{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.9999290604050652, "eval_steps": 500, "global_step": 10572, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 154.0, "learning_rate": 1.5723270440251573e-08, "loss": 1.7056, "step": 1 }, { "epoch": 0.0, "grad_norm": 112.5, "learning_rate": 3.1446540880503146e-08, "loss": 1.5344, "step": 2 }, { "epoch": 0.0, "grad_norm": 127.5, "learning_rate": 4.716981132075472e-08, "loss": 1.5733, "step": 3 }, { "epoch": 0.0, "grad_norm": 116.0, "learning_rate": 6.289308176100629e-08, "loss": 1.6228, "step": 4 }, { "epoch": 0.0, "grad_norm": 126.0, "learning_rate": 7.861635220125787e-08, "loss": 1.745, "step": 5 }, { "epoch": 0.0, "grad_norm": 135.0, "learning_rate": 9.433962264150944e-08, "loss": 1.8331, "step": 6 }, { "epoch": 0.0, "grad_norm": 106.5, "learning_rate": 1.1006289308176101e-07, "loss": 1.5346, "step": 7 }, { "epoch": 0.0, "grad_norm": 140.0, "learning_rate": 1.2578616352201258e-07, "loss": 1.6229, "step": 8 }, { "epoch": 0.0, "grad_norm": 114.5, "learning_rate": 1.4150943396226417e-07, "loss": 1.6129, "step": 9 }, { "epoch": 0.0, "grad_norm": 160.0, "learning_rate": 1.5723270440251575e-07, "loss": 1.8405, "step": 10 }, { "epoch": 0.0, "grad_norm": 116.0, "learning_rate": 1.729559748427673e-07, "loss": 1.6994, "step": 11 }, { "epoch": 0.0, "grad_norm": 107.0, "learning_rate": 1.886792452830189e-07, "loss": 1.5812, "step": 12 }, { "epoch": 0.0, "grad_norm": 145.0, "learning_rate": 2.0440251572327044e-07, "loss": 1.6384, "step": 13 }, { "epoch": 0.0, "grad_norm": 123.5, "learning_rate": 2.2012578616352203e-07, "loss": 1.5081, "step": 14 }, { "epoch": 0.0, "grad_norm": 75.5, "learning_rate": 2.358490566037736e-07, "loss": 1.1546, "step": 15 }, { "epoch": 0.0, "grad_norm": 143.0, "learning_rate": 2.5157232704402517e-07, "loss": 1.7763, "step": 16 }, { "epoch": 0.0, "grad_norm": 95.0, "learning_rate": 2.672955974842768e-07, "loss": 1.3867, "step": 17 }, { "epoch": 0.0, "grad_norm": 108.5, "learning_rate": 2.8301886792452833e-07, "loss": 1.5579, "step": 18 }, { "epoch": 0.0, "grad_norm": 135.0, "learning_rate": 2.987421383647799e-07, "loss": 1.7327, "step": 19 }, { "epoch": 0.0, "grad_norm": 135.0, "learning_rate": 3.144654088050315e-07, "loss": 1.6331, "step": 20 }, { "epoch": 0.0, "grad_norm": 103.0, "learning_rate": 3.3018867924528305e-07, "loss": 1.5425, "step": 21 }, { "epoch": 0.0, "grad_norm": 90.0, "learning_rate": 3.459119496855346e-07, "loss": 1.3659, "step": 22 }, { "epoch": 0.0, "grad_norm": 118.0, "learning_rate": 3.6163522012578617e-07, "loss": 1.433, "step": 23 }, { "epoch": 0.0, "grad_norm": 98.5, "learning_rate": 3.773584905660378e-07, "loss": 1.4436, "step": 24 }, { "epoch": 0.0, "grad_norm": 118.0, "learning_rate": 3.9308176100628933e-07, "loss": 1.3783, "step": 25 }, { "epoch": 0.0, "grad_norm": 102.5, "learning_rate": 4.088050314465409e-07, "loss": 1.3991, "step": 26 }, { "epoch": 0.01, "grad_norm": 79.0, "learning_rate": 4.2452830188679244e-07, "loss": 1.3552, "step": 27 }, { "epoch": 0.01, "grad_norm": 92.5, "learning_rate": 4.4025157232704405e-07, "loss": 1.312, "step": 28 }, { "epoch": 0.01, "grad_norm": 80.5, "learning_rate": 4.559748427672956e-07, "loss": 1.0644, "step": 29 }, { "epoch": 0.01, "grad_norm": 85.5, "learning_rate": 4.716981132075472e-07, "loss": 1.1594, "step": 30 }, { "epoch": 0.01, "grad_norm": 85.5, "learning_rate": 4.874213836477988e-07, "loss": 1.217, "step": 31 }, { "epoch": 0.01, "grad_norm": 102.5, "learning_rate": 5.031446540880503e-07, "loss": 1.1485, "step": 32 }, { "epoch": 0.01, "grad_norm": 86.0, "learning_rate": 5.18867924528302e-07, "loss": 1.146, "step": 33 }, { "epoch": 0.01, "grad_norm": 90.5, "learning_rate": 5.345911949685535e-07, "loss": 1.0805, "step": 34 }, { "epoch": 0.01, "grad_norm": 73.5, "learning_rate": 5.503144654088051e-07, "loss": 0.9344, "step": 35 }, { "epoch": 0.01, "grad_norm": 65.0, "learning_rate": 5.660377358490567e-07, "loss": 0.9574, "step": 36 }, { "epoch": 0.01, "grad_norm": 51.0, "learning_rate": 5.817610062893082e-07, "loss": 0.9265, "step": 37 }, { "epoch": 0.01, "grad_norm": 40.75, "learning_rate": 5.974842767295598e-07, "loss": 0.7112, "step": 38 }, { "epoch": 0.01, "grad_norm": 50.5, "learning_rate": 6.132075471698113e-07, "loss": 0.8787, "step": 39 }, { "epoch": 0.01, "grad_norm": 38.0, "learning_rate": 6.28930817610063e-07, "loss": 0.8841, "step": 40 }, { "epoch": 0.01, "grad_norm": 48.25, "learning_rate": 6.446540880503145e-07, "loss": 0.7773, "step": 41 }, { "epoch": 0.01, "grad_norm": 39.0, "learning_rate": 6.603773584905661e-07, "loss": 0.7662, "step": 42 }, { "epoch": 0.01, "grad_norm": 32.5, "learning_rate": 6.761006289308177e-07, "loss": 0.8099, "step": 43 }, { "epoch": 0.01, "grad_norm": 34.25, "learning_rate": 6.918238993710692e-07, "loss": 0.8475, "step": 44 }, { "epoch": 0.01, "grad_norm": 33.0, "learning_rate": 7.075471698113208e-07, "loss": 0.8157, "step": 45 }, { "epoch": 0.01, "grad_norm": 38.0, "learning_rate": 7.232704402515723e-07, "loss": 0.76, "step": 46 }, { "epoch": 0.01, "grad_norm": 30.625, "learning_rate": 7.389937106918239e-07, "loss": 0.7033, "step": 47 }, { "epoch": 0.01, "grad_norm": 30.5, "learning_rate": 7.547169811320755e-07, "loss": 0.6235, "step": 48 }, { "epoch": 0.01, "grad_norm": 37.5, "learning_rate": 7.704402515723271e-07, "loss": 0.8002, "step": 49 }, { "epoch": 0.01, "grad_norm": 29.125, "learning_rate": 7.861635220125787e-07, "loss": 0.6508, "step": 50 }, { "epoch": 0.01, "grad_norm": 30.0, "learning_rate": 8.018867924528302e-07, "loss": 0.6782, "step": 51 }, { "epoch": 0.01, "grad_norm": 30.75, "learning_rate": 8.176100628930818e-07, "loss": 0.5906, "step": 52 }, { "epoch": 0.01, "grad_norm": 36.5, "learning_rate": 8.333333333333333e-07, "loss": 0.7475, "step": 53 }, { "epoch": 0.01, "grad_norm": 34.0, "learning_rate": 8.490566037735849e-07, "loss": 0.6445, "step": 54 }, { "epoch": 0.01, "grad_norm": 33.25, "learning_rate": 8.647798742138364e-07, "loss": 0.5897, "step": 55 }, { "epoch": 0.01, "grad_norm": 36.0, "learning_rate": 8.805031446540881e-07, "loss": 0.6611, "step": 56 }, { "epoch": 0.01, "grad_norm": 34.0, "learning_rate": 8.962264150943397e-07, "loss": 0.6005, "step": 57 }, { "epoch": 0.01, "grad_norm": 27.375, "learning_rate": 9.119496855345912e-07, "loss": 0.6288, "step": 58 }, { "epoch": 0.01, "grad_norm": 28.25, "learning_rate": 9.276729559748428e-07, "loss": 0.5794, "step": 59 }, { "epoch": 0.01, "grad_norm": 27.625, "learning_rate": 9.433962264150944e-07, "loss": 0.6671, "step": 60 }, { "epoch": 0.01, "grad_norm": 27.125, "learning_rate": 9.59119496855346e-07, "loss": 0.5524, "step": 61 }, { "epoch": 0.01, "grad_norm": 25.125, "learning_rate": 9.748427672955975e-07, "loss": 0.6101, "step": 62 }, { "epoch": 0.01, "grad_norm": 26.75, "learning_rate": 9.90566037735849e-07, "loss": 0.6089, "step": 63 }, { "epoch": 0.01, "grad_norm": 25.25, "learning_rate": 1.0062893081761007e-06, "loss": 0.5772, "step": 64 }, { "epoch": 0.01, "grad_norm": 21.75, "learning_rate": 1.0220125786163524e-06, "loss": 0.4854, "step": 65 }, { "epoch": 0.01, "grad_norm": 25.375, "learning_rate": 1.037735849056604e-06, "loss": 0.5164, "step": 66 }, { "epoch": 0.01, "grad_norm": 31.75, "learning_rate": 1.0534591194968555e-06, "loss": 0.5298, "step": 67 }, { "epoch": 0.01, "grad_norm": 25.75, "learning_rate": 1.069182389937107e-06, "loss": 0.6093, "step": 68 }, { "epoch": 0.01, "grad_norm": 32.0, "learning_rate": 1.0849056603773587e-06, "loss": 0.5012, "step": 69 }, { "epoch": 0.01, "grad_norm": 27.875, "learning_rate": 1.1006289308176102e-06, "loss": 0.5294, "step": 70 }, { "epoch": 0.01, "grad_norm": 29.5, "learning_rate": 1.1163522012578618e-06, "loss": 0.5619, "step": 71 }, { "epoch": 0.01, "grad_norm": 26.125, "learning_rate": 1.1320754716981133e-06, "loss": 0.5473, "step": 72 }, { "epoch": 0.01, "grad_norm": 28.625, "learning_rate": 1.1477987421383649e-06, "loss": 0.5757, "step": 73 }, { "epoch": 0.01, "grad_norm": 35.0, "learning_rate": 1.1635220125786164e-06, "loss": 0.729, "step": 74 }, { "epoch": 0.01, "grad_norm": 33.75, "learning_rate": 1.179245283018868e-06, "loss": 0.5983, "step": 75 }, { "epoch": 0.01, "grad_norm": 33.25, "learning_rate": 1.1949685534591195e-06, "loss": 0.5535, "step": 76 }, { "epoch": 0.01, "grad_norm": 33.0, "learning_rate": 1.210691823899371e-06, "loss": 0.5377, "step": 77 }, { "epoch": 0.01, "grad_norm": 39.25, "learning_rate": 1.2264150943396227e-06, "loss": 0.5558, "step": 78 }, { "epoch": 0.01, "grad_norm": 31.25, "learning_rate": 1.2421383647798742e-06, "loss": 0.5395, "step": 79 }, { "epoch": 0.02, "grad_norm": 23.75, "learning_rate": 1.257861635220126e-06, "loss": 0.4725, "step": 80 }, { "epoch": 0.02, "grad_norm": 27.25, "learning_rate": 1.2735849056603775e-06, "loss": 0.5112, "step": 81 }, { "epoch": 0.02, "grad_norm": 19.625, "learning_rate": 1.289308176100629e-06, "loss": 0.4277, "step": 82 }, { "epoch": 0.02, "grad_norm": 19.125, "learning_rate": 1.3050314465408807e-06, "loss": 0.5338, "step": 83 }, { "epoch": 0.02, "grad_norm": 25.125, "learning_rate": 1.3207547169811322e-06, "loss": 0.5109, "step": 84 }, { "epoch": 0.02, "grad_norm": 27.625, "learning_rate": 1.3364779874213838e-06, "loss": 0.473, "step": 85 }, { "epoch": 0.02, "grad_norm": 35.0, "learning_rate": 1.3522012578616353e-06, "loss": 0.5789, "step": 86 }, { "epoch": 0.02, "grad_norm": 30.625, "learning_rate": 1.3679245283018869e-06, "loss": 0.5569, "step": 87 }, { "epoch": 0.02, "grad_norm": 25.75, "learning_rate": 1.3836477987421384e-06, "loss": 0.626, "step": 88 }, { "epoch": 0.02, "grad_norm": 34.0, "learning_rate": 1.39937106918239e-06, "loss": 0.5303, "step": 89 }, { "epoch": 0.02, "grad_norm": 25.625, "learning_rate": 1.4150943396226415e-06, "loss": 0.45, "step": 90 }, { "epoch": 0.02, "grad_norm": 22.25, "learning_rate": 1.430817610062893e-06, "loss": 0.4815, "step": 91 }, { "epoch": 0.02, "grad_norm": 23.5, "learning_rate": 1.4465408805031447e-06, "loss": 0.4769, "step": 92 }, { "epoch": 0.02, "grad_norm": 19.125, "learning_rate": 1.4622641509433962e-06, "loss": 0.4095, "step": 93 }, { "epoch": 0.02, "grad_norm": 26.625, "learning_rate": 1.4779874213836478e-06, "loss": 0.5006, "step": 94 }, { "epoch": 0.02, "grad_norm": 21.75, "learning_rate": 1.4937106918238995e-06, "loss": 0.3748, "step": 95 }, { "epoch": 0.02, "grad_norm": 26.25, "learning_rate": 1.509433962264151e-06, "loss": 0.4987, "step": 96 }, { "epoch": 0.02, "grad_norm": 23.75, "learning_rate": 1.5251572327044027e-06, "loss": 0.4805, "step": 97 }, { "epoch": 0.02, "grad_norm": 30.125, "learning_rate": 1.5408805031446542e-06, "loss": 0.4682, "step": 98 }, { "epoch": 0.02, "grad_norm": 24.125, "learning_rate": 1.5566037735849058e-06, "loss": 0.4292, "step": 99 }, { "epoch": 0.02, "grad_norm": 23.875, "learning_rate": 1.5723270440251573e-06, "loss": 0.5331, "step": 100 }, { "epoch": 0.02, "grad_norm": 26.875, "learning_rate": 1.5880503144654089e-06, "loss": 0.5881, "step": 101 }, { "epoch": 0.02, "grad_norm": 24.125, "learning_rate": 1.6037735849056604e-06, "loss": 0.3955, "step": 102 }, { "epoch": 0.02, "grad_norm": 22.125, "learning_rate": 1.619496855345912e-06, "loss": 0.4782, "step": 103 }, { "epoch": 0.02, "grad_norm": 26.5, "learning_rate": 1.6352201257861635e-06, "loss": 0.4031, "step": 104 }, { "epoch": 0.02, "grad_norm": 28.625, "learning_rate": 1.650943396226415e-06, "loss": 0.4659, "step": 105 }, { "epoch": 0.02, "grad_norm": 25.625, "learning_rate": 1.6666666666666667e-06, "loss": 0.4377, "step": 106 }, { "epoch": 0.02, "grad_norm": 28.0, "learning_rate": 1.6823899371069182e-06, "loss": 0.5867, "step": 107 }, { "epoch": 0.02, "grad_norm": 28.875, "learning_rate": 1.6981132075471698e-06, "loss": 0.5003, "step": 108 }, { "epoch": 0.02, "grad_norm": 25.875, "learning_rate": 1.7138364779874213e-06, "loss": 0.5406, "step": 109 }, { "epoch": 0.02, "grad_norm": 26.375, "learning_rate": 1.7295597484276729e-06, "loss": 0.3781, "step": 110 }, { "epoch": 0.02, "grad_norm": 28.125, "learning_rate": 1.7452830188679247e-06, "loss": 0.504, "step": 111 }, { "epoch": 0.02, "grad_norm": 31.5, "learning_rate": 1.7610062893081762e-06, "loss": 0.4315, "step": 112 }, { "epoch": 0.02, "grad_norm": 22.75, "learning_rate": 1.7767295597484278e-06, "loss": 0.4829, "step": 113 }, { "epoch": 0.02, "grad_norm": 25.75, "learning_rate": 1.7924528301886793e-06, "loss": 0.4624, "step": 114 }, { "epoch": 0.02, "grad_norm": 24.5, "learning_rate": 1.8081761006289309e-06, "loss": 0.6236, "step": 115 }, { "epoch": 0.02, "grad_norm": 27.25, "learning_rate": 1.8238993710691824e-06, "loss": 0.4962, "step": 116 }, { "epoch": 0.02, "grad_norm": 24.75, "learning_rate": 1.839622641509434e-06, "loss": 0.4383, "step": 117 }, { "epoch": 0.02, "grad_norm": 26.0, "learning_rate": 1.8553459119496855e-06, "loss": 0.502, "step": 118 }, { "epoch": 0.02, "grad_norm": 28.625, "learning_rate": 1.871069182389937e-06, "loss": 0.4168, "step": 119 }, { "epoch": 0.02, "grad_norm": 18.75, "learning_rate": 1.8867924528301889e-06, "loss": 0.4042, "step": 120 }, { "epoch": 0.02, "grad_norm": 27.875, "learning_rate": 1.9025157232704406e-06, "loss": 0.4588, "step": 121 }, { "epoch": 0.02, "grad_norm": 33.25, "learning_rate": 1.918238993710692e-06, "loss": 0.5517, "step": 122 }, { "epoch": 0.02, "grad_norm": 30.375, "learning_rate": 1.9339622641509438e-06, "loss": 0.5322, "step": 123 }, { "epoch": 0.02, "grad_norm": 31.75, "learning_rate": 1.949685534591195e-06, "loss": 0.5831, "step": 124 }, { "epoch": 0.02, "grad_norm": 24.625, "learning_rate": 