 16.75929054054054, + "grad_norm": 0.319103479385376, + "learning_rate": 2.6721e-05, + "loss": 0.0171, + "step": 19843 + }, + { + "epoch": 16.760135135135137, + "grad_norm": 0.6552079319953918, + "learning_rate": 2.6720666666666666e-05, + "loss": 0.03, + "step": 19844 + }, + { + "epoch": 16.76097972972973, + "grad_norm": 0.5965809226036072, + "learning_rate": 2.6720333333333336e-05, + "loss": 0.1394, + "step": 19845 + }, + { + "epoch": 16.761824324324323, + "grad_norm": 0.4394081234931946, + "learning_rate": 2.672e-05, + "loss": 0.0784, + "step": 19846 + }, + { + "epoch": 16.76266891891892, + "grad_norm": 0.3761254847049713, + "learning_rate": 2.6719666666666668e-05, + "loss": 0.043, + "step": 19847 + }, + { + "epoch": 16.763513513513512, + "grad_norm": 0.29615113139152527, + "learning_rate": 2.6719333333333334e-05, + "loss": 0.0717, + "step": 19848 + }, + { + "epoch": 16.76435810810811, + "grad_norm": 0.3463123142719269, + "learning_rate": 2.6719000000000003e-05, + "loss": 0.038, + "step": 19849 + }, + { + "epoch": 16.7652027027027, + "grad_norm": 0.18142695724964142, + "learning_rate": 2.6718666666666665e-05, + "loss": 0.0154, + "step": 19850 + }, + { + "epoch": 16.7660472972973, + "grad_norm": 0.15725241601467133, + "learning_rate": 2.6718333333333335e-05, + "loss": 0.013, + "step": 19851 + }, + { + "epoch": 16.76689189189189, + "grad_norm": 0.20513184368610382, + "learning_rate": 2.6718e-05, + "loss": 0.0166, + "step": 19852 + }, + { + "epoch": 16.767736486486488, + "grad_norm": 0.19831416010856628, + "learning_rate": 2.6717666666666667e-05, + "loss": 0.0125, + "step": 19853 + }, + { + "epoch": 16.76858108108108, + "grad_norm": 0.21716250479221344, + "learning_rate": 2.6717333333333336e-05, + "loss": 0.0183, + "step": 19854 + }, + { + "epoch": 16.769425675675677, + "grad_norm": 0.3545268476009369, + "learning_rate": 2.6717e-05, + "loss": 0.0249, + "step": 19855 + }, + { + "epoch": 16.77027027027027, + "grad_norm": 0.17419587075710297, + "learning_rate": 2.6716666666666668e-05, + "loss": 0.0129, + "step": 19856 + }, + { + "epoch": 16.771114864864863, + "grad_norm": 0.11137746274471283, + "learning_rate": 2.6716333333333334e-05, + "loss": 0.0114, + "step": 19857 + }, + { + "epoch": 16.77195945945946, + "grad_norm": 0.45952704548835754, + "learning_rate": 2.6716e-05, + "loss": 0.0181, + "step": 19858 + }, + { + "epoch": 16.772804054054053, + "grad_norm": 0.22824721038341522, + "learning_rate": 2.6715666666666666e-05, + "loss": 0.0111, + "step": 19859 + }, + { + "epoch": 16.77364864864865, + "grad_norm": 0.13291670382022858, + "learning_rate": 2.6715333333333335e-05, + "loss": 0.0097, + "step": 19860 + }, + { + "epoch": 16.774493243243242, + "grad_norm": 0.25964778661727905, + "learning_rate": 2.6715e-05, + "loss": 0.0115, + "step": 19861 + }, + { + "epoch": 16.77533783783784, + "grad_norm": 0.3328406810760498, + "learning_rate": 2.6714666666666667e-05, + "loss": 0.0222, + "step": 19862 + }, + { + "epoch": 16.77618243243243, + "grad_norm": 0.23009108006954193, + "learning_rate": 2.6714333333333336e-05, + "loss": 0.0116, + "step": 19863 + }, + { + "epoch": 16.777027027027028, + "grad_norm": 0.1729234755039215, + "learning_rate": 2.6714e-05, + "loss": 0.0096, + "step": 19864 + }, + { + "epoch": 16.77787162162162, + "grad_norm": 0.1429482400417328, + "learning_rate": 2.6713666666666668e-05, + "loss": 0.0105, + "step": 19865 + }, + { + "epoch": 16.778716216216218, + "grad_norm": 0.348979115486145, + "learning_rate": 2.6713333333333334e-05, + "loss": 0.0139, + "step": 19866 + }, + { + "epoch": 16.77956081081081, + "grad_norm": 0.2810640335083008, + "learning_rate": 2.6713e-05, + "loss": 0.0162, + "step": 19867 + }, + { + "epoch": 16.780405405405407, + "grad_norm": 0.1785285919904709, + "learning_rate": 2.6712666666666666e-05, + "loss": 0.0113, + "step": 19868 + }, + { + "epoch": 16.78125, + "grad_norm": 0.1674376279115677, + "learning_rate": 2.6712333333333335e-05, + "loss": 0.0151, + "step": 19869 + }, + { + "epoch": 16.782094594594593, + "grad_norm": 0.20797206461429596, + "learning_rate": 2.6712e-05, + "loss": 0.0153, + "step": 19870 + }, + { + "epoch": 16.78293918918919, + "grad_norm": 0.23357583582401276, + "learning_rate": 2.6711666666666667e-05, + "loss": 0.0151, + "step": 19871 + }, + { + "epoch": 16.783783783783782, + "grad_norm": 0.32488173246383667, + "learning_rate": 2.6711333333333336e-05, + "loss": 0.0112, + "step": 19872 + }, + { + "epoch": 16.78462837837838, + "grad_norm": 0.14272315800189972, + "learning_rate": 2.6711e-05, + "loss": 0.0083, + "step": 19873 + }, + { + "epoch": 16.785472972972972, + "grad_norm": 0.21962031722068787, + "learning_rate": 2.6710666666666668e-05, + "loss": 0.0101, + "step": 19874 + }, + { + "epoch": 16.78631756756757, + "grad_norm": 0.20845907926559448, + "learning_rate": 2.6710333333333334e-05, + "loss": 0.0091, + "step": 19875 + }, + { + "epoch": 16.78716216216216, + "grad_norm": 0.3122953474521637, + "learning_rate": 2.671e-05, + "loss": 0.017, + "step": 19876 + }, + { + "epoch": 16.788006756756758, + "grad_norm": 0.5344440937042236, + "learning_rate": 2.670966666666667e-05, + "loss": 0.0178, + "step": 19877 + }, + { + "epoch": 16.78885135135135, + "grad_norm": 0.5828477144241333, + "learning_rate": 2.6709333333333335e-05, + "loss": 0.0256, + "step": 19878 + }, + { + "epoch": 16.789695945945947, + "grad_norm": 0.2901999056339264, + "learning_rate": 2.6709e-05, + "loss": 0.0187, + "step": 19879 + }, + { + "epoch": 16.79054054054054, + "grad_norm": 0.18510428071022034, + "learning_rate": 2.6708666666666667e-05, + "loss": 0.0107, + "step": 19880 + }, + { + "epoch": 16.791385135135137, + "grad_norm": 0.5997622013092041, + "learning_rate": 2.6708333333333333e-05, + "loss": 0.0162, + "step": 19881 + }, + { + "epoch": 16.79222972972973, + "grad_norm": 0.39139029383659363, + "learning_rate": 2.6708e-05, + "loss": 0.0105, + "step": 19882 + }, + { + "epoch": 16.793074324324323, + "grad_norm": 0.23895849287509918, + "learning_rate": 2.670766666666667e-05, + "loss": 0.0145, + "step": 19883 + }, + { + "epoch": 16.79391891891892, + "grad_norm": 0.32030391693115234, + "learning_rate": 2.670733333333333e-05, + "loss": 0.0162, + "step": 19884 + }, + { + "epoch": 16.794763513513512, + "grad_norm": 0.3982110321521759, + "learning_rate": 2.6707e-05, + "loss": 0.0099, + "step": 19885 + }, + { + "epoch": 16.79560810810811, + "grad_norm": 0.28462955355644226, + "learning_rate": 2.670666666666667e-05, + "loss": 0.0205, + "step": 19886 + }, + { + "epoch": 16.7964527027027, + "grad_norm": 0.2977035641670227, + "learning_rate": 2.6706333333333332e-05, + "loss": 0.0195, + "step": 19887 + }, + { + "epoch": 16.7972972972973, + "grad_norm": 0.2967980206012726, + "learning_rate": 2.6706e-05, + "loss": 0.0105, + "step": 19888 + }, + { + "epoch": 16.79814189189189, + "grad_norm": 0.6467311382293701, + "learning_rate": 2.6705666666666667e-05, + "loss": 0.016, + "step": 19889 + }, + { + "epoch": 16.798986486486488, + "grad_norm": 1.0112224817276, + "learning_rate": 2.6705333333333333e-05, + "loss": 0.019, + "step": 19890 + }, + { + "epoch": 16.79983108108108, + "grad_norm": 0.5766909718513489, + "learning_rate": 2.6705e-05, + "loss": 0.0339, + "step": 19891 + }, + { + "epoch": 16.800675675675677, + "grad_norm": 0.32525739073753357, + "learning_rate": 2.670466666666667e-05, + "loss": 0.0164, + "step": 19892 + }, + { + "epoch": 16.80152027027027, + "grad_norm": 1.173047661781311, + "learning_rate": 2.670433333333333e-05, + "loss": 0.0137, + "step": 19893 + }, + { + "epoch": 16.802364864864863, + "grad_norm": 0.46428462862968445, + "learning_rate": 2.6704e-05, + "loss": 0.0262, + "step": 19894 + }, + { + "epoch": 16.80320945945946, + "grad_norm": 0.6751092672348022, + "learning_rate": 2.670366666666667e-05, + "loss": 0.1556, + "step": 19895 + }, + { + "epoch": 16.804054054054053, + "grad_norm": 0.4896838665008545, + "learning_rate": 2.6703333333333332e-05, + "loss": 0.0733, + "step": 19896 + }, + { + "epoch": 16.80489864864865, + "grad_norm": 0.2632107138633728, + "learning_rate": 2.6703e-05, + "loss": 0.0359, + "step": 19897 + }, + { + "epoch": 16.805743243243242, + "grad_norm": 0.2662995159626007, + "learning_rate": 2.6702666666666668e-05, + "loss": 0.021, + "step": 19898 + }, + { + "epoch": 16.80658783783784, + "grad_norm": 0.2725374102592468, + "learning_rate": 2.6702333333333334e-05, + "loss": 0.0155, + "step": 19899 + }, + { + "epoch": 16.80743243243243, + "grad_norm": 0.23870475590229034, + "learning_rate": 2.6702e-05, + "loss": 0.017, + "step": 19900 + }, + { + "epoch": 16.808277027027028, + "grad_norm": 0.3676711916923523, + "learning_rate": 2.670166666666667e-05, + "loss": 0.0237, + "step": 19901 + }, + { + "epoch": 16.80912162162162, + "grad_norm": 0.13160568475723267, + "learning_rate": 2.6701333333333335e-05, + "loss": 0.0085, + "step": 19902 + }, + { + "epoch": 16.809966216216218, + "grad_norm": 0.2909424602985382, + "learning_rate": 2.6701e-05, + "loss": 0.0241, + "step": 19903 + }, + { + "epoch": 16.81081081081081, + "grad_norm": 0.21944056451320648, + "learning_rate": 2.670066666666667e-05, + "loss": 0.0122, + "step": 19904 + }, + { + "epoch": 16.811655405405407, + "grad_norm": 0.3420203626155853, + "learning_rate": 2.6700333333333333e-05, + "loss": 0.0173, + "step": 19905 + }, + { + "epoch": 16.8125, + "grad_norm": 0.23275278508663177, + "learning_rate": 2.6700000000000002e-05, + "loss": 0.0173, + "step": 19906 + }, + { + "epoch": 16.813344594594593, + "grad_norm": 0.2752026915550232, + "learning_rate": 2.6699666666666668e-05, + "loss": 0.0287, + "step": 19907 + }, + { + "epoch": 16.81418918918919, + "grad_norm": 0.4193895161151886, + "learning_rate": 2.6699333333333334e-05, + "loss": 0.0181, + "step": 19908 + }, + { + "epoch": 16.815033783783782, + "grad_norm": 0.21404288709163666, + "learning_rate": 2.6699e-05, + "loss": 0.0135, + "step": 19909 + }, + { + "epoch": 16.81587837837838, + "grad_norm": 0.15694308280944824, + "learning_rate": 2.669866666666667e-05, + "loss": 0.0133, + "step": 19910 + }, + { + "epoch": 16.816722972972972, + "grad_norm": 0.43310311436653137, + "learning_rate": 2.6698333333333335e-05, + "loss": 0.0261, + "step": 19911 + }, + { + "epoch": 16.81756756756757, + "grad_norm": 0.2216680943965912, + "learning_rate": 2.6698e-05, + "loss": 0.0147, + "step": 19912 + }, + { + "epoch": 16.81841216216216, + "grad_norm": 0.21773993968963623, + "learning_rate": 2.6697666666666667e-05, + "loss": 0.0121, + "step": 19913 + }, + { + "epoch": 16.819256756756758, + "grad_norm": 0.2276897132396698, + "learning_rate": 2.6697333333333333e-05, + "loss": 0.0156, + "step": 19914 + }, + { + "epoch": 16.82010135135135, + "grad_norm": 0.19522599875926971, + "learning_rate": 2.6697000000000002e-05, + "loss": 0.0124, + "step": 19915 + }, + { + "epoch": 16.820945945945947, + "grad_norm": 0.20250694453716278, + "learning_rate": 2.6696666666666665e-05, + "loss": 0.0126, + "step": 19916 + }, + { + "epoch": 16.82179054054054, + "grad_norm": 0.22217772901058197, + "learning_rate": 2.6696333333333334e-05, + "loss": 0.0116, + "step": 19917 + }, + { + "epoch": 16.822635135135137, + "grad_norm": 0.24294722080230713, + "learning_rate": 2.6696e-05, + "loss": 0.0091, + "step": 19918 + }, + { + "epoch": 16.82347972972973, + "grad_norm": 0.3351796865463257, + "learning_rate": 2.6695666666666666e-05, + "loss": 0.0104, + "step": 19919 + }, + { + "epoch": 16.824324324324323, + "grad_norm": 0.24664081633090973, + "learning_rate": 2.6695333333333335e-05, + "loss": 0.0172, + "step": 19920 + }, + { + "epoch": 16.82516891891892, + "grad_norm": 0.2910107970237732, + "learning_rate": 2.6695e-05, + "loss": 0.018, + "step": 19921 + }, + { + "epoch": 16.826013513513512, + "grad_norm": 0.2434946447610855, + "learning_rate": 2.6694666666666667e-05, + "loss": 0.0143, + "step": 19922 + }, + { + "epoch": 16.82685810810811, + "grad_norm": 0.18966488540172577, + "learning_rate": 2.6694333333333333e-05, + "loss": 0.0182, + "step": 19923 + }, + { + "epoch": 16.8277027027027, + "grad_norm": 0.18673105537891388, + "learning_rate": 2.6694000000000002e-05, + "loss": 0.0102, + "step": 19924 + }, + { + "epoch": 16.8285472972973, + "grad_norm": 0.1589708775281906, + "learning_rate": 2.6693666666666665e-05, + "loss": 0.0088, + "step": 19925 + }, + { + "epoch": 16.82939189189189, + "grad_norm": 0.21933862566947937, + "learning_rate": 2.6693333333333334e-05, + "loss": 0.0107, + "step": 19926 + }, + { + "epoch": 16.830236486486488, + "grad_norm": 0.1588081568479538, + "learning_rate": 2.6693000000000003e-05, + "loss": 0.0082, + "step": 19927 + }, + { + "epoch": 16.83108108108108, + "grad_norm": 0.41010284423828125, + "learning_rate": 2.6692666666666666e-05, + "loss": 0.0143, + "step": 19928 + }, + { + "epoch": 16.831925675675677, + "grad_norm": 0.29304641485214233, + "learning_rate": 2.6692333333333335e-05, + "loss": 0.0569, + "step": 19929 + }, + { + "epoch": 16.83277027027027, + "grad_norm": 0.24079783260822296, + "learning_rate": 2.6692e-05, + "loss": 0.0098, + "step": 19930 + }, + { + "epoch": 16.833614864864863, + "grad_norm": 0.14954687654972076, + "learning_rate": 2.6691666666666667e-05, + "loss": 0.0097, + "step": 19931 + }, + { + "epoch": 16.83445945945946, + "grad_norm": 0.3907177150249481, + "learning_rate": 2.6691333333333333e-05, + "loss": 0.0202, + "step": 19932 + }, + { + "epoch": 16.835304054054053, + "grad_norm": 0.31470227241516113, + "learning_rate": 2.6691000000000002e-05, + "loss": 0.0129, + "step": 19933 + }, + { + "epoch": 16.83614864864865, + "grad_norm": 0.3693319261074066, + "learning_rate": 2.6690666666666665e-05, + "loss": 0.0244, + "step": 19934 + }, + { + "epoch": 16.836993243243242, + "grad_norm": 0.1764923483133316, + "learning_rate": 2.6690333333333334e-05, + "loss": 0.0096, + "step": 19935 + }, + { + "epoch": 16.83783783783784, + "grad_norm": 0.19967222213745117, + "learning_rate": 2.6690000000000004e-05, + "loss": 0.0097, + "step": 19936 + }, + { + "epoch": 16.83868243243243, + "grad_norm": 0.17449411749839783, + "learning_rate": 2.6689666666666666e-05, + "loss": 0.0055, + "step": 19937 + }, + { + "epoch": 16.839527027027028, + "grad_norm": 0.21984776854515076, + "learning_rate": 2.6689333333333336e-05, + "loss": 0.0104, + "step": 19938 + }, + { + "epoch": 16.84037162162162, + "grad_norm": 0.24789831042289734, + "learning_rate": 2.6689e-05, + "loss": 0.0157, + "step": 19939 + }, + { + "epoch": 16.841216216216218, + "grad_norm": 0.20429621636867523, + "learning_rate": 2.6688666666666667e-05, + "loss": 0.0118, + "step": 19940 + }, + { + "epoch": 16.84206081081081, + "grad_norm": 0.4962853491306305, + "learning_rate": 2.6688333333333333e-05, + "loss": 0.0253, + "step": 19941 + }, + { + "epoch": 16.842905405405407, + "grad_norm": 0.3964400291442871, + "learning_rate": 2.6688e-05, + "loss": 0.0135, + "step": 19942 + }, + { + "epoch": 16.84375, + "grad_norm": 0.23989523947238922, + "learning_rate": 2.6687666666666665e-05, + "loss": 0.0143, + "step": 19943 + }, + { + "epoch": 16.844594594594593, + "grad_norm": 1.4443762302398682, + "learning_rate": 2.6687333333333335e-05, + "loss": 0.0406, + "step": 19944 + }, + { + "epoch": 16.84543918918919, + "grad_norm": 0.5433652400970459, + "learning_rate": 2.6687e-05, + "loss": 0.1443, + "step": 19945 + }, + { + "epoch": 16.846283783783782, + "grad_norm": 0.5007030963897705, + "learning_rate": 2.6686666666666666e-05, + "loss": 0.0801, + "step": 19946 + }, + { + "epoch": 16.84712837837838, + "grad_norm": 0.4605485796928406, + "learning_rate": 2.6686333333333336e-05, + "loss": 0.0685, + "step": 19947 + }, + { + "epoch": 16.847972972972972, + "grad_norm": 0.13533112406730652, + "learning_rate": 2.6685999999999998e-05, + "loss": 0.0117, + "step": 19948 + }, + { + "epoch": 16.84881756756757, + "grad_norm": 0.18544313311576843, + "learning_rate": 2.6685666666666668e-05, + "loss": 0.0276, + "step": 19949 + }, + { + "epoch": 16.84966216216216, + "grad_norm": 0.19326122105121613, + "learning_rate": 2.6685333333333334e-05, + "loss": 0.0181, + "step": 19950 + }, + { + "epoch": 16.850506756756758, + "grad_norm": 0.2280014008283615, + "learning_rate": 2.6685e-05, + "loss": 0.0197, + "step": 19951 + }, + { + "epoch": 16.85135135135135, + "grad_norm": 0.2617165446281433, + "learning_rate": 2.668466666666667e-05, + "loss": 0.0164, + "step": 19952 + }, + { + "epoch": 16.852195945945947, + "grad_norm": 0.24437332153320312, + "learning_rate": 2.6684333333333335e-05, + "loss": 0.014, + "step": 19953 + }, + { + "epoch": 16.85304054054054, + "grad_norm": 0.23829534649848938, + "learning_rate": 2.6684e-05, + "loss": 0.0156, + "step": 19954 + }, + { + "epoch": 16.853885135135137, + "grad_norm": 0.17185892164707184, + "learning_rate": 2.6683666666666667e-05, + "loss": 0.0098, + "step": 19955 + }, + { + "epoch": 16.85472972972973, + "grad_norm": 0.19351117312908173, + "learning_rate": 2.6683333333333336e-05, + "loss": 0.0117, + "step": 19956 + }, + { + "epoch": 16.855574324324323, + "grad_norm": 0.24673205614089966, + "learning_rate": 2.6683e-05, + "loss": 0.0203, + "step": 19957 + }, + { + "epoch": 16.85641891891892, + "grad_norm": 0.19490598142147064, + "learning_rate": 2.6682666666666668e-05, + "loss": 0.0171, + "step": 19958 + }, + { + "epoch": 16.857263513513512, + "grad_norm": 0.3306031823158264, + "learning_rate": 2.6682333333333334e-05, + "loss": 0.0139, + "step": 19959 + }, + { + "epoch": 16.85810810810811, + "grad_norm": 0.18334786593914032, + "learning_rate": 2.6682e-05, + "loss": 0.0159, + "step": 19960 + }, + { + "epoch": 16.8589527027027, + "grad_norm": 0.26239320635795593, + "learning_rate": 2.668166666666667e-05, + "loss": 0.0125, + "step": 19961 + }, + { + "epoch": 16.8597972972973, + "grad_norm": 0.16294997930526733, + "learning_rate": 2.6681333333333335e-05, + "loss": 0.0129, + "step": 19962 + }, + { + "epoch": 16.86064189189189, + "grad_norm": 0.16409802436828613, + "learning_rate": 2.6681e-05, + "loss": 0.0105, + "step": 19963 + }, + { + "epoch": 16.861486486486488, + "grad_norm": 0.2170223444700241, + "learning_rate": 2.6680666666666667e-05, + "loss": 0.01, + "step": 19964 + }, + { + "epoch": 16.86233108108108, + "grad_norm": 0.24858702719211578, + "learning_rate": 2.6680333333333336e-05, + "loss": 0.0245, + "step": 19965 + }, + { + "epoch": 16.863175675675677, + "grad_norm": 0.17400076985359192, + "learning_rate": 2.668e-05, + "loss": 0.0064, + "step": 19966 + }, + { + "epoch": 16.86402027027027, + "grad_norm": 0.29052481055259705, + "learning_rate": 2.6679666666666668e-05, + "loss": 0.0186, + "step": 19967 + }, + { + "epoch": 16.864864864864863, + "grad_norm": 0.26178213953971863, + "learning_rate": 2.6679333333333334e-05, + "loss": 0.0191, + "step": 19968 + }, + { + "epoch": 16.86570945945946, + "grad_norm": 0.5307897329330444, + "learning_rate": 2.6679e-05, + "loss": 0.023, + "step": 19969 + }, + { + "epoch": 16.866554054054053, + "grad_norm": 0.6568804383277893, + "learning_rate": 2.667866666666667e-05, + "loss": 0.0088, + "step": 19970 + }, + { + "epoch": 16.86739864864865, + "grad_norm": 0.22138875722885132, + "learning_rate": 2.6678333333333332e-05, + "loss": 0.0172, + "step": 19971 + }, + { + "epoch": 16.868243243243242, + "grad_norm": 0.3342382311820984, + "learning_rate": 2.6678e-05, + "loss": 0.0105, + "step": 19972 + }, + { + "epoch": 16.86908783783784, + "grad_norm": 0.21693773567676544, + "learning_rate": 2.6677666666666667e-05, + "loss": 0.0068, + "step": 19973 + }, + { + "epoch": 16.86993243243243, + "grad_norm": 0.2861182391643524, + "learning_rate": 2.6677333333333333e-05, + "loss": 0.0191, + "step": 19974 + }, + { + "epoch": 16.870777027027028, + "grad_norm": 0.3985101282596588, + "learning_rate": 2.6677e-05, + "loss": 0.0202, + "step": 19975 + }, + { + "epoch": 16.87162162162162, + "grad_norm": 0.2975236773490906, + "learning_rate": 2.6676666666666668e-05, + "loss": 0.0147, + "step": 19976 + }, + { + "epoch": 16.872466216216218, + "grad_norm": 0.36443501710891724, + "learning_rate": 2.6676333333333334e-05, + "loss": 0.0179, + "step": 19977 + }, + { + "epoch": 16.87331081081081, + "grad_norm": 0.19620215892791748, + "learning_rate": 2.6676e-05, + "loss": 0.0141, + "step": 19978 + }, + { + "epoch": 16.874155405405407, + "grad_norm": 0.23195010423660278, + "learning_rate": 2.667566666666667e-05, + "loss": 0.0174, + "step": 19979 + }, + { + "epoch": 16.875, + "grad_norm": 0.13043281435966492, + "learning_rate": 2.6675333333333332e-05, + "loss": 0.0035, + "step": 19980 + }, + { + "epoch": 16.875844594594593, + "grad_norm": 0.4758123755455017, + "learning_rate": 2.6675e-05, + "loss": 0.0245, + "step": 19981 + }, + { + "epoch": 16.87668918918919, + "grad_norm": 0.20725955069065094, + "learning_rate": 2.6674666666666667e-05, + "loss": 0.0104, + "step": 19982 + }, + { + "epoch": 16.877533783783782, + "grad_norm": 0.215715691447258, + "learning_rate": 2.6674333333333333e-05, + "loss": 0.0096, + "step": 19983 + }, + { + "epoch": 16.87837837837838, + "grad_norm": 0.23844730854034424, + "learning_rate": 2.6674e-05, + "loss": 0.0101, + "step": 19984 + }, + { + "epoch": 16.879222972972972, + "grad_norm": 0.40319564938545227, + "learning_rate": 2.667366666666667e-05, + "loss": 0.0167, + "step": 19985 + }, + { + "epoch": 16.88006756756757, + "grad_norm": 0.2593608498573303, + "learning_rate": 2.6673333333333334e-05, + "loss": 0.0117, + "step": 19986 + }, + { + "epoch": 16.88091216216216, + "grad_norm": 0.3110766112804413, + "learning_rate": 2.6673e-05, + "loss": 0.008, + "step": 19987 + }, + { + "epoch": 16.881756756756758, + "grad_norm": 0.18607020378112793, + "learning_rate": 2.667266666666667e-05, + "loss": 0.0111, + "step": 19988 + }, + { + "epoch": 16.88260135135135, + "grad_norm": 0.3952803313732147, + "learning_rate": 2.6672333333333332e-05, + "loss": 0.013, + "step": 19989 + }, + { + "epoch": 16.883445945945947, + "grad_norm": 0.5607307553291321, + "learning_rate": 2.6672e-05, + "loss": 0.0309, + "step": 19990 + }, + { + "epoch": 16.88429054054054, + "grad_norm": 0.29701492190361023, + "learning_rate": 2.6671666666666667e-05, + "loss": 0.0138, + "step": 19991 + }, + { + "epoch": 16.885135135135137, + "grad_norm": 0.2290623039007187, + "learning_rate": 2.6671333333333333e-05, + "loss": 0.0083, + "step": 19992 + }, + { + "epoch": 16.88597972972973, + "grad_norm": 0.32578712701797485, + "learning_rate": 2.6671e-05, + "loss": 0.0128, + "step": 19993 + }, + { + "epoch": 16.886824324324323, + "grad_norm": 0.49644193053245544, + "learning_rate": 2.667066666666667e-05, + "loss": 0.0073, + "step": 19994 + }, + { + "epoch": 16.88766891891892, + "grad_norm": 0.56746906042099, + "learning_rate": 2.6670333333333335e-05, + "loss": 0.1322, + "step": 19995 + }, + { + "epoch": 16.888513513513512, + "grad_norm": 0.36929282546043396, + "learning_rate": 2.667e-05, + "loss": 0.0674, + "step": 19996 + }, + { + "epoch": 16.88935810810811, + "grad_norm": 0.3056950867176056, + "learning_rate": 2.666966666666667e-05, + "loss": 0.0357, + "step": 19997 + }, + { + "epoch": 16.8902027027027, + "grad_norm": 0.22150343656539917, + "learning_rate": 2.6669333333333332e-05, + "loss": 0.0161, + "step": 19998 + }, + { + "epoch": 16.8910472972973, + "grad_norm": 0.14130879938602448, + "learning_rate": 2.6669e-05, + "loss": 0.0121, + "step": 19999 + }, + { + "epoch": 16.89189189189189, + "grad_norm": 0.6255161762237549, + "learning_rate": 2.6668666666666668e-05, + "loss": 0.0302, + "step": 20000 + }, + { + "epoch": 16.89189189189189, + "eval_cer": 0.08489277969640993, + "eval_loss": 0.2831272482872009, + "eval_runtime": 8.2251, + "eval_samples_per_second": 12.279, + "eval_steps_per_second": 0.486, + "eval_wer": 0.24911816578483245, + "step": 20000 + }, + { + "epoch": 16.892736486486488, + "grad_norm": 0.2244560569524765, + "learning_rate": 2.6668333333333334e-05, + "loss": 0.0151, + "step": 20001 + }, + { + "epoch": 16.89358108108108, + "grad_norm": 0.43013837933540344, + "learning_rate": 2.6668000000000003e-05, + "loss": 0.0275, + "step": 20002 + }, + { + "epoch": 16.894425675675677, + "grad_norm": 0.21038947999477386, + "learning_rate": 2.6667666666666665e-05, + "loss": 0.0109, + "step": 20003 + }, + { + "epoch": 16.89527027027027, + "grad_norm": 0.6402376294136047, + "learning_rate": 2.6667333333333335e-05, + "loss": 0.0104, + "step": 20004 + }, + { + "epoch": 16.896114864864863, + "grad_norm": 0.28540071845054626, + "learning_rate": 2.6667e-05, + "loss": 0.0094, + "step": 20005 + }, + { + "epoch": 16.89695945945946, + "grad_norm": 0.37600550055503845, + "learning_rate": 2.6666666666666667e-05, + "loss": 0.0207, + "step": 20006 + }, + { + "epoch": 16.897804054054053, + "grad_norm": 0.1953336000442505, + "learning_rate": 2.6666333333333333e-05, + "loss": 0.0099, + "step": 20007 + }, + { + "epoch": 16.89864864864865, + "grad_norm": 0.20496110618114471, + "learning_rate": 2.6666000000000002e-05, + "loss": 0.0174, + "step": 20008 + }, + { + "epoch": 16.899493243243242, + "grad_norm": 0.26588237285614014, + "learning_rate": 2.6665666666666664e-05, + "loss": 0.0133, + "step": 20009 + }, + { + "epoch": 16.90033783783784, + "grad_norm": 0.26857027411460876, + "learning_rate": 2.6665333333333334e-05, + "loss": 0.0182, + "step": 20010 + }, + { + "epoch": 16.90118243243243, + "grad_norm": 0.3005896210670471, + "learning_rate": 2.6665000000000003e-05, + "loss": 0.0103, + "step": 20011 + }, + { + "epoch": 16.902027027027028, + "grad_norm": 0.5476837158203125, + "learning_rate": 2.6664666666666666e-05, + "loss": 0.0177, + "step": 20012 + }, + { + "epoch": 16.90287162162162, + "grad_norm": 0.28087282180786133, + "learning_rate": 2.6664333333333335e-05, + "loss": 0.0214, + "step": 20013 + }, + { + "epoch": 16.903716216216218, + "grad_norm": 0.3463473916053772, + "learning_rate": 2.6664e-05, + "loss": 0.0167, + "step": 20014 + }, + { + "epoch": 16.90456081081081, + "grad_norm": 0.192842498421669, + "learning_rate": 2.6663666666666667e-05, + "loss": 0.0232, + "step": 20015 + }, + { + "epoch": 16.905405405405407, + "grad_norm": 0.3336617350578308, + "learning_rate": 2.6663333333333333e-05, + "loss": 0.0124, + "step": 20016 + }, + { + "epoch": 16.90625, + "grad_norm": 0.24223417043685913, + "learning_rate": 2.6663000000000002e-05, + "loss": 0.0109, + "step": 20017 + }, + { + "epoch": 16.907094594594593, + "grad_norm": 0.35107317566871643, + "learning_rate": 2.6662666666666668e-05, + "loss": 0.024, + "step": 20018 + }, + { + "epoch": 16.90793918918919, + "grad_norm": 0.3050956726074219, + "learning_rate": 2.6662333333333334e-05, + "loss": 0.0177, + "step": 20019 + }, + { + "epoch": 16.908783783783782, + "grad_norm": 0.4773622751235962, + "learning_rate": 2.6662000000000003e-05, + "loss": 0.016, + "step": 20020 + }, + { + "epoch": 16.90962837837838, + "grad_norm": 0.17573440074920654, + "learning_rate": 2.6661666666666666e-05, + "loss": 0.0118, + "step": 20021 + }, + { + "epoch": 16.910472972972972, + "grad_norm": 0.2551496922969818, + "learning_rate": 2.6661333333333335e-05, + "loss": 0.0103, + "step": 20022 + }, + { + "epoch": 16.91131756756757, + "grad_norm": 0.17315427958965302, + "learning_rate": 2.6661e-05, + "loss": 0.0058, + "step": 20023 + }, + { + "epoch": 16.91216216216216, + "grad_norm": 0.3371641933917999, + "learning_rate": 2.6660666666666667e-05, + "loss": 0.0117, + "step": 20024 + }, + { + "epoch": 16.913006756756758, + "grad_norm": 0.25144141912460327, + "learning_rate": 2.6660333333333333e-05, + "loss": 0.0071, + "step": 20025 + }, + { + "epoch": 16.91385135135135, + "grad_norm": 0.31509318947792053, + "learning_rate": 2.6660000000000002e-05, + "loss": 0.0129, + "step": 20026 + }, + { + "epoch": 16.914695945945947, + "grad_norm": 0.676478922367096, + "learning_rate": 2.6659666666666668e-05, + "loss": 0.0147, + "step": 20027 + }, + { + "epoch": 16.91554054054054, + "grad_norm": 0.343832790851593, + "learning_rate": 2.6659333333333334e-05, + "loss": 0.0192, + "step": 20028 + }, + { + "epoch": 16.916385135135137, + "grad_norm": 0.4913879334926605, + "learning_rate": 2.6659000000000003e-05, + "loss": 0.0112, + "step": 20029 + }, + { + "epoch": 16.91722972972973, + "grad_norm": 0.1287195384502411, + "learning_rate": 2.6658666666666666e-05, + "loss": 0.0065, + "step": 20030 + }, + { + "epoch": 16.918074324324323, + "grad_norm": 0.40973833203315735, + "learning_rate": 2.6658333333333335e-05, + "loss": 0.0181, + "step": 20031 + }, + { + "epoch": 16.91891891891892, + "grad_norm": 0.3554222583770752, + "learning_rate": 2.6657999999999998e-05, + "loss": 0.0207, + "step": 20032 + }, + { + "epoch": 16.919763513513512, + "grad_norm": 0.21503375470638275, + "learning_rate": 2.6657666666666667e-05, + "loss": 0.0064, + "step": 20033 + }, + { + "epoch": 16.92060810810811, + "grad_norm": 0.3857993185520172, + "learning_rate": 2.6657333333333333e-05, + "loss": 0.0213, + "step": 20034 + }, + { + "epoch": 16.9214527027027, + "grad_norm": 0.14351558685302734, + "learning_rate": 2.6657e-05, + "loss": 0.008, + "step": 20035 + }, + { + "epoch": 16.9222972972973, + "grad_norm": 0.16172772645950317, + "learning_rate": 2.665666666666667e-05, + "loss": 0.009, + "step": 20036 + }, + { + "epoch": 16.92314189189189, + "grad_norm": 0.2448194921016693, + "learning_rate": 2.6656333333333334e-05, + "loss": 0.0173, + "step": 20037 + }, + { + "epoch": 16.923986486486488, + "grad_norm": 2.0985541343688965, + "learning_rate": 2.6656e-05, + "loss": 0.0305, + "step": 20038 + }, + { + "epoch": 16.92483108108108, + "grad_norm": 0.3254096806049347, + "learning_rate": 2.6655666666666666e-05, + "loss": 0.014, + "step": 20039 + }, + { + "epoch": 16.925675675675677, + "grad_norm": 0.28410398960113525, + "learning_rate": 2.6655333333333336e-05, + "loss": 0.0113, + "step": 20040 + }, + { + "epoch": 16.92652027027027, + "grad_norm": 0.32800421118736267, + "learning_rate": 2.6654999999999998e-05, + "loss": 0.014, + "step": 20041 + }, + { + "epoch": 16.927364864864863, + "grad_norm": 0.25139129161834717, + "learning_rate": 2.6654666666666667e-05, + "loss": 0.0181, + "step": 20042 + }, + { + "epoch": 16.92820945945946, + "grad_norm": 0.3486502170562744, + "learning_rate": 2.6654333333333333e-05, + "loss": 0.014, + "step": 20043 + }, + { + "epoch": 16.929054054054053, + "grad_norm": 0.438706636428833, + "learning_rate": 2.6654e-05, + "loss": 0.0213, + "step": 20044 + }, + { + "epoch": 16.92989864864865, + "grad_norm": 0.5643428564071655, + "learning_rate": 2.665366666666667e-05, + "loss": 0.1964, + "step": 20045 + }, + { + "epoch": 16.930743243243242, + "grad_norm": 0.6644043922424316, + "learning_rate": 2.6653333333333335e-05, + "loss": 0.1492, + "step": 20046 + }, + { + "epoch": 16.93158783783784, + "grad_norm": 0.5887123942375183, + "learning_rate": 2.6653e-05, + "loss": 0.0692, + "step": 20047 + }, + { + "epoch": 16.93243243243243, + "grad_norm": 0.3889258801937103, + "learning_rate": 2.6652666666666666e-05, + "loss": 0.042, + "step": 20048 + }, + { + "epoch": 16.933277027027028, + "grad_norm": 0.4638616740703583, + "learning_rate": 2.6652333333333336e-05, + "loss": 0.0308, + "step": 20049 + }, + { + "epoch": 16.93412162162162, + "grad_norm": 0.19639383256435394, + "learning_rate": 2.6651999999999998e-05, + "loss": 0.0128, + "step": 20050 + }, + { + "epoch": 16.934966216216218, + "grad_norm": 0.2499595284461975, + "learning_rate": 2.6651666666666668e-05, + "loss": 0.0147, + "step": 20051 + }, + { + "epoch": 16.93581081081081, + "grad_norm": 0.21518708765506744, + "learning_rate": 2.6651333333333337e-05, + "loss": 0.0187, + "step": 20052 + }, + { + "epoch": 16.936655405405407, + "grad_norm": 0.17744094133377075, + "learning_rate": 2.6651e-05, + "loss": 0.0182, + "step": 20053 + }, + { + "epoch": 16.9375, + "grad_norm": 0.21210232377052307, + "learning_rate": 2.665066666666667e-05, + "loss": 0.0181, + "step": 20054 + }, + { + "epoch": 16.938344594594593, + "grad_norm": 0.26561474800109863, + "learning_rate": 2.6650333333333335e-05, + "loss": 0.0154, + "step": 20055 + }, + { + "epoch": 16.93918918918919, + "grad_norm": 0.2004932016134262, + "learning_rate": 2.665e-05, + "loss": 0.0198, + "step": 20056 + }, + { + "epoch": 16.940033783783782, + "grad_norm": 0.2150963991880417, + "learning_rate": 2.6649666666666667e-05, + "loss": 0.0154, + "step": 20057 + }, + { + "epoch": 16.94087837837838, + "grad_norm": 0.12789951264858246, + "learning_rate": 2.6649333333333336e-05, + "loss": 0.0132, + "step": 20058 + }, + { + "epoch": 16.941722972972972, + "grad_norm": 0.14490222930908203, + "learning_rate": 2.6649e-05, + "loss": 0.0113, + "step": 20059 + }, + { + "epoch": 16.94256756756757, + "grad_norm": 0.41437268257141113, + "learning_rate": 2.6648666666666668e-05, + "loss": 0.0277, + "step": 20060 + }, + { + "epoch": 16.94341216216216, + "grad_norm": 0.22665564715862274, + "learning_rate": 2.6648333333333337e-05, + "loss": 0.0196, + "step": 20061 + }, + { + "epoch": 16.944256756756758, + "grad_norm": 0.2262222021818161, + "learning_rate": 2.6648e-05, + "loss": 0.0155, + "step": 20062 + }, + { + "epoch": 16.94510135135135, + "grad_norm": 0.22596338391304016, + "learning_rate": 2.664766666666667e-05, + "loss": 0.0201, + "step": 20063 + }, + { + "epoch": 16.945945945945947, + "grad_norm": 0.41460713744163513, + "learning_rate": 2.664733333333333e-05, + "loss": 0.0161, + "step": 20064 + }, + { + "epoch": 16.94679054054054, + "grad_norm": 0.2630484402179718, + "learning_rate": 2.6647e-05, + "loss": 0.0151, + "step": 20065 + }, + { + "epoch": 16.947635135135137, + "grad_norm": 0.13225848972797394, + "learning_rate": 2.6646666666666667e-05, + "loss": 0.0147, + "step": 20066 + }, + { + "epoch": 16.94847972972973, + "grad_norm": 0.28023046255111694, + "learning_rate": 2.6646333333333333e-05, + "loss": 0.0112, + "step": 20067 + }, + { + "epoch": 16.949324324324323, + "grad_norm": 0.6903375387191772, + "learning_rate": 2.6646000000000002e-05, + "loss": 0.0112, + "step": 20068 + }, + { + "epoch": 16.95016891891892, + "grad_norm": 0.27051958441734314, + "learning_rate": 2.6645666666666668e-05, + "loss": 0.0157, + "step": 20069 + }, + { + "epoch": 16.951013513513512, + "grad_norm": 0.7947453260421753, + "learning_rate": 2.6645333333333334e-05, + "loss": 0.0224, + "step": 20070 + }, + { + "epoch": 16.95185810810811, + "grad_norm": 0.23388789594173431, + "learning_rate": 2.6645e-05, + "loss": 0.0157, + "step": 20071 + }, + { + "epoch": 16.9527027027027, + "grad_norm": 0.3104005753993988, + "learning_rate": 2.664466666666667e-05, + "loss": 0.0134, + "step": 20072 + }, + { + "epoch": 16.9535472972973, + "grad_norm": 0.1972954124212265, + "learning_rate": 2.6644333333333332e-05, + "loss": 0.0127, + "step": 20073 + }, + { + "epoch": 16.95439189189189, + "grad_norm": 0.32913437485694885, + "learning_rate": 2.6644e-05, + "loss": 0.0167, + "step": 20074 + }, + { + "epoch": 16.955236486486488, + "grad_norm": 0.22725774347782135, + "learning_rate": 2.6643666666666667e-05, + "loss": 0.0135, + "step": 20075 + }, + { + "epoch": 16.95608108108108, + "grad_norm": 0.17595796287059784, + "learning_rate": 2.6643333333333333e-05, + "loss": 0.0106, + "step": 20076 + }, + { + "epoch": 16.956925675675677, + "grad_norm": 0.2102898508310318, + "learning_rate": 2.6643000000000002e-05, + "loss": 0.0069, + "step": 20077 + }, + { + "epoch": 16.95777027027027, + "grad_norm": 0.33570805191993713, + "learning_rate": 2.6642666666666668e-05, + "loss": 0.0161, + "step": 20078 + }, + { + "epoch": 16.958614864864863, + "grad_norm": 0.1631905734539032, + "learning_rate": 2.6642333333333334e-05, + "loss": 0.0068, + "step": 20079 + }, + { + "epoch": 16.95945945945946, + "grad_norm": 0.4876590371131897, + "learning_rate": 2.6642e-05, + "loss": 0.0175, + "step": 20080 + }, + { + "epoch": 16.960304054054053, + "grad_norm": 0.3300284445285797, + "learning_rate": 2.664166666666667e-05, + "loss": 0.018, + "step": 20081 + }, + { + "epoch": 16.96114864864865, + "grad_norm": 0.7499544620513916, + "learning_rate": 2.6641333333333332e-05, + "loss": 0.013, + "step": 20082 + }, + { + "epoch": 16.961993243243242, + "grad_norm": 0.3704514801502228, + "learning_rate": 2.6641e-05, + "loss": 0.022, + "step": 20083 + }, + { + "epoch": 16.96283783783784, + "grad_norm": 0.14465908706188202, + "learning_rate": 2.6640666666666667e-05, + "loss": 0.0066, + "step": 20084 + }, + { + "epoch": 16.96368243243243, + "grad_norm": 0.17181642353534698, + "learning_rate": 2.6640333333333333e-05, + "loss": 0.0092, + "step": 20085 + }, + { + "epoch": 16.964527027027028, + "grad_norm": 0.6381332278251648, + "learning_rate": 2.6640000000000002e-05, + "loss": 0.014, + "step": 20086 + }, + { + "epoch": 16.96537162162162, + "grad_norm": 0.2938053011894226, + "learning_rate": 2.663966666666667e-05, + "loss": 0.013, + "step": 20087 + }, + { + "epoch": 16.966216216216218, + "grad_norm": 0.33718380331993103, + "learning_rate": 2.6639333333333334e-05, + "loss": 0.0135, + "step": 20088 + }, + { + "epoch": 16.96706081081081, + "grad_norm": 0.18733170628547668, + "learning_rate": 2.6639e-05, + "loss": 0.0075, + "step": 20089 + }, + { + "epoch": 16.967905405405407, + "grad_norm": 0.5330472588539124, + "learning_rate": 2.663866666666667e-05, + "loss": 0.0187, + "step": 20090 + }, + { + "epoch": 16.96875, + "grad_norm": 0.1265491247177124, + "learning_rate": 2.6638333333333332e-05, + "loss": 0.0081, + "step": 20091 + }, + { + "epoch": 16.969594594594593, + "grad_norm": 0.49136972427368164, + "learning_rate": 2.6638e-05, + "loss": 0.0175, + "step": 20092 + }, + { + "epoch": 16.97043918918919, + "grad_norm": 0.20019499957561493, + "learning_rate": 2.6637666666666667e-05, + "loss": 0.0116, + "step": 20093 + }, + { + "epoch": 16.971283783783782, + "grad_norm": 0.6746835708618164, + "learning_rate": 2.6637333333333333e-05, + "loss": 0.0413, + "step": 20094 + }, + { + "epoch": 16.97212837837838, + "grad_norm": 0.5174017548561096, + "learning_rate": 2.6637000000000003e-05, + "loss": 0.1045, + "step": 20095 + }, + { + "epoch": 16.972972972972972, + "grad_norm": 0.4460534155368805, + "learning_rate": 2.6636666666666665e-05, + "loss": 0.0493, + "step": 20096 + }, + { + "epoch": 16.97381756756757, + "grad_norm": 0.2835191786289215, + "learning_rate": 2.6636333333333335e-05, + "loss": 0.0213, + "step": 20097 + }, + { + "epoch": 16.97466216216216, + "grad_norm": 0.422150582075119, + "learning_rate": 2.6636e-05, + "loss": 0.0332, + "step": 20098 + }, + { + "epoch": 16.975506756756758, + "grad_norm": 0.2686139643192291, + "learning_rate": 2.6635666666666666e-05, + "loss": 0.0147, + "step": 20099 + }, + { + "epoch": 16.97635135135135, + "grad_norm": 0.19637292623519897, + "learning_rate": 2.6635333333333332e-05, + "loss": 0.0119, + "step": 20100 + }, + { + "epoch": 16.977195945945947, + "grad_norm": 0.181899756193161, + "learning_rate": 2.6635e-05, + "loss": 0.0135, + "step": 20101 + }, + { + "epoch": 16.97804054054054, + "grad_norm": 0.19136008620262146, + "learning_rate": 2.6634666666666668e-05, + "loss": 0.0111, + "step": 20102 + }, + { + "epoch": 16.978885135135137, + "grad_norm": 0.349818617105484, + "learning_rate": 2.6634333333333334e-05, + "loss": 0.02, + "step": 20103 + }, + { + "epoch": 16.97972972972973, + "grad_norm": 0.21760112047195435, + "learning_rate": 2.6634000000000003e-05, + "loss": 0.0145, + "step": 20104 + }, + { + "epoch": 16.980574324324323, + "grad_norm": 0.20051340758800507, + "learning_rate": 2.6633666666666665e-05, + "loss": 0.013, + "step": 20105 + }, + { + "epoch": 16.98141891891892, + "grad_norm": 0.23199766874313354, + "learning_rate": 2.6633333333333335e-05, + "loss": 0.022, + "step": 20106 + }, + { + "epoch": 16.982263513513512, + "grad_norm": 0.3167756497859955, + "learning_rate": 2.6633e-05, + "loss": 0.0126, + "step": 20107 + }, + { + "epoch": 16.98310810810811, + "grad_norm": 0.1948697566986084, + "learning_rate": 2.6632666666666667e-05, + "loss": 0.0166, + "step": 20108 + }, + { + "epoch": 16.9839527027027, + "grad_norm": 0.21488851308822632, + "learning_rate": 2.6632333333333333e-05, + "loss": 0.0102, + "step": 20109 + }, + { + "epoch": 16.9847972972973, + "grad_norm": 0.31340262293815613, + "learning_rate": 2.6632000000000002e-05, + "loss": 0.022, + "step": 20110 + }, + { + "epoch": 16.98564189189189, + "grad_norm": 0.12886296212673187, + "learning_rate": 2.6631666666666668e-05, + "loss": 0.0104, + "step": 20111 + }, + { + "epoch": 16.986486486486488, + "grad_norm": 0.1539335399866104, + "learning_rate": 2.6631333333333334e-05, + "loss": 0.014, + "step": 20112 + }, + { + "epoch": 16.98733108108108, + "grad_norm": 0.28587284684181213, + "learning_rate": 2.6631000000000003e-05, + "loss": 0.02, + "step": 20113 + }, + { + "epoch": 16.988175675675677, + "grad_norm": 0.1766311228275299, + "learning_rate": 2.6630666666666666e-05, + "loss": 0.01, + "step": 20114 + }, + { + "epoch": 16.98902027027027, + "grad_norm": 0.3747064173221588, + "learning_rate": 2.6630333333333335e-05, + "loss": 0.0136, + "step": 20115 + }, + { + "epoch": 16.989864864864863, + "grad_norm": 0.24400730431079865, + "learning_rate": 2.663e-05, + "loss": 0.0116, + "step": 20116 + }, + { + "epoch": 16.99070945945946, + "grad_norm": 0.12934622168540955, + "learning_rate": 2.6629666666666667e-05, + "loss": 0.0069, + "step": 20117 + }, + { + "epoch": 16.991554054054053, + "grad_norm": 0.2072007805109024, + "learning_rate": 2.6629333333333336e-05, + "loss": 0.0107, + "step": 20118 + }, + { + "epoch": 16.99239864864865, + "grad_norm": 0.21586066484451294, + "learning_rate": 2.6629000000000002e-05, + "loss": 0.0158, + "step": 20119 + }, + { + "epoch": 16.993243243243242, + "grad_norm": 0.2305668443441391, + "learning_rate": 2.6628666666666668e-05, + "loss": 0.0117, + "step": 20120 + }, + { + "epoch": 16.99408783783784, + "grad_norm": 0.2926592528820038, + "learning_rate": 2.6628333333333334e-05, + "loss": 0.0144, + "step": 20121 + }, + { + "epoch": 16.99493243243243, + "grad_norm": 0.23455952107906342, + "learning_rate": 2.6628e-05, + "loss": 0.0108, + "step": 20122 + }, + { + "epoch": 16.995777027027028, + "grad_norm": 0.29359182715415955, + "learning_rate": 2.6627666666666666e-05, + "loss": 0.0086, + "step": 20123 + }, + { + "epoch": 16.99662162162162, + "grad_norm": 0.20437262952327728, + "learning_rate": 2.6627333333333335e-05, + "loss": 0.0186, + "step": 20124 + }, + { + "epoch": 16.997466216216218, + "grad_norm": 0.5768082141876221, + "learning_rate": 2.6626999999999998e-05, + "loss": 0.017, + "step": 20125 + }, + { + "epoch": 16.99831081081081, + "grad_norm": 0.6345168352127075, + "learning_rate": 2.6626666666666667e-05, + "loss": 0.0165, + "step": 20126 + }, + { + "epoch": 16.999155405405407, + "grad_norm": 0.49087461829185486, + "learning_rate": 2.6626333333333336e-05, + "loss": 0.0164, + "step": 20127 + }, + { + "epoch": 17.0, + "grad_norm": 0.7058737277984619, + "learning_rate": 2.6626e-05, + "loss": 0.0341, + "step": 20128 + }, + { + "epoch": 17.000844594594593, + "grad_norm": 0.6041871905326843, + "learning_rate": 2.6625666666666668e-05, + "loss": 0.1231, + "step": 20129 + }, + { + "epoch": 17.00168918918919, + "grad_norm": 0.32385048270225525, + "learning_rate": 2.6625333333333334e-05, + "loss": 0.0719, + "step": 20130 + }, + { + "epoch": 17.002533783783782, + "grad_norm": 0.37581324577331543, + "learning_rate": 2.6625e-05, + "loss": 0.0285, + "step": 20131 + }, + { + "epoch": 17.00337837837838, + "grad_norm": 0.41581472754478455, + "learning_rate": 2.6624666666666666e-05, + "loss": 