1.965408805031447e-06, "loss": 0.4179, "step": 125 }, { "epoch": 0.02, "grad_norm": 21.375, "learning_rate": 1.981132075471698e-06, "loss": 0.4637, "step": 126 }, { "epoch": 0.02, "grad_norm": 21.75, "learning_rate": 1.99685534591195e-06, "loss": 0.4354, "step": 127 }, { "epoch": 0.02, "grad_norm": 25.375, "learning_rate": 2.0125786163522013e-06, "loss": 0.4512, "step": 128 }, { "epoch": 0.02, "grad_norm": 22.875, "learning_rate": 2.028301886792453e-06, "loss": 0.4658, "step": 129 }, { "epoch": 0.02, "grad_norm": 26.875, "learning_rate": 2.044025157232705e-06, "loss": 0.4747, "step": 130 }, { "epoch": 0.02, "grad_norm": 21.875, "learning_rate": 2.059748427672956e-06, "loss": 0.4106, "step": 131 }, { "epoch": 0.02, "grad_norm": 20.375, "learning_rate": 2.075471698113208e-06, "loss": 0.5149, "step": 132 }, { "epoch": 0.03, "grad_norm": 29.625, "learning_rate": 2.0911949685534593e-06, "loss": 0.551, "step": 133 }, { "epoch": 0.03, "grad_norm": 23.0, "learning_rate": 2.106918238993711e-06, "loss": 0.4205, "step": 134 }, { "epoch": 0.03, "grad_norm": 34.0, "learning_rate": 2.1226415094339624e-06, "loss": 0.557, "step": 135 }, { "epoch": 0.03, "grad_norm": 22.75, "learning_rate": 2.138364779874214e-06, "loss": 0.4591, "step": 136 }, { "epoch": 0.03, "grad_norm": 16.0, "learning_rate": 2.1540880503144655e-06, "loss": 0.4063, "step": 137 }, { "epoch": 0.03, "grad_norm": 31.0, "learning_rate": 2.1698113207547173e-06, "loss": 0.4804, "step": 138 }, { "epoch": 0.03, "grad_norm": 25.75, "learning_rate": 2.1855345911949687e-06, "loss": 0.3588, "step": 139 }, { "epoch": 0.03, "grad_norm": 24.875, "learning_rate": 2.2012578616352204e-06, "loss": 0.386, "step": 140 }, { "epoch": 0.03, "grad_norm": 27.625, "learning_rate": 2.2169811320754718e-06, "loss": 0.4766, "step": 141 }, { "epoch": 0.03, "grad_norm": 22.0, "learning_rate": 2.2327044025157235e-06, "loss": 0.4239, "step": 142 }, { "epoch": 0.03, "grad_norm": 22.0, "learning_rate": 2.248427672955975e-06, "loss": 0.3294, "step": 143 }, { "epoch": 0.03, "grad_norm": 26.375, "learning_rate": 2.2641509433962266e-06, "loss": 0.4473, "step": 144 }, { "epoch": 0.03, "grad_norm": 26.125, "learning_rate": 2.2798742138364784e-06, "loss": 0.4707, "step": 145 }, { "epoch": 0.03, "grad_norm": 18.5, "learning_rate": 2.2955974842767298e-06, "loss": 0.3358, "step": 146 }, { "epoch": 0.03, "grad_norm": 25.5, "learning_rate": 2.3113207547169815e-06, "loss": 0.4452, "step": 147 }, { "epoch": 0.03, "grad_norm": 29.375, "learning_rate": 2.327044025157233e-06, "loss": 0.4103, "step": 148 }, { "epoch": 0.03, "grad_norm": 30.25, "learning_rate": 2.3427672955974846e-06, "loss": 0.5534, "step": 149 }, { "epoch": 0.03, "grad_norm": 31.125, "learning_rate": 2.358490566037736e-06, "loss": 0.4634, "step": 150 }, { "epoch": 0.03, "grad_norm": 26.25, "learning_rate": 2.3742138364779878e-06, "loss": 0.4506, "step": 151 }, { "epoch": 0.03, "grad_norm": 23.75, "learning_rate": 2.389937106918239e-06, "loss": 0.4663, "step": 152 }, { "epoch": 0.03, "grad_norm": 28.875, "learning_rate": 2.405660377358491e-06, "loss": 0.5679, "step": 153 }, { "epoch": 0.03, "grad_norm": 28.25, "learning_rate": 2.421383647798742e-06, "loss": 0.5078, "step": 154 }, { "epoch": 0.03, "grad_norm": 22.25, "learning_rate": 2.437106918238994e-06, "loss": 0.4684, "step": 155 }, { "epoch": 0.03, "grad_norm": 26.75, "learning_rate": 2.4528301886792453e-06, "loss": 0.5082, "step": 156 }, { "epoch": 0.03, "grad_norm": 34.0, "learning_rate": 2.468553459119497e-06, "loss": 0.5108, "step": 157 }, { "epoch": 0.03, "grad_norm": 21.25, "learning_rate": 2.4842767295597484e-06, "loss": 0.4825, "step": 158 }, { "epoch": 0.03, "grad_norm": 17.75, "learning_rate": 2.5e-06, "loss": 0.3955, "step": 159 }, { "epoch": 0.03, "grad_norm": 28.75, "learning_rate": 2.515723270440252e-06, "loss": 0.4903, "step": 160 }, { "epoch": 0.03, "grad_norm": 22.75, "learning_rate": 2.5314465408805033e-06, "loss": 0.3555, "step": 161 }, { "epoch": 0.03, "grad_norm": 19.75, "learning_rate": 2.547169811320755e-06, "loss": 0.4296, "step": 162 }, { "epoch": 0.03, "grad_norm": 20.0, "learning_rate": 2.5628930817610064e-06, "loss": 0.3517, "step": 163 }, { "epoch": 0.03, "grad_norm": 21.625, "learning_rate": 2.578616352201258e-06, "loss": 0.4307, "step": 164 }, { "epoch": 0.03, "grad_norm": 22.125, "learning_rate": 2.5943396226415095e-06, "loss": 0.5274, "step": 165 }, { "epoch": 0.03, "grad_norm": 21.875, "learning_rate": 2.6100628930817613e-06, "loss": 0.4943, "step": 166 }, { "epoch": 0.03, "grad_norm": 23.25, "learning_rate": 2.6257861635220127e-06, "loss": 0.3928, "step": 167 }, { "epoch": 0.03, "grad_norm": 27.875, "learning_rate": 2.6415094339622644e-06, "loss": 0.4462, "step": 168 }, { "epoch": 0.03, "grad_norm": 28.375, "learning_rate": 2.6572327044025158e-06, "loss": 0.4013, "step": 169 }, { "epoch": 0.03, "grad_norm": 18.625, "learning_rate": 2.6729559748427675e-06, "loss": 0.4021, "step": 170 }, { "epoch": 0.03, "grad_norm": 24.25, "learning_rate": 2.688679245283019e-06, "loss": 0.3648, "step": 171 }, { "epoch": 0.03, "grad_norm": 20.5, "learning_rate": 2.7044025157232706e-06, "loss": 0.4156, "step": 172 }, { "epoch": 0.03, "grad_norm": 27.5, "learning_rate": 2.720125786163522e-06, "loss": 0.4162, "step": 173 }, { "epoch": 0.03, "grad_norm": 26.125, "learning_rate": 2.7358490566037738e-06, "loss": 0.4839, "step": 174 }, { "epoch": 0.03, "grad_norm": 29.5, "learning_rate": 2.7515723270440255e-06, "loss": 0.4664, "step": 175 }, { "epoch": 0.03, "grad_norm": 17.0, "learning_rate": 2.767295597484277e-06, "loss": 0.3771, "step": 176 }, { "epoch": 0.03, "grad_norm": 17.0, "learning_rate": 2.7830188679245286e-06, "loss": 0.3413, "step": 177 }, { "epoch": 0.03, "grad_norm": 25.25, "learning_rate": 2.79874213836478e-06, "loss": 0.5413, "step": 178 }, { "epoch": 0.03, "grad_norm": 28.375, "learning_rate": 2.8144654088050318e-06, "loss": 0.3885, "step": 179 }, { "epoch": 0.03, "grad_norm": 21.375, "learning_rate": 2.830188679245283e-06, "loss": 0.5146, "step": 180 }, { "epoch": 0.03, "grad_norm": 33.5, "learning_rate": 2.845911949685535e-06, "loss": 0.3563, "step": 181 }, { "epoch": 0.03, "grad_norm": 22.625, "learning_rate": 2.861635220125786e-06, "loss": 0.4598, "step": 182 }, { "epoch": 0.03, "grad_norm": 28.875, "learning_rate": 2.877358490566038e-06, "loss": 0.406, "step": 183 }, { "epoch": 0.03, "grad_norm": 18.5, "learning_rate": 2.8930817610062893e-06, "loss": 0.3892, "step": 184 }, { "epoch": 0.03, "grad_norm": 18.5, "learning_rate": 2.908805031446541e-06, "loss": 0.3769, "step": 185 }, { "epoch": 0.04, "grad_norm": 15.9375, "learning_rate": 2.9245283018867924e-06, "loss": 0.3572, "step": 186 }, { "epoch": 0.04, "grad_norm": 20.75, "learning_rate": 2.940251572327044e-06, "loss": 0.4337, "step": 187 }, { "epoch": 0.04, "grad_norm": 20.0, "learning_rate": 2.9559748427672955e-06, "loss": 0.4221, "step": 188 }, { "epoch": 0.04, "grad_norm": 18.0, "learning_rate": 2.9716981132075473e-06, "loss": 0.4219, "step": 189 }, { "epoch": 0.04, "grad_norm": 17.5, "learning_rate": 2.987421383647799e-06, "loss": 0.3966, "step": 190 }, { "epoch": 0.04, "grad_norm": 19.625, "learning_rate": 3.0031446540880504e-06, "loss": 0.5014, "step": 191 }, { "epoch": 0.04, "grad_norm": 17.875, "learning_rate": 3.018867924528302e-06, "loss": 0.4032, "step": 192 }, { "epoch": 0.04, "grad_norm": 17.25, "learning_rate": 3.0345911949685535e-06, "loss": 0.4115, "step": 193 }, { "epoch": 0.04, "grad_norm": 27.875, "learning_rate": 3.0503144654088053e-06, "loss": 0.4038, "step": 194 }, { "epoch": 0.04, "grad_norm": 21.75, "learning_rate": 3.0660377358490567e-06, "loss": 0.4118, "step": 195 }, { "epoch": 0.04, "grad_norm": 23.0, "learning_rate": 3.0817610062893084e-06, "loss": 0.4172, "step": 196 }, { "epoch": 0.04, "grad_norm": 22.875, "learning_rate": 3.0974842767295598e-06, "loss": 0.4853, "step": 197 }, { "epoch": 0.04, "grad_norm": 23.375, "learning_rate": 3.1132075471698115e-06, "loss": 0.3626, "step": 198 }, { "epoch": 0.04, "grad_norm": 21.625, "learning_rate": 3.128930817610063e-06, "loss": 0.376, "step": 199 }, { "epoch": 0.04, "grad_norm": 18.625, "learning_rate": 3.1446540880503146e-06, "loss": 0.3621, "step": 200 }, { "epoch": 0.04, "grad_norm": 17.75, "learning_rate": 3.160377358490566e-06, "loss": 0.4225, "step": 201 }, { "epoch": 0.04, "grad_norm": 17.375, "learning_rate": 3.1761006289308178e-06, "loss": 0.343, "step": 202 }, { "epoch": 0.04, "grad_norm": 21.375, "learning_rate": 3.191823899371069e-06, "loss": 0.4365, "step": 203 }, { "epoch": 0.04, "grad_norm": 20.875, "learning_rate": 3.207547169811321e-06, "loss": 0.3967, "step": 204 }, { "epoch": 0.04, "grad_norm": 23.25, "learning_rate": 3.2232704402515726e-06, "loss": 0.3415, "step": 205 }, { "epoch": 0.04, "grad_norm": 20.875, "learning_rate": 3.238993710691824e-06, "loss": 0.4564, "step": 206 }, { "epoch": 0.04, "grad_norm": 35.0, "learning_rate": 3.2547169811320758e-06, "loss": 0.4755, "step": 207 }, { "epoch": 0.04, "grad_norm": 33.25, "learning_rate": 3.270440251572327e-06, "loss": 0.404, "step": 208 }, { "epoch": 0.04, "grad_norm": 34.25, "learning_rate": 3.286163522012579e-06, "loss": 0.4672, "step": 209 }, { "epoch": 0.04, "grad_norm": 18.75, "learning_rate": 3.30188679245283e-06, "loss": 0.3921, "step": 210 }, { "epoch": 0.04, "grad_norm": 19.375, "learning_rate": 3.317610062893082e-06, "loss": 0.4339, "step": 211 }, { "epoch": 0.04, "grad_norm": 24.375, "learning_rate": 3.3333333333333333e-06, "loss": 0.3702, "step": 212 }, { "epoch": 0.04, "grad_norm": 20.375, "learning_rate": 3.349056603773585e-06, "loss": 0.4247, "step": 213 }, { "epoch": 0.04, "grad_norm": 16.5, "learning_rate": 3.3647798742138364e-06, "loss": 0.4095, "step": 214 }, { "epoch": 0.04, "grad_norm": 16.25, "learning_rate": 3.380503144654088e-06, "loss": 0.4095, "step": 215 }, { "epoch": 0.04, "grad_norm": 15.6875, "learning_rate": 3.3962264150943395e-06, "loss": 0.3489, "step": 216 }, { "epoch": 0.04, "grad_norm": 21.75, "learning_rate": 3.4119496855345913e-06, "loss": 0.3766, "step": 217 }, { "epoch": 0.04, "grad_norm": 14.375, "learning_rate": 3.4276729559748427e-06, "loss": 0.3304, "step": 218 }, { "epoch": 0.04, "grad_norm": 18.125, "learning_rate": 3.4433962264150944e-06, "loss": 0.3874, "step": 219 }, { "epoch": 0.04, "grad_norm": 20.375, "learning_rate": 3.4591194968553458e-06, "loss": 0.372, "step": 220 }, { "epoch": 0.04, "grad_norm": 19.25, "learning_rate": 3.4748427672955975e-06, "loss": 0.405, "step": 221 }, { "epoch": 0.04, "grad_norm": 26.125, "learning_rate": 3.4905660377358493e-06, "loss": 0.4138, "step": 222 }, { "epoch": 0.04, "grad_norm": 19.625, "learning_rate": 3.5062893081761007e-06, "loss": 0.3923, "step": 223 }, { "epoch": 0.04, "grad_norm": 18.875, "learning_rate": 3.5220125786163524e-06, "loss": 0.3736, "step": 224 }, { "epoch": 0.04, "grad_norm": 21.375, "learning_rate": 3.5377358490566038e-06, "loss": 0.4285, "step": 225 }, { "epoch": 0.04, "grad_norm": 21.875, "learning_rate": 3.5534591194968555e-06, "loss": 0.3933, "step": 226 }, { "epoch": 0.04, "grad_norm": 19.875, "learning_rate": 3.569182389937107e-06, "loss": 0.3964, "step": 227 }, { "epoch": 0.04, "grad_norm": 16.875, "learning_rate": 3.5849056603773586e-06, "loss": 0.4118, "step": 228 }, { "epoch": 0.04, "grad_norm": 26.75, "learning_rate": 3.60062893081761e-06, "loss": 0.3828, "step": 229 }, { "epoch": 0.04, "grad_norm": 26.125, "learning_rate": 3.6163522012578618e-06, "loss": 0.4133, "step": 230 }, { "epoch": 0.04, "grad_norm": 20.125, "learning_rate": 3.632075471698113e-06, "loss": 0.3447, "step": 231 }, { "epoch": 0.04, "grad_norm": 21.25, "learning_rate": 3.647798742138365e-06, "loss": 0.3851, "step": 232 }, { "epoch": 0.04, "grad_norm": 17.25, "learning_rate": 3.6635220125786162e-06, "loss": 0.4931, "step": 233 }, { "epoch": 0.04, "grad_norm": 21.125, "learning_rate": 3.679245283018868e-06, "loss": 0.3914, "step": 234 }, { "epoch": 0.04, "grad_norm": 30.375, "learning_rate": 3.6949685534591193e-06, "loss": 0.4266, "step": 235 }, { "epoch": 0.04, "grad_norm": 18.0, "learning_rate": 3.710691823899371e-06, "loss": 0.352, "step": 236 }, { "epoch": 0.04, "grad_norm": 25.0, "learning_rate": 3.726415094339623e-06, "loss": 0.4071, "step": 237 }, { "epoch": 0.05, "grad_norm": 25.5, "learning_rate": 3.742138364779874e-06, "loss": 0.3569, "step": 238 }, { "epoch": 0.05, "grad_norm": 23.5, "learning_rate": 3.7578616352201264e-06, "loss": 0.4584, "step": 239 }, { "epoch": 0.05, "grad_norm": 20.25, "learning_rate": 3.7735849056603777e-06, "loss": 0.3881, "step": 240 }, { "epoch": 0.05, "grad_norm": 24.5, "learning_rate": 3.7893081761006295e-06, "loss": 0.4608, "step": 241 }, { "epoch": 0.05, "grad_norm": 18.5, "learning_rate": 3.8050314465408813e-06, "loss": 0.3557, "step": 242 }, { "epoch": 0.05, "grad_norm": 25.375, "learning_rate": 3.820754716981133e-06, "loss": 0.5502, "step": 243 }, { "epoch": 0.05, "grad_norm": 17.5, "learning_rate": 3.836477987421384e-06, "loss": 0.3618, "step": 244 }, { "epoch": 0.05, "grad_norm": 29.75, "learning_rate": 3.852201257861636e-06, "loss": 0.3911, "step": 245 }, { "epoch": 0.05, "grad_norm": 14.9375, "learning_rate": 3.8679245283018875e-06, "loss": 0.3085, "step": 246 }, { "epoch": 0.05, "grad_norm": 37.25, "learning_rate": 3.883647798742139e-06, "loss": 0.305, "step": 247 }, { "epoch": 0.05, "grad_norm": 23.0, "learning_rate": 3.89937106918239e-06, "loss": 0.3474, "step": 248 }, { "epoch": 0.05, "grad_norm": 16.75, "learning_rate": 3.915094339622642e-06, "loss": 