0.0301, + "step": 20132 + }, + { + "epoch": 17.004222972972972, + "grad_norm": 0.3024970591068268, + "learning_rate": 2.6624333333333335e-05, + "loss": 0.0226, + "step": 20133 + }, + { + "epoch": 17.00506756756757, + "grad_norm": 0.4046271741390228, + "learning_rate": 2.6623999999999998e-05, + "loss": 0.0175, + "step": 20134 + }, + { + "epoch": 17.00591216216216, + "grad_norm": 0.1722588986158371, + "learning_rate": 2.6623666666666667e-05, + "loss": 0.0092, + "step": 20135 + }, + { + "epoch": 17.006756756756758, + "grad_norm": 0.24441012740135193, + "learning_rate": 2.6623333333333336e-05, + "loss": 0.0109, + "step": 20136 + }, + { + "epoch": 17.00760135135135, + "grad_norm": 0.18708369135856628, + "learning_rate": 2.6623e-05, + "loss": 0.0162, + "step": 20137 + }, + { + "epoch": 17.008445945945947, + "grad_norm": 0.1894211769104004, + "learning_rate": 2.662266666666667e-05, + "loss": 0.0093, + "step": 20138 + }, + { + "epoch": 17.00929054054054, + "grad_norm": 0.38540154695510864, + "learning_rate": 2.6622333333333334e-05, + "loss": 0.0122, + "step": 20139 + }, + { + "epoch": 17.010135135135137, + "grad_norm": 0.18914148211479187, + "learning_rate": 2.6622e-05, + "loss": 0.0073, + "step": 20140 + }, + { + "epoch": 17.01097972972973, + "grad_norm": 0.23595339059829712, + "learning_rate": 2.6621666666666666e-05, + "loss": 0.0133, + "step": 20141 + }, + { + "epoch": 17.011824324324323, + "grad_norm": 0.24989885091781616, + "learning_rate": 2.6621333333333335e-05, + "loss": 0.0138, + "step": 20142 + }, + { + "epoch": 17.01266891891892, + "grad_norm": 0.2462436556816101, + "learning_rate": 2.6621e-05, + "loss": 0.011, + "step": 20143 + }, + { + "epoch": 17.013513513513512, + "grad_norm": 0.2605535387992859, + "learning_rate": 2.6620666666666667e-05, + "loss": 0.0163, + "step": 20144 + }, + { + "epoch": 17.01435810810811, + "grad_norm": 0.17166320979595184, + "learning_rate": 2.6620333333333337e-05, + "loss": 0.0087, + "step": 20145 + }, + { + "epoch": 17.0152027027027, + "grad_norm": 0.176632821559906, + "learning_rate": 2.662e-05, + "loss": 0.0071, + "step": 20146 + }, + { + "epoch": 17.0160472972973, + "grad_norm": 0.2900724411010742, + "learning_rate": 2.661966666666667e-05, + "loss": 0.0171, + "step": 20147 + }, + { + "epoch": 17.01689189189189, + "grad_norm": 0.2495499551296234, + "learning_rate": 2.6619333333333334e-05, + "loss": 0.0085, + "step": 20148 + }, + { + "epoch": 17.017736486486488, + "grad_norm": 0.23831647634506226, + "learning_rate": 2.6619e-05, + "loss": 0.0186, + "step": 20149 + }, + { + "epoch": 17.01858108108108, + "grad_norm": 0.19439086318016052, + "learning_rate": 2.6618666666666666e-05, + "loss": 0.0095, + "step": 20150 + }, + { + "epoch": 17.019425675675677, + "grad_norm": 0.169192835688591, + "learning_rate": 2.6618333333333336e-05, + "loss": 0.0092, + "step": 20151 + }, + { + "epoch": 17.02027027027027, + "grad_norm": 0.13217000663280487, + "learning_rate": 2.6618e-05, + "loss": 0.0099, + "step": 20152 + }, + { + "epoch": 17.021114864864863, + "grad_norm": 0.8068461418151855, + "learning_rate": 2.6617666666666668e-05, + "loss": 0.0126, + "step": 20153 + }, + { + "epoch": 17.02195945945946, + "grad_norm": 0.2592713236808777, + "learning_rate": 2.6617333333333334e-05, + "loss": 0.0177, + "step": 20154 + }, + { + "epoch": 17.022804054054053, + "grad_norm": 0.14271385967731476, + "learning_rate": 2.6617e-05, + "loss": 0.008, + "step": 20155 + }, + { + "epoch": 17.02364864864865, + "grad_norm": 0.2310892790555954, + "learning_rate": 2.661666666666667e-05, + "loss": 0.0147, + "step": 20156 + }, + { + "epoch": 17.024493243243242, + "grad_norm": 0.1316346377134323, + "learning_rate": 2.661633333333333e-05, + "loss": 0.0062, + "step": 20157 + }, + { + "epoch": 17.02533783783784, + "grad_norm": 0.15854795277118683, + "learning_rate": 2.6616e-05, + "loss": 0.0088, + "step": 20158 + }, + { + "epoch": 17.02618243243243, + "grad_norm": 0.3853559195995331, + "learning_rate": 2.6615666666666667e-05, + "loss": 0.0188, + "step": 20159 + }, + { + "epoch": 17.027027027027028, + "grad_norm": 0.34495699405670166, + "learning_rate": 2.6615333333333333e-05, + "loss": 0.0088, + "step": 20160 + }, + { + "epoch": 17.02787162162162, + "grad_norm": 0.4098280072212219, + "learning_rate": 2.6615000000000002e-05, + "loss": 0.0101, + "step": 20161 + }, + { + "epoch": 17.028716216216218, + "grad_norm": 0.22125931084156036, + "learning_rate": 2.6614666666666668e-05, + "loss": 0.011, + "step": 20162 + }, + { + "epoch": 17.02956081081081, + "grad_norm": 0.21908894181251526, + "learning_rate": 2.6614333333333334e-05, + "loss": 0.0145, + "step": 20163 + }, + { + "epoch": 17.030405405405407, + "grad_norm": 0.2604563534259796, + "learning_rate": 2.6614e-05, + "loss": 0.0084, + "step": 20164 + }, + { + "epoch": 17.03125, + "grad_norm": 0.2131306231021881, + "learning_rate": 2.661366666666667e-05, + "loss": 0.0092, + "step": 20165 + }, + { + "epoch": 17.032094594594593, + "grad_norm": 0.15561892092227936, + "learning_rate": 2.661333333333333e-05, + "loss": 0.0119, + "step": 20166 + }, + { + "epoch": 17.03293918918919, + "grad_norm": 0.17081762850284576, + "learning_rate": 2.6613e-05, + "loss": 0.0161, + "step": 20167 + }, + { + "epoch": 17.033783783783782, + "grad_norm": 0.3303912281990051, + "learning_rate": 2.661266666666667e-05, + "loss": 0.0169, + "step": 20168 + }, + { + "epoch": 17.03462837837838, + "grad_norm": 0.2117329090833664, + "learning_rate": 2.6612333333333333e-05, + "loss": 0.014, + "step": 20169 + }, + { + "epoch": 17.035472972972972, + "grad_norm": 0.1359056830406189, + "learning_rate": 2.6612000000000002e-05, + "loss": 0.007, + "step": 20170 + }, + { + "epoch": 17.03631756756757, + "grad_norm": 0.3711645007133484, + "learning_rate": 2.6611666666666668e-05, + "loss": 0.0175, + "step": 20171 + }, + { + "epoch": 17.03716216216216, + "grad_norm": 0.4506279528141022, + "learning_rate": 2.6611333333333334e-05, + "loss": 0.0077, + "step": 20172 + }, + { + "epoch": 17.038006756756758, + "grad_norm": 0.3369710445404053, + "learning_rate": 2.6611e-05, + "loss": 0.0114, + "step": 20173 + }, + { + "epoch": 17.03885135135135, + "grad_norm": 0.18756629526615143, + "learning_rate": 2.661066666666667e-05, + "loss": 0.0079, + "step": 20174 + }, + { + "epoch": 17.039695945945947, + "grad_norm": 0.2657172977924347, + "learning_rate": 2.6610333333333332e-05, + "loss": 0.0128, + "step": 20175 + }, + { + "epoch": 17.04054054054054, + "grad_norm": 0.16879186034202576, + "learning_rate": 2.661e-05, + "loss": 0.0075, + "step": 20176 + }, + { + "epoch": 17.041385135135137, + "grad_norm": 0.6305025219917297, + "learning_rate": 2.660966666666667e-05, + "loss": 0.0152, + "step": 20177 + }, + { + "epoch": 17.04222972972973, + "grad_norm": 2.0196361541748047, + "learning_rate": 2.6609333333333333e-05, + "loss": 0.0354, + "step": 20178 + }, + { + "epoch": 17.043074324324323, + "grad_norm": 0.49041950702667236, + "learning_rate": 2.6609000000000002e-05, + "loss": 0.1276, + "step": 20179 + }, + { + "epoch": 17.04391891891892, + "grad_norm": 0.44338056445121765, + "learning_rate": 2.6608666666666668e-05, + "loss": 0.0682, + "step": 20180 + }, + { + "epoch": 17.044763513513512, + "grad_norm": 0.2562679946422577, + "learning_rate": 2.6608333333333334e-05, + "loss": 0.0328, + "step": 20181 + }, + { + "epoch": 17.04560810810811, + "grad_norm": 0.42151883244514465, + "learning_rate": 2.6608e-05, + "loss": 0.0632, + "step": 20182 + }, + { + "epoch": 17.0464527027027, + "grad_norm": 0.23186181485652924, + "learning_rate": 2.6607666666666666e-05, + "loss": 0.0163, + "step": 20183 + }, + { + "epoch": 17.0472972972973, + "grad_norm": 0.3624137341976166, + "learning_rate": 2.6607333333333332e-05, + "loss": 0.0114, + "step": 20184 + }, + { + "epoch": 17.04814189189189, + "grad_norm": 0.18667669594287872, + "learning_rate": 2.6607e-05, + "loss": 0.0147, + "step": 20185 + }, + { + "epoch": 17.048986486486488, + "grad_norm": 0.18735341727733612, + "learning_rate": 2.6606666666666667e-05, + "loss": 0.0114, + "step": 20186 + }, + { + "epoch": 17.04983108108108, + "grad_norm": 0.3408220708370209, + "learning_rate": 2.6606333333333333e-05, + "loss": 0.0118, + "step": 20187 + }, + { + "epoch": 17.050675675675677, + "grad_norm": 0.2787860333919525, + "learning_rate": 2.6606000000000002e-05, + "loss": 0.0149, + "step": 20188 + }, + { + "epoch": 17.05152027027027, + "grad_norm": 0.19399508833885193, + "learning_rate": 2.6605666666666665e-05, + "loss": 0.0112, + "step": 20189 + }, + { + "epoch": 17.052364864864863, + "grad_norm": 0.2102925032377243, + "learning_rate": 2.6605333333333334e-05, + "loss": 0.01, + "step": 20190 + }, + { + "epoch": 17.05320945945946, + "grad_norm": 0.18191348016262054, + "learning_rate": 2.6605e-05, + "loss": 0.0088, + "step": 20191 + }, + { + "epoch": 17.054054054054053, + "grad_norm": 0.17187096178531647, + "learning_rate": 2.6604666666666666e-05, + "loss": 0.008, + "step": 20192 + }, + { + "epoch": 17.05489864864865, + "grad_norm": 0.3537081778049469, + "learning_rate": 2.6604333333333335e-05, + "loss": 0.019, + "step": 20193 + }, + { + "epoch": 17.055743243243242, + "grad_norm": 0.23860837519168854, + "learning_rate": 2.6604e-05, + "loss": 0.0123, + "step": 20194 + }, + { + "epoch": 17.05658783783784, + "grad_norm": 0.17747297883033752, + "learning_rate": 2.6603666666666667e-05, + "loss": 0.0075, + "step": 20195 + }, + { + "epoch": 17.05743243243243, + "grad_norm": 0.138861745595932, + "learning_rate": 2.6603333333333333e-05, + "loss": 0.0053, + "step": 20196 + }, + { + "epoch": 17.058277027027028, + "grad_norm": 0.3731328845024109, + "learning_rate": 2.6603000000000003e-05, + "loss": 0.0084, + "step": 20197 + }, + { + "epoch": 17.05912162162162, + "grad_norm": 0.23131364583969116, + "learning_rate": 2.6602666666666665e-05, + "loss": 0.0143, + "step": 20198 + }, + { + "epoch": 17.059966216216218, + "grad_norm": 0.19110111892223358, + "learning_rate": 2.6602333333333334e-05, + "loss": 0.0102, + "step": 20199 + }, + { + "epoch": 17.06081081081081, + "grad_norm": 0.4159303605556488, + "learning_rate": 2.6602e-05, + "loss": 0.011, + "step": 20200 + }, + { + "epoch": 17.061655405405407, + "grad_norm": 0.1666741669178009, + "learning_rate": 2.6601666666666666e-05, + "loss": 0.0107, + "step": 20201 + }, + { + "epoch": 17.0625, + "grad_norm": 0.2152222841978073, + "learning_rate": 2.6601333333333336e-05, + "loss": 0.0107, + "step": 20202 + }, + { + "epoch": 17.063344594594593, + "grad_norm": 0.10519618541002274, + "learning_rate": 2.6601e-05, + "loss": 0.0065, + "step": 20203 + }, + { + "epoch": 17.06418918918919, + "grad_norm": 0.29367807507514954, + "learning_rate": 2.6600666666666668e-05, + "loss": 0.0118, + "step": 20204 + }, + { + "epoch": 17.065033783783782, + "grad_norm": 0.19786328077316284, + "learning_rate": 2.6600333333333333e-05, + "loss": 0.0075, + "step": 20205 + }, + { + "epoch": 17.06587837837838, + "grad_norm": 0.12654486298561096, + "learning_rate": 2.6600000000000003e-05, + "loss": 0.0076, + "step": 20206 + }, + { + "epoch": 17.066722972972972, + "grad_norm": 0.229843869805336, + "learning_rate": 2.6599666666666665e-05, + "loss": 0.0079, + "step": 20207 + }, + { + "epoch": 17.06756756756757, + "grad_norm": 0.5784358382225037, + "learning_rate": 2.6599333333333335e-05, + "loss": 0.0239, + "step": 20208 + }, + { + "epoch": 17.06841216216216, + "grad_norm": 0.22884877026081085, + "learning_rate": 2.6599e-05, + "loss": 0.0068, + "step": 20209 + }, + { + "epoch": 17.069256756756758, + "grad_norm": 0.1869702935218811, + "learning_rate": 2.6598666666666667e-05, + "loss": 0.013, + "step": 20210 + }, + { + "epoch": 17.07010135135135, + "grad_norm": 0.18947292864322662, + "learning_rate": 2.6598333333333336e-05, + "loss": 0.008, + "step": 20211 + }, + { + "epoch": 17.070945945945947, + "grad_norm": 0.20225340127944946, + "learning_rate": 2.6598000000000002e-05, + "loss": 0.0094, + "step": 20212 + }, + { + "epoch": 17.07179054054054, + "grad_norm": 0.24393349885940552, + "learning_rate": 2.6597666666666668e-05, + "loss": 0.0138, + "step": 20213 + }, + { + "epoch": 17.072635135135137, + "grad_norm": 0.28008660674095154, + "learning_rate": 2.6597333333333334e-05, + "loss": 0.0074, + "step": 20214 + }, + { + "epoch": 17.07347972972973, + "grad_norm": 0.23399963974952698, + "learning_rate": 2.6597e-05, + "loss": 0.0115, + "step": 20215 + }, + { + "epoch": 17.074324324324323, + "grad_norm": 0.1339341700077057, + "learning_rate": 2.6596666666666666e-05, + "loss": 0.007, + "step": 20216 + }, + { + "epoch": 17.07516891891892, + "grad_norm": 0.27926889061927795, + "learning_rate": 2.6596333333333335e-05, + "loss": 0.0091, + "step": 20217 + }, + { + "epoch": 17.076013513513512, + "grad_norm": 0.2617841064929962, + "learning_rate": 2.6596e-05, + "loss": 0.0094, + "step": 20218 + }, + { + "epoch": 17.07685810810811, + "grad_norm": 0.2338234931230545, + "learning_rate": 2.6595666666666667e-05, + "loss": 0.0073, + "step": 20219 + }, + { + "epoch": 17.0777027027027, + "grad_norm": 0.14003416895866394, + "learning_rate": 2.6595333333333336e-05, + "loss": 0.0056, + "step": 20220 + }, + { + "epoch": 17.0785472972973, + "grad_norm": 0.20715689659118652, + "learning_rate": 2.6595e-05, + "loss": 0.0109, + "step": 20221 + }, + { + "epoch": 17.07939189189189, + "grad_norm": 0.2273479849100113, + "learning_rate": 2.6594666666666668e-05, + "loss": 0.0108, + "step": 20222 + }, + { + "epoch": 17.080236486486488, + "grad_norm": 0.3076590597629547, + "learning_rate": 2.6594333333333334e-05, + "loss": 0.0051, + "step": 20223 + }, + { + "epoch": 17.08108108108108, + "grad_norm": 0.3282453417778015, + "learning_rate": 2.6594e-05, + "loss": 0.0098, + "step": 20224 + }, + { + "epoch": 17.081925675675677, + "grad_norm": 0.16735850274562836, + "learning_rate": 2.6593666666666666e-05, + "loss": 0.0061, + "step": 20225 + }, + { + "epoch": 17.08277027027027, + "grad_norm": 0.23229104280471802, + "learning_rate": 2.6593333333333335e-05, + "loss": 0.0111, + "step": 20226 + }, + { + "epoch": 17.083614864864863, + "grad_norm": 0.19499698281288147, + "learning_rate": 2.6593e-05, + "loss": 0.0131, + "step": 20227 + }, + { + "epoch": 17.08445945945946, + "grad_norm": 0.41815346479415894, + "learning_rate": 2.6592666666666667e-05, + "loss": 0.0088, + "step": 20228 + }, + { + "epoch": 17.085304054054053, + "grad_norm": 0.5538463592529297, + "learning_rate": 2.6592333333333336e-05, + "loss": 0.1269, + "step": 20229 + }, + { + "epoch": 17.08614864864865, + "grad_norm": 0.5415225625038147, + "learning_rate": 2.6592e-05, + "loss": 0.0644, + "step": 20230 + }, + { + "epoch": 17.086993243243242, + "grad_norm": 0.21717402338981628, + "learning_rate": 2.6591666666666668e-05, + "loss": 0.0286, + "step": 20231 + }, + { + "epoch": 17.08783783783784, + "grad_norm": 0.18255630135536194, + "learning_rate": 2.6591333333333334e-05, + "loss": 0.0154, + "step": 20232 + }, + { + "epoch": 17.08868243243243, + "grad_norm": 0.25863033533096313, + "learning_rate": 2.6591e-05, + "loss": 0.0224, + "step": 20233 + }, + { + "epoch": 17.089527027027028, + "grad_norm": 0.20607520639896393, + "learning_rate": 2.6590666666666666e-05, + "loss": 0.0184, + "step": 20234 + }, + { + "epoch": 17.09037162162162, + "grad_norm": 0.22720056772232056, + "learning_rate": 2.6590333333333335e-05, + "loss": 0.012, + "step": 20235 + }, + { + "epoch": 17.091216216216218, + "grad_norm": 0.19037403166294098, + "learning_rate": 2.659e-05, + "loss": 0.0115, + "step": 20236 + }, + { + "epoch": 17.09206081081081, + "grad_norm": 0.11522271484136581, + "learning_rate": 2.6589666666666667e-05, + "loss": 0.0089, + "step": 20237 + }, + { + "epoch": 17.092905405405407, + "grad_norm": 0.1625034213066101, + "learning_rate": 2.6589333333333336e-05, + "loss": 0.0106, + "step": 20238 + }, + { + "epoch": 17.09375, + "grad_norm": 0.369968444108963, + "learning_rate": 2.6589e-05, + "loss": 0.0128, + "step": 20239 + }, + { + "epoch": 17.094594594594593, + "grad_norm": 0.16537347435951233, + "learning_rate": 2.658866666666667e-05, + "loss": 0.0091, + "step": 20240 + }, + { + "epoch": 17.09543918918919, + "grad_norm": 0.18321946263313293, + "learning_rate": 2.6588333333333334e-05, + "loss": 0.0136, + "step": 20241 + }, + { + "epoch": 17.096283783783782, + "grad_norm": 0.17315304279327393, + "learning_rate": 2.6588e-05, + "loss": 0.0111, + "step": 20242 + }, + { + "epoch": 17.09712837837838, + "grad_norm": 0.32963430881500244, + "learning_rate": 2.658766666666667e-05, + "loss": 0.0112, + "step": 20243 + }, + { + "epoch": 17.097972972972972, + "grad_norm": 0.478508859872818, + "learning_rate": 2.6587333333333332e-05, + "loss": 0.0162, + "step": 20244 + }, + { + "epoch": 17.09881756756757, + "grad_norm": 0.36420774459838867, + "learning_rate": 2.6587e-05, + "loss": 0.0127, + "step": 20245 + }, + { + "epoch": 17.09966216216216, + "grad_norm": 0.2580210268497467, + "learning_rate": 2.6586666666666667e-05, + "loss": 0.0101, + "step": 20246 + }, + { + "epoch": 17.100506756756758, + "grad_norm": 0.6462010741233826, + "learning_rate": 2.6586333333333333e-05, + "loss": 0.0117, + "step": 20247 + }, + { + "epoch": 17.10135135135135, + "grad_norm": 0.21760697662830353, + "learning_rate": 2.6586e-05, + "loss": 0.0299, + "step": 20248 + }, + { + "epoch": 17.102195945945947, + "grad_norm": 0.23505331575870514, + "learning_rate": 2.658566666666667e-05, + "loss": 0.0088, + "step": 20249 + }, + { + "epoch": 17.10304054054054, + "grad_norm": 0.15167048573493958, + "learning_rate": 2.658533333333333e-05, + "loss": 0.0108, + "step": 20250 + }, + { + "epoch": 17.103885135135137, + "grad_norm": 0.18130037188529968, + "learning_rate": 2.6585e-05, + "loss": 0.01, + "step": 20251 + }, + { + "epoch": 17.10472972972973, + "grad_norm": 0.16603076457977295, + "learning_rate": 2.658466666666667e-05, + "loss": 0.0085, + "step": 20252 + }, + { + "epoch": 17.105574324324323, + "grad_norm": 0.3219764530658722, + "learning_rate": 2.6584333333333332e-05, + "loss": 0.0072, + "step": 20253 + }, + { + "epoch": 17.10641891891892, + "grad_norm": 0.10029509663581848, + "learning_rate": 2.6584e-05, + "loss": 0.0053, + "step": 20254 + }, + { + "epoch": 17.107263513513512, + "grad_norm": 0.15830521285533905, + "learning_rate": 2.6583666666666668e-05, + "loss": 0.0046, + "step": 20255 + }, + { + "epoch": 17.10810810810811, + "grad_norm": 0.11417665332555771, + "learning_rate": 2.6583333333333333e-05, + "loss": 0.006, + "step": 20256 + }, + { + "epoch": 17.1089527027027, + "grad_norm": 0.1423196941614151, + "learning_rate": 2.6583e-05, + "loss": 0.0063, + "step": 20257 + }, + { + "epoch": 17.1097972972973, + "grad_norm": 0.2846553921699524, + "learning_rate": 2.658266666666667e-05, + "loss": 0.0106, + "step": 20258 + }, + { + "epoch": 17.11064189189189, + "grad_norm": 0.1169080063700676, + "learning_rate": 2.658233333333333e-05, + "loss": 0.006, + "step": 20259 + }, + { + "epoch": 17.111486486486488, + "grad_norm": 0.3856663107872009, + "learning_rate": 2.6582e-05, + "loss": 0.0106, + "step": 20260 + }, + { + "epoch": 17.11233108108108, + "grad_norm": 0.25285160541534424, + "learning_rate": 2.658166666666667e-05, + "loss": 0.0086, + "step": 20261 + }, + { + "epoch": 17.113175675675677, + "grad_norm": 0.22419503331184387, + "learning_rate": 2.6581333333333332e-05, + "loss": 0.0118, + "step": 20262 + }, + { + "epoch": 17.11402027027027, + "grad_norm": 0.3103466033935547, + "learning_rate": 2.6581000000000002e-05, + "loss": 0.0115, + "step": 20263 + }, + { + "epoch": 17.114864864864863, + "grad_norm": 0.19858258962631226, + "learning_rate": 2.6580666666666668e-05, + "loss": 0.009, + "step": 20264 + }, + { + "epoch": 17.11570945945946, + "grad_norm": 0.28578606247901917, + "learning_rate": 2.6580333333333334e-05, + "loss": 0.0125, + "step": 20265 + }, + { + "epoch": 17.116554054054053, + "grad_norm": 0.425554484128952, + "learning_rate": 2.658e-05, + "loss": 0.0128, + "step": 20266 + }, + { + "epoch": 17.11739864864865, + "grad_norm": 0.32265686988830566, + "learning_rate": 2.657966666666667e-05, + "loss": 0.0214, + "step": 20267 + }, + { + "epoch": 17.118243243243242, + "grad_norm": 0.40737655758857727, + "learning_rate": 2.6579333333333335e-05, + "loss": 0.0132, + "step": 20268 + }, + { + "epoch": 17.11908783783784, + "grad_norm": 0.6275595426559448, + "learning_rate": 2.6579e-05, + "loss": 0.0155, + "step": 20269 + }, + { + "epoch": 17.11993243243243, + "grad_norm": 0.2721918821334839, + "learning_rate": 2.657866666666667e-05, + "loss": 0.004, + "step": 20270 + }, + { + "epoch": 17.120777027027028, + "grad_norm": 0.36668339371681213, + "learning_rate": 2.6578333333333333e-05, + "loss": 0.0221, + "step": 20271 + }, + { + "epoch": 17.12162162162162, + "grad_norm": 0.2655080258846283, + "learning_rate": 2.6578000000000002e-05, + "loss": 0.0093, + "step": 20272 + }, + { + "epoch": 17.122466216216218, + "grad_norm": 0.1653519570827484, + "learning_rate": 2.6577666666666668e-05, + "loss": 0.007, + "step": 20273 + }, + { + "epoch": 17.12331081081081, + "grad_norm": 0.37068042159080505, + "learning_rate": 2.6577333333333334e-05, + "loss": 0.0181, + "step": 20274 + }, + { + "epoch": 17.124155405405407, + "grad_norm": 0.5553783178329468, + "learning_rate": 2.6577e-05, + "loss": 0.0071, + "step": 20275 + }, + { + "epoch": 17.125, + "grad_norm": 0.4286063313484192, + "learning_rate": 2.6576666666666666e-05, + "loss": 0.0169, + "step": 20276 + }, + { + "epoch": 17.125844594594593, + "grad_norm": 0.4956154525279999, + "learning_rate": 2.6576333333333335e-05, + "loss": 0.0243, + "step": 20277 + }, + { + "epoch": 17.12668918918919, + "grad_norm": 0.3405512571334839, + "learning_rate": 2.6576e-05, + "loss": 0.0172, + "step": 