0.3963, "step": 249 }, { "epoch": 0.05, "grad_norm": 17.625, "learning_rate": 3.930817610062894e-06, "loss": 0.4431, "step": 250 }, { "epoch": 0.05, "grad_norm": 29.5, "learning_rate": 3.9465408805031455e-06, "loss": 0.4411, "step": 251 }, { "epoch": 0.05, "grad_norm": 23.375, "learning_rate": 3.962264150943396e-06, "loss": 0.3562, "step": 252 }, { "epoch": 0.05, "grad_norm": 24.875, "learning_rate": 3.977987421383648e-06, "loss": 0.3915, "step": 253 }, { "epoch": 0.05, "grad_norm": 21.25, "learning_rate": 3.9937106918239e-06, "loss": 0.4699, "step": 254 }, { "epoch": 0.05, "grad_norm": 19.5, "learning_rate": 4.009433962264152e-06, "loss": 0.4316, "step": 255 }, { "epoch": 0.05, "grad_norm": 21.5, "learning_rate": 4.025157232704403e-06, "loss": 0.4561, "step": 256 }, { "epoch": 0.05, "grad_norm": 18.75, "learning_rate": 4.040880503144654e-06, "loss": 0.2755, "step": 257 }, { "epoch": 0.05, "grad_norm": 16.875, "learning_rate": 4.056603773584906e-06, "loss": 0.4289, "step": 258 }, { "epoch": 0.05, "grad_norm": 22.75, "learning_rate": 4.072327044025158e-06, "loss": 0.4211, "step": 259 }, { "epoch": 0.05, "grad_norm": 18.875, "learning_rate": 4.08805031446541e-06, "loss": 0.4213, "step": 260 }, { "epoch": 0.05, "grad_norm": 18.125, "learning_rate": 4.103773584905661e-06, "loss": 0.3896, "step": 261 }, { "epoch": 0.05, "grad_norm": 18.625, "learning_rate": 4.119496855345912e-06, "loss": 0.457, "step": 262 }, { "epoch": 0.05, "grad_norm": 17.25, "learning_rate": 4.135220125786164e-06, "loss": 0.3403, "step": 263 }, { "epoch": 0.05, "grad_norm": 21.125, "learning_rate": 4.150943396226416e-06, "loss": 0.5224, "step": 264 }, { "epoch": 0.05, "grad_norm": 14.5625, "learning_rate": 4.166666666666667e-06, "loss": 0.3405, "step": 265 }, { "epoch": 0.05, "grad_norm": 30.0, "learning_rate": 4.182389937106919e-06, "loss": 0.4169, "step": 266 }, { "epoch": 0.05, "grad_norm": 19.125, "learning_rate": 4.19811320754717e-06, "loss": 0.4014, "step": 267 }, { "epoch": 0.05, "grad_norm": 20.75, "learning_rate": 4.213836477987422e-06, "loss": 0.4174, "step": 268 }, { "epoch": 0.05, "grad_norm": 21.25, "learning_rate": 4.229559748427673e-06, "loss": 0.3997, "step": 269 }, { "epoch": 0.05, "grad_norm": 27.0, "learning_rate": 4.245283018867925e-06, "loss": 0.5649, "step": 270 }, { "epoch": 0.05, "grad_norm": 14.0, "learning_rate": 4.261006289308177e-06, "loss": 0.3284, "step": 271 }, { "epoch": 0.05, "grad_norm": 12.375, "learning_rate": 4.276729559748428e-06, "loss": 0.4192, "step": 272 }, { "epoch": 0.05, "grad_norm": 21.125, "learning_rate": 4.29245283018868e-06, "loss": 0.3718, "step": 273 }, { "epoch": 0.05, "grad_norm": 17.625, "learning_rate": 4.308176100628931e-06, "loss": 0.3621, "step": 274 }, { "epoch": 0.05, "grad_norm": 18.875, "learning_rate": 4.323899371069183e-06, "loss": 0.3881, "step": 275 }, { "epoch": 0.05, "grad_norm": 15.5, "learning_rate": 4.339622641509435e-06, "loss": 0.2994, "step": 276 }, { "epoch": 0.05, "grad_norm": 17.25, "learning_rate": 4.355345911949686e-06, "loss": 0.3705, "step": 277 }, { "epoch": 0.05, "grad_norm": 18.75, "learning_rate": 4.371069182389937e-06, "loss": 0.5117, "step": 278 }, { "epoch": 0.05, "grad_norm": 14.4375, "learning_rate": 4.386792452830189e-06, "loss": 0.4011, "step": 279 }, { "epoch": 0.05, "grad_norm": 14.9375, "learning_rate": 4.402515723270441e-06, "loss": 0.4229, "step": 280 }, { "epoch": 0.05, "grad_norm": 22.5, "learning_rate": 4.418238993710693e-06, "loss": 0.3715, "step": 281 }, { "epoch": 0.05, "grad_norm": 14.3125, "learning_rate": 4.4339622641509435e-06, "loss": 0.424, "step": 282 }, { "epoch": 0.05, "grad_norm": 14.9375, "learning_rate": 4.449685534591195e-06, "loss": 0.4098, "step": 283 }, { "epoch": 0.05, "grad_norm": 14.8125, "learning_rate": 4.465408805031447e-06, "loss": 0.3201, "step": 284 }, { "epoch": 0.05, "grad_norm": 27.0, "learning_rate": 4.481132075471699e-06, "loss": 0.4254, "step": 285 }, { "epoch": 0.05, "grad_norm": 19.875, "learning_rate": 4.49685534591195e-06, "loss": 0.4035, "step": 286 }, { "epoch": 0.05, "grad_norm": 15.0, "learning_rate": 4.5125786163522015e-06, "loss": 0.4354, "step": 287 }, { "epoch": 0.05, "grad_norm": 18.25, "learning_rate": 4.528301886792453e-06, "loss": 0.344, "step": 288 }, { "epoch": 0.05, "grad_norm": 16.0, "learning_rate": 4.544025157232705e-06, "loss": 0.4122, "step": 289 }, { "epoch": 0.05, "grad_norm": 20.875, "learning_rate": 4.559748427672957e-06, "loss": 0.3948, "step": 290 }, { "epoch": 0.06, "grad_norm": 13.5625, "learning_rate": 4.575471698113208e-06, "loss": 0.3416, "step": 291 }, { "epoch": 0.06, "grad_norm": 14.625, "learning_rate": 4.5911949685534595e-06, "loss": 0.4063, "step": 292 }, { "epoch": 0.06, "grad_norm": 13.4375, "learning_rate": 4.606918238993711e-06, "loss": 0.3446, "step": 293 }, { "epoch": 0.06, "grad_norm": 16.875, "learning_rate": 4.622641509433963e-06, "loss": 0.3559, "step": 294 }, { "epoch": 0.06, "grad_norm": 21.625, "learning_rate": 4.638364779874214e-06, "loss": 0.3656, "step": 295 }, { "epoch": 0.06, "grad_norm": 17.25, "learning_rate": 4.654088050314466e-06, "loss": 0.367, "step": 296 }, { "epoch": 0.06, "grad_norm": 13.375, "learning_rate": 4.6698113207547175e-06, "loss": 0.3824, "step": 297 }, { "epoch": 0.06, "grad_norm": 24.125, "learning_rate": 4.685534591194969e-06, "loss": 0.4078, "step": 298 }, { "epoch": 0.06, "grad_norm": 13.5, "learning_rate": 4.70125786163522e-06, "loss": 0.4113, "step": 299 }, { "epoch": 0.06, "grad_norm": 17.0, "learning_rate": 4.716981132075472e-06, "loss": 0.4041, "step": 300 }, { "epoch": 0.06, "grad_norm": 16.625, "learning_rate": 4.732704402515724e-06, "loss": 0.3579, "step": 301 }, { "epoch": 0.06, "grad_norm": 21.75, "learning_rate": 4.7484276729559755e-06, "loss": 0.453, "step": 302 }, { "epoch": 0.06, "grad_norm": 16.0, "learning_rate": 4.764150943396227e-06, "loss": 0.3141, "step": 303 }, { "epoch": 0.06, "grad_norm": 17.25, "learning_rate": 4.779874213836478e-06, "loss": 0.4763, "step": 304 }, { "epoch": 0.06, "grad_norm": 21.625, "learning_rate": 4.79559748427673e-06, "loss": 0.4064, "step": 305 }, { "epoch": 0.06, "grad_norm": 19.375, "learning_rate": 4.811320754716982e-06, "loss": 0.43, "step": 306 }, { "epoch": 0.06, "grad_norm": 17.125, "learning_rate": 4.8270440251572335e-06, "loss": 0.4345, "step": 307 }, { "epoch": 0.06, "grad_norm": 14.5, "learning_rate": 4.842767295597484e-06, "loss": 0.4404, "step": 308 }, { "epoch": 0.06, "grad_norm": 18.0, "learning_rate": 4.858490566037736e-06, "loss": 0.4211, "step": 309 }, { "epoch": 0.06, "grad_norm": 23.375, "learning_rate": 4.874213836477988e-06, "loss": 0.4317, "step": 310 }, { "epoch": 0.06, "grad_norm": 23.25, "learning_rate": 4.88993710691824e-06, "loss": 0.4662, "step": 311 }, { "epoch": 0.06, "grad_norm": 15.3125, "learning_rate": 4.905660377358491e-06, "loss": 0.4321, "step": 312 }, { "epoch": 0.06, "grad_norm": 14.4375, "learning_rate": 4.921383647798742e-06, "loss": 0.3522, "step": 313 }, { "epoch": 0.06, "grad_norm": 15.3125, "learning_rate": 4.937106918238994e-06, "loss": 0.3414, "step": 314 }, { "epoch": 0.06, "grad_norm": 23.125, "learning_rate": 4.952830188679246e-06, "loss": 0.3923, "step": 315 }, { "epoch": 0.06, "grad_norm": 17.875, "learning_rate": 4.968553459119497e-06, "loss": 0.3885, "step": 316 }, { "epoch": 0.06, "grad_norm": 11.625, "learning_rate": 4.984276729559749e-06, "loss": 0.328, "step": 317 }, { "epoch": 0.06, "grad_norm": 14.1875, "learning_rate": 5e-06, "loss": 0.3607, "step": 318 }, { "epoch": 0.06, "grad_norm": 25.5, "learning_rate": 4.999999882666202e-06, "loss": 0.4078, "step": 319 }, { "epoch": 0.06, "grad_norm": 18.5, "learning_rate": 4.999999530664819e-06, "loss": 0.4382, "step": 320 }, { "epoch": 0.06, "grad_norm": 18.5, "learning_rate": 4.999998943995884e-06, "loss": 0.3511, "step": 321 }, { "epoch": 0.06, "grad_norm": 19.75, "learning_rate": 4.999998122659452e-06, "loss": 0.4333, "step": 322 }, { "epoch": 0.06, "grad_norm": 15.0, "learning_rate": 4.9999970666556e-06, "loss": 0.4233, "step": 323 }, { "epoch": 0.06, "grad_norm": 13.75, "learning_rate": 4.999995775984428e-06, "loss": 0.3908, "step": 324 }, { "epoch": 0.06, "grad_norm": 17.0, "learning_rate": 4.999994250646055e-06, "loss": 0.3711, "step": 325 }, { "epoch": 0.06, "grad_norm": 15.125, "learning_rate": 4.999992490640627e-06, "loss": 0.3942, "step": 326 }, { "epoch": 0.06, "grad_norm": 13.875, "learning_rate": 4.999990495968307e-06, "loss": 0.4082, "step": 327 }, { "epoch": 0.06, "grad_norm": 21.625, "learning_rate": 4.999988266629283e-06, "loss": 0.3754, "step": 328 }, { "epoch": 0.06, "grad_norm": 16.625, "learning_rate": 4.999985802623765e-06, "loss": 0.3468, "step": 329 }, { "epoch": 0.06, "grad_norm": 18.75, "learning_rate": 4.999983103951983e-06, "loss": 0.3714, "step": 330 }, { "epoch": 0.06, "grad_norm": 13.8125, "learning_rate": 4.999980170614191e-06, "loss": 0.3571, "step": 331 }, { "epoch": 0.06, "grad_norm": 16.375, "learning_rate": 4.999977002610664e-06, "loss": 0.457, "step": 332 }, { "epoch": 0.06, "grad_norm": 16.125, "learning_rate": 4.999973599941701e-06, "loss": 0.3813, "step": 333 }, { "epoch": 0.06, "grad_norm": 14.3125, "learning_rate": 4.999969962607618e-06, "loss": 0.4148, "step": 334 }, { "epoch": 0.06, "grad_norm": 18.75, "learning_rate": 4.99996609060876e-06, "loss": 0.3746, "step": 335 }, { "epoch": 0.06, "grad_norm": 14.1875, "learning_rate": 4.999961983945488e-06, "loss": 0.4341, "step": 336 }, { "epoch": 0.06, "grad_norm": 13.5625, "learning_rate": 4.999957642618188e-06, "loss": 0.389, "step": 337 }, { "epoch": 0.06, "grad_norm": 17.75, "learning_rate": 4.99995306662727e-06, "loss": 0.4204, "step": 338 }, { "epoch": 0.06, "grad_norm": 16.375, "learning_rate": 4.999948255973159e-06, "loss": 0.5364, "step": 339 }, { "epoch": 0.06, "grad_norm": 25.125, "learning_rate": 4.999943210656311e-06, "loss": 0.4162, "step": 340 }, { "epoch": 0.06, "grad_norm": 17.0, "learning_rate": 4.999937930677196e-06, "loss": 0.3731, "step": 341 }, { "epoch": 0.06, "grad_norm": 15.5, "learning_rate": 4.9999324160363115e-06, "loss": 0.3723, "step": 342 }, { "epoch": 0.06, "grad_norm": 17.75, "learning_rate": 4.999926666734175e-06, "loss": 0.4074, "step": 343 }, { "epoch": 0.07, "grad_norm": 19.625, "learning_rate": 4.9999206827713265e-06, "loss": 0.5059, "step": 344 }, { "epoch": 0.07, "grad_norm": 17.75, "learning_rate": 4.999914464148327e-06, "loss": 0.3604, "step": 345 }, { "epoch": 0.07, "grad_norm": 18.5, "learning_rate": 4.99990801086576e-06, "loss": 0.3497, "step": 346 }, { "epoch": 0.07, "grad_norm": 20.0, "learning_rate": 4.999901322924232e-06, "loss": 0.363, "step": 347 }, { "epoch": 0.07, "grad_norm": 13.9375, "learning_rate": 4.999894400324371e-06, "loss": 0.3568, "step": 348 }, { "epoch": 0.07, "grad_norm": 12.9375, "learning_rate": 4.999887243066826e-06, "loss": 0.4255, "step": 349 }, { "epoch": 0.07, "grad_norm": 17.75, "learning_rate": 4.999879851152268e-06, "loss": 0.444, "step": 350 }, { "epoch": 0.07, "grad_norm": 21.625, "learning_rate": 4.999872224581393e-06, "loss": 0.4492, "step": 351 }, { "epoch": 0.07, "grad_norm": 19.0, "learning_rate": 4.999864363354915e-06, "loss": 0.4042, "step": 352 }, { "epoch": 0.07, "grad_norm": 18.625, "learning_rate": 4.999856267473573e-06, "loss": 0.415, "step": 353 }, { "epoch": 0.07, "grad_norm": 17.0, "learning_rate": 4.999847936938127e-06, "loss": 0.3095, "step": 354 }, { "epoch": 0.07, "grad_norm": 11.8125, "learning_rate": 4.999839371749359e-06, "loss": 0.3651, "step": 355 }, { "epoch": 0.07, "grad_norm": 12.625, "learning_rate": 4.999830571908072e-06, "loss": 0.3397, "step": 356 }, { "epoch": 0.07, "grad_norm": 16.25, "learning_rate": 4.999821537415092e-06, "loss": 0.4235, "step": 357 }, { "epoch": 0.07, "grad_norm": 15.1875, "learning_rate": 4.999812268271269e-06, "loss": 0.3942, "step": 358 }, { "epoch": 0.07, "grad_norm": 13.25, "learning_rate": 4.999802764477472e-06, "loss": 0.3885, "step": 359 }, { "epoch": 0.07, "grad_norm": 22.375, "learning_rate": 4.999793026034592e-06, "loss": 0.4057, "step": 360 }, { "epoch": 0.07, "grad_norm": 25.25, "learning_rate": 4.999783052943544e-06, "loss": 0.4194, "step": 361 }, { "epoch": 0.07, "grad_norm": 12.875, "learning_rate": 4.999772845205264e-06, "loss": 0.3824, "step": 362 }, { "epoch": 0.07, "grad_norm": 19.5, "learning_rate": 4.99976240282071e-06, "loss": 0.3944, "step": 363 }, { "epoch": 0.07, "grad_norm": 14.625, "learning_rate": 4.999751725790864e-06, "loss": 0.3015, "step": 364 }, { "epoch": 0.07, "grad_norm": 12.6875, "learning_rate": 4.999740814116725e-06, "loss": 0.3054, "step": 365 }, { "epoch": 0.07, "grad_norm": 14.3125, "learning_rate": 4.9997296677993195e-06, "loss": 0.3647, "step": 366 }, { "epoch": 0.07, "grad_norm": 29.625, "learning_rate": 4.999718286839693e-06, "loss": 0.4288, "step": 367 }, { "epoch": 0.07, "grad_norm": 20.0, "learning_rate": 4.999706671238915e-06, "loss": 0.3659, "step": 368 }, { "epoch": 0.07, "grad_norm": 22.75, "learning_rate": 4.999694820998074e-06, "loss": 0.426, "step": 369 }, { "epoch": 0.07, "grad_norm": 10.875, "learning_rate": 4.999682736118283e-06, "loss": 0.3323, "step": 370 }, { "epoch": 0.07, "grad_norm": 15.0, "learning_rate": 4.999670416600677e-06, "loss": 0.4083, "step": 371 }, { "epoch": 0.07, "grad_norm": 17.375, "learning_rate": 4.999657862446412e-06, "loss": 0.3613, "step": 372 }, { "epoch": 0.07, "grad_norm": 19.625, "learning_rate": 4.9996450736566656e-06, "loss": 0.3451, "step": 373 }, { "epoch": 0.07, "grad_norm": 