20278 + }, + { + "epoch": 17.127533783783782, + "grad_norm": 0.4092283844947815, + "learning_rate": 2.6575666666666667e-05, + "loss": 0.1105, + "step": 20279 + }, + { + "epoch": 17.12837837837838, + "grad_norm": 0.43650075793266296, + "learning_rate": 2.6575333333333333e-05, + "loss": 0.068, + "step": 20280 + }, + { + "epoch": 17.129222972972972, + "grad_norm": 0.3508577048778534, + "learning_rate": 2.6575000000000002e-05, + "loss": 0.0471, + "step": 20281 + }, + { + "epoch": 17.13006756756757, + "grad_norm": 0.2075316607952118, + "learning_rate": 2.6574666666666665e-05, + "loss": 0.0118, + "step": 20282 + }, + { + "epoch": 17.13091216216216, + "grad_norm": 0.15551355481147766, + "learning_rate": 2.6574333333333334e-05, + "loss": 0.0156, + "step": 20283 + }, + { + "epoch": 17.131756756756758, + "grad_norm": 0.24844759702682495, + "learning_rate": 2.6574e-05, + "loss": 0.0158, + "step": 20284 + }, + { + "epoch": 17.13260135135135, + "grad_norm": 0.21501430869102478, + "learning_rate": 2.6573666666666666e-05, + "loss": 0.0121, + "step": 20285 + }, + { + "epoch": 17.133445945945947, + "grad_norm": 0.1699712872505188, + "learning_rate": 2.6573333333333335e-05, + "loss": 0.009, + "step": 20286 + }, + { + "epoch": 17.13429054054054, + "grad_norm": 0.19900958240032196, + "learning_rate": 2.6573e-05, + "loss": 0.0151, + "step": 20287 + }, + { + "epoch": 17.135135135135137, + "grad_norm": 0.11189257353544235, + "learning_rate": 2.6572666666666667e-05, + "loss": 0.0061, + "step": 20288 + }, + { + "epoch": 17.13597972972973, + "grad_norm": 0.17104758322238922, + "learning_rate": 2.6572333333333333e-05, + "loss": 0.0091, + "step": 20289 + }, + { + "epoch": 17.136824324324323, + "grad_norm": 0.6208055019378662, + "learning_rate": 2.6572000000000002e-05, + "loss": 0.024, + "step": 20290 + }, + { + "epoch": 17.13766891891892, + "grad_norm": 0.15833780169487, + "learning_rate": 2.6571666666666665e-05, + "loss": 0.0107, + "step": 20291 + }, + { + "epoch": 17.138513513513512, + "grad_norm": 0.35154616832733154, + "learning_rate": 2.6571333333333334e-05, + "loss": 0.018, + "step": 20292 + }, + { + "epoch": 17.13935810810811, + "grad_norm": 0.1817365139722824, + "learning_rate": 2.6571000000000004e-05, + "loss": 0.012, + "step": 20293 + }, + { + "epoch": 17.1402027027027, + "grad_norm": 0.21757617592811584, + "learning_rate": 2.6570666666666666e-05, + "loss": 0.0134, + "step": 20294 + }, + { + "epoch": 17.1410472972973, + "grad_norm": 0.14699578285217285, + "learning_rate": 2.6570333333333335e-05, + "loss": 0.0084, + "step": 20295 + }, + { + "epoch": 17.14189189189189, + "grad_norm": 0.13946564495563507, + "learning_rate": 2.657e-05, + "loss": 0.0085, + "step": 20296 + }, + { + "epoch": 17.142736486486488, + "grad_norm": 0.22959914803504944, + "learning_rate": 2.6569666666666667e-05, + "loss": 0.0101, + "step": 20297 + }, + { + "epoch": 17.14358108108108, + "grad_norm": 0.20295336842536926, + "learning_rate": 2.6569333333333333e-05, + "loss": 0.0086, + "step": 20298 + }, + { + "epoch": 17.144425675675677, + "grad_norm": 0.19881072640419006, + "learning_rate": 2.6569000000000003e-05, + "loss": 0.0106, + "step": 20299 + }, + { + "epoch": 17.14527027027027, + "grad_norm": 0.23064066469669342, + "learning_rate": 2.6568666666666665e-05, + "loss": 0.0121, + "step": 20300 + }, + { + "epoch": 17.146114864864863, + "grad_norm": 0.24535097181797028, + "learning_rate": 2.6568333333333334e-05, + "loss": 0.0156, + "step": 20301 + }, + { + "epoch": 17.14695945945946, + "grad_norm": 0.18076524138450623, + "learning_rate": 2.6568000000000004e-05, + "loss": 0.0098, + "step": 20302 + }, + { + "epoch": 17.147804054054053, + "grad_norm": 0.20539769530296326, + "learning_rate": 2.6567666666666666e-05, + "loss": 0.0096, + "step": 20303 + }, + { + "epoch": 17.14864864864865, + "grad_norm": 0.15901750326156616, + "learning_rate": 2.6567333333333336e-05, + "loss": 0.0114, + "step": 20304 + }, + { + "epoch": 17.149493243243242, + "grad_norm": 0.2709020972251892, + "learning_rate": 2.6566999999999998e-05, + "loss": 0.0096, + "step": 20305 + }, + { + "epoch": 17.15033783783784, + "grad_norm": 0.2385926991701126, + "learning_rate": 2.6566666666666668e-05, + "loss": 0.0087, + "step": 20306 + }, + { + "epoch": 17.15118243243243, + "grad_norm": 0.32948848605155945, + "learning_rate": 2.6566333333333333e-05, + "loss": 0.0128, + "step": 20307 + }, + { + "epoch": 17.152027027027028, + "grad_norm": 0.9273222088813782, + "learning_rate": 2.6566e-05, + "loss": 0.0113, + "step": 20308 + }, + { + "epoch": 17.15287162162162, + "grad_norm": 0.2575879693031311, + "learning_rate": 2.6565666666666665e-05, + "loss": 0.0162, + "step": 20309 + }, + { + "epoch": 17.153716216216218, + "grad_norm": 0.19275788962841034, + "learning_rate": 2.6565333333333335e-05, + "loss": 0.0107, + "step": 20310 + }, + { + "epoch": 17.15456081081081, + "grad_norm": 0.2417689710855484, + "learning_rate": 2.6565e-05, + "loss": 0.0101, + "step": 20311 + }, + { + "epoch": 17.155405405405407, + "grad_norm": 0.1526051163673401, + "learning_rate": 2.6564666666666667e-05, + "loss": 0.0057, + "step": 20312 + }, + { + "epoch": 17.15625, + "grad_norm": 0.26356518268585205, + "learning_rate": 2.6564333333333336e-05, + "loss": 0.0156, + "step": 20313 + }, + { + "epoch": 17.157094594594593, + "grad_norm": 0.290452241897583, + "learning_rate": 2.6564e-05, + "loss": 0.013, + "step": 20314 + }, + { + "epoch": 17.15793918918919, + "grad_norm": 0.16343852877616882, + "learning_rate": 2.6563666666666668e-05, + "loss": 0.0103, + "step": 20315 + }, + { + "epoch": 17.158783783783782, + "grad_norm": 0.10342618077993393, + "learning_rate": 2.6563333333333334e-05, + "loss": 0.0064, + "step": 20316 + }, + { + "epoch": 17.15962837837838, + "grad_norm": 0.2500399053096771, + "learning_rate": 2.6563e-05, + "loss": 0.0112, + "step": 20317 + }, + { + "epoch": 17.160472972972972, + "grad_norm": 0.34626391530036926, + "learning_rate": 2.656266666666667e-05, + "loss": 0.0127, + "step": 20318 + }, + { + "epoch": 17.16131756756757, + "grad_norm": 0.19894586503505707, + "learning_rate": 2.6562333333333335e-05, + "loss": 0.0088, + "step": 20319 + }, + { + "epoch": 17.16216216216216, + "grad_norm": 0.18376125395298004, + "learning_rate": 2.6562e-05, + "loss": 0.0088, + "step": 20320 + }, + { + "epoch": 17.163006756756758, + "grad_norm": 0.18578658998012543, + "learning_rate": 2.6561666666666667e-05, + "loss": 0.0182, + "step": 20321 + }, + { + "epoch": 17.16385135135135, + "grad_norm": 0.2835889756679535, + "learning_rate": 2.6561333333333336e-05, + "loss": 0.0095, + "step": 20322 + }, + { + "epoch": 17.164695945945947, + "grad_norm": 0.1919495016336441, + "learning_rate": 2.6561e-05, + "loss": 0.0098, + "step": 20323 + }, + { + "epoch": 17.16554054054054, + "grad_norm": 0.1850907951593399, + "learning_rate": 2.6560666666666668e-05, + "loss": 0.0082, + "step": 20324 + }, + { + "epoch": 17.166385135135137, + "grad_norm": 0.3374785780906677, + "learning_rate": 2.6560333333333334e-05, + "loss": 0.0253, + "step": 20325 + }, + { + "epoch": 17.16722972972973, + "grad_norm": 0.25779688358306885, + "learning_rate": 2.656e-05, + "loss": 0.0128, + "step": 20326 + }, + { + "epoch": 17.168074324324323, + "grad_norm": 0.3115796446800232, + "learning_rate": 2.655966666666667e-05, + "loss": 0.0096, + "step": 20327 + }, + { + "epoch": 17.16891891891892, + "grad_norm": 0.09219354391098022, + "learning_rate": 2.6559333333333335e-05, + "loss": 0.0067, + "step": 20328 + }, + { + "epoch": 17.169763513513512, + "grad_norm": 0.5950193405151367, + "learning_rate": 2.6559e-05, + "loss": 0.1216, + "step": 20329 + }, + { + "epoch": 17.17060810810811, + "grad_norm": 0.338037371635437, + "learning_rate": 2.6558666666666667e-05, + "loss": 0.0636, + "step": 20330 + }, + { + "epoch": 17.1714527027027, + "grad_norm": 0.3653857707977295, + "learning_rate": 2.6558333333333336e-05, + "loss": 0.0601, + "step": 20331 + }, + { + "epoch": 17.1722972972973, + "grad_norm": 1.2306207418441772, + "learning_rate": 2.6558e-05, + "loss": 0.02, + "step": 20332 + }, + { + "epoch": 17.17314189189189, + "grad_norm": 0.35086575150489807, + "learning_rate": 2.6557666666666668e-05, + "loss": 0.0351, + "step": 20333 + }, + { + "epoch": 17.173986486486488, + "grad_norm": 0.18714265525341034, + "learning_rate": 2.655733333333333e-05, + "loss": 0.0104, + "step": 20334 + }, + { + "epoch": 17.17483108108108, + "grad_norm": 0.18710090219974518, + "learning_rate": 2.6557e-05, + "loss": 0.0138, + "step": 20335 + }, + { + "epoch": 17.175675675675677, + "grad_norm": 0.5565236210823059, + "learning_rate": 2.655666666666667e-05, + "loss": 0.0208, + "step": 20336 + }, + { + "epoch": 17.17652027027027, + "grad_norm": 0.12414928525686264, + "learning_rate": 2.6556333333333332e-05, + "loss": 0.0068, + "step": 20337 + }, + { + "epoch": 17.177364864864863, + "grad_norm": 0.17727400362491608, + "learning_rate": 2.6556e-05, + "loss": 0.0123, + "step": 20338 + }, + { + "epoch": 17.17820945945946, + "grad_norm": 0.29555395245552063, + "learning_rate": 2.6555666666666667e-05, + "loss": 0.0133, + "step": 20339 + }, + { + "epoch": 17.179054054054053, + "grad_norm": 0.2858080267906189, + "learning_rate": 2.6555333333333333e-05, + "loss": 0.0244, + "step": 20340 + }, + { + "epoch": 17.17989864864865, + "grad_norm": 0.28708958625793457, + "learning_rate": 2.6555e-05, + "loss": 0.0105, + "step": 20341 + }, + { + "epoch": 17.180743243243242, + "grad_norm": 0.17631205916404724, + "learning_rate": 2.655466666666667e-05, + "loss": 0.0189, + "step": 20342 + }, + { + "epoch": 17.18158783783784, + "grad_norm": 0.20683811604976654, + "learning_rate": 2.6554333333333334e-05, + "loss": 0.0094, + "step": 20343 + }, + { + "epoch": 17.18243243243243, + "grad_norm": 0.25454196333885193, + "learning_rate": 2.6554e-05, + "loss": 0.0188, + "step": 20344 + }, + { + "epoch": 17.183277027027028, + "grad_norm": 0.7854971289634705, + "learning_rate": 2.655366666666667e-05, + "loss": 0.0109, + "step": 20345 + }, + { + "epoch": 17.18412162162162, + "grad_norm": 0.3386842608451843, + "learning_rate": 2.6553333333333332e-05, + "loss": 0.0136, + "step": 20346 + }, + { + "epoch": 17.184966216216218, + "grad_norm": 0.38398218154907227, + "learning_rate": 2.6553e-05, + "loss": 0.0176, + "step": 20347 + }, + { + "epoch": 17.18581081081081, + "grad_norm": 0.5003837943077087, + "learning_rate": 2.6552666666666667e-05, + "loss": 0.0336, + "step": 20348 + }, + { + "epoch": 17.186655405405407, + "grad_norm": 0.24233287572860718, + "learning_rate": 2.6552333333333333e-05, + "loss": 0.009, + "step": 20349 + }, + { + "epoch": 17.1875, + "grad_norm": 0.6681084632873535, + "learning_rate": 2.6552e-05, + "loss": 0.009, + "step": 20350 + }, + { + "epoch": 17.188344594594593, + "grad_norm": 0.215255007147789, + "learning_rate": 2.655166666666667e-05, + "loss": 0.0153, + "step": 20351 + }, + { + "epoch": 17.18918918918919, + "grad_norm": 0.2041672021150589, + "learning_rate": 2.6551333333333334e-05, + "loss": 0.0102, + "step": 20352 + }, + { + "epoch": 17.190033783783782, + "grad_norm": 0.22052207589149475, + "learning_rate": 2.6551e-05, + "loss": 0.0109, + "step": 20353 + }, + { + "epoch": 17.19087837837838, + "grad_norm": 0.2690315842628479, + "learning_rate": 2.655066666666667e-05, + "loss": 0.0198, + "step": 20354 + }, + { + "epoch": 17.191722972972972, + "grad_norm": 0.25280874967575073, + "learning_rate": 2.6550333333333332e-05, + "loss": 0.0117, + "step": 20355 + }, + { + "epoch": 17.19256756756757, + "grad_norm": 0.10344468802213669, + "learning_rate": 2.655e-05, + "loss": 0.0051, + "step": 20356 + }, + { + "epoch": 17.19341216216216, + "grad_norm": 0.25324687361717224, + "learning_rate": 2.6549666666666668e-05, + "loss": 0.0104, + "step": 20357 + }, + { + "epoch": 17.194256756756758, + "grad_norm": 0.3379001319408417, + "learning_rate": 2.6549333333333333e-05, + "loss": 0.0127, + "step": 20358 + }, + { + "epoch": 17.19510135135135, + "grad_norm": 0.25589677691459656, + "learning_rate": 2.6549e-05, + "loss": 0.0103, + "step": 20359 + }, + { + "epoch": 17.195945945945947, + "grad_norm": 0.2078464776277542, + "learning_rate": 2.654866666666667e-05, + "loss": 0.0115, + "step": 20360 + }, + { + "epoch": 17.19679054054054, + "grad_norm": 0.18879163265228271, + "learning_rate": 2.6548333333333335e-05, + "loss": 0.0088, + "step": 20361 + }, + { + "epoch": 17.197635135135137, + "grad_norm": 0.20126481354236603, + "learning_rate": 2.6548e-05, + "loss": 0.0138, + "step": 20362 + }, + { + "epoch": 17.19847972972973, + "grad_norm": 0.17240756750106812, + "learning_rate": 2.654766666666667e-05, + "loss": 0.0097, + "step": 20363 + }, + { + "epoch": 17.199324324324323, + "grad_norm": 0.42361727356910706, + "learning_rate": 2.6547333333333332e-05, + "loss": 0.0243, + "step": 20364 + }, + { + "epoch": 17.20016891891892, + "grad_norm": 0.3594917356967926, + "learning_rate": 2.6547000000000002e-05, + "loss": 0.0139, + "step": 20365 + }, + { + "epoch": 17.201013513513512, + "grad_norm": 0.2893505096435547, + "learning_rate": 2.6546666666666664e-05, + "loss": 0.0117, + "step": 20366 + }, + { + "epoch": 17.20185810810811, + "grad_norm": 0.3156919479370117, + "learning_rate": 2.6546333333333334e-05, + "loss": 0.0111, + "step": 20367 + }, + { + "epoch": 17.2027027027027, + "grad_norm": 0.24552924931049347, + "learning_rate": 2.6546000000000003e-05, + "loss": 0.0145, + "step": 20368 + }, + { + "epoch": 17.2035472972973, + "grad_norm": 0.19411513209342957, + "learning_rate": 2.6545666666666666e-05, + "loss": 0.0097, + "step": 20369 + }, + { + "epoch": 17.20439189189189, + "grad_norm": 0.12929219007492065, + "learning_rate": 2.6545333333333335e-05, + "loss": 0.0061, + "step": 20370 + }, + { + "epoch": 17.205236486486488, + "grad_norm": 0.34146854281425476, + "learning_rate": 2.6545e-05, + "loss": 0.0144, + "step": 20371 + }, + { + "epoch": 17.20608108108108, + "grad_norm": 0.23447854816913605, + "learning_rate": 2.6544666666666667e-05, + "loss": 0.008, + "step": 20372 + }, + { + "epoch": 17.206925675675677, + "grad_norm": 0.41355565190315247, + "learning_rate": 2.6544333333333333e-05, + "loss": 0.0145, + "step": 20373 + }, + { + "epoch": 17.20777027027027, + "grad_norm": 0.4422474801540375, + "learning_rate": 2.6544000000000002e-05, + "loss": 0.024, + "step": 20374 + }, + { + "epoch": 17.208614864864863, + "grad_norm": 0.31623077392578125, + "learning_rate": 2.6543666666666665e-05, + "loss": 0.0148, + "step": 20375 + }, + { + "epoch": 17.20945945945946, + "grad_norm": 0.3024877607822418, + "learning_rate": 2.6543333333333334e-05, + "loss": 0.0116, + "step": 20376 + }, + { + "epoch": 17.210304054054053, + "grad_norm": 0.4629150629043579, + "learning_rate": 2.6543000000000003e-05, + "loss": 0.0103, + "step": 20377 + }, + { + "epoch": 17.21114864864865, + "grad_norm": 0.40071552991867065, + "learning_rate": 2.6542666666666666e-05, + "loss": 0.0619, + "step": 20378 + }, + { + "epoch": 17.211993243243242, + "grad_norm": 0.42134353518486023, + "learning_rate": 2.6542333333333335e-05, + "loss": 0.1196, + "step": 20379 + }, + { + "epoch": 17.21283783783784, + "grad_norm": 0.3000779151916504, + "learning_rate": 2.6542e-05, + "loss": 0.0667, + "step": 20380 + }, + { + "epoch": 17.21368243243243, + "grad_norm": 0.28437915444374084, + "learning_rate": 2.6541666666666667e-05, + "loss": 0.0381, + "step": 20381 + }, + { + "epoch": 17.214527027027028, + "grad_norm": 0.1850946843624115, + "learning_rate": 2.6541333333333333e-05, + "loss": 0.0168, + "step": 20382 + }, + { + "epoch": 17.21537162162162, + "grad_norm": 0.5032756924629211, + "learning_rate": 2.6541000000000002e-05, + "loss": 0.0212, + "step": 20383 + }, + { + "epoch": 17.216216216216218, + "grad_norm": 0.23962369561195374, + "learning_rate": 2.6540666666666665e-05, + "loss": 0.016, + "step": 20384 + }, + { + "epoch": 17.21706081081081, + "grad_norm": 0.26995623111724854, + "learning_rate": 2.6540333333333334e-05, + "loss": 0.0271, + "step": 20385 + }, + { + "epoch": 17.217905405405407, + "grad_norm": 0.5191651582717896, + "learning_rate": 2.6540000000000003e-05, + "loss": 0.0179, + "step": 20386 + }, + { + "epoch": 17.21875, + "grad_norm": 0.2574825882911682, + "learning_rate": 2.6539666666666666e-05, + "loss": 0.019, + "step": 20387 + }, + { + "epoch": 17.219594594594593, + "grad_norm": 0.29421529173851013, + "learning_rate": 2.6539333333333335e-05, + "loss": 0.0112, + "step": 20388 + }, + { + "epoch": 17.22043918918919, + "grad_norm": 0.25338250398635864, + "learning_rate": 2.6539e-05, + "loss": 0.0143, + "step": 20389 + }, + { + "epoch": 17.221283783783782, + "grad_norm": 0.14838910102844238, + "learning_rate": 2.6538666666666667e-05, + "loss": 0.008, + "step": 20390 + }, + { + "epoch": 17.22212837837838, + "grad_norm": 0.18577806651592255, + "learning_rate": 2.6538333333333333e-05, + "loss": 0.01, + "step": 20391 + }, + { + "epoch": 17.222972972972972, + "grad_norm": 0.3176786005496979, + "learning_rate": 2.6538000000000002e-05, + "loss": 0.0198, + "step": 20392 + }, + { + "epoch": 17.22381756756757, + "grad_norm": 0.2308070808649063, + "learning_rate": 2.653766666666667e-05, + "loss": 0.0071, + "step": 20393 + }, + { + "epoch": 17.22466216216216, + "grad_norm": 0.29385942220687866, + "learning_rate": 2.6537333333333334e-05, + "loss": 0.0176, + "step": 20394 + }, + { + "epoch": 17.225506756756758, + "grad_norm": 0.1695120632648468, + "learning_rate": 2.6537e-05, + "loss": 0.0053, + "step": 20395 + }, + { + "epoch": 17.22635135135135, + "grad_norm": 0.24720852077007294, + "learning_rate": 2.6536666666666666e-05, + "loss": 0.0208, + "step": 20396 + }, + { + "epoch": 17.227195945945947, + "grad_norm": 0.18561868369579315, + "learning_rate": 2.6536333333333335e-05, + "loss": 0.01, + "step": 20397 + }, + { + "epoch": 17.22804054054054, + "grad_norm": 0.2250462770462036, + "learning_rate": 2.6535999999999998e-05, + "loss": 0.0084, + "step": 20398 + }, + { + "epoch": 17.228885135135137, + "grad_norm": 0.2371789962053299, + "learning_rate": 2.6535666666666667e-05, + "loss": 0.0135, + "step": 20399 + }, + { + "epoch": 17.22972972972973, + "grad_norm": 0.14925821125507355, + "learning_rate": 2.6535333333333333e-05, + "loss": 0.0101, + "step": 20400 + }, + { + "epoch": 17.230574324324323, + "grad_norm": 0.20757333934307098, + "learning_rate": 2.6535e-05, + "loss": 0.0075, + "step": 20401 + }, + { + "epoch": 17.23141891891892, + "grad_norm": 0.5566330552101135, + "learning_rate": 2.653466666666667e-05, + "loss": 0.0153, + "step": 20402 + }, + { + "epoch": 17.232263513513512, + "grad_norm": 0.24205972254276276, + "learning_rate": 2.6534333333333334e-05, + "loss": 0.0101, + "step": 20403 + }, + { + "epoch": 17.23310810810811, + "grad_norm": 0.24057233333587646, + "learning_rate": 2.6534e-05, + "loss": 0.0076, + "step": 20404 + }, + { + "epoch": 17.2339527027027, + "grad_norm": 0.1967032253742218, + "learning_rate": 2.6533666666666666e-05, + "loss": 0.0088, + "step": 20405 + }, + { + "epoch": 17.2347972972973, + "grad_norm": 0.13730430603027344, + "learning_rate": 2.6533333333333336e-05, + "loss": 0.0058, + "step": 20406 + }, + { + "epoch": 17.23564189189189, + "grad_norm": 0.10688403248786926, + "learning_rate": 2.6532999999999998e-05, + "loss": 0.0055, + "step": 20407 + }, + { + "epoch": 17.236486486486488, + "grad_norm": 0.2029438316822052, + "learning_rate": 2.6532666666666668e-05, + "loss": 0.0083, + "step": 20408 + }, + { + "epoch": 17.23733108108108, + "grad_norm": 0.37790605425834656, + "learning_rate": 2.6532333333333333e-05, + "loss": 0.0157, + "step": 20409 + }, + { + "epoch": 17.238175675675677, + "grad_norm": 0.22681301832199097, + "learning_rate": 2.6532e-05, + "loss": 0.0081, + "step": 20410 + }, + { + "epoch": 17.23902027027027, + "grad_norm": 0.19151008129119873, + "learning_rate": 2.653166666666667e-05, + "loss": 0.0134, + "step": 20411 + }, + { + "epoch": 17.239864864864863, + "grad_norm": 0.11708272993564606, + "learning_rate": 2.6531333333333335e-05, + "loss": 0.0058, + "step": 20412 + }, + { + "epoch": 17.24070945945946, + "grad_norm": 0.27826759219169617, + "learning_rate": 2.6531e-05, + "loss": 0.0067, + "step": 20413 + }, + { + "epoch": 17.241554054054053, + "grad_norm": 0.635734498500824, + "learning_rate": 2.6530666666666667e-05, + "loss": 0.0175, + "step": 20414 + }, + { + "epoch": 17.24239864864865, + "grad_norm": 0.30018502473831177, + "learning_rate": 2.6530333333333336e-05, + "loss": 0.0173, + "step": 20415 + }, + { + "epoch": 17.243243243243242, + "grad_norm": 0.3056379556655884, + "learning_rate": 2.653e-05, + "loss": 0.0146, + "step": 20416 + }, + { + "epoch": 17.24408783783784, + "grad_norm": 0.11865570396184921, + "learning_rate": 2.6529666666666668e-05, + "loss": 0.0113, + "step": 20417 + }, + { + "epoch": 17.24493243243243, + "grad_norm": 0.16726505756378174, + "learning_rate": 2.6529333333333337e-05, + "loss": 0.0055, + "step": 20418 + }, + { + "epoch": 17.245777027027028, + "grad_norm": 0.23691020905971527, + "learning_rate": 2.6529e-05, + "loss": 0.0165, + "step": 20419 + }, + { + "epoch": 17.24662162162162, + "grad_norm": 0.21415579319000244, + "learning_rate": 2.652866666666667e-05, + "loss": 0.0069, + "step": 20420 + }, + { + "epoch": 17.247466216216218, + "grad_norm": 0.3596586585044861, + "learning_rate": 2.6528333333333335e-05, + "loss": 0.0128, + "step": 20421 + }, + { + "epoch": 17.24831081081081, + "grad_norm": 0.22204391658306122, + "learning_rate": 2.6528e-05, + "loss": 0.0078, + "step": 20422 + }, + { + "epoch": 17.249155405405407, + "grad_norm": 0.2174736112356186, + "learning_rate": 2.6527666666666667e-05, + "loss": 0.0092, + "step": 20423 + }, + { + "epoch": 17.25, + "grad_norm": 