16.875, "learning_rate": 4.99963205023264e-06, "loss": 0.3668, "step": 374 }, { "epoch": 0.07, "grad_norm": 17.75, "learning_rate": 4.999618792175556e-06, "loss": 0.4158, "step": 375 }, { "epoch": 0.07, "grad_norm": 25.375, "learning_rate": 4.99960529948666e-06, "loss": 0.4149, "step": 376 }, { "epoch": 0.07, "grad_norm": 12.3125, "learning_rate": 4.999591572167215e-06, "loss": 0.3506, "step": 377 }, { "epoch": 0.07, "grad_norm": 27.5, "learning_rate": 4.999577610218514e-06, "loss": 0.4411, "step": 378 }, { "epoch": 0.07, "grad_norm": 18.0, "learning_rate": 4.999563413641864e-06, "loss": 0.3827, "step": 379 }, { "epoch": 0.07, "grad_norm": 20.5, "learning_rate": 4.9995489824386e-06, "loss": 0.4852, "step": 380 }, { "epoch": 0.07, "grad_norm": 12.0625, "learning_rate": 4.999534316610075e-06, "loss": 0.4315, "step": 381 }, { "epoch": 0.07, "grad_norm": 21.0, "learning_rate": 4.999519416157666e-06, "loss": 0.3881, "step": 382 }, { "epoch": 0.07, "grad_norm": 12.0625, "learning_rate": 4.999504281082773e-06, "loss": 0.3562, "step": 383 }, { "epoch": 0.07, "grad_norm": 28.0, "learning_rate": 4.999488911386815e-06, "loss": 0.2601, "step": 384 }, { "epoch": 0.07, "grad_norm": 18.875, "learning_rate": 4.9994733070712355e-06, "loss": 0.4264, "step": 385 }, { "epoch": 0.07, "grad_norm": 20.875, "learning_rate": 4.9994574681374985e-06, "loss": 0.3586, "step": 386 }, { "epoch": 0.07, "grad_norm": 19.625, "learning_rate": 4.999441394587091e-06, "loss": 0.4381, "step": 387 }, { "epoch": 0.07, "grad_norm": 13.75, "learning_rate": 4.999425086421523e-06, "loss": 0.3465, "step": 388 }, { "epoch": 0.07, "grad_norm": 20.5, "learning_rate": 4.999408543642325e-06, "loss": 0.3614, "step": 389 }, { "epoch": 0.07, "grad_norm": 23.875, "learning_rate": 4.999391766251047e-06, "loss": 0.4537, "step": 390 }, { "epoch": 0.07, "grad_norm": 11.8125, "learning_rate": 4.999374754249268e-06, "loss": 0.3396, "step": 391 }, { "epoch": 0.07, "grad_norm": 13.9375, "learning_rate": 4.999357507638582e-06, "loss": 0.3269, "step": 392 }, { "epoch": 0.07, "grad_norm": 12.25, "learning_rate": 4.9993400264206085e-06, "loss": 0.3576, "step": 393 }, { "epoch": 0.07, "grad_norm": 16.625, "learning_rate": 4.999322310596989e-06, "loss": 0.4011, "step": 394 }, { "epoch": 0.07, "grad_norm": 14.375, "learning_rate": 4.9993043601693866e-06, "loss": 0.3592, "step": 395 }, { "epoch": 0.07, "grad_norm": 13.9375, "learning_rate": 4.999286175139485e-06, "loss": 0.3936, "step": 396 }, { "epoch": 0.08, "grad_norm": 17.0, "learning_rate": 4.999267755508993e-06, "loss": 0.3033, "step": 397 }, { "epoch": 0.08, "grad_norm": 25.25, "learning_rate": 4.999249101279637e-06, "loss": 0.3291, "step": 398 }, { "epoch": 0.08, "grad_norm": 14.875, "learning_rate": 4.999230212453171e-06, "loss": 0.4415, "step": 399 }, { "epoch": 0.08, "grad_norm": 12.625, "learning_rate": 4.999211089031366e-06, "loss": 0.3507, "step": 400 }, { "epoch": 0.08, "grad_norm": 16.375, "learning_rate": 4.999191731016018e-06, "loss": 0.3471, "step": 401 }, { "epoch": 0.08, "grad_norm": 17.25, "learning_rate": 4.999172138408944e-06, "loss": 0.3367, "step": 402 }, { "epoch": 0.08, "grad_norm": 23.625, "learning_rate": 4.9991523112119825e-06, "loss": 0.3976, "step": 403 }, { "epoch": 0.08, "grad_norm": 26.75, "learning_rate": 4.999132249426994e-06, "loss": 0.4122, "step": 404 }, { "epoch": 0.08, "grad_norm": 21.25, "learning_rate": 4.999111953055864e-06, "loss": 0.333, "step": 405 }, { "epoch": 0.08, "grad_norm": 13.125, "learning_rate": 4.999091422100495e-06, "loss": 0.3591, "step": 406 }, { "epoch": 0.08, "grad_norm": 17.5, "learning_rate": 4.9990706565628165e-06, "loss": 0.4385, "step": 407 }, { "epoch": 0.08, "grad_norm": 16.75, "learning_rate": 4.999049656444777e-06, "loss": 0.4209, "step": 408 }, { "epoch": 0.08, "grad_norm": 18.0, "learning_rate": 4.9990284217483455e-06, "loss": 0.3832, "step": 409 }, { "epoch": 0.08, "grad_norm": 15.6875, "learning_rate": 4.999006952475519e-06, "loss": 0.3687, "step": 410 }, { "epoch": 0.08, "grad_norm": 16.75, "learning_rate": 4.998985248628309e-06, "loss": 0.3986, "step": 411 }, { "epoch": 0.08, "grad_norm": 11.4375, "learning_rate": 4.998963310208756e-06, "loss": 0.3576, "step": 412 }, { "epoch": 0.08, "grad_norm": 14.625, "learning_rate": 4.998941137218918e-06, "loss": 0.3812, "step": 413 }, { "epoch": 0.08, "grad_norm": 15.4375, "learning_rate": 4.9989187296608765e-06, "loss": 0.3933, "step": 414 }, { "epoch": 0.08, "grad_norm": 15.5, "learning_rate": 4.998896087536734e-06, "loss": 0.429, "step": 415 }, { "epoch": 0.08, "grad_norm": 13.25, "learning_rate": 4.998873210848616e-06, "loss": 0.4397, "step": 416 }, { "epoch": 0.08, "grad_norm": 17.125, "learning_rate": 4.998850099598671e-06, "loss": 0.3933, "step": 417 }, { "epoch": 0.08, "grad_norm": 11.5625, "learning_rate": 4.998826753789067e-06, "loss": 0.2912, "step": 418 }, { "epoch": 0.08, "grad_norm": 53.75, "learning_rate": 4.998803173421997e-06, "loss": 0.3189, "step": 419 }, { "epoch": 0.08, "grad_norm": 13.25, "learning_rate": 4.9987793584996725e-06, "loss": 0.3884, "step": 420 }, { "epoch": 0.08, "grad_norm": 13.625, "learning_rate": 4.99875530902433e-06, "loss": 0.3546, "step": 421 }, { "epoch": 0.08, "grad_norm": 20.625, "learning_rate": 4.998731024998228e-06, "loss": 0.3842, "step": 422 }, { "epoch": 0.08, "grad_norm": 14.125, "learning_rate": 4.998706506423644e-06, "loss": 0.3934, "step": 423 }, { "epoch": 0.08, "grad_norm": 20.125, "learning_rate": 4.998681753302881e-06, "loss": 0.3537, "step": 424 }, { "epoch": 0.08, "grad_norm": 16.25, "learning_rate": 4.99865676563826e-06, "loss": 0.3595, "step": 425 }, { "epoch": 0.08, "grad_norm": 17.375, "learning_rate": 4.9986315434321304e-06, "loss": 0.3813, "step": 426 }, { "epoch": 0.08, "grad_norm": 12.1875, "learning_rate": 4.998606086686857e-06, "loss": 0.3775, "step": 427 }, { "epoch": 0.08, "grad_norm": 16.75, "learning_rate": 4.998580395404829e-06, "loss": 0.3875, "step": 428 }, { "epoch": 0.08, "grad_norm": 17.125, "learning_rate": 4.9985544695884615e-06, "loss": 0.3738, "step": 429 }, { "epoch": 0.08, "grad_norm": 22.875, "learning_rate": 4.9985283092401836e-06, "loss": 0.4331, "step": 430 }, { "epoch": 0.08, "grad_norm": 14.0, "learning_rate": 4.9985019143624535e-06, "loss": 0.3493, "step": 431 }, { "epoch": 0.08, "grad_norm": 16.875, "learning_rate": 4.998475284957748e-06, "loss": 0.361, "step": 432 }, { "epoch": 0.08, "grad_norm": 12.5625, "learning_rate": 4.998448421028567e-06, "loss": 0.328, "step": 433 }, { "epoch": 0.08, "grad_norm": 20.25, "learning_rate": 4.998421322577432e-06, "loss": 0.3858, "step": 434 }, { "epoch": 0.08, "grad_norm": 18.75, "learning_rate": 4.998393989606887e-06, "loss": 0.4, "step": 435 }, { "epoch": 0.08, "grad_norm": 19.625, "learning_rate": 4.9983664221194975e-06, "loss": 0.2882, "step": 436 }, { "epoch": 0.08, "grad_norm": 12.25, "learning_rate": 4.998338620117849e-06, "loss": 0.2701, "step": 437 }, { "epoch": 0.08, "grad_norm": 14.4375, "learning_rate": 4.998310583604556e-06, "loss": 0.4055, "step": 438 }, { "epoch": 0.08, "grad_norm": 26.375, "learning_rate": 4.9982823125822465e-06, "loss": 0.3863, "step": 439 }, { "epoch": 0.08, "grad_norm": 20.0, "learning_rate": 4.998253807053575e-06, "loss": 0.4652, "step": 440 }, { "epoch": 0.08, "grad_norm": 18.625, "learning_rate": 4.998225067021218e-06, "loss": 0.3723, "step": 441 }, { "epoch": 0.08, "grad_norm": 15.0625, "learning_rate": 4.998196092487871e-06, "loss": 0.3984, "step": 442 }, { "epoch": 0.08, "grad_norm": 17.375, "learning_rate": 4.998166883456257e-06, "loss": 0.3481, "step": 443 }, { "epoch": 0.08, "grad_norm": 12.75, "learning_rate": 4.998137439929115e-06, "loss": 0.4029, "step": 444 }, { "epoch": 0.08, "grad_norm": 14.625, "learning_rate": 4.998107761909211e-06, "loss": 0.4708, "step": 445 }, { "epoch": 0.08, "grad_norm": 28.25, "learning_rate": 4.998077849399329e-06, "loss": 0.33, "step": 446 }, { "epoch": 0.08, "grad_norm": 13.375, "learning_rate": 4.998047702402278e-06, "loss": 0.3246, "step": 447 }, { "epoch": 0.08, "grad_norm": 11.0625, "learning_rate": 4.998017320920886e-06, "loss": 0.3389, "step": 448 }, { "epoch": 0.08, "grad_norm": 15.0, "learning_rate": 4.997986704958008e-06, "loss": 0.3648, "step": 449 }, { "epoch": 0.09, "grad_norm": 11.5, "learning_rate": 4.997955854516515e-06, "loss": 0.3702, "step": 450 }, { "epoch": 0.09, "grad_norm": 24.75, "learning_rate": 4.997924769599304e-06, "loss": 0.3726, "step": 451 }, { "epoch": 0.09, "grad_norm": 14.5625, "learning_rate": 4.997893450209292e-06, "loss": 0.3104, "step": 452 }, { "epoch": 0.09, "grad_norm": 12.75, "learning_rate": 4.9978618963494205e-06, "loss": 0.3725, "step": 453 }, { "epoch": 0.09, "grad_norm": 21.375, "learning_rate": 4.997830108022651e-06, "loss": 0.3619, "step": 454 }, { "epoch": 0.09, "grad_norm": 15.0, "learning_rate": 4.9977980852319654e-06, "loss": 0.3957, "step": 455 }, { "epoch": 0.09, "grad_norm": 22.75, "learning_rate": 4.997765827980371e-06, "loss": 0.3492, "step": 456 }, { "epoch": 0.09, "grad_norm": 11.5, "learning_rate": 4.997733336270896e-06, "loss": 0.3183, "step": 457 }, { "epoch": 0.09, "grad_norm": 13.0, "learning_rate": 4.99770061010659e-06, "loss": 0.3854, "step": 458 }, { "epoch": 0.09, "grad_norm": 18.25, "learning_rate": 4.9976676494905256e-06, "loss": 0.3625, "step": 459 }, { "epoch": 0.09, "grad_norm": 18.25, "learning_rate": 4.997634454425794e-06, "loss": 0.324, "step": 460 }, { "epoch": 0.09, "grad_norm": 19.875, "learning_rate": 4.997601024915515e-06, "loss": 0.4757, "step": 461 }, { "epoch": 0.09, "grad_norm": 11.9375, "learning_rate": 4.997567360962824e-06, "loss": 0.3075, "step": 462 }, { "epoch": 0.09, "grad_norm": 13.125, "learning_rate": 4.99753346257088e-06, "loss": 0.3819, "step": 463 }, { "epoch": 0.09, "grad_norm": 11.9375, "learning_rate": 4.997499329742869e-06, "loss": 0.372, "step": 464 }, { "epoch": 0.09, "grad_norm": 11.0, "learning_rate": 4.997464962481991e-06, "loss": 0.371, "step": 465 }, { "epoch": 0.09, "grad_norm": 18.0, "learning_rate": 4.997430360791473e-06, "loss": 0.3122, "step": 466 }, { "epoch": 0.09, "grad_norm": 13.1875, "learning_rate": 4.997395524674564e-06, "loss": 0.338, "step": 467 }, { "epoch": 0.09, "grad_norm": 29.625, "learning_rate": 4.997360454134535e-06, "loss": 0.3877, "step": 468 }, { "epoch": 0.09, "grad_norm": 21.75, "learning_rate": 4.9973251491746735e-06, "loss": 0.3394, "step": 469 }, { "epoch": 0.09, "grad_norm": 14.75, "learning_rate": 4.997289609798298e-06, "loss": 0.355, "step": 470 }, { "epoch": 0.09, "grad_norm": 22.625, "learning_rate": 4.997253836008742e-06, "loss": 0.3539, "step": 471 }, { "epoch": 0.09, "grad_norm": 18.75, "learning_rate": 4.997217827809364e-06, "loss": 0.3849, "step": 472 }, { "epoch": 0.09, "grad_norm": 11.625, "learning_rate": 4.997181585203545e-06, "loss": 0.3825, "step": 473 }, { "epoch": 0.09, "grad_norm": 13.5625, "learning_rate": 4.997145108194685e-06, "loss": 0.2735, "step": 474 }, { "epoch": 0.09, "grad_norm": 18.25, "learning_rate": 4.997108396786211e-06, "loss": 0.3584, "step": 475 }, { "epoch": 0.09, "grad_norm": 12.0, "learning_rate": 4.997071450981565e-06, "loss": 0.3736, "step": 476 }, { "epoch": 0.09, "grad_norm": 14.875, "learning_rate": 4.997034270784219e-06, "loss": 0.3421, "step": 477 }, { "epoch": 0.09, "grad_norm": 13.8125, "learning_rate": 4.9969968561976604e-06, "loss": 0.4446, "step": 478 }, { "epoch": 0.09, "grad_norm": 10.9375, "learning_rate": 4.996959207225402e-06, "loss": 0.2661, "step": 479 }, { "epoch": 0.09, "grad_norm": 13.8125, "learning_rate": 4.996921323870978e-06, "loss": 0.3855, "step": 480 }, { "epoch": 0.09, "grad_norm": 13.375, "learning_rate": 4.996883206137943e-06, "loss": 0.417, "step": 481 }, { "epoch": 0.09, "grad_norm": 17.125, "learning_rate": 4.996844854029878e-06, "loss": 0.4135, "step": 482 }, { "epoch": 0.09, "grad_norm": 11.5, "learning_rate": 4.996806267550379e-06, "loss": 0.2814, "step": 483 }, { "epoch": 0.09, "grad_norm": 15.875, "learning_rate": 4.996767446703071e-06, "loss": 0.4692, "step": 484 }, { "epoch": 0.09, "grad_norm": 14.125, "learning_rate": 4.996728391491598e-06, "loss": 0.3589, "step": 485 }, { "epoch": 0.09, "grad_norm": 10.0, "learning_rate": 4.996689101919624e-06, "loss": 0.3169, "step": 486 }, { "epoch": 0.09, "grad_norm": 11.75, "learning_rate": 4.996649577990838e-06, "loss": 0.3424, "step": 487 }, { "epoch": 0.09, "grad_norm": 17.625, "learning_rate": 4.996609819708952e-06, "loss": 0.4506, "step": 488 }, { "epoch": 0.09, "grad_norm": 18.75, "learning_rate": 4.996569827077695e-06, "loss": 0.35, "step": 489 }, { "epoch": 0.09, "grad_norm": 18.25, "learning_rate": 4.996529600100822e-06, "loss": 0.3566, "step": 490 }, { "epoch": 0.09, "grad_norm": 12.1875, "learning_rate": 4.9964891387821084e-06, "loss": 0.3284, "step": 491 }, { "epoch": 0.09, "grad_norm": 12.0, "learning_rate": 4.996448443125354e-06, "loss": 0.3578, "step": 492 }, { "epoch": 0.09, "grad_norm": 23.125, "learning_rate": 4.996407513134377e-06, "loss": 0.4357, "step": 493 }, { "epoch": 0.09, "grad_norm": 12.125, "learning_rate": 4.99636634881302e-06, "loss": 0.3672, "step": 494 }, { "epoch": 0.09, "grad_norm": 15.5, "learning_rate": 4.996324950165148e-06, "loss": 0.3269, "step": 495 }, { "epoch": 0.09, "grad_norm": 11.6875, "learning_rate": 4.996283317194644e-06, "loss": 0.342, "step": 496 }, { "epoch": 0.09, "grad_norm": 15.0625, "learning_rate": 4.996241449905419e-06, "loss": 0.3577, "step": 497 }, { "epoch": 0.09, "grad_norm": 10.875, "learning_rate": 