0.29763564467430115, + "learning_rate": 2.6527333333333336e-05, + "loss": 0.0074, + "step": 20424 + }, + { + "epoch": 17.250844594594593, + "grad_norm": 0.304736852645874, + "learning_rate": 2.6527e-05, + "loss": 0.0172, + "step": 20425 + }, + { + "epoch": 17.25168918918919, + "grad_norm": 0.18094255030155182, + "learning_rate": 2.6526666666666668e-05, + "loss": 0.0056, + "step": 20426 + }, + { + "epoch": 17.252533783783782, + "grad_norm": 0.31220948696136475, + "learning_rate": 2.6526333333333334e-05, + "loss": 0.0138, + "step": 20427 + }, + { + "epoch": 17.25337837837838, + "grad_norm": 0.6055034399032593, + "learning_rate": 2.6526e-05, + "loss": 0.0117, + "step": 20428 + }, + { + "epoch": 17.254222972972972, + "grad_norm": 0.621918797492981, + "learning_rate": 2.652566666666667e-05, + "loss": 0.1496, + "step": 20429 + }, + { + "epoch": 17.25506756756757, + "grad_norm": 0.4824022054672241, + "learning_rate": 2.652533333333333e-05, + "loss": 0.0913, + "step": 20430 + }, + { + "epoch": 17.25591216216216, + "grad_norm": 0.3605186343193054, + "learning_rate": 2.6525e-05, + "loss": 0.0435, + "step": 20431 + }, + { + "epoch": 17.256756756756758, + "grad_norm": 0.2308012843132019, + "learning_rate": 2.6524666666666667e-05, + "loss": 0.024, + "step": 20432 + }, + { + "epoch": 17.25760135135135, + "grad_norm": 0.14759600162506104, + "learning_rate": 2.6524333333333333e-05, + "loss": 0.0132, + "step": 20433 + }, + { + "epoch": 17.258445945945947, + "grad_norm": 0.20050382614135742, + "learning_rate": 2.6524e-05, + "loss": 0.0146, + "step": 20434 + }, + { + "epoch": 17.25929054054054, + "grad_norm": 0.31572622060775757, + "learning_rate": 2.6523666666666668e-05, + "loss": 0.0137, + "step": 20435 + }, + { + "epoch": 17.260135135135137, + "grad_norm": 0.23352336883544922, + "learning_rate": 2.6523333333333334e-05, + "loss": 0.0217, + "step": 20436 + }, + { + "epoch": 17.26097972972973, + "grad_norm": 0.256930410861969, + "learning_rate": 2.6523e-05, + "loss": 0.0126, + "step": 20437 + }, + { + "epoch": 17.261824324324323, + "grad_norm": 0.205433189868927, + "learning_rate": 2.652266666666667e-05, + "loss": 0.0085, + "step": 20438 + }, + { + "epoch": 17.26266891891892, + "grad_norm": 0.4241963326931, + "learning_rate": 2.6522333333333332e-05, + "loss": 0.0211, + "step": 20439 + }, + { + "epoch": 17.263513513513512, + "grad_norm": 0.2779706120491028, + "learning_rate": 2.6522e-05, + "loss": 0.0131, + "step": 20440 + }, + { + "epoch": 17.26435810810811, + "grad_norm": 0.14130163192749023, + "learning_rate": 2.6521666666666667e-05, + "loss": 0.008, + "step": 20441 + }, + { + "epoch": 17.2652027027027, + "grad_norm": 0.2573670744895935, + "learning_rate": 2.6521333333333333e-05, + "loss": 0.0101, + "step": 20442 + }, + { + "epoch": 17.2660472972973, + "grad_norm": 0.18389399349689484, + "learning_rate": 2.6521000000000002e-05, + "loss": 0.0113, + "step": 20443 + }, + { + "epoch": 17.26689189189189, + "grad_norm": 0.34110161662101746, + "learning_rate": 2.652066666666667e-05, + "loss": 0.0104, + "step": 20444 + }, + { + "epoch": 17.267736486486488, + "grad_norm": 0.22807173430919647, + "learning_rate": 2.6520333333333334e-05, + "loss": 0.0091, + "step": 20445 + }, + { + "epoch": 17.26858108108108, + "grad_norm": 0.2062566578388214, + "learning_rate": 2.652e-05, + "loss": 0.0091, + "step": 20446 + }, + { + "epoch": 17.269425675675677, + "grad_norm": 0.23174327611923218, + "learning_rate": 2.651966666666667e-05, + "loss": 0.0078, + "step": 20447 + }, + { + "epoch": 17.27027027027027, + "grad_norm": 0.11749245971441269, + "learning_rate": 2.6519333333333332e-05, + "loss": 0.0082, + "step": 20448 + }, + { + "epoch": 17.271114864864863, + "grad_norm": 0.30307701230049133, + "learning_rate": 2.6519e-05, + "loss": 0.0127, + "step": 20449 + }, + { + "epoch": 17.27195945945946, + "grad_norm": 0.13305765390396118, + "learning_rate": 2.6518666666666667e-05, + "loss": 0.0098, + "step": 20450 + }, + { + "epoch": 17.272804054054053, + "grad_norm": 0.12141695618629456, + "learning_rate": 2.6518333333333333e-05, + "loss": 0.0069, + "step": 20451 + }, + { + "epoch": 17.27364864864865, + "grad_norm": 0.21871912479400635, + "learning_rate": 2.6518000000000003e-05, + "loss": 0.0155, + "step": 20452 + }, + { + "epoch": 17.274493243243242, + "grad_norm": 0.11915320158004761, + "learning_rate": 2.651766666666667e-05, + "loss": 0.0082, + "step": 20453 + }, + { + "epoch": 17.27533783783784, + "grad_norm": 0.35001835227012634, + "learning_rate": 2.6517333333333334e-05, + "loss": 0.0081, + "step": 20454 + }, + { + "epoch": 17.27618243243243, + "grad_norm": 0.27684256434440613, + "learning_rate": 2.6517e-05, + "loss": 0.0086, + "step": 20455 + }, + { + "epoch": 17.277027027027028, + "grad_norm": 0.15060876309871674, + "learning_rate": 2.6516666666666666e-05, + "loss": 0.0098, + "step": 20456 + }, + { + "epoch": 17.27787162162162, + "grad_norm": 0.2054673135280609, + "learning_rate": 2.6516333333333332e-05, + "loss": 0.0089, + "step": 20457 + }, + { + "epoch": 17.278716216216218, + "grad_norm": 0.6167739033699036, + "learning_rate": 2.6516e-05, + "loss": 0.0134, + "step": 20458 + }, + { + "epoch": 17.27956081081081, + "grad_norm": 0.47434699535369873, + "learning_rate": 2.6515666666666668e-05, + "loss": 0.0164, + "step": 20459 + }, + { + "epoch": 17.280405405405407, + "grad_norm": 0.1571979820728302, + "learning_rate": 2.6515333333333333e-05, + "loss": 0.0075, + "step": 20460 + }, + { + "epoch": 17.28125, + "grad_norm": 0.22525066137313843, + "learning_rate": 2.6515000000000003e-05, + "loss": 0.0129, + "step": 20461 + }, + { + "epoch": 17.282094594594593, + "grad_norm": 0.7347282767295837, + "learning_rate": 2.6514666666666665e-05, + "loss": 0.0192, + "step": 20462 + }, + { + "epoch": 17.28293918918919, + "grad_norm": 0.37630653381347656, + "learning_rate": 2.6514333333333335e-05, + "loss": 0.017, + "step": 20463 + }, + { + "epoch": 17.283783783783782, + "grad_norm": 0.5980290770530701, + "learning_rate": 2.6514e-05, + "loss": 0.0509, + "step": 20464 + }, + { + "epoch": 17.28462837837838, + "grad_norm": 0.47835466265678406, + "learning_rate": 2.6513666666666667e-05, + "loss": 0.0087, + "step": 20465 + }, + { + "epoch": 17.285472972972972, + "grad_norm": 0.42532142996788025, + "learning_rate": 2.6513333333333332e-05, + "loss": 0.056, + "step": 20466 + }, + { + "epoch": 17.28631756756757, + "grad_norm": 0.3051443099975586, + "learning_rate": 2.6513000000000002e-05, + "loss": 0.0086, + "step": 20467 + }, + { + "epoch": 17.28716216216216, + "grad_norm": 0.3666139841079712, + "learning_rate": 2.6512666666666668e-05, + "loss": 0.0122, + "step": 20468 + }, + { + "epoch": 17.288006756756758, + "grad_norm": 0.4239935278892517, + "learning_rate": 2.6512333333333334e-05, + "loss": 0.0202, + "step": 20469 + }, + { + "epoch": 17.28885135135135, + "grad_norm": 0.20952686667442322, + "learning_rate": 2.6512000000000003e-05, + "loss": 0.0051, + "step": 20470 + }, + { + "epoch": 17.289695945945947, + "grad_norm": 0.43912598490715027, + "learning_rate": 2.6511666666666666e-05, + "loss": 0.0148, + "step": 20471 + }, + { + "epoch": 17.29054054054054, + "grad_norm": 0.34033915400505066, + "learning_rate": 2.6511333333333335e-05, + "loss": 0.0225, + "step": 20472 + }, + { + "epoch": 17.291385135135137, + "grad_norm": 0.4291020631790161, + "learning_rate": 2.6511e-05, + "loss": 0.0148, + "step": 20473 + }, + { + "epoch": 17.29222972972973, + "grad_norm": 0.3336700201034546, + "learning_rate": 2.6510666666666667e-05, + "loss": 0.0197, + "step": 20474 + }, + { + "epoch": 17.293074324324323, + "grad_norm": 0.40077510476112366, + "learning_rate": 2.6510333333333333e-05, + "loss": 0.0085, + "step": 20475 + }, + { + "epoch": 17.29391891891892, + "grad_norm": 0.2189883440732956, + "learning_rate": 2.6510000000000002e-05, + "loss": 0.014, + "step": 20476 + }, + { + "epoch": 17.294763513513512, + "grad_norm": 0.5309166312217712, + "learning_rate": 2.6509666666666668e-05, + "loss": 0.015, + "step": 20477 + }, + { + "epoch": 17.29560810810811, + "grad_norm": 0.359415203332901, + "learning_rate": 2.6509333333333334e-05, + "loss": 0.0079, + "step": 20478 + }, + { + "epoch": 17.2964527027027, + "grad_norm": 0.5345059037208557, + "learning_rate": 2.6509000000000003e-05, + "loss": 0.1205, + "step": 20479 + }, + { + "epoch": 17.2972972972973, + "grad_norm": 0.3074483275413513, + "learning_rate": 2.6508666666666666e-05, + "loss": 0.0943, + "step": 20480 + }, + { + "epoch": 17.29814189189189, + "grad_norm": 0.24229341745376587, + "learning_rate": 2.6508333333333335e-05, + "loss": 0.029, + "step": 20481 + }, + { + "epoch": 17.298986486486488, + "grad_norm": 0.191271111369133, + "learning_rate": 2.6508e-05, + "loss": 0.018, + "step": 20482 + }, + { + "epoch": 17.29983108108108, + "grad_norm": 0.2741510272026062, + "learning_rate": 2.6507666666666667e-05, + "loss": 0.0244, + "step": 20483 + }, + { + "epoch": 17.300675675675677, + "grad_norm": 0.31537917256355286, + "learning_rate": 2.6507333333333333e-05, + "loss": 0.02, + "step": 20484 + }, + { + "epoch": 17.30152027027027, + "grad_norm": 0.3752654194831848, + "learning_rate": 2.6507e-05, + "loss": 0.0183, + "step": 20485 + }, + { + "epoch": 17.302364864864863, + "grad_norm": 0.19962863624095917, + "learning_rate": 2.6506666666666668e-05, + "loss": 0.012, + "step": 20486 + }, + { + "epoch": 17.30320945945946, + "grad_norm": 0.19973620772361755, + "learning_rate": 2.6506333333333334e-05, + "loss": 0.0144, + "step": 20487 + }, + { + "epoch": 17.304054054054053, + "grad_norm": 0.227559432387352, + "learning_rate": 2.6506e-05, + "loss": 0.0184, + "step": 20488 + }, + { + "epoch": 17.30489864864865, + "grad_norm": 0.1492932140827179, + "learning_rate": 2.6505666666666666e-05, + "loss": 0.0073, + "step": 20489 + }, + { + "epoch": 17.305743243243242, + "grad_norm": 0.228739395737648, + "learning_rate": 2.6505333333333335e-05, + "loss": 0.0156, + "step": 20490 + }, + { + "epoch": 17.30658783783784, + "grad_norm": 0.1769954264163971, + "learning_rate": 2.6504999999999998e-05, + "loss": 0.012, + "step": 20491 + }, + { + "epoch": 17.30743243243243, + "grad_norm": 0.22534126043319702, + "learning_rate": 2.6504666666666667e-05, + "loss": 0.0137, + "step": 20492 + }, + { + "epoch": 17.308277027027028, + "grad_norm": 0.23723377287387848, + "learning_rate": 2.6504333333333336e-05, + "loss": 0.0104, + "step": 20493 + }, + { + "epoch": 17.30912162162162, + "grad_norm": 0.1972476691007614, + "learning_rate": 2.6504e-05, + "loss": 0.0072, + "step": 20494 + }, + { + "epoch": 17.309966216216218, + "grad_norm": 0.6749029755592346, + "learning_rate": 2.650366666666667e-05, + "loss": 0.0149, + "step": 20495 + }, + { + "epoch": 17.31081081081081, + "grad_norm": 0.3090088665485382, + "learning_rate": 2.6503333333333334e-05, + "loss": 0.0164, + "step": 20496 + }, + { + "epoch": 17.311655405405407, + "grad_norm": 0.2631024718284607, + "learning_rate": 2.6503e-05, + "loss": 0.0143, + "step": 20497 + }, + { + "epoch": 17.3125, + "grad_norm": 0.2584463655948639, + "learning_rate": 2.6502666666666666e-05, + "loss": 0.0187, + "step": 20498 + }, + { + "epoch": 17.313344594594593, + "grad_norm": 0.17583072185516357, + "learning_rate": 2.6502333333333335e-05, + "loss": 0.0166, + "step": 20499 + }, + { + "epoch": 17.31418918918919, + "grad_norm": 0.15618085861206055, + "learning_rate": 2.6501999999999998e-05, + "loss": 0.0087, + "step": 20500 + }, + { + "epoch": 17.315033783783782, + "grad_norm": 0.1427364945411682, + "learning_rate": 2.6501666666666667e-05, + "loss": 0.0071, + "step": 20501 + }, + { + "epoch": 17.31587837837838, + "grad_norm": 0.14500176906585693, + "learning_rate": 2.6501333333333337e-05, + "loss": 0.0077, + "step": 20502 + }, + { + "epoch": 17.316722972972972, + "grad_norm": 0.13213184475898743, + "learning_rate": 2.6501e-05, + "loss": 0.005, + "step": 20503 + }, + { + "epoch": 17.31756756756757, + "grad_norm": 0.08394637703895569, + "learning_rate": 2.650066666666667e-05, + "loss": 0.0059, + "step": 20504 + }, + { + "epoch": 17.31841216216216, + "grad_norm": 0.36265498399734497, + "learning_rate": 2.6500333333333334e-05, + "loss": 0.0184, + "step": 20505 + }, + { + "epoch": 17.319256756756758, + "grad_norm": 0.142057865858078, + "learning_rate": 2.65e-05, + "loss": 0.0062, + "step": 20506 + }, + { + "epoch": 17.32010135135135, + "grad_norm": 0.2130260020494461, + "learning_rate": 2.6499666666666666e-05, + "loss": 0.0146, + "step": 20507 + }, + { + "epoch": 17.320945945945947, + "grad_norm": 0.2841265797615051, + "learning_rate": 2.6499333333333336e-05, + "loss": 0.0166, + "step": 20508 + }, + { + "epoch": 17.32179054054054, + "grad_norm": 0.33034124970436096, + "learning_rate": 2.6499e-05, + "loss": 0.0146, + "step": 20509 + }, + { + "epoch": 17.322635135135137, + "grad_norm": 0.3607029914855957, + "learning_rate": 2.6498666666666668e-05, + "loss": 0.0139, + "step": 20510 + }, + { + "epoch": 17.32347972972973, + "grad_norm": 0.37483838200569153, + "learning_rate": 2.6498333333333337e-05, + "loss": 0.0151, + "step": 20511 + }, + { + "epoch": 17.324324324324323, + "grad_norm": 0.1517849713563919, + "learning_rate": 2.6498e-05, + "loss": 0.0099, + "step": 20512 + }, + { + "epoch": 17.32516891891892, + "grad_norm": 0.18873140215873718, + "learning_rate": 2.649766666666667e-05, + "loss": 0.0119, + "step": 20513 + }, + { + "epoch": 17.326013513513512, + "grad_norm": 0.10838600993156433, + "learning_rate": 2.6497333333333335e-05, + "loss": 0.004, + "step": 20514 + }, + { + "epoch": 17.32685810810811, + "grad_norm": 0.11829833686351776, + "learning_rate": 2.6497e-05, + "loss": 0.0062, + "step": 20515 + }, + { + "epoch": 17.3277027027027, + "grad_norm": 0.28796905279159546, + "learning_rate": 2.6496666666666667e-05, + "loss": 0.0138, + "step": 20516 + }, + { + "epoch": 17.3285472972973, + "grad_norm": 0.18416175246238708, + "learning_rate": 2.6496333333333332e-05, + "loss": 0.0085, + "step": 20517 + }, + { + "epoch": 17.32939189189189, + "grad_norm": 0.253917396068573, + "learning_rate": 2.6496000000000002e-05, + "loss": 0.0143, + "step": 20518 + }, + { + "epoch": 17.330236486486488, + "grad_norm": 0.20322836935520172, + "learning_rate": 2.6495666666666668e-05, + "loss": 0.0074, + "step": 20519 + }, + { + "epoch": 17.33108108108108, + "grad_norm": 0.4480322599411011, + "learning_rate": 2.6495333333333334e-05, + "loss": 0.0138, + "step": 20520 + }, + { + "epoch": 17.331925675675677, + "grad_norm": 0.31149041652679443, + "learning_rate": 2.6495e-05, + "loss": 0.0172, + "step": 20521 + }, + { + "epoch": 17.33277027027027, + "grad_norm": 0.3184524178504944, + "learning_rate": 2.649466666666667e-05, + "loss": 0.0103, + "step": 20522 + }, + { + "epoch": 17.333614864864863, + "grad_norm": 0.3240010142326355, + "learning_rate": 2.649433333333333e-05, + "loss": 0.0066, + "step": 20523 + }, + { + "epoch": 17.33445945945946, + "grad_norm": 0.2994574010372162, + "learning_rate": 2.6494e-05, + "loss": 0.0131, + "step": 20524 + }, + { + "epoch": 17.335304054054053, + "grad_norm": 1.6286109685897827, + "learning_rate": 2.6493666666666667e-05, + "loss": 0.0287, + "step": 20525 + }, + { + "epoch": 17.33614864864865, + "grad_norm": 0.1726675033569336, + "learning_rate": 2.6493333333333333e-05, + "loss": 0.0076, + "step": 20526 + }, + { + "epoch": 17.336993243243242, + "grad_norm": 0.26660940051078796, + "learning_rate": 2.6493000000000002e-05, + "loss": 0.0121, + "step": 20527 + }, + { + "epoch": 17.33783783783784, + "grad_norm": 0.46246469020843506, + "learning_rate": 2.6492666666666668e-05, + "loss": 0.0259, + "step": 20528 + }, + { + "epoch": 17.33868243243243, + "grad_norm": 0.5077871084213257, + "learning_rate": 2.6492333333333334e-05, + "loss": 0.1425, + "step": 20529 + }, + { + "epoch": 17.339527027027028, + "grad_norm": 0.8744062781333923, + "learning_rate": 2.6492e-05, + "loss": 0.0831, + "step": 20530 + }, + { + "epoch": 17.34037162162162, + "grad_norm": 0.34192806482315063, + "learning_rate": 2.649166666666667e-05, + "loss": 0.0286, + "step": 20531 + }, + { + "epoch": 17.341216216216218, + "grad_norm": 0.3162730634212494, + "learning_rate": 2.649133333333333e-05, + "loss": 0.0214, + "step": 20532 + }, + { + "epoch": 17.34206081081081, + "grad_norm": 0.22991056740283966, + "learning_rate": 2.6491e-05, + "loss": 0.0257, + "step": 20533 + }, + { + "epoch": 17.342905405405407, + "grad_norm": 0.2565244734287262, + "learning_rate": 2.6490666666666667e-05, + "loss": 0.0156, + "step": 20534 + }, + { + "epoch": 17.34375, + "grad_norm": 0.10767596960067749, + "learning_rate": 2.6490333333333333e-05, + "loss": 0.0091, + "step": 20535 + }, + { + "epoch": 17.344594594594593, + "grad_norm": 0.19227057695388794, + "learning_rate": 2.6490000000000002e-05, + "loss": 0.0198, + "step": 20536 + }, + { + "epoch": 17.34543918918919, + "grad_norm": 1.7474899291992188, + "learning_rate": 2.6489666666666668e-05, + "loss": 0.0146, + "step": 20537 + }, + { + "epoch": 17.346283783783782, + "grad_norm": 0.17401619255542755, + "learning_rate": 2.6489333333333334e-05, + "loss": 0.013, + "step": 20538 + }, + { + "epoch": 17.34712837837838, + "grad_norm": 0.15080483257770538, + "learning_rate": 2.6489e-05, + "loss": 0.0067, + "step": 20539 + }, + { + "epoch": 17.347972972972972, + "grad_norm": 0.43833720684051514, + "learning_rate": 2.648866666666667e-05, + "loss": 0.0142, + "step": 20540 + }, + { + "epoch": 17.34881756756757, + "grad_norm": 0.23635916411876678, + "learning_rate": 2.6488333333333332e-05, + "loss": 0.0178, + "step": 20541 + }, + { + "epoch": 17.34966216216216, + "grad_norm": 0.4035707414150238, + "learning_rate": 2.6488e-05, + "loss": 0.0141, + "step": 20542 + }, + { + "epoch": 17.350506756756758, + "grad_norm": 0.2556866407394409, + "learning_rate": 2.648766666666667e-05, + "loss": 0.0127, + "step": 20543 + }, + { + "epoch": 17.35135135135135, + "grad_norm": 0.2717675566673279, + "learning_rate": 2.6487333333333333e-05, + "loss": 0.0156, + "step": 20544 + }, + { + "epoch": 17.352195945945947, + "grad_norm": 0.2710951864719391, + "learning_rate": 2.6487000000000002e-05, + "loss": 0.0104, + "step": 20545 + }, + { + "epoch": 17.35304054054054, + "grad_norm": 0.2245398312807083, + "learning_rate": 2.6486666666666665e-05, + "loss": 0.0131, + "step": 20546 + }, + { + "epoch": 17.353885135135137, + "grad_norm": 0.3178161382675171, + "learning_rate": 2.6486333333333334e-05, + "loss": 0.0171, + "step": 20547 + }, + { + "epoch": 17.35472972972973, + "grad_norm": 0.25188910961151123, + "learning_rate": 2.6486e-05, + "loss": 0.0163, + "step": 20548 + }, + { + "epoch": 17.355574324324323, + "grad_norm": 0.15526887774467468, + "learning_rate": 2.6485666666666666e-05, + "loss": 0.01, + "step": 20549 + }, + { + "epoch": 17.35641891891892, + "grad_norm": 0.1465648114681244, + "learning_rate": 2.6485333333333332e-05, + "loss": 0.0091, + "step": 20550 + }, + { + "epoch": 17.357263513513512, + "grad_norm": 0.1647321581840515, + "learning_rate": 2.6485e-05, + "loss": 0.0083, + "step": 20551 + }, + { + "epoch": 17.35810810810811, + "grad_norm": 0.1468493491411209, + "learning_rate": 2.6484666666666667e-05, + "loss": 0.0094, + "step": 20552 + }, + { + "epoch": 17.3589527027027, + "grad_norm": 0.36203503608703613, + "learning_rate": 2.6484333333333333e-05, + "loss": 0.0237, + "step": 20553 + }, + { + "epoch": 17.3597972972973, + "grad_norm": 1.0798137187957764, + "learning_rate": 2.6484000000000003e-05, + "loss": 0.0268, + "step": 20554 + }, + { + "epoch": 17.36064189189189, + "grad_norm": 0.40265801548957825, + "learning_rate": 2.6483666666666665e-05, + "loss": 0.0154, + "step": 20555 + }, + { + "epoch": 17.361486486486488, + "grad_norm": 0.18619601428508759, + "learning_rate": 2.6483333333333334e-05, + "loss": 0.0094, + "step": 20556 + }, + { + "epoch": 17.36233108108108, + "grad_norm": 0.165811225771904, + "learning_rate": 2.6483e-05, + "loss": 0.0106, + "step": 20557 + }, + { + "epoch": 17.363175675675677, + "grad_norm": 0.21329733729362488, + "learning_rate": 2.6482666666666666e-05, + "loss": 0.0129, + "step": 20558 + }, + { + "epoch": 17.36402027027027, + "grad_norm": 0.44660860300064087, + "learning_rate": 2.6482333333333336e-05, + "loss": 0.014, + "step": 20559 + }, + { + "epoch": 17.364864864864863, + "grad_norm": 0.24582244455814362, + "learning_rate": 2.6482e-05, + "loss": 0.0097, + "step": 20560 + }, + { + "epoch": 17.36570945945946, + "grad_norm": 0.19197596609592438, + "learning_rate": 2.6481666666666668e-05, + "loss": 0.0109, + "step": 20561 + }, + { + "epoch": 17.366554054054053, + "grad_norm": 0.5558735132217407, + "learning_rate": 2.6481333333333333e-05, + "loss": 0.0097, + "step": 20562 + }, + { + "epoch": 17.36739864864865, + "grad_norm": 0.4921469986438751, + "learning_rate": 2.6481000000000003e-05, + "loss": 0.0152, + "step": 20563 + }, + { + "epoch": 17.368243243243242, + "grad_norm": 0.15410391986370087, + "learning_rate": 2.6480666666666665e-05, + "loss": 0.018, + "step": 20564 + }, + { + "epoch": 17.36908783783784, + "grad_norm": 0.24461831152439117, + "learning_rate": 2.6480333333333335e-05, + "loss": 0.0057, + "step": 20565 + }, + { + "epoch": 17.36993243243243, + "grad_norm": 0.3766486644744873, + "learning_rate": 2.648e-05, + "loss": 0.0501, + "step": 20566 + }, + { + "epoch": 17.370777027027028, + "grad_norm": 0.1259014904499054, + "learning_rate": 2.6479666666666667e-05, + "loss": 0.0079, + "step": 20567 + }, + { + "epoch": 17.37162162162162, + "grad_norm": 0.326213002204895, + "learning_rate": 2.6479333333333336e-05, + "loss": 0.0154, + "step": 20568 + }, + { + "epoch": 17.372466216216218, + "grad_norm": 0.40943342447280884, + "learning_rate": 2.6479000000000002e-05, + "loss": 0.0152, + "step": 20569 + }, + { + "epoch": 17.37331081081081, + "grad_norm": 2.947329044342041, + "learning_rate": 2.6478666666666668e-05, + "loss": 0.0211, + "step": 20570 + }, + { + "epoch": 17.374155405405407, + "grad_norm": 1.7236050367355347, + "learning_rate": 2.6478333333333334e-05, + "loss": 0.0274, + "step": 20571 + }, + { + "epoch": 17.375, + "grad_norm": 0.5120203495025635, + "learning_rate": 2.6478000000000003e-05, + "loss": 0.0163, + "step": 20572 + }, + { + "epoch": 17.375844594594593, + "grad_norm": 0.6267961859703064, + "learning_rate": 2.6477666666666666e-05, + "loss": 0.0099, + "step": 20573 + }, + { + "epoch": 17.37668918918919, + "grad_norm": 0.32202962040901184, + "learning_rate": 2.6477333333333335e-05, + "loss": 0.0112, + "step": 20574 + }, + { + "epoch": 17.377533783783782, + "grad_norm": 0.9648950099945068, + "learning_rate": 2.6477e-05, + "loss": 0.0163, + "step": 20575 + }, + { + "epoch": 17.37837837837838, + "grad_norm": 0.5467977523803711, + "learning_rate": 2.6476666666666667e-05, + "loss": 0.0154, + "step": 20576 + }, + { + "epoch": 17.379222972972972, + "grad_norm": 0.228271946310997, + "learning_rate": 2.6476333333333336e-05, + "loss": 0.0091, + "step": 20577 + }, + { + "epoch": 17.38006756756757, + "grad_norm": 0.256086528301239, + "learning_rate": 2.6476e-05, + "loss": 0.0133, + "step": 20578 + }, + { + "epoch": 17.38091216216216, + "grad_norm": 0.4769381880760193, + "learning_rate": 2.6475666666666668e-05, + "loss": 0.1279, + "step": 20579 + }, + { + "epoch": 17.381756756756758, + "grad_norm": 0.6084849834442139, + "learning_rate": 2.6475333333333334e-05, + "loss": 0.1021, + "step": 20580 + }, + { + "epoch": 17.38260135135135, + "grad_norm": 0.7088574767112732, + "learning_rate": 2.6475e-05, + "loss": 0.0523, + "step": 20581 + }, + { + "epoch": 17.383445945945947, + "grad_norm": 0.2721990644931793, + "learning_rate": 2.6474666666666666e-05, + "loss": 0.0305, + "step": 20582 + }, + { + "epoch": 17.38429054054054, + "grad_norm": 0.3083021938800812, + "learning_rate": 2.6474333333333335e-05, + "loss": 0.0273, + "step": 20583 + }, + { + "epoch": 17.385135135135137, + "grad_norm": 0.173218235373497, + "learning_rate": 2.6474e-05, + "loss": 0.0151, + "step": 20584 + }, + { + "epoch": 17.38597972972973, + "grad_norm": 0.13868167996406555, + "learning_rate": 2.6473666666666667e-05, + "loss": 0.0084, + "step": 20585 + }, + { + "epoch": 17.386824324324323, + "grad_norm": 0.2512985169887543, + "learning_rate": 2.6473333333333336e-05, + "loss": 0.018, + "step": 20586 + }, + { + "epoch": 17.38766891891892, + "grad_norm": 0.12547625601291656, + "learning_rate": 2.6473e-05, + "loss": 0.0048, + "step": 20587 + }, + { + "epoch": 17.388513513513512, + "grad_norm": 0.2517460286617279, + "learning_rate": 2.6472666666666668e-05, + "loss": 0.0187, + "step": 20588 + }, + { + "epoch": 17.38935810810811, + "grad_norm": 0.17424330115318298, + "learning_rate": 2.6472333333333334e-05, + "loss": 0.0178, + "step": 20589 + }, + { + "epoch": 17.3902027027027, + "grad_norm": 0.29650208353996277, + "learning_rate": 2.6472e-05, + "loss": 0.0107, + "step": 20590 + }, + { + "epoch": 17.3910472972973, + "grad_norm": 0.39756691455841064, + "learning_rate": 2.6471666666666666e-05, + "loss": 0.0163, + "step": 20591 + }, + { + "epoch": 17.39189189189189, + "grad_norm": 0.18240810930728912, + "learning_rate": 2.6471333333333335e-05, + "loss": 0.0102, + "step": 20592 + }, + { + "epoch": 17.392736486486488, + "grad_norm": 0.16572245955467224, + "learning_rate": 2.6471e-05, + "loss": 0.0091, + "step": 20593 + }, + { + "epoch": 17.39358108108108, + "grad_norm": 0.1648866832256317, + "learning_rate": 2.6470666666666667e-05, + "loss": 0.0072, + "step": 20594 + }, + { + "epoch": 17.394425675675677, + "grad_norm": 0.40182584524154663, + "learning_rate": 2.6470333333333336e-05, + "loss": 0.0129, + "step": 20595 + }, + { + "epoch": 17.39527027027027, + "grad_norm": 0.22673490643501282, + "learning_rate": 2.647e-05, + "loss": 0.0192, + "step": 20596 + }, + { + "epoch": 17.396114864864863, + "grad_norm": 0.22085948288440704, + "learning_rate": 2.646966666666667e-05, + "loss": 0.0118, + "step": 20597 + }, + { + "epoch": 17.39695945945946, + "grad_norm": 0.3394702672958374, + "learning_rate": 2.6469333333333334e-05, + "loss": 0.0127, + "step": 20598 + }, + { + "epoch": 17.397804054054053, + "grad_norm": 0.24910466372966766, + "learning_rate": 2.6469e-05, + "loss": 0.0114, + "step": 20599 + }, + { + "epoch": 17.39864864864865, + "grad_norm": 0.3066386878490448, + "learning_rate": 2.6468666666666666e-05, + "loss": 0.0209, + "step": 20600 + }, + { + "epoch": 17.399493243243242, + "grad_norm": 0.3546934425830841, + "learning_rate": 2.6468333333333335e-05, + "loss": 0.0166, + "step": 20601 + }, + { + "epoch": 17.40033783783784, + "grad_norm": 0.13742294907569885, + "learning_rate": 2.6468e-05, + "loss": 0.0058, + "step": 20602 + }, + { + "epoch": 17.40118243243243, + "grad_norm": 0.3559055030345917, + "learning_rate": 2.6467666666666667e-05, + "loss": 0.0132, + "step": 20603 + }, + { + "epoch": 17.402027027027028, + "grad_norm": 0.17230427265167236, + "learning_rate": 2.6467333333333337e-05, + "loss": 0.0128, + "step": 20604 + }, + { + "epoch": 17.40287162162162, + "grad_norm": 0.6913607120513916, + "learning_rate": 2.6467e-05, + "loss": 0.0327, + "step": 20605 + }, + { + "epoch": 17.403716216216218, + "grad_norm": 0.3593735992908478, + "learning_rate": 2.646666666666667e-05, + "loss": 0.0121, + "step": 20606 + }, + { + "epoch": 17.40456081081081, + "grad_norm": 0.14633548259735107, + "learning_rate": 2.646633333333333e-05, + "loss": 0.0093, + "step": 20607 + }, + { + "epoch": 17.405405405405407, + "grad_norm": 0.18272662162780762, + "learning_rate": 2.6466e-05, + "loss": 0.0103, + "step": 20608 + }, + { + "epoch": 17.40625, + "grad_norm": 0.24498340487480164, + "learning_rate": 2.646566666666667e-05, + "loss": 0.011, + "step": 20609 + }, + { + "epoch": 17.407094594594593, + "grad_norm": 0.430119127035141, + "learning_rate": 2.6465333333333332e-05, + "loss": 0.0112, + "step": 20610 + }, + { + "epoch": 17.40793918918919, + "grad_norm": 0.2454308420419693, + "learning_rate": 2.6465e-05, + "loss": 0.0121, + "step": 20611 + }, + { + "epoch": 17.408783783783782, + "grad_norm": 0.25617069005966187, + "learning_rate": 2.6464666666666668e-05, + "loss": 0.0184, + "step": 20612 + }, + { + "epoch": 17.40962837837838, + "grad_norm": 0.2920222580432892, + "learning_rate": 2.6464333333333333e-05, + "loss": 0.0139, + "step": 20613 + }, + { + "epoch": 17.410472972972972, + "grad_norm": 0.8605881929397583, + "learning_rate": 2.6464e-05, + "loss": 0.0208, + "step": 20614 + }, + { + "epoch": 17.41131756756757, + "grad_norm": 2.5094170570373535, + "learning_rate": 2.646366666666667e-05, + "loss": 0.0187, + "step": 20615 + }, + { + "epoch": 17.41216216216216, + "grad_norm": 0.18703167140483856, + "learning_rate": 2.646333333333333e-05, + "loss": 0.0096, + "step": 20616 + }, + { + "epoch": 17.413006756756758, + "grad_norm": 0.2725328803062439, + "learning_rate": 2.6463e-05, + "loss": 0.009, + "step": 20617 + }, + { + "epoch": 17.41385135135135, + "grad_norm": 0.24266965687274933, + "learning_rate": 2.646266666666667e-05, + "loss": 0.0089, + "step": 20618 + }, + { + "epoch": 17.414695945945947, + "grad_norm": 0.2259119749069214, + "learning_rate": 2.6462333333333332e-05, + "loss": 0.0069, + "step": 20619 + }, + { + "epoch": 17.41554054054054, + "grad_norm": 0.18740448355674744, + "learning_rate": 2.6462000000000002e-05, + "loss": 0.0079, + "step": 20620 + }, + { + "epoch": 17.416385135135137, + "grad_norm": 0.3145260810852051, + "learning_rate": 2.6461666666666668e-05, + "loss": 0.0133, + "step": 20621 + }, + { + "epoch": 17.41722972972973, + "grad_norm": 0.32436439394950867, + "learning_rate": 2.6461333333333334e-05, + "loss": 0.0127, + "step": 20622 + }, + { + "epoch": 17.418074324324323, + "grad_norm": 0.32832592725753784, + "learning_rate": 2.6461e-05, + "loss": 0.007, + "step": 20623 + }, + { + "epoch": 17.41891891891892, + "grad_norm": 0.4586365520954132, + "learning_rate": 2.646066666666667e-05, + "loss": 0.0168, + "step": 20624 + }, + { + "epoch": 17.419763513513512, + "grad_norm": 0.47244229912757874, + "learning_rate": 2.646033333333333e-05, + "loss": 0.011, + "step": 20625 + }, + { + "epoch": 17.42060810810811, + "grad_norm": 0.18137982487678528, + "learning_rate": 2.646e-05, + "loss": 0.0078, + "step": 20626 + }, + { + "epoch": 17.4214527027027, + "grad_norm": 0.49655911326408386, + "learning_rate": 2.645966666666667e-05, + "loss": 0.0113, + "step": 20627 + }, + { + "epoch": 17.4222972972973, + "grad_norm": 2.584980010986328, + "learning_rate": 2.6459333333333333e-05, + "loss": 0.0456, + "step": 20628 + }, + { + "epoch": 17.42314189189189, + "grad_norm": 0.9432194828987122, + "learning_rate": 2.6459000000000002e-05, + "loss": 0.1877, + "step": 20629 + }, + { + "epoch": 17.423986486486488, + "grad_norm": 0.6166984438896179, + "learning_rate": 2.6458666666666668e-05, + "loss": 0.0662, + "step": 20630 + }, + { + "epoch": 17.42483108108108, + "grad_norm": 0.4136921465396881, + "learning_rate": 2.6458333333333334e-05, + "loss": 0.056, + "step": 20631 + }, + { + "epoch": 17.425675675675677, + "grad_norm": 1.1228253841400146, + "learning_rate": 2.6458e-05, + "loss": 0.0339, + "step": 20632 + }, + { + "epoch": 17.42652027027027, + "grad_norm": 0.3739668130874634, + "learning_rate": 2.645766666666667e-05, + "loss": 0.0318, + "step": 20633 + }, + { + "epoch": 17.427364864864863, + "grad_norm": 0.20248132944107056, + "learning_rate": 2.6457333333333335e-05, + "loss": 0.0147, + "step": 20634 + }, + { + "epoch": 17.42820945945946, + "grad_norm": 0.20806202292442322, + "learning_rate": 2.6457e-05, + "loss": 0.0172, + "step": 20635 + }, + { + "epoch": 17.429054054054053, + "grad_norm": 0.5197327136993408, + "learning_rate": 2.645666666666667e-05, + "loss": 0.0167, + "step": 20636 + }, + { + "epoch": 17.42989864864865, + "grad_norm": 0.2265409231185913, + "learning_rate": 2.6456333333333333e-05, + "loss": 0.017, + "step": 20637 + }, + { + "epoch": 17.430743243243242, + "grad_norm": 0.13301056623458862, + "learning_rate": 2.6456000000000002e-05, + "loss": 0.0073, + "step": 20638 + }, + { + "epoch": 17.43158783783784, + "grad_norm": 0.25533390045166016, + "learning_rate": 2.6455666666666665e-05, + "loss": 0.0243, + "step": 20639 + }, + { + "epoch": 17.43243243243243, + "grad_norm": 0.46214190125465393, + "learning_rate": 2.6455333333333334e-05, + "loss": 0.0261, + "step": 20640 + }, + { + "epoch": 17.433277027027028, + "grad_norm": 1.8725234270095825, + "learning_rate": 2.6455e-05, + "loss": 0.0147, + "step": 20641 + }, + { + "epoch": 17.43412162162162, + "grad_norm": 0.25336772203445435, + "learning_rate": 2.6454666666666666e-05, + "loss": 0.0169, + "step": 20642 + }, + { + "epoch": 17.434966216216218, + "grad_norm": 0.1848335564136505, + "learning_rate": 2.6454333333333335e-05, + "loss": 0.0096, + "step": 20643 + }, + { + "epoch": 17.43581081081081, + "grad_norm": 0.3574279546737671, + "learning_rate": 2.6454e-05, + "loss": 0.0177, + "step": 20644 + }, + { + "epoch": 17.436655405405407, + "grad_norm": 0.23895427584648132, + "learning_rate": 2.6453666666666667e-05, + "loss": 0.017, + "step": 20645 + }, + { + "epoch": 17.4375, + "grad_norm": 0.2590904235839844, + "learning_rate": 2.6453333333333333e-05, + "loss": 0.0183, + "step": 20646 + }, + { + "epoch": 17.438344594594593, + "grad_norm": 0.143774151802063, + "learning_rate": 2.6453000000000002e-05, + "loss": 0.0111, + "step": 20647 + }, + { + "epoch": 17.43918918918919, + "grad_norm": 0.3097897171974182, + "learning_rate": 2.6452666666666665e-05, + "loss": 0.0154, + "step": 20648 + }, + { + "epoch": 17.440033783783782, + "grad_norm": 0.1876155436038971, + "learning_rate": 2.6452333333333334e-05, + "loss": 0.0124, + "step": 20649 + }, + { + "epoch": 17.44087837837838, + "grad_norm": 0.14419768750667572, + "learning_rate": 2.6452e-05, + "loss": 0.009, + "step": 20650 + }, + { + "epoch": 17.441722972972972, + "grad_norm": 0.1548696905374527, + "learning_rate": 2.6451666666666666e-05, + "loss": 0.0092, + "step": 20651 + }, + { + "epoch": 17.44256756756757, + "grad_norm": 0.1774548888206482, + "learning_rate": 2.6451333333333335e-05, + "loss": 0.0088, + "step": 20652 + }, + { + "epoch": 17.44341216216216, + "grad_norm": 0.15099366009235382, + "learning_rate": 2.6451e-05, + "loss": 0.008, + "step": 20653 + }, + { + "epoch": 17.444256756756758, + "grad_norm": 0.33433443307876587, + "learning_rate": 2.6450666666666667e-05, + "loss": 0.0091, + "step": 20654 + }, + { + "epoch": 17.44510135135135, + "grad_norm": 0.6528444886207581, + "learning_rate": 2.6450333333333333e-05, + "loss": 0.0109, + "step": 20655 + }, + { + "epoch": 17.445945945945947, + "grad_norm": 0.12172052264213562, + "learning_rate": 2.6450000000000003e-05, + "loss": 0.0079, + "step": 20656 + }, + { + "epoch": 17.44679054054054, + "grad_norm": 0.28363630175590515, + "learning_rate": 2.6449666666666665e-05, + "loss": 0.0127, + "step": 20657 + }, + { + "epoch": 17.447635135135137, + "grad_norm": 0.2681371867656708, + "learning_rate": 2.6449333333333334e-05, + "loss": 0.016, + "step": 20658 + }, + { + "epoch": 17.44847972972973, + "grad_norm": 0.136338010430336, + "learning_rate": 2.6449000000000004e-05, + "loss": 0.0073, + "step": 20659 + }, + { + "epoch": 17.449324324324323, + "grad_norm": 0.20175482332706451, + "learning_rate": 2.6448666666666666e-05, + "loss": 0.0158, + "step": 20660 + }, + { + "epoch": 17.45016891891892, + "grad_norm": 0.1747310757637024, + "learning_rate": 2.6448333333333336e-05, + "loss": 0.0107, + "step": 20661 + }, + { + "epoch": 17.451013513513512, + "grad_norm": 0.39465638995170593, + "learning_rate": 2.6448e-05, + "loss": 0.011, + "step": 20662 + }, + { + "epoch": 17.45185810810811, + "grad_norm": 0.17524173855781555, + "learning_rate": 2.6447666666666668e-05, + "loss": 0.0067, + "step": 20663 + }, + { + "epoch": 17.4527027027027, + "grad_norm": 0.11352553963661194, + "learning_rate": 2.6447333333333333e-05, + "loss": 0.0051, + "step": 20664 + }, + { + "epoch": 17.4535472972973, + "grad_norm": 0.3708215355873108, + "learning_rate": 2.6447000000000003e-05, + "loss": 0.0203, + "step": 20665 + }, + { + "epoch": 17.45439189189189, + "grad_norm": 0.2930543124675751, + "learning_rate": 2.6446666666666665e-05, + "loss": 0.0061, + "step": 20666 + }, + { + "epoch": 17.455236486486488, + "grad_norm": 0.4574612081050873, + "learning_rate": 2.6446333333333335e-05, + "loss": 0.0139, + "step": 20667 + }, + { + "epoch": 17.45608108108108, + "grad_norm": 0.3253757953643799, + "learning_rate": 2.6446e-05, + "loss": 0.0121, + "step": 20668 + }, + { + "epoch": 17.456925675675677, + "grad_norm": 0.547715961933136, + "learning_rate": 2.6445666666666667e-05, + "loss": 0.0151, + "step": 20669 + }, + { + "epoch": 17.45777027027027, + "grad_norm": 0.26766836643218994, + "learning_rate": 2.6445333333333336e-05, + "loss": 0.0135, + "step": 20670 + }, + { + "epoch": 17.458614864864863, + "grad_norm": 0.46905744075775146, + "learning_rate": 2.6445e-05, + "loss": 0.0174, + "step": 20671 + }, + { + "epoch": 17.45945945945946, + "grad_norm": 0.3205924332141876, + "learning_rate": 2.6444666666666668e-05, + "loss": 0.0504, + "step": 20672 + }, + { + "epoch": 17.460304054054053, + "grad_norm": 0.4876959025859833, + "learning_rate": 2.6444333333333334e-05, + "loss": 0.019, + "step": 20673 + }, + { + "epoch": 17.46114864864865, + "grad_norm": 0.4080825448036194, + "learning_rate": 2.6444e-05, + "loss": 0.0121, + "step": 20674 + }, + { + "epoch": 17.461993243243242, + "grad_norm": 0.7119101285934448, + "learning_rate": 2.6443666666666666e-05, + "loss": 0.0194, + "step": 20675 + }, + { + "epoch": 17.46283783783784, + "grad_norm": 0.2520095705986023, + "learning_rate": 2.6443333333333335e-05, + "loss": 0.0068, + "step": 20676 + }, + { + "epoch": 17.46368243243243, + "grad_norm": 0.21045003831386566, + "learning_rate": 2.6443e-05, + "loss": 0.0095, + "step": 20677 + }, + { + "epoch": 17.464527027027028, + "grad_norm": 0.5142418742179871, + "learning_rate": 2.6442666666666667e-05, + "loss": 0.0157, + "step": 20678 + }, + { + "epoch": 17.46537162162162, + "grad_norm": 0.6112800240516663, + "learning_rate": 2.6442333333333336e-05, + "loss": 0.1608, + "step": 20679 + }, + { + "epoch": 17.466216216216218, + "grad_norm": 0.6490424275398254, + "learning_rate": 2.6442e-05, + "loss": 0.0954, + "step": 20680 + }, + { + "epoch": 17.46706081081081, + "grad_norm": 0.32868462800979614, + "learning_rate": 2.6441666666666668e-05, + "loss": 0.0362, + "step": 20681 + }, + { + "epoch": 17.467905405405407, + "grad_norm": 0.2865753769874573, + "learning_rate": 2.6441333333333334e-05, + "loss": 0.0148, + "step": 20682 + }, + { + "epoch": 17.46875, + "grad_norm": 0.3348965644836426, + "learning_rate": 2.6441e-05, + "loss": 0.013, + "step": 20683 + }, + { + "epoch": 17.469594594594593, + "grad_norm": 0.3107299208641052, + "learning_rate": 2.644066666666667e-05, + "loss": 0.0228, + "step": 20684 + }, + { + "epoch": 17.47043918918919, + "grad_norm": 0.29906854033470154, + "learning_rate": 2.6440333333333335e-05, + "loss": 0.0134, + "step": 20685 + }, + { + "epoch": 17.471283783783782, + "grad_norm": 0.201941579580307, + "learning_rate": 2.644e-05, + "loss": 0.0109, + "step": 20686 + }, + { + "epoch": 17.47212837837838, + "grad_norm": 0.14478357136249542, + "learning_rate": 2.6439666666666667e-05, + "loss": 0.0106, + "step": 20687 + }, + { + "epoch": 17.472972972972972, + "grad_norm": 0.20126831531524658, + "learning_rate": 2.6439333333333336e-05, + "loss": 0.0086, + "step": 20688 + }, + { + "epoch": 17.47381756756757, + "grad_norm": 0.2021738737821579, + "learning_rate": 2.6439e-05, + "loss": 0.0132, + "step": 20689 + }, + { + "epoch": 17.47466216216216, + "grad_norm": 0.2185145914554596, + "learning_rate": 2.6438666666666668e-05, + "loss": 0.0093, + "step": 20690 + }, + { + "epoch": 17.475506756756758, + "grad_norm": 0.1460707038640976, + "learning_rate": 2.6438333333333334e-05, + "loss": 0.0127, + "step": 20691 + }, + { + "epoch": 17.47635135135135, + "grad_norm": 0.28131070733070374, + "learning_rate": 2.6438e-05, + "loss": 0.0141, + "step": 20692 + }, + { + "epoch": 17.477195945945947, + "grad_norm": 0.1853073686361313, + "learning_rate": 2.643766666666667e-05, + "loss": 0.0084, + "step": 20693 + }, + { + "epoch": 17.47804054054054, + "grad_norm": 0.17978936433792114, + "learning_rate": 2.6437333333333335e-05, + "loss": 0.0098, + "step": 20694 + }, + { + "epoch": 17.478885135135137, + "grad_norm": 0.288775235414505, + "learning_rate": 2.6437e-05, + "loss": 0.0171, + "step": 20695 + }, + { + "epoch": 17.47972972972973, + "grad_norm": 0.2722279727458954, + "learning_rate": 2.6436666666666667e-05, + "loss": 0.0132, + "step": 20696 + }, + { + "epoch": 17.480574324324323, + "grad_norm": 0.26961997151374817, + "learning_rate": 2.6436333333333333e-05, + "loss": 0.0118, + "step": 20697 + }, + { + "epoch": 17.48141891891892, + "grad_norm": 0.21593157947063446, + "learning_rate": 2.6436e-05, + "loss": 0.0101, + "step": 20698 + }, + { + "epoch": 17.482263513513512, + "grad_norm": 0.2569977045059204, + "learning_rate": 2.6435666666666668e-05, + "loss": 0.0141, + "step": 20699 + }, + { + "epoch": 17.48310810810811, + "grad_norm": 0.16596809029579163, + "learning_rate": 2.643533333333333e-05, + "loss": 0.0077, + "step": 20700 + }, + { + "epoch": 17.4839527027027, + "grad_norm": 0.2539590895175934, + "learning_rate": 2.6435e-05, + "loss": 0.0093, + "step": 20701 + }, + { + "epoch": 17.4847972972973, + "grad_norm": 0.18592669069766998, + "learning_rate": 2.643466666666667e-05, + "loss": 0.0132, + "step": 20702 + }, + { + "epoch": 17.48564189189189, + "grad_norm": 0.5167093873023987, + "learning_rate": 2.6434333333333332e-05, + "loss": 0.0054, + "step": 20703 + }, + { + "epoch": 17.486486486486488, + "grad_norm": 0.3958689868450165, + "learning_rate": 2.6434e-05, + "loss": 0.0186, + "step": 20704 + }, + { + "epoch": 17.48733108108108, + "grad_norm": 0.28176501393318176, + "learning_rate": 2.6433666666666667e-05, + "loss": 0.0152, + "step": 20705 + }, + { + "epoch": 17.488175675675677, + "grad_norm": 0.25665244460105896, + "learning_rate": 2.6433333333333333e-05, + "loss": 0.0154, + "step": 20706 + }, + { + "epoch": 17.48902027027027, + "grad_norm": 0.15955661237239838, + "learning_rate": 2.6433e-05, + "loss": 0.0114, + "step": 20707 + }, + { + "epoch": 17.489864864864863, + "grad_norm": 0.32034996151924133, + "learning_rate": 2.643266666666667e-05, + "loss": 0.0156, + "step": 20708 + }, + { + "epoch": 17.49070945945946, + "grad_norm": 0.2556689977645874, + "learning_rate": 2.6432333333333334e-05, + "loss": 0.0125, + "step": 20709 + }, + { + "epoch": 17.491554054054053, + "grad_norm": 0.2613679766654968, + "learning_rate": 2.6432e-05, + "loss": 0.0125, + "step": 20710 + }, + { + "epoch": 17.49239864864865, + "grad_norm": 0.26443400979042053, + "learning_rate": 2.643166666666667e-05, + "loss": 0.0091, + "step": 20711 + }, + { + "epoch": 17.493243243243242, + "grad_norm": 0.17690728604793549, + "learning_rate": 2.6431333333333332e-05, + "loss": 0.0097, + "step": 20712 + }, + { + "epoch": 17.49408783783784, + "grad_norm": 0.17782212793827057, + "learning_rate": 2.6431e-05, + "loss": 0.0082, + "step": 20713 + }, + { + "epoch": 17.49493243243243, + "grad_norm": 1.523199439048767, + "learning_rate": 2.6430666666666667e-05, + "loss": 0.0362, + "step": 20714 + }, + { + "epoch": 17.495777027027028, + "grad_norm": 0.19988350570201874, + "learning_rate": 2.6430333333333333e-05, + "loss": 0.0149, + "step": 20715 + }, + { + "epoch": 17.49662162162162, + "grad_norm": 0.6350082159042358, + "learning_rate": 2.643e-05, + "loss": 0.0231, + "step": 20716 + }, + { + "epoch": 17.497466216216218, + "grad_norm": 2.366553783416748, + "learning_rate": 2.642966666666667e-05, + "loss": 0.0212, + "step": 20717 + }, + { + "epoch": 17.49831081081081, + "grad_norm": 0.17301975190639496, + "learning_rate": 2.6429333333333335e-05, + "loss": 0.0115, + "step": 20718 + }, + { + "epoch": 17.499155405405407, + "grad_norm": 0.21815255284309387, + "learning_rate": 2.6429e-05, + "loss": 0.0119, + "step": 20719 + }, + { + "epoch": 17.5, + "grad_norm": 0.21818026900291443, + "learning_rate": 2.642866666666667e-05, + "loss": 0.0103, + "step": 20720 + }, + { + "epoch": 17.500844594594593, + "grad_norm": 0.4053456783294678, + "learning_rate": 2.6428333333333332e-05, + "loss": 0.009, + "step": 20721 + }, + { + "epoch": 17.50168918918919, + "grad_norm": 0.17908960580825806, + "learning_rate": 2.6428000000000002e-05, + "loss": 0.0071, + "step": 20722 + }, + { + "epoch": 17.502533783783782, + "grad_norm": 0.44426003098487854, + "learning_rate": 2.6427666666666668e-05, + "loss": 0.0118, + "step": 20723 + }, + { + "epoch": 17.50337837837838, + "grad_norm": 0.3092028498649597, + "learning_rate": 2.6427333333333334e-05, + "loss": 0.0116, + "step": 20724 + }, + { + "epoch": 17.504222972972972, + "grad_norm": 0.21633481979370117, + "learning_rate": 2.6427e-05, + "loss": 0.0095, + "step": 20725 + }, + { + "epoch": 17.50506756756757, + "grad_norm": 0.2885088324546814, + "learning_rate": 2.642666666666667e-05, + "loss": 0.0098, + "step": 20726 + }, + { + "epoch": 17.50591216216216, + "grad_norm": 0.19575940072536469, + "learning_rate": 2.6426333333333335e-05, + "loss": 0.008, + "step": 20727 + }, + { + "epoch": 17.506756756756758, + "grad_norm": 0.3908848762512207, + "learning_rate": 2.6426e-05, + "loss": 0.0244, + "step": 20728 + }, + { + "epoch": 17.50760135135135, + "grad_norm": 0.862577497959137, + "learning_rate": 2.6425666666666667e-05, + "loss": 0.1868, + "step": 20729 + }, + { + "epoch": 17.508445945945947, + "grad_norm": 0.3443557918071747, + "learning_rate": 2.6425333333333333e-05, + "loss": 0.0632, + "step": 20730 + }, + { + "epoch": 17.50929054054054, + "grad_norm": 0.2471914440393448, + "learning_rate": 2.6425000000000002e-05, + "loss": 0.0289, + "step": 20731 + }, + { + "epoch": 17.510135135135137, + "grad_norm": 0.4581805169582367, + "learning_rate": 2.6424666666666665e-05, + "loss": 0.019, + "step": 20732 + }, + { + "epoch": 17.51097972972973, + "grad_norm": 0.19791658222675323, + "learning_rate": 2.6424333333333334e-05, + "loss": 0.0125, + "step": 20733 + }, + { + "epoch": 17.511824324324323, + "grad_norm": 0.2548249661922455, + "learning_rate": 2.6424000000000003e-05, + "loss": 0.0248, + "step": 20734 + }, + { + "epoch": 17.51266891891892, + "grad_norm": 0.24959029257297516, + "learning_rate": 2.6423666666666666e-05, + "loss": 0.011, + "step": 20735 + }, + { + "epoch": 17.513513513513512, + "grad_norm": 0.5749578475952148, + "learning_rate": 2.6423333333333335e-05, + "loss": 0.0324, + "step": 20736 + }, + { + "epoch": 17.51435810810811, + "grad_norm": 0.2619737982749939, + "learning_rate": 2.6423e-05, + "loss": 0.0065, + "step": 20737 + }, + { + "epoch": 17.5152027027027, + "grad_norm": 0.23592950403690338, + "learning_rate": 2.6422666666666667e-05, + "loss": 0.0135, + "step": 20738 + }, + { + "epoch": 17.5160472972973, + "grad_norm": 0.3409147560596466, + "learning_rate": 2.6422333333333333e-05, + "loss": 0.0269, + "step": 20739 + }, + { + "epoch": 17.51689189189189, + "grad_norm": 0.23252339661121368, + "learning_rate": 2.6422000000000002e-05, + "loss": 0.0147, + "step": 20740 + }, + { + "epoch": 17.517736486486488, + "grad_norm": 0.21938350796699524, + "learning_rate": 2.6421666666666665e-05, + "loss": 0.0096, + "step": 20741 + }, + { + "epoch": 17.51858108108108, + "grad_norm": 0.27923208475112915, + "learning_rate": 2.6421333333333334e-05, + "loss": 0.013, + "step": 20742 + }, + { + "epoch": 17.519425675675677, + "grad_norm": 0.17907334864139557, + "learning_rate": 2.6421000000000003e-05, + "loss": 0.0106, + "step": 20743 + }, + { + "epoch": 17.52027027027027, + "grad_norm": 0.16884510219097137, + "learning_rate": 2.6420666666666666e-05, + "loss": 0.0144, + "step": 20744 + }, + { + "epoch": 17.521114864864863, + "grad_norm": 0.16306504607200623, + "learning_rate": 2.6420333333333335e-05, + "loss": 0.0095, + "step": 20745 + }, + { + "epoch": 17.52195945945946, + "grad_norm": 0.4227568507194519, + "learning_rate": 2.642e-05, + "loss": 0.0214, + "step": 20746 + }, + { + "epoch": 17.522804054054053, + "grad_norm": 0.24797409772872925, + "learning_rate": 2.6419666666666667e-05, + "loss": 0.0098, + "step": 20747 + }, + { + "epoch": 17.52364864864865, + "grad_norm": 0.2447805553674698, + "learning_rate": 2.6419333333333333e-05, + "loss": 0.0127, + "step": 20748 + }, + { + "epoch": 17.524493243243242, + "grad_norm": 0.21598666906356812, + "learning_rate": 2.6419000000000002e-05, + "loss": 0.0129, + "step": 20749 + }, + { + "epoch": 17.52533783783784, + "grad_norm": 0.21201534569263458, + "learning_rate": 2.6418666666666665e-05, + "loss": 0.0083, + "step": 20750 + }, + { + "epoch": 17.52618243243243, + "grad_norm": 0.23117311298847198, + "learning_rate": 2.6418333333333334e-05, + "loss": 0.0077, + "step": 20751 + }, + { + "epoch": 17.527027027027028, + "grad_norm": 0.21022477746009827, + "learning_rate": 2.6418000000000004e-05, + "loss": 0.0105, + "step": 20752 + }, + { + "epoch": 17.52787162162162, + "grad_norm": 0.1640971451997757, + "learning_rate": 2.6417666666666666e-05, + "loss": 0.0057, + "step": 20753 + }, + { + "epoch": 17.528716216216218, + "grad_norm": 0.6233765482902527, + "learning_rate": 2.6417333333333335e-05, + "loss": 0.009, + "step": 20754 + }, + { + "epoch": 17.52956081081081, + "grad_norm": 0.7212936282157898, + "learning_rate": 2.6417e-05, + "loss": 0.0198, + "step": 20755 + }, + { + "epoch": 17.530405405405407, + "grad_norm": 0.18771342933177948, + "learning_rate": 2.6416666666666667e-05, + "loss": 0.0088, + "step": 20756 + }, + { + "epoch": 17.53125, + "grad_norm": 0.18339656293392181, + "learning_rate": 2.6416333333333333e-05, + "loss": 0.0092, + "step": 20757 + }, + { + "epoch": 17.532094594594593, + "grad_norm": 0.2192794531583786, + "learning_rate": 2.6416e-05, + "loss": 0.011, + "step": 20758 + }, + { + "epoch": 17.53293918918919, + "grad_norm": 0.5544326901435852, + "learning_rate": 2.641566666666667e-05, + "loss": 0.0083, + "step": 20759 + }, + { + "epoch": 17.533783783783782, + "grad_norm": 0.15295548737049103, + "learning_rate": 2.6415333333333334e-05, + "loss": 0.008, + "step": 20760 + }, + { + "epoch": 17.53462837837838, + "grad_norm": 0.12260312587022781, + "learning_rate": 2.6415e-05, + "loss": 0.0073, + "step": 20761 + }, + { + "epoch": 17.535472972972972, + "grad_norm": 0.10024204850196838, + "learning_rate": 2.6414666666666666e-05, + "loss": 0.0094, + "step": 20762 + }, + { + "epoch": 17.53631756756757, + "grad_norm": 0.20104697346687317, + "learning_rate": 2.6414333333333336e-05, + "loss": 0.0049, + "step": 20763 + }, + { + "epoch": 17.53716216216216, + "grad_norm": 0.1397590935230255, + "learning_rate": 2.6413999999999998e-05, + "loss": 0.0088, + "step": 20764 + }, + { + "epoch": 17.538006756756758, + "grad_norm": 0.24135734140872955, + "learning_rate": 2.6413666666666667e-05, + "loss": 0.0088, + "step": 20765 + }, + { + "epoch": 17.53885135135135, + "grad_norm": 0.26334288716316223, + "learning_rate": 2.6413333333333333e-05, + "loss": 0.0153, + "step": 20766 + }, + { + "epoch": 17.539695945945947, + "grad_norm": 0.834331214427948, + "learning_rate": 2.6413e-05, + "loss": 0.0156, + "step": 20767 + }, + { + "epoch": 17.54054054054054, + "grad_norm": 0.8391607403755188, + "learning_rate": 2.641266666666667e-05, + "loss": 0.0163, + "step": 20768 + }, + { + "epoch": 17.541385135135137, + "grad_norm": 0.16577647626399994, + "learning_rate": 2.6412333333333335e-05, + "loss": 0.0122, + "step": 20769 + }, + { + "epoch": 17.54222972972973, + "grad_norm": 0.30826130509376526, + "learning_rate": 2.6412e-05, + "loss": 0.0168, + "step": 20770 + }, + { + "epoch": 17.543074324324323, + "grad_norm": 0.29160767793655396, + "learning_rate": 2.6411666666666666e-05, + "loss": 0.0098, + "step": 20771 + }, + { + "epoch": 17.54391891891892, + "grad_norm": 0.19680596888065338, + "learning_rate": 2.6411333333333336e-05, + "loss": 0.0153, + "step": 20772 + }, + { + "epoch": 17.544763513513512, + "grad_norm": 0.4271756112575531, + "learning_rate": 2.6411e-05, + "loss": 0.0164, + "step": 20773 + }, + { + "epoch": 17.54560810810811, + "grad_norm": 0.45256364345550537, + "learning_rate": 2.6410666666666668e-05, + "loss": 0.0118, + "step": 20774 + }, + { + "epoch": 17.5464527027027, + "grad_norm": 0.7663348913192749, + "learning_rate": 2.6410333333333334e-05, + "loss": 0.0227, + "step": 20775 + }, + { + "epoch": 17.5472972972973, + "grad_norm": 0.5973339676856995, + "learning_rate": 2.641e-05, + "loss": 0.021, + "step": 20776 + }, + { + "epoch": 17.54814189189189, + "grad_norm": NaN, + "learning_rate": 2.641e-05, + "loss": 0.0123, + "step": 20777 + }, + { + "epoch": 17.548986486486488, + "grad_norm": 1.1691111326217651, + "learning_rate": 2.640966666666667e-05, + "loss": 0.0356, + "step": 20778 + }, + { + "epoch": 17.54983108108108, + "grad_norm": 0.6847699880599976, + "learning_rate": 2.6409333333333335e-05, + "loss": 0.1713, + "step": 20779 + }, + { + "epoch": 17.550675675675677, + "grad_norm": 0.5635061264038086, + "learning_rate": 2.6409e-05, + "loss": 0.0979, + "step": 20780 + }, + { + "epoch": 17.55152027027027, + "grad_norm": 0.4156847894191742, + "learning_rate": 2.6408666666666667e-05, + "loss": 0.0436, + "step": 20781 + }, + { + "epoch": 17.552364864864863, + "grad_norm": 0.18556198477745056, + "learning_rate": 2.6408333333333336e-05, + "loss": 0.0188, + "step": 20782 + }, + { + "epoch": 17.55320945945946, + "grad_norm": 0.6231811046600342, + "learning_rate": 2.6408e-05, + "loss": 0.0315, + "step": 20783 + }, + { + "epoch": 17.554054054054053, + "grad_norm": 0.28696227073669434, + "learning_rate": 2.6407666666666668e-05, + "loss": 0.0208, + "step": 20784 + }, + { + "epoch": 17.55489864864865, + "grad_norm": 0.21915635466575623, + "learning_rate": 2.6407333333333337e-05, + "loss": 0.0169, + "step": 20785 + }, + { + "epoch": 17.555743243243242, + "grad_norm": 0.22715015709400177, + "learning_rate": 2.6407e-05, + "loss": 0.019, + "step": 20786 + }, + { + "epoch": 17.55658783783784, + "grad_norm": 0.5981148481369019, + "learning_rate": 2.640666666666667e-05, + "loss": 0.0211, + "step": 20787 + }, + { + "epoch": 17.55743243243243, + "grad_norm": 0.21869048476219177, + "learning_rate": 2.6406333333333335e-05, + "loss": 0.0218, + "step": 20788 + }, + { + "epoch": 17.558277027027028, + "grad_norm": 0.1755819022655487, + "learning_rate": 2.6406e-05, + "loss": 0.0133, + "step": 20789 + }, + { + "epoch": 17.55912162162162, + "grad_norm": 0.22500477731227875, + "learning_rate": 2.6405666666666667e-05, + "loss": 0.0139, + "step": 20790 + }, + { + "epoch": 17.559966216216218, + "grad_norm": 0.2670654058456421, + "learning_rate": 2.6405333333333333e-05, + "loss": 0.0153, + "step": 20791 + }, + { + "epoch": 17.56081081081081, + "grad_norm": 0.4157698452472687, + "learning_rate": 2.6405e-05, + "loss": 0.0186, + "step": 20792 + }, + { + "epoch": 17.561655405405407, + "grad_norm": 0.2710672914981842, + "learning_rate": 2.6404666666666668e-05, + "loss": 0.0135, + "step": 20793 + }, + { + "epoch": 17.5625, + "grad_norm": 0.24603556096553802, + "learning_rate": 2.6404333333333334e-05, + "loss": 0.0085, + "step": 20794 + }, + { + "epoch": 17.563344594594593, + "grad_norm": 0.22678954899311066, + "learning_rate": 2.6404e-05, + "loss": 0.0121, + "step": 20795 + }, + { + "epoch": 17.56418918918919, + "grad_norm": 0.13531376421451569, + "learning_rate": 2.640366666666667e-05, + "loss": 0.0069, + "step": 20796 + }, + { + "epoch": 17.565033783783782, + "grad_norm": 0.2313874363899231, + "learning_rate": 2.6403333333333332e-05, + "loss": 0.0128, + "step": 20797 + }, + { + "epoch": 17.56587837837838, + "grad_norm": 0.2511146664619446, + "learning_rate": 2.6403e-05, + "loss": 0.0142, + "step": 20798 + }, + { + "epoch": 17.566722972972972, + "grad_norm": 0.132111594080925, + "learning_rate": 2.6402666666666667e-05, + "loss": 0.0069, + "step": 20799 + }, + { + "epoch": 17.56756756756757, + "grad_norm": 0.2572878897190094, + "learning_rate": 2.6402333333333333e-05, + "loss": 0.0106, + "step": 20800 + }, + { + "epoch": 17.56841216216216, + "grad_norm": 0.1711820811033249, + "learning_rate": 2.6402e-05, + "loss": 0.0096, + "step": 20801 + }, + { + "epoch": 17.569256756756758, + "grad_norm": 0.2212677150964737, + "learning_rate": 2.6401666666666668e-05, + "loss": 0.0085, + "step": 20802 + }, + { + "epoch": 17.57010135135135, + "grad_norm": 0.3692474067211151, + "learning_rate": 2.6401333333333334e-05, + "loss": 0.0205, + "step": 20803 + }, + { + "epoch": 17.570945945945947, + "grad_norm": 0.2941652834415436, + "learning_rate": 2.6401e-05, + "loss": 0.0157, + "step": 20804 + }, + { + "epoch": 17.57179054054054, + "grad_norm": 0.23016346991062164, + "learning_rate": 2.640066666666667e-05, + "loss": 0.012, + "step": 20805 + }, + { + "epoch": 17.572635135135137, + "grad_norm": 0.22416357696056366, + "learning_rate": 2.6400333333333332e-05, + "loss": 0.0086, + "step": 20806 + }, + { + "epoch": 17.57347972972973, + "grad_norm": 0.18813005089759827, + "learning_rate": 2.64e-05, + "loss": 0.0113, + "step": 20807 + }, + { + "epoch": 17.574324324324323, + "grad_norm": 0.20724265277385712, + "learning_rate": 2.6399666666666667e-05, + "loss": 0.0111, + "step": 20808 + }, + { + "epoch": 17.57516891891892, + "grad_norm": 0.3196503520011902, + "learning_rate": 2.6399333333333333e-05, + "loss": 0.0109, + "step": 20809 + }, + { + "epoch": 17.576013513513512, + "grad_norm": 0.3423687219619751, + "learning_rate": 2.6399000000000003e-05, + "loss": 0.0087, + "step": 20810 + }, + { + "epoch": 17.57685810810811, + "grad_norm": 0.8766198754310608, + "learning_rate": 2.639866666666667e-05, + "loss": 0.0153, + "step": 20811 + }, + { + "epoch": 17.5777027027027, + "grad_norm": 0.45733293890953064, + "learning_rate": 2.6398333333333334e-05, + "loss": 0.0603, + "step": 20812 + }, + { + "epoch": 17.5785472972973, + "grad_norm": 0.15109418332576752, + "learning_rate": 2.6398e-05, + "loss": 0.0043, + "step": 20813 + }, + { + "epoch": 17.57939189189189, + "grad_norm": 0.30062350630760193, + "learning_rate": 2.639766666666667e-05, + "loss": 0.0125, + "step": 20814 + }, + { + "epoch": 17.580236486486488, + "grad_norm": 0.23005589842796326, + "learning_rate": 2.6397333333333332e-05, + "loss": 0.0073, + "step": 20815 + }, + { + "epoch": 17.58108108108108, + "grad_norm": 0.23879346251487732, + "learning_rate": 2.6397e-05, + "loss": 0.0087, + "step": 20816 + }, + { + "epoch": 17.581925675675677, + "grad_norm": 0.2752740681171417, + "learning_rate": 2.6396666666666667e-05, + "loss": 0.0205, + "step": 20817 + }, + { + "epoch": 17.58277027027027, + "grad_norm": 0.5164676904678345, + "learning_rate": 2.6396333333333333e-05, + "loss": 0.0115, + "step": 20818 + }, + { + "epoch": 17.583614864864863, + "grad_norm": 0.26094913482666016, + "learning_rate": 2.6396000000000003e-05, + "loss": 0.0141, + "step": 20819 + }, + { + "epoch": 17.58445945945946, + "grad_norm": 0.19640225172042847, + "learning_rate": 2.6395666666666665e-05, + "loss": 0.0116, + "step": 20820 + }, + { + "epoch": 17.585304054054053, + "grad_norm": 0.3209109306335449, + "learning_rate": 2.6395333333333335e-05, + "loss": 0.0137, + "step": 20821 + }, + { + "epoch": 17.58614864864865, + "grad_norm": 1.9793503284454346, + "learning_rate": 2.6395e-05, + "loss": 0.0165, + "step": 20822 + }, + { + "epoch": 17.586993243243242, + "grad_norm": 0.44660088419914246, + "learning_rate": 2.6394666666666666e-05, + "loss": 0.0221, + "step": 20823 + }, + { + "epoch": 17.58783783783784, + "grad_norm": 0.5379143357276917, + "learning_rate": 2.6394333333333332e-05, + "loss": 0.0179, + "step": 20824 + }, + { + "epoch": 17.58868243243243, + "grad_norm": 0.236058309674263, + "learning_rate": 2.6394000000000002e-05, + "loss": 0.0134, + "step": 20825 + }, + { + "epoch": 17.589527027027028, + "grad_norm": 0.8796201348304749, + "learning_rate": 2.6393666666666664e-05, + "loss": 0.0208, + "step": 20826 + }, + { + "epoch": 17.59037162162162, + "grad_norm": 0.45541051030158997, + "learning_rate": 2.6393333333333334e-05, + "loss": 0.0116, + "step": 20827 + }, + { + "epoch": 17.591216216216218, + "grad_norm": 0.3161376118659973, + "learning_rate": 2.6393000000000003e-05, + "loss": 0.0243, + "step": 20828 + }, + { + "epoch": 17.59206081081081, + "grad_norm": 1.4070327281951904, + "learning_rate": 2.6392666666666665e-05, + "loss": 0.1929, + "step": 20829 + }, + { + "epoch": 17.592905405405407, + "grad_norm": 0.5413599014282227, + "learning_rate": 2.6392333333333335e-05, + "loss": 0.0949, + "step": 20830 + }, + { + "epoch": 17.59375, + "grad_norm": 0.30426886677742004, + "learning_rate": 2.6392e-05, + "loss": 0.0403, + "step": 20831 + }, + { + "epoch": 17.594594594594593, + "grad_norm": 0.2879432439804077, + "learning_rate": 2.6391666666666667e-05, + "loss": 0.0261, + "step": 20832 + }, + { + "epoch": 17.59543918918919, + "grad_norm": 0.2659459412097931, + "learning_rate": 2.6391333333333333e-05, + "loss": 0.0185, + "step": 20833 + }, + { + "epoch": 17.596283783783782, + "grad_norm": 0.657185435295105, + "learning_rate": 2.6391000000000002e-05, + "loss": 0.0229, + "step": 20834 + }, + { + "epoch": 17.59712837837838, + "grad_norm": 0.40828558802604675, + "learning_rate": 2.6390666666666668e-05, + "loss": 0.0311, + "step": 20835 + }, + { + "epoch": 17.597972972972972, + "grad_norm": 0.29948100447654724, + "learning_rate": 2.6390333333333334e-05, + "loss": 0.0191, + "step": 20836 + }, + { + "epoch": 17.59881756756757, + "grad_norm": 0.23363107442855835, + "learning_rate": 2.6390000000000003e-05, + "loss": 0.0147, + "step": 20837 + }, + { + "epoch": 17.59966216216216, + "grad_norm": 0.28153181076049805, + "learning_rate": 2.6389666666666666e-05, + "loss": 0.0192, + "step": 20838 + }, + { + "epoch": 17.600506756756758, + "grad_norm": 0.3299662172794342, + "learning_rate": 2.6389333333333335e-05, + "loss": 0.0284, + "step": 20839 + }, + { + "epoch": 17.60135135135135, + "grad_norm": 0.2445039004087448, + "learning_rate": 2.6389e-05, + "loss": 0.014, + "step": 20840 + }, + { + "epoch": 17.602195945945947, + "grad_norm": 0.24896672368049622, + "learning_rate": 2.6388666666666667e-05, + "loss": 0.0167, + "step": 20841 + }, + { + "epoch": 17.60304054054054, + "grad_norm": 0.1318225860595703, + "learning_rate": 2.6388333333333333e-05, + "loss": 0.0087, + "step": 20842 + }, + { + "epoch": 17.603885135135137, + "grad_norm": 0.35848468542099, + "learning_rate": 2.6388000000000002e-05, + "loss": 0.0158, + "step": 20843 + }, + { + "epoch": 17.60472972972973, + "grad_norm": 0.194313645362854, + "learning_rate": 2.6387666666666668e-05, + "loss": 0.0106, + "step": 20844 + }, + { + "epoch": 17.605574324324323, + "grad_norm": 0.2713698744773865, + "learning_rate": 2.6387333333333334e-05, + "loss": 0.0161, + "step": 20845 + }, + { + "epoch": 17.60641891891892, + "grad_norm": 0.24133527278900146, + "learning_rate": 2.6387000000000003e-05, + "loss": 0.0101, + "step": 20846 + }, + { + "epoch": 17.607263513513512, + "grad_norm": 0.26018601655960083, + "learning_rate": 2.6386666666666666e-05, + "loss": 0.0175, + "step": 20847 + }, + { + "epoch": 17.60810810810811, + "grad_norm": 0.34841838479042053, + "learning_rate": 2.6386333333333335e-05, + "loss": 0.012, + "step": 20848 + }, + { + "epoch": 17.6089527027027, + "grad_norm": 0.4992193877696991, + "learning_rate": 2.6385999999999998e-05, + "loss": 0.0233, + "step": 20849 + }, + { + "epoch": 17.6097972972973, + "grad_norm": 0.18042796850204468, + "learning_rate": 2.6385666666666667e-05, + "loss": 0.0075, + "step": 20850 + }, + { + "epoch": 17.61064189189189, + "grad_norm": 0.3787729740142822, + "learning_rate": 2.6385333333333333e-05, + "loss": 0.0208, + "step": 20851 + }, + { + "epoch": 17.611486486486488, + "grad_norm": 0.17324109375476837, + "learning_rate": 2.6385e-05, + "loss": 0.0106, + "step": 20852 + }, + { + "epoch": 17.61233108108108, + "grad_norm": 0.2548243999481201, + "learning_rate": 2.6384666666666668e-05, + "loss": 0.0143, + "step": 20853 + }, + { + "epoch": 17.613175675675677, + "grad_norm": 0.6117156147956848, + "learning_rate": 2.6384333333333334e-05, + "loss": 0.0255, + "step": 20854 + }, + { + "epoch": 17.61402027027027, + "grad_norm": 0.18313778936862946, + "learning_rate": 2.6384e-05, + "loss": 0.0106, + "step": 20855 + }, + { + "epoch": 17.614864864864863, + "grad_norm": 0.13265404105186462, + "learning_rate": 2.6383666666666666e-05, + "loss": 0.0099, + "step": 20856 + }, + { + "epoch": 17.61570945945946, + "grad_norm": 0.3182343542575836, + "learning_rate": 2.6383333333333335e-05, + "loss": 0.0137, + "step": 20857 + }, + { + "epoch": 17.616554054054053, + "grad_norm": 0.1302572637796402, + "learning_rate": 2.6382999999999998e-05, + "loss": 0.0101, + "step": 20858 + }, + { + "epoch": 17.61739864864865, + "grad_norm": 0.1588195264339447, + "learning_rate": 2.6382666666666667e-05, + "loss": 0.0082, + "step": 20859 + }, + { + "epoch": 17.618243243243242, + "grad_norm": 0.3034918010234833, + "learning_rate": 2.6382333333333337e-05, + "loss": 0.0085, + "step": 20860 + }, + { + "epoch": 17.61908783783784, + "grad_norm": 1.140015721321106, + "learning_rate": 2.6382e-05, + "loss": 0.0152, + "step": 20861 + }, + { + "epoch": 17.61993243243243, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.638166666666667e-05, + "loss": 0.0079, + "step": 20862 + }, + { + "epoch": 17.620777027027028, + "grad_norm": 0.43122079968452454, + "learning_rate": 2.6381333333333334e-05, + "loss": 0.0222, + "step": 20863 + }, + { + "epoch": 17.62162162162162, + "grad_norm": 0.33543068170547485, + "learning_rate": 2.6381e-05, + "loss": 0.0089, + "step": 20864 + }, + { + "epoch": 17.622466216216218, + "grad_norm": 0.36672091484069824, + "learning_rate": 2.6380666666666666e-05, + "loss": 0.0126, + "step": 20865 + }, + { + "epoch": 17.62331081081081, + "grad_norm": 0.40313756465911865, + "learning_rate": 2.6380333333333336e-05, + "loss": 0.0246, + "step": 20866 + }, + { + "epoch": 17.624155405405407, + "grad_norm": 0.33679142594337463, + "learning_rate": 2.6379999999999998e-05, + "loss": 0.0107, + "step": 20867 + }, + { + "epoch": 17.625, + "grad_norm": 0.3919258117675781, + "learning_rate": 2.6379666666666667e-05, + "loss": 0.0151, + "step": 20868 + }, + { + "epoch": 17.625844594594593, + "grad_norm": 0.5508133769035339, + "learning_rate": 2.6379333333333337e-05, + "loss": 0.011, + "step": 20869 + }, + { + "epoch": 17.62668918918919, + "grad_norm": 0.2531143128871918, + "learning_rate": 2.6379e-05, + "loss": 0.0135, + "step": 20870 + }, + { + "epoch": 17.627533783783782, + "grad_norm": 0.32026392221450806, + "learning_rate": 2.637866666666667e-05, + "loss": 0.0187, + "step": 20871 + }, + { + "epoch": 17.62837837837838, + "grad_norm": 0.38399097323417664, + "learning_rate": 2.6378333333333335e-05, + "loss": 0.0262, + "step": 20872 + }, + { + "epoch": 17.629222972972972, + "grad_norm": 0.31859585642814636, + "learning_rate": 2.6378e-05, + "loss": 0.0112, + "step": 20873 + }, + { + "epoch": 17.63006756756757, + "grad_norm": 0.20379793643951416, + "learning_rate": 2.6377666666666666e-05, + "loss": 0.011, + "step": 20874 + }, + { + "epoch": 17.63091216216216, + "grad_norm": 0.11857039481401443, + "learning_rate": 2.6377333333333336e-05, + "loss": 0.0051, + "step": 20875 + }, + { + "epoch": 17.631756756756758, + "grad_norm": 0.5244336724281311, + "learning_rate": 2.6377e-05, + "loss": 0.0211, + "step": 20876 + }, + { + "epoch": 17.63260135135135, + "grad_norm": 0.3444940447807312, + "learning_rate": 2.6376666666666668e-05, + "loss": 0.0183, + "step": 20877 + }, + { + "epoch": 17.633445945945947, + "grad_norm": 0.6328725218772888, + "learning_rate": 2.6376333333333337e-05, + "loss": 0.0159, + "step": 20878 + }, + { + "epoch": 17.63429054054054, + "grad_norm": 0.6296855807304382, + "learning_rate": 2.6376e-05, + "loss": 0.1475, + "step": 20879 + }, + { + "epoch": 17.635135135135137, + "grad_norm": 0.5874374508857727, + "learning_rate": 2.637566666666667e-05, + "loss": 0.0843, + "step": 20880 + }, + { + "epoch": 17.63597972972973, + "grad_norm": 0.4349276125431061, + "learning_rate": 2.637533333333333e-05, + "loss": 0.0358, + "step": 20881 + }, + { + "epoch": 17.636824324324323, + "grad_norm": 0.40585577487945557, + "learning_rate": 2.6375e-05, + "loss": 0.0327, + "step": 20882 + }, + { + "epoch": 17.63766891891892, + "grad_norm": 0.23415222764015198, + "learning_rate": 2.6374666666666667e-05, + "loss": 0.0119, + "step": 20883 + }, + { + "epoch": 17.638513513513512, + "grad_norm": 0.28614699840545654, + "learning_rate": 2.6374333333333333e-05, + "loss": 0.0193, + "step": 20884 + }, + { + "epoch": 17.63935810810811, + "grad_norm": 0.7101491093635559, + "learning_rate": 2.6374000000000002e-05, + "loss": 0.0242, + "step": 20885 + }, + { + "epoch": 17.6402027027027, + "grad_norm": 0.22274573147296906, + "learning_rate": 2.6373666666666668e-05, + "loss": 0.0169, + "step": 20886 + }, + { + "epoch": 17.6410472972973, + "grad_norm": 0.5324739217758179, + "learning_rate": 2.6373333333333334e-05, + "loss": 0.0168, + "step": 20887 + }, + { + "epoch": 17.64189189189189, + "grad_norm": 0.5348934531211853, + "learning_rate": 2.6373e-05, + "loss": 0.0146, + "step": 20888 + }, + { + "epoch": 17.642736486486488, + "grad_norm": 0.26771798729896545, + "learning_rate": 2.637266666666667e-05, + "loss": 0.0135, + "step": 20889 + }, + { + "epoch": 17.64358108108108, + "grad_norm": 0.25944992899894714, + "learning_rate": 2.637233333333333e-05, + "loss": 0.0129, + "step": 20890 + }, + { + "epoch": 17.644425675675677, + "grad_norm": 0.25216561555862427, + "learning_rate": 2.6372e-05, + "loss": 0.0081, + "step": 20891 + }, + { + "epoch": 17.64527027027027, + "grad_norm": 0.3250722885131836, + "learning_rate": 2.6371666666666667e-05, + "loss": 0.0255, + "step": 20892 + }, + { + "epoch": 17.646114864864863, + "grad_norm": 0.15841186046600342, + "learning_rate": 2.6371333333333333e-05, + "loss": 0.0097, + "step": 20893 + }, + { + "epoch": 17.64695945945946, + "grad_norm": 0.6339001059532166, + "learning_rate": 2.6371000000000002e-05, + "loss": 0.0176, + "step": 20894 + }, + { + "epoch": 17.647804054054053, + "grad_norm": 0.22647935152053833, + "learning_rate": 2.6370666666666668e-05, + "loss": 0.0136, + "step": 20895 + }, + { + "epoch": 17.64864864864865, + "grad_norm": 0.24350978434085846, + "learning_rate": 2.6370333333333334e-05, + "loss": 0.0151, + "step": 20896 + }, + { + "epoch": 17.649493243243242, + "grad_norm": 0.15609924495220184, + "learning_rate": 2.637e-05, + "loss": 0.0101, + "step": 20897 + }, + { + "epoch": 17.65033783783784, + "grad_norm": 0.17289835214614868, + "learning_rate": 2.636966666666667e-05, + "loss": 0.0089, + "step": 20898 + }, + { + "epoch": 17.65118243243243, + "grad_norm": 0.16700829565525055, + "learning_rate": 2.6369333333333332e-05, + "loss": 0.0128, + "step": 20899 + }, + { + "epoch": 17.652027027027028, + "grad_norm": 0.3554334044456482, + "learning_rate": 2.6369e-05, + "loss": 0.0165, + "step": 20900 + }, + { + "epoch": 17.65287162162162, + "grad_norm": 0.21712392568588257, + "learning_rate": 2.6368666666666667e-05, + "loss": 0.0112, + "step": 20901 + }, + { + "epoch": 17.653716216216218, + "grad_norm": 0.30865150690078735, + "learning_rate": 2.6368333333333333e-05, + "loss": 0.0518, + "step": 20902 + }, + { + "epoch": 17.65456081081081, + "grad_norm": 0.2527311146259308, + "learning_rate": 2.6368000000000002e-05, + "loss": 0.0137, + "step": 20903 + }, + { + "epoch": 17.655405405405407, + "grad_norm": 0.21036787331104279, + "learning_rate": 2.6367666666666668e-05, + "loss": 0.01, + "step": 20904 + }, + { + "epoch": 17.65625, + "grad_norm": 0.4648321568965912, + "learning_rate": 2.6367333333333334e-05, + "loss": 0.0279, + "step": 20905 + }, + { + "epoch": 17.657094594594593, + "grad_norm": 0.20537298917770386, + "learning_rate": 2.6367e-05, + "loss": 0.0096, + "step": 20906 + }, + { + "epoch": 17.65793918918919, + "grad_norm": 0.26143935322761536, + "learning_rate": 2.636666666666667e-05, + "loss": 0.0116, + "step": 20907 + }, + { + "epoch": 17.658783783783782, + "grad_norm": 0.15570256114006042, + "learning_rate": 2.6366333333333332e-05, + "loss": 0.009, + "step": 20908 + }, + { + "epoch": 17.65962837837838, + "grad_norm": 0.14238035678863525, + "learning_rate": 2.6366e-05, + "loss": 0.0077, + "step": 20909 + }, + { + "epoch": 17.660472972972972, + "grad_norm": 0.16819718480110168, + "learning_rate": 2.6365666666666667e-05, + "loss": 0.0129, + "step": 20910 + }, + { + "epoch": 17.66131756756757, + "grad_norm": 0.2546752095222473, + "learning_rate": 2.6365333333333333e-05, + "loss": 0.0114, + "step": 20911 + }, + { + "epoch": 17.66216216216216, + "grad_norm": 0.1421748697757721, + "learning_rate": 2.6365000000000003e-05, + "loss": 0.012, + "step": 20912 + }, + { + "epoch": 17.663006756756758, + "grad_norm": 0.22306598722934723, + "learning_rate": 2.6364666666666665e-05, + "loss": 0.0092, + "step": 20913 + }, + { + "epoch": 17.66385135135135, + "grad_norm": 0.28009846806526184, + "learning_rate": 2.6364333333333334e-05, + "loss": 0.0098, + "step": 20914 + }, + { + "epoch": 17.664695945945947, + "grad_norm": 0.20216356217861176, + "learning_rate": 2.6364e-05, + "loss": 0.0054, + "step": 20915 + }, + { + "epoch": 17.66554054054054, + "grad_norm": 0.5850732922554016, + "learning_rate": 2.6363666666666666e-05, + "loss": 0.0221, + "step": 20916 + }, + { + "epoch": 17.666385135135137, + "grad_norm": 0.1498252898454666, + "learning_rate": 2.6363333333333332e-05, + "loss": 0.0087, + "step": 20917 + }, + { + "epoch": 17.66722972972973, + "grad_norm": 1.0491892099380493, + "learning_rate": 2.6363e-05, + "loss": 0.0117, + "step": 20918 + }, + { + "epoch": 17.668074324324323, + "grad_norm": 0.1687413454055786, + "learning_rate": 2.6362666666666667e-05, + "loss": 0.0105, + "step": 20919 + }, + { + "epoch": 17.66891891891892, + "grad_norm": 0.19696325063705444, + "learning_rate": 2.6362333333333333e-05, + "loss": 0.01, + "step": 20920 + }, + { + "epoch": 17.669763513513512, + "grad_norm": 0.2815086841583252, + "learning_rate": 2.6362000000000003e-05, + "loss": 0.0135, + "step": 20921 + }, + { + "epoch": 17.67060810810811, + "grad_norm": 0.19921058416366577, + "learning_rate": 2.6361666666666665e-05, + "loss": 0.0073, + "step": 20922 + }, + { + "epoch": 17.6714527027027, + "grad_norm": 0.1002109944820404, + "learning_rate": 2.6361333333333335e-05, + "loss": 0.0071, + "step": 20923 + }, + { + "epoch": 17.6722972972973, + "grad_norm": 0.41328170895576477, + "learning_rate": 2.6361e-05, + "loss": 0.0145, + "step": 20924 + }, + { + "epoch": 17.67314189189189, + "grad_norm": 0.21477660536766052, + "learning_rate": 2.6360666666666666e-05, + "loss": 0.0112, + "step": 20925 + }, + { + "epoch": 17.673986486486488, + "grad_norm": 1.0916844606399536, + "learning_rate": 2.6360333333333332e-05, + "loss": 0.0292, + "step": 20926 + }, + { + "epoch": 17.67483108108108, + "grad_norm": 0.8928534984588623, + "learning_rate": 2.6360000000000002e-05, + "loss": 0.0308, + "step": 20927 + }, + { + "epoch": 17.675675675675677, + "grad_norm": 0.6323763728141785, + "learning_rate": 2.6359666666666668e-05, + "loss": 0.0122, + "step": 20928 + }, + { + "epoch": 17.67652027027027, + "grad_norm": 0.6790310740470886, + "learning_rate": 2.6359333333333334e-05, + "loss": 0.1587, + "step": 20929 + }, + { + "epoch": 17.677364864864863, + "grad_norm": 0.32405906915664673, + "learning_rate": 2.6359000000000003e-05, + "loss": 0.0756, + "step": 20930 + }, + { + "epoch": 17.67820945945946, + "grad_norm": 0.4462611973285675, + "learning_rate": 2.6358666666666665e-05, + "loss": 0.0317, + "step": 20931 + }, + { + "epoch": 17.679054054054053, + "grad_norm": 0.27312466502189636, + "learning_rate": 2.6358333333333335e-05, + "loss": 0.0231, + "step": 20932 + }, + { + "epoch": 17.67989864864865, + "grad_norm": 0.32002371549606323, + "learning_rate": 2.6358e-05, + "loss": 0.0288, + "step": 20933 + }, + { + "epoch": 17.680743243243242, + "grad_norm": 0.17449548840522766, + "learning_rate": 2.6357666666666667e-05, + "loss": 0.0143, + "step": 20934 + }, + { + "epoch": 17.68158783783784, + "grad_norm": 0.3216214179992676, + "learning_rate": 2.6357333333333336e-05, + "loss": 0.0155, + "step": 20935 + }, + { + "epoch": 17.68243243243243, + "grad_norm": 0.2846514582633972, + "learning_rate": 2.6357000000000002e-05, + "loss": 0.015, + "step": 20936 + }, + { + "epoch": 17.683277027027028, + "grad_norm": 0.15905873477458954, + "learning_rate": 2.6356666666666668e-05, + "loss": 0.0119, + "step": 20937 + }, + { + "epoch": 17.68412162162162, + "grad_norm": 0.3699643313884735, + "learning_rate": 2.6356333333333334e-05, + "loss": 0.0156, + "step": 20938 + }, + { + "epoch": 17.684966216216218, + "grad_norm": 0.24816583096981049, + "learning_rate": 2.6356000000000003e-05, + "loss": 0.0164, + "step": 20939 + }, + { + "epoch": 17.68581081081081, + "grad_norm": 0.18796153366565704, + "learning_rate": 2.6355666666666666e-05, + "loss": 0.0111, + "step": 20940 + }, + { + "epoch": 17.686655405405407, + "grad_norm": 0.3563583791255951, + "learning_rate": 2.6355333333333335e-05, + "loss": 0.0129, + "step": 20941 + }, + { + "epoch": 17.6875, + "grad_norm": 0.25091326236724854, + "learning_rate": 2.6354999999999998e-05, + "loss": 0.0149, + "step": 20942 + }, + { + "epoch": 17.688344594594593, + "grad_norm": 0.22505199909210205, + "learning_rate": 2.6354666666666667e-05, + "loss": 0.0092, + "step": 20943 + }, + { + "epoch": 17.68918918918919, + "grad_norm": 0.37840354442596436, + "learning_rate": 2.6354333333333336e-05, + "loss": 0.0161, + "step": 20944 + }, + { + "epoch": 17.690033783783782, + "grad_norm": 0.21484175324440002, + "learning_rate": 2.6354e-05, + "loss": 0.0139, + "step": 20945 + }, + { + "epoch": 17.69087837837838, + "grad_norm": 0.16037195920944214, + "learning_rate": 2.6353666666666668e-05, + "loss": 0.01, + "step": 20946 + }, + { + "epoch": 17.691722972972972, + "grad_norm": 0.28431400656700134, + "learning_rate": 2.6353333333333334e-05, + "loss": 0.0115, + "step": 20947 + }, + { + "epoch": 17.69256756756757, + "grad_norm": 0.3532794415950775, + "learning_rate": 2.6353e-05, + "loss": 0.0136, + "step": 20948 + }, + { + "epoch": 17.69341216216216, + "grad_norm": 0.24329517781734467, + "learning_rate": 2.6352666666666666e-05, + "loss": 0.0165, + "step": 20949 + }, + { + "epoch": 17.694256756756758, + "grad_norm": 0.3704681694507599, + "learning_rate": 2.6352333333333335e-05, + "loss": 0.0115, + "step": 20950 + }, + { + "epoch": 17.69510135135135, + "grad_norm": 0.16448453068733215, + "learning_rate": 2.6351999999999998e-05, + "loss": 0.0107, + "step": 20951 + }, + { + "epoch": 17.695945945945947, + "grad_norm": 0.2558770775794983, + "learning_rate": 2.6351666666666667e-05, + "loss": 0.0119, + "step": 20952 + }, + { + "epoch": 17.69679054054054, + "grad_norm": 0.2552342116832733, + "learning_rate": 2.6351333333333336e-05, + "loss": 0.014, + "step": 20953 + }, + { + "epoch": 17.697635135135137, + "grad_norm": 0.19492553174495697, + "learning_rate": 2.6351e-05, + "loss": 0.0087, + "step": 20954 + }, + { + "epoch": 17.69847972972973, + "grad_norm": 0.5612860918045044, + "learning_rate": 2.6350666666666668e-05, + "loss": 0.0227, + "step": 20955 + }, + { + "epoch": 17.699324324324323, + "grad_norm": 0.1355871856212616, + "learning_rate": 2.6350333333333334e-05, + "loss": 0.0041, + "step": 20956 + }, + { + "epoch": 17.70016891891892, + "grad_norm": 0.1641593873500824, + "learning_rate": 2.635e-05, + "loss": 0.007, + "step": 20957 + }, + { + "epoch": 17.701013513513512, + "grad_norm": 0.1708894968032837, + "learning_rate": 2.6349666666666666e-05, + "loss": 0.0083, + "step": 20958 + }, + { + "epoch": 17.70185810810811, + "grad_norm": 0.35563698410987854, + "learning_rate": 2.6349333333333335e-05, + "loss": 0.0168, + "step": 20959 + }, + { + "epoch": 17.7027027027027, + "grad_norm": 0.3192192018032074, + "learning_rate": 2.6349e-05, + "loss": 0.0213, + "step": 20960 + }, + { + "epoch": 17.7035472972973, + "grad_norm": 0.14646045863628387, + "learning_rate": 2.6348666666666667e-05, + "loss": 0.0071, + "step": 20961 + }, + { + "epoch": 17.70439189189189, + "grad_norm": 0.22081106901168823, + "learning_rate": 2.6348333333333337e-05, + "loss": 0.015, + "step": 20962 + }, + { + "epoch": 17.705236486486488, + "grad_norm": 0.13321782648563385, + "learning_rate": 2.6348e-05, + "loss": 0.008, + "step": 20963 + }, + { + "epoch": 17.70608108108108, + "grad_norm": 0.37075743079185486, + "learning_rate": 2.634766666666667e-05, + "loss": 0.0191, + "step": 20964 + }, + { + "epoch": 17.706925675675677, + "grad_norm": 0.20611344277858734, + "learning_rate": 2.6347333333333334e-05, + "loss": 0.0056, + "step": 20965 + }, + { + "epoch": 17.70777027027027, + "grad_norm": 0.26614752411842346, + "learning_rate": 2.6347e-05, + "loss": 0.013, + "step": 20966 + }, + { + "epoch": 17.708614864864863, + "grad_norm": 0.26533761620521545, + "learning_rate": 2.6346666666666666e-05, + "loss": 0.0078, + "step": 20967 + }, + { + "epoch": 17.70945945945946, + "grad_norm": 0.30549031496047974, + "learning_rate": 2.6346333333333336e-05, + "loss": 0.01, + "step": 20968 + }, + { + "epoch": 17.710304054054053, + "grad_norm": 0.20035716891288757, + "learning_rate": 2.6346e-05, + "loss": 0.0083, + "step": 20969 + }, + { + "epoch": 17.71114864864865, + "grad_norm": 1.1439440250396729, + "learning_rate": 2.6345666666666667e-05, + "loss": 0.0193, + "step": 20970 + }, + { + "epoch": 17.711993243243242, + "grad_norm": 0.4439351260662079, + "learning_rate": 2.6345333333333333e-05, + "loss": 0.0119, + "step": 20971 + }, + { + "epoch": 17.71283783783784, + "grad_norm": 0.2353433519601822, + "learning_rate": 2.6345e-05, + "loss": 0.011, + "step": 20972 + }, + { + "epoch": 17.71368243243243, + "grad_norm": 0.25316399335861206, + "learning_rate": 2.634466666666667e-05, + "loss": 0.0082, + "step": 20973 + }, + { + "epoch": 17.714527027027028, + "grad_norm": 0.4664192795753479, + "learning_rate": 2.634433333333333e-05, + "loss": 0.0221, + "step": 20974 + }, + { + "epoch": 17.71537162162162, + "grad_norm": 0.25289633870124817, + "learning_rate": 2.6344e-05, + "loss": 0.0175, + "step": 20975 + }, + { + "epoch": 17.716216216216218, + "grad_norm": 0.17929992079734802, + "learning_rate": 2.6343666666666666e-05, + "loss": 0.0078, + "step": 20976 + }, + { + "epoch": 17.71706081081081, + "grad_norm": 0.24484238028526306, + "learning_rate": 2.6343333333333332e-05, + "loss": 0.0084, + "step": 20977 + }, + { + "epoch": 17.717905405405407, + "grad_norm": 0.6317779421806335, + "learning_rate": 2.6343000000000002e-05, + "loss": 0.0165, + "step": 20978 + }, + { + "epoch": 17.71875, + "grad_norm": 1.1716458797454834, + "learning_rate": 2.6342666666666668e-05, + "loss": 0.1484, + "step": 20979 + }, + { + "epoch": 17.719594594594593, + "grad_norm": 0.4629574418067932, + "learning_rate": 2.6342333333333334e-05, + "loss": 0.0783, + "step": 20980 + }, + { + "epoch": 17.72043918918919, + "grad_norm": 0.3760409355163574, + "learning_rate": 2.6342e-05, + "loss": 0.0402, + "step": 20981 + }, + { + "epoch": 17.721283783783782, + "grad_norm": 0.20498594641685486, + "learning_rate": 2.634166666666667e-05, + "loss": 0.019, + "step": 20982 + }, + { + "epoch": 17.72212837837838, + "grad_norm": 0.28492504358291626, + "learning_rate": 2.634133333333333e-05, + "loss": 0.0202, + "step": 20983 + }, + { + "epoch": 17.722972972972972, + "grad_norm": 0.2952600419521332, + "learning_rate": 2.6341e-05, + "loss": 0.0249, + "step": 20984 + }, + { + "epoch": 17.72381756756757, + "grad_norm": 0.18890249729156494, + "learning_rate": 2.634066666666667e-05, + "loss": 0.0153, + "step": 20985 + }, + { + "epoch": 17.72466216216216, + "grad_norm": 0.21034176647663116, + "learning_rate": 2.6340333333333333e-05, + "loss": 0.0186, + "step": 20986 + }, + { + "epoch": 17.725506756756758, + "grad_norm": 0.24816644191741943, + "learning_rate": 2.6340000000000002e-05, + "loss": 0.0156, + "step": 20987 + }, + { + "epoch": 17.72635135135135, + "grad_norm": 0.209193155169487, + "learning_rate": 2.6339666666666668e-05, + "loss": 0.0105, + "step": 20988 + }, + { + "epoch": 17.727195945945947, + "grad_norm": 0.21840432286262512, + "learning_rate": 2.6339333333333334e-05, + "loss": 0.0092, + "step": 20989 + }, + { + "epoch": 17.72804054054054, + "grad_norm": 0.3973141312599182, + "learning_rate": 2.6339e-05, + "loss": 0.0106, + "step": 20990 + }, + { + "epoch": 17.728885135135137, + "grad_norm": 0.2834128141403198, + "learning_rate": 2.633866666666667e-05, + "loss": 0.0298, + "step": 20991 + }, + { + "epoch": 17.72972972972973, + "grad_norm": 0.6159055233001709, + "learning_rate": 2.633833333333333e-05, + "loss": 0.0129, + "step": 20992 + }, + { + "epoch": 17.730574324324323, + "grad_norm": 0.18883153796195984, + "learning_rate": 2.6338e-05, + "loss": 0.0158, + "step": 20993 + }, + { + "epoch": 17.73141891891892, + "grad_norm": 0.21322472393512726, + "learning_rate": 2.633766666666667e-05, + "loss": 0.0086, + "step": 20994 + }, + { + "epoch": 17.732263513513512, + "grad_norm": 0.27108797430992126, + "learning_rate": 2.6337333333333333e-05, + "loss": 0.0119, + "step": 20995 + }, + { + "epoch": 17.73310810810811, + "grad_norm": 0.28088414669036865, + "learning_rate": 2.6337000000000002e-05, + "loss": 0.0117, + "step": 20996 + }, + { + "epoch": 17.7339527027027, + "grad_norm": 0.2459288090467453, + "learning_rate": 2.6336666666666668e-05, + "loss": 0.0152, + "step": 20997 + }, + { + "epoch": 17.7347972972973, + "grad_norm": 0.21286064386367798, + "learning_rate": 2.6336333333333334e-05, + "loss": 0.0147, + "step": 20998 + }, + { + "epoch": 17.73564189189189, + "grad_norm": 0.8490746021270752, + "learning_rate": 2.6336e-05, + "loss": 0.015, + "step": 20999 + }, + { + "epoch": 17.736486486486488, + "grad_norm": 0.2126060277223587, + "learning_rate": 2.6335666666666666e-05, + "loss": 0.0115, + "step": 21000 + }, + { + "epoch": 17.736486486486488, + "eval_cer": 0.09196048510159827, + "eval_loss": 0.2965540587902069, + "eval_runtime": 8.566, + "eval_samples_per_second": 11.791, + "eval_steps_per_second": 0.467, + "eval_wer": 0.2751322751322751, + "step": 21000 + }, + { + "epoch": 17.736486486486488, + "step": 21000, + "total_flos": 1.85831881720076e+20, + "train_loss": 0.19500272537244573, + "train_runtime": 26407.434, + "train_samples_per_second": 242.356, + "train_steps_per_second": 3.787 + } + ], + "logging_steps": 1.0, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 85, + "save_steps": 1000, + "stateful_callbacks": { + "EarlyStoppingCallback": { + "args": { + "early_stopping_patience": 5, + "early_stopping_threshold": 0.0 + }, + "attributes": { + "early_stopping_patience_counter": 5 + } + }, + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.85831881720076e+20, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}