4.9961993483014026e-06, "loss": 0.3592, "step": 498 }, { "epoch": 0.09, "grad_norm": 14.1875, "learning_rate": 4.996157012386545e-06, "loss": 0.3732, "step": 499 }, { "epoch": 0.09, "grad_norm": 15.5625, "learning_rate": 4.9961144421648224e-06, "loss": 0.3256, "step": 500 }, { "epoch": 0.09, "grad_norm": 13.75, "learning_rate": 4.996071637640229e-06, "loss": 0.3776, "step": 501 }, { "epoch": 0.09, "grad_norm": 17.625, "learning_rate": 4.996028598816782e-06, "loss": 0.3351, "step": 502 }, { "epoch": 0.1, "grad_norm": 15.125, "learning_rate": 4.9959853256985245e-06, "loss": 0.3651, "step": 503 }, { "epoch": 0.1, "grad_norm": 16.625, "learning_rate": 4.995941818289516e-06, "loss": 0.4948, "step": 504 }, { "epoch": 0.1, "grad_norm": 16.875, "learning_rate": 4.99589807659384e-06, "loss": 0.3708, "step": 505 }, { "epoch": 0.1, "grad_norm": 15.8125, "learning_rate": 4.995854100615604e-06, "loss": 0.4067, "step": 506 }, { "epoch": 0.1, "grad_norm": 15.125, "learning_rate": 4.995809890358935e-06, "loss": 0.3973, "step": 507 }, { "epoch": 0.1, "grad_norm": 14.25, "learning_rate": 4.995765445827984e-06, "loss": 0.3555, "step": 508 }, { "epoch": 0.1, "grad_norm": 17.375, "learning_rate": 4.99572076702692e-06, "loss": 0.3218, "step": 509 }, { "epoch": 0.1, "grad_norm": 11.9375, "learning_rate": 4.995675853959941e-06, "loss": 0.3127, "step": 510 }, { "epoch": 0.1, "grad_norm": 10.5625, "learning_rate": 4.995630706631259e-06, "loss": 0.3286, "step": 511 }, { "epoch": 0.1, "grad_norm": 11.875, "learning_rate": 4.995585325045115e-06, "loss": 0.3827, "step": 512 }, { "epoch": 0.1, "grad_norm": 16.125, "learning_rate": 4.995539709205767e-06, "loss": 0.2958, "step": 513 }, { "epoch": 0.1, "grad_norm": 12.125, "learning_rate": 4.995493859117496e-06, "loss": 0.2619, "step": 514 }, { "epoch": 0.1, "grad_norm": 14.0625, "learning_rate": 4.995447774784609e-06, "loss": 0.3395, "step": 515 }, { "epoch": 0.1, "grad_norm": 12.125, "learning_rate": 4.995401456211429e-06, "loss": 0.3106, "step": 516 }, { "epoch": 0.1, "grad_norm": 11.0, "learning_rate": 4.995354903402304e-06, "loss": 0.3268, "step": 517 }, { "epoch": 0.1, "grad_norm": 16.625, "learning_rate": 4.995308116361605e-06, "loss": 0.3433, "step": 518 }, { "epoch": 0.1, "grad_norm": 16.75, "learning_rate": 4.995261095093723e-06, "loss": 0.4043, "step": 519 }, { "epoch": 0.1, "grad_norm": 19.5, "learning_rate": 4.995213839603071e-06, "loss": 0.4076, "step": 520 }, { "epoch": 0.1, "grad_norm": 12.0, "learning_rate": 4.995166349894087e-06, "loss": 0.3396, "step": 521 }, { "epoch": 0.1, "grad_norm": 15.125, "learning_rate": 4.995118625971227e-06, "loss": 0.2927, "step": 522 }, { "epoch": 0.1, "grad_norm": 15.1875, "learning_rate": 4.995070667838971e-06, "loss": 0.3422, "step": 523 }, { "epoch": 0.1, "grad_norm": 12.125, "learning_rate": 4.99502247550182e-06, "loss": 0.3879, "step": 524 }, { "epoch": 0.1, "grad_norm": 13.75, "learning_rate": 4.9949740489643e-06, "loss": 0.3282, "step": 525 }, { "epoch": 0.1, "grad_norm": 17.875, "learning_rate": 4.994925388230954e-06, "loss": 0.3885, "step": 526 }, { "epoch": 0.1, "grad_norm": 13.4375, "learning_rate": 4.994876493306351e-06, "loss": 0.3487, "step": 527 }, { "epoch": 0.1, "grad_norm": 14.0625, "learning_rate": 4.99482736419508e-06, "loss": 0.3939, "step": 528 }, { "epoch": 0.1, "grad_norm": 31.75, "learning_rate": 4.994778000901754e-06, "loss": 0.3237, "step": 529 }, { "epoch": 0.1, "grad_norm": 19.625, "learning_rate": 4.994728403431004e-06, "loss": 0.3247, "step": 530 }, { "epoch": 0.1, "grad_norm": 14.125, "learning_rate": 4.994678571787488e-06, "loss": 0.3588, "step": 531 }, { "epoch": 0.1, "grad_norm": 12.3125, "learning_rate": 4.994628505975882e-06, "loss": 0.3154, "step": 532 }, { "epoch": 0.1, "grad_norm": 18.5, "learning_rate": 4.994578206000886e-06, "loss": 0.3904, "step": 533 }, { "epoch": 0.1, "grad_norm": 64.5, "learning_rate": 4.994527671867223e-06, "loss": 0.3605, "step": 534 }, { "epoch": 0.1, "grad_norm": 17.625, "learning_rate": 4.994476903579634e-06, "loss": 0.3813, "step": 535 }, { "epoch": 0.1, "grad_norm": 21.625, "learning_rate": 4.994425901142887e-06, "loss": 0.4391, "step": 536 }, { "epoch": 0.1, "grad_norm": 14.9375, "learning_rate": 4.994374664561766e-06, "loss": 0.3995, "step": 537 }, { "epoch": 0.1, "grad_norm": 11.3125, "learning_rate": 4.994323193841084e-06, "loss": 0.3816, "step": 538 }, { "epoch": 0.1, "grad_norm": 13.25, "learning_rate": 4.99427148898567e-06, "loss": 0.3969, "step": 539 }, { "epoch": 0.1, "grad_norm": 14.375, "learning_rate": 4.994219550000379e-06, "loss": 0.437, "step": 540 }, { "epoch": 0.1, "grad_norm": 17.0, "learning_rate": 4.994167376890086e-06, "loss": 0.3988, "step": 541 }, { "epoch": 0.1, "grad_norm": 10.3125, "learning_rate": 4.994114969659687e-06, "loss": 0.3556, "step": 542 }, { "epoch": 0.1, "grad_norm": 12.25, "learning_rate": 4.994062328314103e-06, "loss": 0.446, "step": 543 }, { "epoch": 0.1, "grad_norm": 10.75, "learning_rate": 4.994009452858274e-06, "loss": 0.2753, "step": 544 }, { "epoch": 0.1, "grad_norm": 11.875, "learning_rate": 4.993956343297164e-06, "loss": 0.3203, "step": 545 }, { "epoch": 0.1, "grad_norm": 13.875, "learning_rate": 4.9939029996357585e-06, "loss": 0.3805, "step": 546 }, { "epoch": 0.1, "grad_norm": 17.0, "learning_rate": 4.993849421879064e-06, "loss": 0.4285, "step": 547 }, { "epoch": 0.1, "grad_norm": 12.8125, "learning_rate": 4.99379561003211e-06, "loss": 0.3592, "step": 548 }, { "epoch": 0.1, "grad_norm": 18.5, "learning_rate": 4.993741564099948e-06, "loss": 0.3801, "step": 549 }, { "epoch": 0.1, "grad_norm": 10.0, "learning_rate": 4.99368728408765e-06, "loss": 0.3051, "step": 550 }, { "epoch": 0.1, "grad_norm": 11.9375, "learning_rate": 4.9936327700003126e-06, "loss": 0.3886, "step": 551 }, { "epoch": 0.1, "grad_norm": 11.6875, "learning_rate": 4.993578021843052e-06, "loss": 0.2922, "step": 552 }, { "epoch": 0.1, "grad_norm": 13.6875, "learning_rate": 4.993523039621007e-06, "loss": 0.3161, "step": 553 }, { "epoch": 0.1, "grad_norm": 11.25, "learning_rate": 4.99346782333934e-06, "loss": 0.3679, "step": 554 }, { "epoch": 0.1, "grad_norm": 9.5, "learning_rate": 4.993412373003232e-06, "loss": 0.3168, "step": 555 }, { "epoch": 0.11, "grad_norm": 9.1875, "learning_rate": 4.993356688617889e-06, "loss": 0.293, "step": 556 }, { "epoch": 0.11, "grad_norm": 14.4375, "learning_rate": 4.993300770188539e-06, "loss": 0.3598, "step": 557 }, { "epoch": 0.11, "grad_norm": 10.0, "learning_rate": 4.993244617720429e-06, "loss": 0.3444, "step": 558 }, { "epoch": 0.11, "grad_norm": 14.8125, "learning_rate": 4.993188231218829e-06, "loss": 0.3492, "step": 559 }, { "epoch": 0.11, "grad_norm": 23.375, "learning_rate": 4.993131610689036e-06, "loss": 0.3199, "step": 560 }, { "epoch": 0.11, "grad_norm": 12.0, "learning_rate": 4.993074756136361e-06, "loss": 0.3614, "step": 561 }, { "epoch": 0.11, "grad_norm": 12.0625, "learning_rate": 4.9930176675661425e-06, "loss": 0.3706, "step": 562 }, { "epoch": 0.11, "grad_norm": 18.375, "learning_rate": 4.9929603449837394e-06, "loss": 0.2779, "step": 563 }, { "epoch": 0.11, "grad_norm": 12.0625, "learning_rate": 4.992902788394531e-06, "loss": 0.306, "step": 564 }, { "epoch": 0.11, "grad_norm": 12.0625, "learning_rate": 4.992844997803921e-06, "loss": 0.3775, "step": 565 }, { "epoch": 0.11, "grad_norm": 11.375, "learning_rate": 4.992786973217333e-06, "loss": 0.3394, "step": 566 }, { "epoch": 0.11, "grad_norm": 15.3125, "learning_rate": 4.992728714640216e-06, "loss": 0.337, "step": 567 }, { "epoch": 0.11, "grad_norm": 10.875, "learning_rate": 4.992670222078036e-06, "loss": 0.2771, "step": 568 }, { "epoch": 0.11, "grad_norm": 21.875, "learning_rate": 4.992611495536284e-06, "loss": 0.327, "step": 569 }, { "epoch": 0.11, "grad_norm": 13.0, "learning_rate": 4.992552535020474e-06, "loss": 0.3363, "step": 570 }, { "epoch": 0.11, "grad_norm": 12.25, "learning_rate": 4.992493340536139e-06, "loss": 0.3731, "step": 571 }, { "epoch": 0.11, "grad_norm": 12.5625, "learning_rate": 4.992433912088835e-06, "loss": 0.3433, "step": 572 }, { "epoch": 0.11, "grad_norm": 14.375, "learning_rate": 4.9923742496841425e-06, "loss": 0.3662, "step": 573 }, { "epoch": 0.11, "grad_norm": 16.75, "learning_rate": 4.99231435332766e-06, "loss": 0.3528, "step": 574 }, { "epoch": 0.11, "grad_norm": 16.5, "learning_rate": 4.99225422302501e-06, "loss": 0.4044, "step": 575 }, { "epoch": 0.11, "grad_norm": 19.5, "learning_rate": 4.9921938587818374e-06, "loss": 0.345, "step": 576 }, { "epoch": 0.11, "grad_norm": 14.1875, "learning_rate": 4.992133260603807e-06, "loss": 0.3512, "step": 577 }, { "epoch": 0.11, "grad_norm": 14.625, "learning_rate": 4.992072428496611e-06, "loss": 0.3903, "step": 578 }, { "epoch": 0.11, "grad_norm": 15.6875, "learning_rate": 4.992011362465954e-06, "loss": 0.3514, "step": 579 }, { "epoch": 0.11, "grad_norm": 14.75, "learning_rate": 4.991950062517571e-06, "loss": 0.4789, "step": 580 }, { "epoch": 0.11, "grad_norm": 13.625, "learning_rate": 4.991888528657216e-06, "loss": 0.3722, "step": 581 }, { "epoch": 0.11, "grad_norm": 17.25, "learning_rate": 4.991826760890665e-06, "loss": 0.377, "step": 582 }, { "epoch": 0.11, "grad_norm": 12.5, "learning_rate": 4.991764759223715e-06, "loss": 0.3963, "step": 583 }, { "epoch": 0.11, "grad_norm": 13.1875, "learning_rate": 4.9917025236621874e-06, "loss": 0.3859, "step": 584 }, { "epoch": 0.11, "grad_norm": 14.0, "learning_rate": 4.991640054211923e-06, "loss": 0.3871, "step": 585 }, { "epoch": 0.11, "grad_norm": 10.5, "learning_rate": 4.991577350878786e-06, "loss": 0.3647, "step": 586 }, { "epoch": 0.11, "grad_norm": 11.125, "learning_rate": 4.9915144136686614e-06, "loss": 0.3218, "step": 587 }, { "epoch": 0.11, "grad_norm": 9.5, "learning_rate": 4.991451242587458e-06, "loss": 0.4118, "step": 588 }, { "epoch": 0.11, "grad_norm": 10.375, "learning_rate": 4.991387837641106e-06, "loss": 0.2578, "step": 589 }, { "epoch": 0.11, "grad_norm": 11.0625, "learning_rate": 4.991324198835555e-06, "loss": 0.3569, "step": 590 }, { "epoch": 0.11, "grad_norm": 26.0, "learning_rate": 4.99126032617678e-06, "loss": 0.4393, "step": 591 }, { "epoch": 0.11, "grad_norm": 18.625, "learning_rate": 4.991196219670776e-06, "loss": 0.3691, "step": 592 }, { "epoch": 0.11, "grad_norm": 21.75, "learning_rate": 4.991131879323561e-06, "loss": 0.4472, "step": 593 }, { "epoch": 0.11, "grad_norm": 16.125, "learning_rate": 4.991067305141174e-06, "loss": 0.2585, "step": 594 }, { "epoch": 0.11, "grad_norm": 10.4375, "learning_rate": 4.991002497129677e-06, "loss": 0.3916, "step": 595 }, { "epoch": 0.11, "grad_norm": 20.875, "learning_rate": 4.990937455295152e-06, "loss": 0.4249, "step": 596 }, { "epoch": 0.11, "grad_norm": 12.0, "learning_rate": 4.990872179643705e-06, "loss": 0.3467, "step": 597 }, { "epoch": 0.11, "grad_norm": 14.5625, "learning_rate": 4.9908066701814645e-06, "loss": 0.3603, "step": 598 }, { "epoch": 0.11, "grad_norm": 13.25, "learning_rate": 4.990740926914578e-06, "loss": 0.3236, "step": 599 }, { "epoch": 0.11, "grad_norm": 14.0625, "learning_rate": 4.990674949849217e-06, "loss": 0.4536, "step": 600 }, { "epoch": 0.11, "grad_norm": 14.5, "learning_rate": 4.990608738991575e-06, "loss": 0.3333, "step": 601 }, { "epoch": 0.11, "grad_norm": 13.25, "learning_rate": 4.990542294347867e-06, "loss": 0.3805, "step": 602 }, { "epoch": 0.11, "grad_norm": 11.25, "learning_rate": 4.99047561592433e-06, "loss": 0.2805, "step": 603 }, { "epoch": 0.11, "grad_norm": 16.625, "learning_rate": 4.990408703727222e-06, "loss": 0.4181, "step": 604 }, { "epoch": 0.11, "grad_norm": 18.0, "learning_rate": 4.9903415577628245e-06, "loss": 0.422, "step": 605 }, { "epoch": 0.11, "grad_norm": 10.5625, "learning_rate": 4.990274178037441e-06, "loss": 0.2674, "step": 606 }, { "epoch": 0.11, "grad_norm": 13.375, "learning_rate": 4.990206564557395e-06, "loss": 0.2841, "step": 607 }, { "epoch": 0.12, "grad_norm": 11.5625, "learning_rate": 4.990138717329034e-06, "loss": 0.3446, "step": 608 }, { "epoch": 0.12, "grad_norm": 15.0625, "learning_rate": 4.990070636358726e-06, "loss": 0.4293, "step": 609 }, { "epoch": 0.12, "grad_norm": 12.125, "learning_rate": 4.990002321652862e-06, "loss": 0.3318, "step": 610 }, { "epoch": 0.12, "grad_norm": 14.4375, "learning_rate": 4.989933773217854e-06, "loss": 0.4894, "step": 611 }, { "epoch": 0.12, "grad_norm": 15.125, "learning_rate": 4.989864991060137e-06, "loss": 0.4268, "step": 612 }, { "epoch": 0.12, "grad_norm": 10.3125, "learning_rate": 4.989795975186168e-06, "loss": 0.3934, "step": 613 }, { "epoch": 0.12, "grad_norm": 13.0, "learning_rate": 4.9897267256024235e-06, "loss": 0.451, "step": 614 }, { "epoch": 0.12, "grad_norm": 18.75, "learning_rate": 4.989657242315406e-06, "loss": 0.3972, "step": 615 }, { "epoch": 0.12, "grad_norm": 12.0625, "learning_rate": 4.9895875253316354e-06, "loss": 0.3818, "step": 616 }, { "epoch": 0.12, "grad_norm": 8.1875, "learning_rate": 4.9895175746576566e-06, "loss": 0.3396, "step": 617 }, { "epoch": 0.12, "grad_norm": 13.8125, "learning_rate": 4.989447390300036e-06, "loss": 0.3889, "step": 618 }, { "epoch": 0.12, "grad_norm": 11.75, "learning_rate": 4.989376972265363e-06, "loss": 0.2871, "step": 619 }, { "epoch": 0.12, "grad_norm": 14.25, "learning_rate": 4.989306320560245e-06, "loss": 0.3345, "step": 620 }, { "epoch": 0.12, "grad_norm": 15.75, "learning_rate": 4.9892354351913154e-06, "loss": 0.4411, "step": 621 }, { "epoch": 0.12, "grad_norm": 11.3125, "learning_rate": 4.989164316165227e-06, "loss": 0.2941, "step": 622 }, { "epoch": 0.12, "grad_norm": 13.0, "learning_rate": 4.989092963488656e-06, "loss": 0.415, "step": 623 }, { "epoch": 0.12, "grad_norm": 16.125, "learning_rate": 4.9890213771683e-06, "loss": 0.3778, "step": 624 }, { "epoch": 0.12, "grad_norm": 11.875, "learning_rate": 4.988949557210879e-06, "loss": 0.3106, "step": 625 }, { "epoch": 0.12, "grad_norm": 10.9375, "learning_rate": 4.988877503623135e-06, "loss": 0.2892, "step": 626 }, { "epoch": 0.12, "grad_norm": 15.5625, "learning_rate": 4.988805216411831e-06, "loss": 0.4058, "step": 627 }, { "epoch": 0.12, "grad_norm": 12.625, "learning_rate": 4.9887326955837504e-06, "loss": 0.3149, "step": 628 }, { "epoch": 0.12, "grad_norm": 11.6875, "learning_rate": 4.988659941145704e-06, "loss": 0.307, "step": 629 }, { "epoch": 0.12, "grad_norm": 11.5, "learning_rate": 4.9885869531045175e-06, "loss": 0.3421, "step": 630 }, { "epoch": 0.12, "grad_norm": 18.25, "learning_rate": 4.988513731467045e-06, "loss": 0.3951, "step": 631 }, { "epoch": 0.12, "grad_norm": 15.4375, "learning_rate": 4.988440276240159e-06, "loss": 0.3743, "step": 632 }, { "epoch": 0.12, "grad_norm": 10.9375, "learning_rate": 4.988366587430752e-06, "loss": 0.3602, "step": 633 }, { "epoch": 0.12, "grad_norm": 12.875, "learning_rate": 4.988292665045744e-06, "loss": 0.3257, "step": 634 }, { "epoch": 0.12, "grad_norm": 12.125, "learning_rate": 4.988218509092073e-06, "loss": 0.2919, "step": 635 }, { "epoch": 0.12, "grad_norm": 13.0625, "learning_rate": 4.9881441195767e-06, "loss": 0.322, "step": 636 }, { "epoch": 0.12, "grad_norm": 10.375, "learning_rate": 4.988069496506607e-06, "loss": 0.2828, "step": 637 }, { "epoch": 0.12, "grad_norm": 13.5, "learning_rate": 4.987994639888798e-06, "loss": 0.4011, "step": 638 }, { "epoch": 0.12, "grad_norm": 15.75, "learning_rate": 4.9879195497303014e-06, "loss": 0.3633, "step": 639 }, { "epoch": 0.12, "grad_norm": 18.125, "learning_rate": 4.987844226038164e-06, "loss": 0.3175, "step": 640 }, { "epoch": 0.12, "grad_norm": 18.5, "learning_rate": 4.987768668819458e-06, "loss": 0.3318, "step": 641 }, { "epoch": 0.12, "grad_norm": 14.6875, "learning_rate": 4.987692878081275e-06, "loss": 0.2896, "step": 642 }, { "epoch": 0.12, "grad_norm": 15.1875, "learning_rate": 4.987616853830728e-06, "loss": 0.4224, "step": 643 }, { "epoch": 0.12, "grad_norm": 10.25, "learning_rate": 4.987540596074954e-06, "loss": 0.2842, "step": 644 }, { "epoch": 0.12, "grad_norm": 14.9375, "learning_rate": 4.987464104821112e-06, "loss": 0.413, "step": 645 }, { "epoch": 0.12, "grad_norm": 16.25, "learning_rate": 4.9873873800763815e-06, "loss": 0.3981, "step": 646 }, { "epoch": 0.12, "grad_norm": 9.625, "learning_rate": 4.987310421847965e-06, "loss": 0.3083, "step": 647 }, { "epoch": 0.12, "grad_norm": 11.75, "learning_rate": 4.987233230143084e-06, "loss": 0.4365, "step": 648 }, { "epoch": 0.12, "grad_norm": 15.3125, "learning_rate": 4.987155804968986e-06, "loss": 0.332, "step": 649 }, { "epoch": 0.12, "grad_norm": 13.1875, "learning_rate": 4.987078146332938e-06, "loss": 0.3055, "step": 650 }, { "epoch": 0.12, "grad_norm": 10.8125, "learning_rate": 4.987000254242232e-06, "loss": 0.3638, "step": 651 }, { "epoch": 0.12, "grad_norm": 14.75, "learning_rate": 4.986922128704176e-06, "loss": 0.3219, "step": 652 }, { "epoch": 0.12, "grad_norm": 8.5, "learning_rate": 4.9868437697261056e-06, "loss": 0.2493, "step": 653 }, { "epoch": 0.12, "grad_norm": 13.375, "learning_rate": 4.986765177315375e-06, "loss": 0.2912, "step": 654 }, { "epoch": 0.12, "grad_norm": 13.625, "learning_rate": 4.986686351479362e-06, "loss": 0.2614, "step": 655 }, { "epoch": 0.12, "grad_norm": 12.0625, "learning_rate": 4.986607292225466e-06, "loss": 0.3851, "step": 656 }, { "epoch": 0.12, "grad_norm": 15.8125, "learning_rate": 4.9865279995611086e-06, "loss": 0.4166, "step": 657 }, { "epoch": 0.12, "grad_norm": 9.625, "learning_rate": 4.98644847349373e-06, "loss": 0.2997, "step": 658 }, { "epoch": 0.12, "grad_norm": 12.0, "learning_rate": 4.986368714030798e-06, "loss": 0.4279, "step": 659 }, { "epoch": 0.12, "grad_norm": 11.8125, "learning_rate": 4.986288721179798e-06, "loss": 0.3673, "step": 660 }, { "epoch": 0.13, "grad_norm": 15.5625, "learning_rate": 4.98620849494824e-06, "loss": 0.3923, "step": 661 }, { "epoch": 0.13, "grad_norm": 20.5, "learning_rate": 4.986128035343653e-06, "loss": 0.338, "step": 662 }, { "epoch": 0.13, "grad_norm": 20.5, "learning_rate": 4.98604734237359e-06, "loss": 0.3881, "step": 663 }, { "epoch": 0.13, "grad_norm": 13.5, "learning_rate": 4.985966416045626e-06, "loss": 0.3357, "step": 664 }, { "epoch": 0.13, "grad_norm": 10.0, "learning_rate": 4.985885256367356e-06, "loss": 0.3956, "step": 665 }, { "epoch": 0.13, "grad_norm": 10.3125, "learning_rate": 4.9858038633464e-06, "loss": 0.2903, "step": 666 }, { "epoch": 0.13, "grad_norm": 16.125, "learning_rate": 4.985722236990398e-06, "loss": 0.3326, "step": 667 }, { "epoch": 0.13, "grad_norm": 12.1875, "learning_rate": 4.985640377307009e-06, "loss": 0.3964, "step": 668 }, { "epoch": 0.13, "grad_norm": 13.6875, "learning_rate": 4.985558284303921e-06, "loss": 0.3688, "step": 669 }, { "epoch": 0.13, "grad_norm": 11.875, "learning_rate": 4.985475957988838e-06, "loss": 0.3419, "step": 670 }, { "epoch": 0.13, "grad_norm": 14.0625, "learning_rate": 4.9853933983694865e-06, "loss": 0.3418, "step": 671 }, { "epoch": 0.13, "grad_norm": 15.5, "learning_rate": 4.985310605453618e-06, "loss": 0.315, "step": 672 }, { "epoch": 0.13, "grad_norm": 16.5, "learning_rate": 4.985227579249004e-06, "loss": 0.3984, "step": 673 }, { "epoch": 0.13, "grad_norm": 44.0, "learning_rate": 4.985144319763436e-06, "loss": 0.3807, "step": 674 }, { "epoch": 0.13, "grad_norm": 13.5, "learning_rate": 4.985060827004732e-06, "loss": 0.3411, "step": 675 }, { "epoch": 0.13, "grad_norm": 17.5, "learning_rate": 4.984977100980727e-06, "loss": 0.2852, "step": 676 }, { "epoch": 0.13, "grad_norm": 14.3125, "learning_rate": 4.9848931416992815e-06, "loss": 0.3265, "step": 677 }, { "epoch": 0.13, "grad_norm": 13.9375, "learning_rate": 4.984808949168276e-06, "loss": 0.328, "step": 678 }, { "epoch": 0.13, "grad_norm": 18.625, "learning_rate": 4.9847245233956135e-06, "loss": 0.3738, "step": 679 }, { "epoch": 0.13, "grad_norm": 9.6875, "learning_rate": 4.984639864389218e-06, "loss": 0.2832, "step": 680 }, { "epoch": 0.13, "grad_norm": 10.0625, "learning_rate": 4.984554972157038e-06, "loss": 0.3306, "step": 681 }, { "epoch": 0.13, "grad_norm": 17.75, "learning_rate": 4.984469846707039e-06, "loss": 0.4315, "step": 682 }, { "epoch": 0.13, "grad_norm": 11.0625, "learning_rate": 4.984384488047215e-06, "loss": 0.3945, "step": 683 }, { "epoch": 0.13, "grad_norm": 13.125, "learning_rate": 4.984298896185576e-06, "loss": 0.443, "step": 684 }, { "epoch": 0.13, "grad_norm": 17.375, "learning_rate": 4.984213071130157e-06, "loss": 0.3595, "step": 685 }, { "epoch": 0.13, "grad_norm": 12.0, "learning_rate": 4.984127012889015e-06, "loss": 0.3527, "step": 686 }, { "epoch": 0.13, "grad_norm": 16.875, "learning_rate": 4.984040721470227e-06, "loss": 0.3535, "step": 687 }, { "epoch": 0.13, "grad_norm": 9.375, "learning_rate": 4.983954196881893e-06, "loss": 0.3096, "step": 688 }, { "epoch": 0.13, "grad_norm": 16.625, "learning_rate": 4.983867439132135e-06, "loss": 0.3951, "step": 689 }, { "epoch": 0.13, "grad_norm": 15.6875, "learning_rate": 4.983780448229096e-06, "loss": 0.4149, "step": 690 }, { "epoch": 0.13, "grad_norm": 11.5, "learning_rate": 4.9836932241809426e-06, "loss": 0.3837, "step": 691 }, { "epoch": 0.13, "grad_norm": 13.375, "learning_rate": 4.983605766995862e-06, "loss": 0.3896, "step": 692 }, { "epoch": 0.13, "grad_norm": 16.25, "learning_rate": 4.983518076682064e-06, "loss": 0.4233, "step": 693 }, { "epoch": 0.13, "grad_norm": 8.4375, "learning_rate": 4.983430153247778e-06, "loss": 0.3598, "step": 694 }, { "epoch": 0.13, "grad_norm": 14.1875, "learning_rate": 4.98334199670126e-06, "loss": 0.3427, "step": 695 }, { "epoch": 0.13, "grad_norm": 13.0625, "learning_rate": 4.983253607050782e-06, "loss": 0.3975, "step": 696 }, { "epoch": 0.13, "grad_norm": 13.8125, "learning_rate": 4.983164984304642e-06, "loss": 0.3602, "step": 697 }, { "epoch": 0.13, "grad_norm": 18.125, "learning_rate": 4.983076128471161e-06, "loss": 0.3974, "step": 698 }, { "epoch": 0.13, "grad_norm": 8.875, "learning_rate": 4.9829870395586755e-06, "loss": 0.3114, "step": 699 }, { "epoch": 0.13, "grad_norm": 15.125, "learning_rate": 4.982897717575551e-06, "loss": 0.2818, "step": 700 }, { "epoch": 0.13, "grad_norm": 14.0625, "learning_rate": 4.98280816253017e-06, "loss": 0.3996, "step": 701 }, { "epoch": 0.13, "grad_norm": 13.25, "learning_rate": 4.982718374430941e-06, "loss": 0.2481, "step": 702 }, { "epoch": 0.13, "grad_norm": 16.375, "learning_rate": 4.982628353286291e-06, "loss": 0.3907, "step": 703 }, { "epoch": 0.13, "grad_norm": 11.625, "learning_rate": 4.98253809910467e-06, "loss": 0.3996, "step": 704 }, { "epoch": 0.13, "grad_norm": 17.25, "learning_rate": 4.98244761189455e-06, "loss": 0.408, "step": 705 }, { "epoch": 0.13, "grad_norm": 10.4375, "learning_rate": 4.982356891664424e-06, "loss": 0.3267, "step": 706 }, { "epoch": 0.13, "grad_norm": 11.6875, "learning_rate": 4.982265938422808e-06, "loss": 0.3146, "step": 707 }, { "epoch": 0.13, "grad_norm": 14.9375, "learning_rate": 4.9821747521782405e-06, "loss": 0.386, "step": 708 }, { "epoch": 0.13, "grad_norm": 12.4375, "learning_rate": 4.982083332939279e-06, "loss": 0.3953, "step": 709 }, { "epoch": 0.13, "grad_norm": 18.375, "learning_rate": 4.981991680714507e-06, "loss": 0.4194, "step": 710 }, { "epoch": 0.13, "grad_norm": 12.6875, "learning_rate": 4.981899795512527e-06, "loss": 0.3658, "step": 711 }, { "epoch": 0.13, "grad_norm": 14.625, "learning_rate": 4.9818076773419625e-06, "loss": 0.382, "step": 712 }, { "epoch": 0.13, "grad_norm": 32.25, "learning_rate": 4.981715326211462e-06, "loss": 0.3598, "step": 713 }, { "epoch": 0.14, "grad_norm": 10.8125, "learning_rate": 4.981622742129692e-06, "loss": 0.3402, "step": 714 }, { "epoch": 0.14, "grad_norm": 12.625, "learning_rate": 4.981529925105347e-06, "loss": 0.38, "step": 715 }, { "epoch": 0.14, "grad_norm": 10.625, "learning_rate": 4.981436875147136e-06, "loss": 0.4021, "step": 716 }, { "epoch": 0.14, "grad_norm": 12.875, "learning_rate": 4.981343592263795e-06, "loss": 0.3386, "step": 717 }, { "epoch": 0.14, "grad_norm": 11.9375, "learning_rate": 4.981250076464079e-06, "loss": 0.3366, "step": 718 }, { "epoch": 0.14, "grad_norm": 14.3125, "learning_rate": 4.981156327756767e-06, "loss": 0.3925, "step": 719 }, { "epoch": 0.14, "grad_norm": 11.9375, "learning_rate": 4.9810623461506576e-06, "loss": 0.3125, "step": 720 }, { "epoch": 0.14, "grad_norm": 13.9375, "learning_rate": 4.980968131654575e-06, "loss": 0.3709, "step": 721 }, { "epoch": 0.14, "grad_norm": 13.4375, "learning_rate": 4.980873684277362e-06, "loss": 0.3515, "step": 722 }, { "epoch": 0.14, "grad_norm": 11.4375, "learning_rate": 4.980779004027882e-06, "loss": 0.3643, "step": 723 }, { "epoch": 0.14, "grad_norm": 12.0625, "learning_rate": 4.980684090915024e-06, "loss": 0.298, "step": 724 }, { "epoch": 0.14, "grad_norm": 14.1875, "learning_rate": 4.9805889449476985e-06, "loss": 0.3173, "step": 725 }, { "epoch": 0.14, "grad_norm": 11.75, "learning_rate": 4.980493566134834e-06, "loss": 0.4124, "step": 726 }, { "epoch": 0.14, "grad_norm": 9.625, "learning_rate": 4.980397954485385e-06, "loss": 0.3337, "step": 727 }, { "epoch": 0.14, "grad_norm": 13.3125, "learning_rate": 4.980302110008326e-06, "loss": 0.3243, "step": 728 }, { "epoch": 0.14, "grad_norm": 17.5, "learning_rate": 4.980206032712654e-06, "loss": 0.372, "step": 729 }, { "epoch": 0.14, "grad_norm": 10.5, "learning_rate": 4.980109722607386e-06, "loss": 0.335, "step": 730 }, { "epoch": 0.14, "grad_norm": 13.875, "learning_rate": 4.980013179701563e-06, "loss": 0.3585, "step": 731 }, { "epoch": 0.14, "grad_norm": 11.9375, "learning_rate": 4.979916404004249e-06, "loss": 0.2753, "step": 732 }, { "epoch": 0.14, "grad_norm": 16.25, "learning_rate": 4.9798193955245255e-06, "loss": 0.3912, "step": 733 }, { "epoch": 0.14, "grad_norm": 36.0, "learning_rate": 4.9797221542715e-06, "loss": 0.3138, "step": 734 }, { "epoch": 0.14, "grad_norm": 29.0, "learning_rate": 4.979624680254299e-06, "loss": 0.2742, "step": 735 }, { "epoch": 0.14, "grad_norm": 14.3125, "learning_rate": 4.979526973482074e-06, "loss": 0.4128, "step": 736 }, { "epoch": 0.14, "grad_norm": 9.3125, "learning_rate": 4.979429033963995e-06, "loss": 0.312, "step": 737 }, { "epoch": 0.14, "grad_norm": 16.25, "learning_rate": 4.979330861709255e-06, "loss": 0.3557, "step": 738 }, { "epoch": 0.14, "grad_norm": 9.375, "learning_rate": 4.97923245672707e-06, "loss": 0.2857, "step": 739 }, { "epoch": 0.14, "grad_norm": 12.5625, "learning_rate": 4.979133819026677e-06, "loss": 0.2957, "step": 740 }, { "epoch": 0.14, "grad_norm": 13.125, "learning_rate": 4.979034948617335e-06, "loss": 0.2958, "step": 741 }, { "epoch": 0.14, "grad_norm": 14.6875, "learning_rate": 4.978935845508323e-06, "loss": 0.3901, "step": 742 }, { "epoch": 0.14, "grad_norm": 14.875, "learning_rate": 4.978836509708945e-06, "loss": 0.3804, "step": 743 }, { "epoch": 0.14, "grad_norm": 11.6875, "learning_rate": 4.978736941228526e-06, "loss": 0.3579, "step": 744 }, { "epoch": 0.14, "grad_norm": 10.625, "learning_rate": 4.9786371400764105e-06, "loss": 0.3682, "step": 745 }, { "epoch": 0.14, "grad_norm": 12.625, "learning_rate": 4.978537106261967e-06, "loss": 0.3837, "step": 746 }, { "epoch": 0.14, "grad_norm": 15.625, "learning_rate": 4.978436839794586e-06, "loss": 0.3093, "step": 747 }, { "epoch": 0.14, "grad_norm": 13.0, "learning_rate": 4.97833634068368e-06, "loss": 0.359, "step": 748 }, { "epoch": 0.14, "grad_norm": 12.3125, "learning_rate": 4.97823560893868e-06, "loss": 0.3584, "step": 749 }, { "epoch": 0.14, "grad_norm": 20.125, "learning_rate": 4.978134644569044e-06, "loss": 0.4109, "step": 750 }, { "epoch": 0.14, "grad_norm": 12.9375, "learning_rate": 4.978033447584247e-06, "loss": 0.3243, "step": 751 }, { "epoch": 0.14, "grad_norm": 13.3125, "learning_rate": 4.97793201799379e-06, "loss": 0.3459, "step": 752 }, { "epoch": 0.14, "grad_norm": 16.0, "learning_rate": 4.977830355807193e-06, "loss": 0.3243, "step": 753 }, { "epoch": 0.14, "grad_norm": 16.875, "learning_rate": 4.977728461033998e-06, "loss": 0.3666, "step": 754 }, { "epoch": 0.14, "grad_norm": 13.3125, "learning_rate": 4.977626333683771e-06, "loss": 0.4001, "step": 755 }, { "epoch": 0.14, "grad_norm": 12.6875, "learning_rate": 4.977523973766097e-06, "loss": 0.3291, "step": 756 }, { "epoch": 0.14, "grad_norm": 13.3125, "learning_rate": 4.977421381290586e-06, "loss": 0.377, "step": 757 }, { "epoch": 0.14, "grad_norm": 15.125, "learning_rate": 4.977318556266867e-06, "loss": 0.4274, "step": 758 }, { "epoch": 0.14, "grad_norm": 16.5, "learning_rate": 4.9772154987045905e-06, "loss": 0.4091, "step": 759 }, { "epoch": 0.14, "grad_norm": 12.4375, "learning_rate": 4.977112208613433e-06, "loss": 0.2957, "step": 760 }, { "epoch": 0.14, "grad_norm": 11.875, "learning_rate": 4.977008686003088e-06, "loss": 0.3354, "step": 761 }, { "epoch": 0.14, "grad_norm": 8.9375, "learning_rate": 4.976904930883274e-06, "loss": 0.2713, "step": 762 }, { "epoch": 0.14, "grad_norm": 12.625, "learning_rate": 4.976800943263729e-06, "loss": 0.31, "step": 763 }, { "epoch": 0.14, "grad_norm": 9.4375, "learning_rate": 4.976696723154215e-06, "loss": 0.2863, "step": 764 }, { "epoch": 0.14, "grad_norm": 13.3125, "learning_rate": 4.9765922705645145e-06, "loss": 0.3319, "step": 765 }, { "epoch": 0.14, "grad_norm": 12.6875, "learning_rate": 4.976487585504432e-06, "loss": 0.3233, "step": 766 }, { "epoch": 0.15, "grad_norm": 18.125, "learning_rate": 4.976382667983794e-06, "loss": 0.4375, "step": 767 }, { "epoch": 0.15, "grad_norm": 16.625, "learning_rate": 4.976277518012449e-06, "loss": 0.3688, "step": 768 }, { "epoch": 0.15, "grad_norm": 14.875, "learning_rate": 4.976172135600268e-06, "loss": 0.362, "step": 769 }, { "epoch": 0.15, "grad_norm": 11.125, "learning_rate": 4.976066520757141e-06, "loss": 0.3755, "step": 770 }, { "epoch": 0.15, "grad_norm": 10.0, "learning_rate": 4.975960673492983e-06, "loss": 0.3224, "step": 771 }, { "epoch": 0.15, "grad_norm": 11.4375, "learning_rate": 4.975854593817729e-06, "loss": 0.2809, "step": 772 }, { "epoch": 0.15, "grad_norm": 15.5625, "learning_rate": 4.975748281741337e-06, "loss": 0.4302, "step": 773 }, { "epoch": 0.15, "grad_norm": 12.3125, "learning_rate": 4.975641737273787e-06, "loss": 0.3732, "step": 774 }, { "epoch": 0.15, "grad_norm": 9.375, "learning_rate": 4.9755349604250765e-06, "loss": 0.3575, "step": 775 }, { "epoch": 0.15, "grad_norm": 13.5, "learning_rate": 4.975427951205233e-06, "loss": 0.3369, "step": 776 }, { "epoch": 0.15, "grad_norm": 12.0625, "learning_rate": 4.975320709624297e-06, "loss": 0.3979, "step": 777 }, { "epoch": 0.15, "grad_norm": 8.8125, "learning_rate": 4.9752132356923385e-06, "loss": 0.3062, "step": 778 }, { "epoch": 0.15, "grad_norm": 16.75, "learning_rate": 4.975105529419443e-06, "loss": 0.3821, "step": 779 }, { "epoch": 0.15, "grad_norm": 12.0625, "learning_rate": 4.974997590815722e-06, "loss": 0.2956, "step": 780 }, { "epoch": 0.15, "grad_norm": 15.75, "learning_rate": 4.974889419891307e-06, "loss": 0.403, "step": 781 }, { "epoch": 0.15, "grad_norm": 13.75, "learning_rate": 4.974781016656351e-06, "loss": 0.356, "step": 782 }, { "epoch": 0.15, "grad_norm": 14.125, "learning_rate": 4.974672381121032e-06, "loss": 0.3858, "step": 783 }, { "epoch": 0.15, "grad_norm": 10.1875, "learning_rate": 4.974563513295544e-06, "loss": 0.38, "step": 784 }, { "epoch": 0.15, "grad_norm": 12.875, "learning_rate": 4.974454413190108e-06, "loss": 0.3524, "step": 785 }, { "epoch": 0.15, "grad_norm": 12.3125, "learning_rate": 4.974345080814965e-06, "loss": 0.3478, "step": 786 }, { "epoch": 0.15, "grad_norm": 7.96875, "learning_rate": 4.974235516180377e-06, "loss": 0.2854, "step": 787 }, { "epoch": 0.15, "grad_norm": 10.8125, "learning_rate": 4.9741257192966286e-06, "loss": 0.3944, "step": 788 }, { "epoch": 0.15, "grad_norm": 12.375, "learning_rate": 4.9740156901740256e-06, "loss": 0.321, "step": 789 }, { "epoch": 0.15, "grad_norm": 20.0, "learning_rate": 4.973905428822897e-06, "loss": 0.3308, "step": 790 }, { "epoch": 0.15, "grad_norm": 11.8125, "learning_rate": 4.973794935253593e-06, "loss": 0.403, "step": 791 }, { "epoch": 0.15, "grad_norm": 14.625, "learning_rate": 4.973684209476485e-06, "loss": 0.3113, "step": 792 }, { "epoch": 0.15, "grad_norm": 12.0625, "learning_rate": 4.973573251501965e-06, "loss": 0.2801, "step": 793 }, { "epoch": 0.15, "grad_norm": 8.8125, "learning_rate": 4.973462061340451e-06, "loss": 0.2901, "step": 794 }, { "epoch": 0.15, "grad_norm": 13.4375, "learning_rate": 4.9733506390023775e-06, "loss": 0.3667, "step": 795 }, { "epoch": 0.15, "grad_norm": 15.5, "learning_rate": 4.973238984498205e-06, "loss": 0.3057, "step": 796 }, { "epoch": 0.15, "grad_norm": 12.125, "learning_rate": 4.973127097838414e-06, "loss": 0.344, "step": 797 }, { "epoch": 0.15, "grad_norm": 10.125, "learning_rate": 4.973014979033506e-06, "loss": 0.3155, "step": 798 }, { "epoch": 0.15, "grad_norm": 18.5, "learning_rate": 4.972902628094006e-06, "loss": 0.459, "step": 799 }, { "epoch": 0.15, "grad_norm": 13.375, "learning_rate": 4.97279004503046e-06, "loss": 0.3241, "step": 800 }, { "epoch": 0.15, "grad_norm": 11.0, "learning_rate": 4.972677229853437e-06, "loss": 0.2849, "step": 801 }, { "epoch": 0.15, "grad_norm": 15.0, "learning_rate": 4.972564182573524e-06, "loss": 0.3548, "step": 802 }, { "epoch": 0.15, "grad_norm": 17.125, "learning_rate": 4.972450903201334e-06, "loss": 0.3299, "step": 803 }, { "epoch": 0.15, "grad_norm": 13.4375, "learning_rate": 4.9723373917475e-06, "loss": 0.4141, "step": 804 }, { "epoch": 0.15, "grad_norm": 11.125, "learning_rate": 4.972223648222678e-06, "loss": 0.3254, "step": 805 }, { "epoch": 0.15, "grad_norm": 12.375, "learning_rate": 4.972109672637544e-06, "loss": 0.2824, "step": 806 }, { "epoch": 0.15, "grad_norm": 9.375, "learning_rate": 4.971995465002795e-06, "loss": 0.2933, "step": 807 }, { "epoch": 0.15, "grad_norm": 15.0, "learning_rate": 4.9718810253291535e-06, "loss": 0.4091, "step": 808 }, { "epoch": 0.15, "grad_norm": 13.25, "learning_rate": 4.971766353627361e-06, "loss": 0.3147, "step": 809 }, { "epoch": 0.15, "grad_norm": 9.8125, "learning_rate": 4.97165144990818e-06, "loss": 0.347, "step": 810 }, { "epoch": 0.15, "grad_norm": 9.9375, "learning_rate": 4.971536314182398e-06, "loss": 0.3447, "step": 811 }, { "epoch": 0.15, "grad_norm": 27.875, "learning_rate": 4.971420946460822e-06, "loss": 0.3689, "step": 812 }, { "epoch": 0.15, "grad_norm": 9.8125, "learning_rate": 4.971305346754282e-06, "loss": 0.4151, "step": 813 }, { "epoch": 0.15, "grad_norm": 12.75, "learning_rate": 4.971189515073627e-06, "loss": 0.3222, "step": 814 }, { "epoch": 0.15, "grad_norm": 20.875, "learning_rate": 4.97107345142973e-06, "loss": 0.3212, "step": 815 }, { "epoch": 0.15, "grad_norm": 13.625, "learning_rate": 4.970957155833488e-06, "loss": 0.2793, "step": 816 }, { "epoch": 0.15, "grad_norm": 16.0, "learning_rate": 4.970840628295815e-06, "loss": 0.2788, "step": 817 }, { "epoch": 0.15, "grad_norm": 14.0, "learning_rate": 4.97072386882765e-06, "loss": 0.3632, "step": 818 }, { "epoch": 0.15, "grad_norm": 12.4375, "learning_rate": 4.970606877439952e-06, "loss": 0.3406, "step": 819 }, { "epoch": 0.16, "grad_norm": 11.125, "learning_rate": 4.9704896541437034e-06, "loss": 0.3461, "step": 820 }, { "epoch": 0.16, "grad_norm": 20.75, "learning_rate": 4.9703721989499085e-06, "loss": 0.3237, "step": 821 }, { "epoch": 0.16, "grad_norm": 25.625, "learning_rate": 4.970254511869591e-06, "loss": 0.4454, "step": 822 }, { "epoch": 0.16, "grad_norm": 13.6875, "learning_rate": 4.970136592913798e-06, "loss": 0.3402, "step": 823 }, { "epoch": 0.16, "grad_norm": 12.0625, "learning_rate": 4.9700184420935985e-06, "loss": 0.3314, "step": 824 }, { "epoch": 0.16, "grad_norm": 12.625, "learning_rate": 4.969900059420083e-06, "loss": 0.3352, "step": 825 }, { "epoch": 0.16, "grad_norm": 12.75, "learning_rate": 4.969781444904364e-06, "loss": 0.3111, "step": 826 }, { "epoch": 0.16, "grad_norm": 14.25, "learning_rate": 4.969662598557575e-06, "loss": 0.2945, "step": 827 }, { "epoch": 0.16, "grad_norm": 15.625, "learning_rate": 4.969543520390872e-06, "loss": 0.2954, "step": 828 }, { "epoch": 0.16, "grad_norm": 20.625, "learning_rate": 4.969424210415432e-06, "loss": 0.3676, "step": 829 }, { "epoch": 0.16, "grad_norm": 14.1875, "learning_rate": 4.9693046686424554e-06, "loss": 0.2441, "step": 830 }, { "epoch": 0.16, "grad_norm": 9.25, "learning_rate": 4.969184895083162e-06, "loss": 0.3372, "step": 831 }, { "epoch": 0.16, "grad_norm": 10.75, "learning_rate": 4.969064889748796e-06, "loss": 0.3589, "step": 832 }, { "epoch": 0.16, "grad_norm": 13.25, "learning_rate": 4.96894465265062e-06, "loss": 0.4421, "step": 833 }, { "epoch": 0.16, "grad_norm": 18.625, "learning_rate": 4.968824183799923e-06, "loss": 0.3605, "step": 834 }, { "epoch": 0.16, "grad_norm": 14.75, "learning_rate": 4.96870348320801e-06, "loss": 0.3674, "step": 835 }, { "epoch": 0.16, "grad_norm": 16.375, "learning_rate": 4.968582550886213e-06, "loss": 0.2817, "step": 836 }, { "epoch": 0.16, "grad_norm": 12.375, "learning_rate": 4.968461386845882e-06, "loss": 0.3458, "step": 837 }, { "epoch": 0.16, "grad_norm": 14.1875, "learning_rate": 4.9683399910983925e-06, "loss": 0.3517, "step": 838 }, { "epoch": 0.16, "grad_norm": 13.125, "learning_rate": 4.9682183636551376e-06, "loss": 0.2798, "step": 839 }, { "epoch": 0.16, "grad_norm": 10.0, "learning_rate": 4.968096504527535e-06, "loss": 0.3072, "step": 840 }, { "epoch": 0.16, "grad_norm": 12.9375, "learning_rate": 4.967974413727021e-06, "loss": 0.4212, "step": 841 }, { "epoch": 0.16, "grad_norm": 12.5, "learning_rate": 4.96785209126506e-06, "loss": 0.3162, "step": 842 }, { "epoch": 0.16, "grad_norm": 10.0625, "learning_rate": 4.967729537153131e-06, "loss": 0.3237, "step": 843 }, { "epoch": 0.16, "grad_norm": 11.375, "learning_rate": 4.967606751402739e-06, "loss": 0.334, "step": 844 }, { "epoch": 0.16, "grad_norm": 13.75, "learning_rate": 4.967483734025409e-06, "loss": 0.3911, "step": 845 }, { "epoch": 0.16, "grad_norm": 11.3125, "learning_rate": 4.967360485032689e-06, "loss": 0.2799, "step": 846 }, { "epoch": 0.16, "grad_norm": 8.75, "learning_rate": 4.967237004436147e-06, "loss": 0.2535, "step": 847 }, { "epoch": 0.16, "grad_norm": 10.9375, "learning_rate": 4.967113292247375e-06, "loss": 0.3595, "step": 848 }, { "epoch": 0.16, "grad_norm": 12.8125, "learning_rate": 4.966989348477985e-06, "loss": 0.2753, "step": 849 }, { "epoch": 0.16, "grad_norm": 13.25, "learning_rate": 4.96686517313961e-06, "loss": 0.381, "step": 850 }, { "epoch": 0.16, "grad_norm": 9.9375, "learning_rate": 4.9667407662439075e-06, "loss": 0.3436, "step": 851 }, { "epoch": 0.16, "grad_norm": 17.375, "learning_rate": 4.966616127802555e-06, "loss": 0.4713, "step": 852 }, { "epoch": 0.16, "grad_norm": 13.75, "learning_rate": 4.96649125782725e-06, "loss": 0.367, "step": 853 }, { "epoch": 0.16, "grad_norm": 10.0, "learning_rate": 4.966366156329717e-06, "loss": 0.3183, "step": 854 }, { "epoch": 0.16, "grad_norm": 12.875, "learning_rate": 4.9662408233216976e-06, "loss": 0.3234, "step": 855 }, { "epoch": 0.16, "grad_norm": 17.625, "learning_rate": 4.966115258814956e-06, "loss": 0.3371, "step": 856 }, { "epoch": 0.16, "grad_norm": 14.5625, "learning_rate": 4.965989462821277e-06, "loss": 0.3992, "step": 857 }, { "epoch": 0.16, "grad_norm": 14.6875, "learning_rate": 4.965863435352471e-06, "loss": 0.4056, "step": 858 }, { "epoch": 0.16, "grad_norm": 12.4375, "learning_rate": 4.965737176420368e-06, "loss": 0.365, "step": 859 }, { "epoch": 0.16, "grad_norm": 18.0, "learning_rate": 4.965610686036818e-06, "loss": 0.3479, "step": 860 }, { "epoch": 0.16, "grad_norm": 11.625, "learning_rate": 4.9654839642136946e-06, "loss": 0.3079, "step": 861 }, { "epoch": 0.16, "grad_norm": 13.375, "learning_rate": 4.965357010962894e-06, "loss": 0.424, "step": 862 }, { "epoch": 0.16, "grad_norm": 15.75, "learning_rate": 4.965229826296332e-06, "loss": 0.3199, "step": 863 }, { "epoch": 0.16, "grad_norm": 10.9375, "learning_rate": 4.965102410225946e-06, "loss": 0.3634, "step": 864 }, { "epoch": 0.16, "grad_norm": 11.75, "learning_rate": 4.964974762763699e-06, "loss": 0.3569, "step": 865 }, { "epoch": 0.16, "grad_norm": 9.8125, "learning_rate": 4.96484688392157e-06, "loss": 0.373, "step": 866 }, { "epoch": 0.16, "grad_norm": 11.375, "learning_rate": 4.964718773711565e-06, "loss": 0.3107, "step": 867 }, { "epoch": 0.16, "grad_norm": 11.9375, "learning_rate": 4.964590432145707e-06, "loss": 0.3372, "step": 868 }, { "epoch": 0.16, "grad_norm": 9.0625, "learning_rate": 4.964461859236045e-06, "loss": 0.2517, "step": 869 }, { "epoch": 0.16, "grad_norm": 11.875, "learning_rate": 4.964333054994647e-06, "loss": 0.3446, "step": 870 }, { "epoch": 0.16, "grad_norm": 15.25, "learning_rate": 4.964204019433603e-06, "loss": 0.2751, "step": 871 }, { "epoch": 0.16, "grad_norm": 10.5, "learning_rate": 4.9640747525650255e-06, "loss": 0.3358, "step": 872 }, { "epoch": 0.17, "grad_norm": 12.6875, "learning_rate": 4.963945254401048e-06, "loss": 0.3603, "step": 873 }, { "epoch": 0.17, "grad_norm": 12.375, "learning_rate": 4.9638155249538275e-06, "loss": 0.4372, "step": 874 }, { "epoch": 0.17, "grad_norm": 9.6875, "learning_rate": 4.9636855642355405e-06, "loss": 0.2952, "step": 875 }, { "epoch": 0.17, "grad_norm": 12.9375, "learning_rate": 4.963555372258385e-06, "loss": 0.3806, "step": 876 }, { "epoch": 0.17, "grad_norm": 16.0, "learning_rate": 4.963424949034584e-06, "loss": 0.3083, "step": 877 }, { "epoch": 0.17, "grad_norm": 9.0625, "learning_rate": 4.963294294576378e-06, "loss": 0.2697, "step": 878 }, { "epoch": 0.17, "grad_norm": 14.0625, "learning_rate": 4.9631634088960325e-06, "loss": 0.2958, "step": 879 }, { "epoch": 0.17, "grad_norm": 14.1875, "learning_rate": 4.963032292005831e-06, "loss": 0.3472, "step": 880 }, { "epoch": 0.17, "grad_norm": 15.375, "learning_rate": 4.962900943918084e-06, "loss": 0.2877, "step": 881 }, { "epoch": 0.17, "grad_norm": 13.875, "learning_rate": 4.962769364645119e-06, "loss": 0.3603, "step": 882 }, { "epoch": 0.17, "grad_norm": 14.4375, "learning_rate": 4.962637554199288e-06, "loss": 0.4381, "step": 883 }, { "epoch": 0.17, "grad_norm": 10.5625, "learning_rate": 4.962505512592962e-06, "loss": 0.3451, "step": 884 }, { "epoch": 0.17, "grad_norm": 15.9375, "learning_rate": 4.962373239838537e-06, "loss": 0.3383, "step": 885 }, { "epoch": 0.17, "grad_norm": 10.6875, "learning_rate": 4.962240735948429e-06, "loss": 0.3369, "step": 886 }, { "epoch": 0.17, "grad_norm": 14.5625, "learning_rate": 4.962108000935075e-06, "loss": 0.3177, "step": 887 }, { "epoch": 0.17, "grad_norm": 14.8125, "learning_rate": 4.961975034810934e-06, "loss": 0.3408, "step": 888 }, { "epoch": 0.17, "grad_norm": 9.5, "learning_rate": 4.961841837588488e-06, "loss": 0.3465, "step": 889 }, { "epoch": 0.17, "grad_norm": 19.625, "learning_rate": 4.961708409280241e-06, "loss": 0.3357, "step": 890 }, { "epoch": 0.17, "grad_norm": 9.75, "learning_rate": 4.9615747498987145e-06, "loss": 0.3252, "step": 891 }, { "epoch": 0.17, "grad_norm": 14.4375, "learning_rate": 4.9614408594564565e-06, "loss": 0.286, "step": 892 }, { "epoch": 0.17, "grad_norm": 15.0625, "learning_rate": 4.961306737966035e-06, "loss": 0.3358, "step": 893 }, { "epoch": 0.17, "grad_norm": 12.1875, "learning_rate": 4.96117238544004e-06, "loss": 0.3512, "step": 894 }, { "epoch": 0.17, "grad_norm": 12.875, "learning_rate": 4.961037801891081e-06, "loss": 0.3142, "step": 895 }, { "epoch": 0.17, "grad_norm": 11.625, "learning_rate": 4.960902987331792e-06, "loss": 0.2976, "step": 896 }, { "epoch": 0.17, "grad_norm": 11.5625, "learning_rate": 4.960767941774829e-06, "loss": 0.2868, "step": 897 }, { "epoch": 0.17, "grad_norm": 13.8125, "learning_rate": 4.960632665232866e-06, "loss": 0.3298, "step": 898 }, { "epoch": 0.17, "grad_norm": 11.9375, "learning_rate": 4.960497157718602e-06, "loss": 0.3032, "step": 899 }, { "epoch": 0.17, "grad_norm": 13.4375, "learning_rate": 4.960361419244758e-06, "loss": 0.3438, "step": 900 }, { "epoch": 0.17, "grad_norm": 10.25, "learning_rate": 4.960225449824073e-06, "loss": 0.3918, "step": 901 }, { "epoch": 0.17, "grad_norm": 7.8125, "learning_rate": 4.960089249469312e-06, "loss": 0.3146, "step": 902 }, { "epoch": 0.17, "grad_norm": 9.6875, "learning_rate": 4.959952818193258e-06, "loss": 0.3172, "step": 903 }, { "epoch": 0.17, "grad_norm": 23.375, "learning_rate": 4.959816156008719e-06, "loss": 0.3398, "step": 904 }, { "epoch": 0.17, "grad_norm": 7.78125, "learning_rate": 4.959679262928521e-06, "loss": 0.3354, "step": 905 }, { "epoch": 0.17, "grad_norm": 9.4375, "learning_rate": 4.959542138965517e-06, "loss": 0.2569, "step": 906 }, { "epoch": 0.17, "grad_norm": 11.4375, "learning_rate": 4.959404784132576e-06, "loss": 0.4417, "step": 907 }, { "epoch": 0.17, "grad_norm": 8.5, "learning_rate": 4.95926719844259e-06, "loss": 0.3546, "step": 908 }, { "epoch": 0.17, "grad_norm": 14.25, "learning_rate": 4.959129381908477e-06, "loss": 0.3448, "step": 909 }, { "epoch": 0.17, "grad_norm": 12.3125, "learning_rate": 4.958991334543171e-06, "loss": 0.3002, "step": 910 }, { "epoch": 0.17, "grad_norm": 12.4375, "learning_rate": 4.958853056359632e-06, "loss": 0.3247, "step": 911 }, { "epoch": 0.17, "grad_norm": 11.125, "learning_rate": 4.958714547370838e-06, "loss": 0.3195, "step": 912 }, { "epoch": 0.17, "grad_norm": 9.375, "learning_rate": 4.95857580758979e-06, "loss": 0.2452, "step": 913 }, { "epoch": 0.17, "grad_norm": 13.0625, "learning_rate": 4.9584368370295135e-06, "loss": 0.3434, "step": 914 }, { "epoch": 0.17, "grad_norm": 16.125, "learning_rate": 4.958297635703052e-06, "loss": 0.4226, "step": 915 }, { "epoch": 0.17, "grad_norm": 12.4375, "learning_rate": 4.9581582036234725e-06, "loss": 0.3297, "step": 916 }, { "epoch": 0.17, "grad_norm": 12.8125, "learning_rate": 4.95801854080386e-06, "loss": 0.3991, "step": 917 }, { "epoch": 0.17, "grad_norm": 14.75, "learning_rate": 4.957878647257329e-06, "loss": 0.3337, "step": 918 }, { "epoch": 0.17, "grad_norm": 10.625, "learning_rate": 4.9577385229970075e-06, "loss": 0.3365, "step": 919 }, { "epoch": 0.17, "grad_norm": 10.625, "learning_rate": 4.95759816803605e-06, "loss": 0.2928, "step": 920 }, { "epoch": 0.17, "grad_norm": 12.0, "learning_rate": 4.95745758238763e-06, "loss": 0.3391, "step": 921 }, { "epoch": 0.17, "grad_norm": 13.6875, "learning_rate": 4.957316766064945e-06, "loss": 0.3353, "step": 922 }, { "epoch": 0.17, "grad_norm": 12.3125, "learning_rate": 4.957175719081214e-06, "loss": 0.2849, "step": 923 }, { "epoch": 0.17, "grad_norm": 18.0, "learning_rate": 4.957034441449675e-06, "loss": 0.4535, "step": 924 }, { "epoch": 0.17, "grad_norm": 13.5625, "learning_rate": 4.956892933183589e-06, "loss": 0.4558, "step": 925 }, { "epoch": 0.18, "grad_norm": 12.875, "learning_rate": 4.956751194296239e-06, "loss": 0.3778, "step": 926 }, { "epoch": 0.18, "grad_norm": 11.125, "learning_rate": 4.956609224800931e-06, "loss": 0.2913, "step": 927 }, { "epoch": 0.18, "grad_norm": 10.625, "learning_rate": 4.956467024710991e-06, "loss": 0.3544, "step": 928 }, { "epoch": 0.18, "grad_norm": 17.375, "learning_rate": 4.956324594039765e-06, "loss": 0.3422, "step": 929 }, { "epoch": 0.18, "grad_norm": 13.125, "learning_rate": 4.956181932800626e-06, "loss": 0.3983, "step": 930 }, { "epoch": 0.18, "grad_norm": 8.5625, "learning_rate": 4.956039041006961e-06, "loss": 0.3318, "step": 931 }, { "epoch": 0.18, "grad_norm": 9.25, "learning_rate": 4.955895918672186e-06, "loss": 0.3111, "step": 932 }, { "epoch": 0.18, "grad_norm": 14.875, "learning_rate": 4.955752565809734e-06, "loss": 0.345, "step": 933 }, { "epoch": 0.18, "grad_norm": 10.375, "learning_rate": 4.955608982433062e-06, "loss": 0.3535, "step": 934 }, { "epoch": 0.18, "grad_norm": 8.5625, "learning_rate": 4.955465168555647e-06, "loss": 0.299, "step": 935 }, { "epoch": 0.18, "grad_norm": 16.125, "learning_rate": 4.955321124190989e-06, "loss": 0.3267, "step": 936 }, { "epoch": 0.18, "grad_norm": 11.0, "learning_rate": 4.955176849352609e-06, "loss": 0.316, "step": 937 }, { "epoch": 0.18, "grad_norm": 21.25, "learning_rate": 4.955032344054048e-06, "loss": 0.3614, "step": 938 }, { "epoch": 0.18, "grad_norm": 14.8125, "learning_rate": 4.954887608308873e-06, "loss": 0.3282, "step": 939 }, { "epoch": 0.18, "grad_norm": 11.9375, "learning_rate": 4.954742642130668e-06, "loss": 0.2747, "step": 940 }, { "epoch": 0.18, "grad_norm": 14.0625, "learning_rate": 4.95459744553304e-06, "loss": 0.2643, "step": 941 }, { "epoch": 0.18, "grad_norm": 12.6875, "learning_rate": 4.954452018529621e-06, "loss": 0.2547, "step": 942 }, { "epoch": 0.18, "grad_norm": 25.5, "learning_rate": 4.954306361134059e-06, "loss": 0.332, "step": 943 }, { "epoch": 0.18, "grad_norm": 9.0, "learning_rate": 4.954160473360028e-06, "loss": 0.2529, "step": 944 }, { "epoch": 0.18, "grad_norm": 10.5625, "learning_rate": 4.954014355221221e-06, "loss": 0.3175, "step": 945 }, { "epoch": 0.18, "grad_norm": 14.875, "learning_rate": 4.953868006731355e-06, "loss": 0.2721, "step": 946 }, { "epoch": 0.18, "grad_norm": 8.75, "learning_rate": 4.953721427904166e-06, "loss": 0.2773, "step": 947 }, { "epoch": 0.18, "grad_norm": 11.5, "learning_rate": 4.9535746187534124e-06, "loss": 0.3106, "step": 948 }, { "epoch": 0.18, "grad_norm": 20.25, "learning_rate": 4.953427579292878e-06, "loss": 0.371, "step": 949 }, { "epoch": 0.18, "grad_norm": 16.375, "learning_rate": 4.953280309536362e-06, "loss": 0.3652, "step": 950 }, { "epoch": 0.18, "grad_norm": 11.3125, "learning_rate": 4.953132809497688e-06, "loss": 0.3353, "step": 951 }, { "epoch": 0.18, "grad_norm": 16.75, "learning_rate": 4.952985079190702e-06, "loss": 0.3957, "step": 952 }, { "epoch": 0.18, "grad_norm": 18.25, "learning_rate": 4.952837118629272e-06, "loss": 0.2512, "step": 953 }, { "epoch": 0.18, "grad_norm": 11.5625, "learning_rate": 4.952688927827287e-06, "loss": 0.2903, "step": 954 }, { "epoch": 0.18, "grad_norm": 8.75, "learning_rate": 4.9525405067986544e-06, "loss": 0.3684, "step": 955 }, { "epoch": 0.18, "grad_norm": 11.8125, "learning_rate": 4.9523918555573096e-06, "loss": 0.4053, "step": 956 }, { "epoch": 0.18, "grad_norm": 17.625, "learning_rate": 4.952242974117203e-06, "loss": 0.3219, "step": 957 }, { "epoch": 0.18, "grad_norm": 12.375, "learning_rate": 4.952093862492311e-06, "loss": 0.3307, "step": 958 }, { "epoch": 0.18, "grad_norm": 8.8125, "learning_rate": 4.95194452069663e-06, "loss": 0.3029, "step": 959 }, { "epoch": 0.18, "grad_norm": 10.125, "learning_rate": 4.951794948744179e-06, "loss": 0.3231, "step": 960 }, { "epoch": 0.18, "grad_norm": 11.0, "learning_rate": 4.951645146648998e-06, "loss": 0.3128, "step": 961 }, { "epoch": 0.18, "grad_norm": 12.6875, "learning_rate": 4.951495114425147e-06, "loss": 0.3079, "step": 962 }, { "epoch": 0.18, "grad_norm": 15.375, "learning_rate": 4.951344852086709e-06, "loss": 0.3319, "step": 963 }, { "epoch": 0.18, "grad_norm": 11.1875, "learning_rate": 4.951194359647791e-06, "loss": 0.2645, "step": 964 }, { "epoch": 0.18, "grad_norm": 9.4375, "learning_rate": 4.951043637122517e-06, "loss": 0.3441, "step": 965 }, { "epoch": 0.18, "grad_norm": 11.8125, "learning_rate": 4.950892684525036e-06, "loss": 0.355, "step": 966 }, { "epoch": 0.18, "grad_norm": 9.625, "learning_rate": 4.950741501869517e-06, "loss": 0.3469, "step": 967 }, { "epoch": 0.18, "grad_norm": 19.625, "learning_rate": 4.950590089170152e-06, "loss": 0.3886, "step": 968 }, { "epoch": 0.18, "grad_norm": 13.5625, "learning_rate": 4.950438446441152e-06, "loss": 0.3374, "step": 969 }, { "epoch": 0.18, "grad_norm": 12.6875, "learning_rate": 4.950286573696752e-06, "loss": 0.3632, "step": 970 }, { "epoch": 0.18, "grad_norm": 11.75, "learning_rate": 4.950134470951209e-06, "loss": 0.3485, "step": 971 }, { "epoch": 0.18, "grad_norm": 12.875, "learning_rate": 4.949982138218798e-06, "loss": 0.4325, "step": 972 }, { "epoch": 0.18, "grad_norm": 19.0, "learning_rate": 4.949829575513822e-06, "loss": 0.2591, "step": 973 }, { "epoch": 0.18, "grad_norm": 11.6875, "learning_rate": 4.949676782850597e-06, "loss": 0.3616, "step": 974 }, { "epoch": 0.18, "grad_norm": 12.5, "learning_rate": 4.949523760243467e-06, "loss": 0.4185, "step": 975 }, { "epoch": 0.18, "grad_norm": 15.0, "learning_rate": 4.949370507706798e-06, "loss": 0.311, "step": 976 }, { "epoch": 0.18, "grad_norm": 15.375, "learning_rate": 4.949217025254972e-06, "loss": 0.3558, "step": 977 }, { "epoch": 0.19, "grad_norm": 11.3125, "learning_rate": 4.949063312902398e-06, "loss": 0.3347, "step": 978 }, { "epoch": 0.19, "grad_norm": 16.5, "learning_rate": 4.948909370663504e-06, "loss": 0.413, "step": 979 }, { "epoch": 0.19, "grad_norm": 13.875, "learning_rate": 4.948755198552739e-06, "loss": 0.3906, "step": 980 }, { "epoch": 0.19, "grad_norm": 11.6875, "learning_rate": 4.948600796584577e-06, "loss": 0.2798, "step": 981 }, { "epoch": 0.19, "grad_norm": 12.1875, "learning_rate": 4.94844616477351e-06, "loss": 0.3715, "step": 982 }, { "epoch": 0.19, "grad_norm": 17.125, "learning_rate": 4.948291303134051e-06, "loss": 0.3869, "step": 983 }, { "epoch": 0.19, "grad_norm": 14.625, "learning_rate": 4.9481362116807395e-06, "loss": 0.301, "step": 984 }, { "epoch": 0.19, "grad_norm": 20.5, "learning_rate": 4.947980890428131e-06, "loss": 0.4078, "step": 985 }, { "epoch": 0.19, "grad_norm": 12.5625, "learning_rate": 4.9478253393908074e-06, "loss": 0.3251, "step": 986 }, { "epoch": 0.19, "grad_norm": 10.6875, "learning_rate": 4.9476695585833685e-06, "loss": 0.2831, "step": 987 }, { "epoch": 0.19, "grad_norm": 12.375, "learning_rate": 4.9475135480204364e-06, "loss": 0.3329, "step": 988 }, { "epoch": 0.19, "grad_norm": 18.25, "learning_rate": 4.947357307716657e-06, "loss": 0.3143, "step": 989 }, { "epoch": 0.19, "grad_norm": 18.75, "learning_rate": 4.947200837686694e-06, "loss": 0.3067, "step": 990 }, { "epoch": 0.19, "grad_norm": 13.875, "learning_rate": 4.947044137945237e-06, "loss": 0.3246, "step": 991 }, { "epoch": 0.19, "grad_norm": 12.6875, "learning_rate": 4.946887208506993e-06, "loss": 0.322, "step": 992 }, { "epoch": 0.19, "grad_norm": 11.5, "learning_rate": 4.946730049386695e-06, "loss": 0.3302, "step": 993 }, { "epoch": 0.19, "grad_norm": 11.4375, "learning_rate": 4.946572660599091e-06, "loss": 0.2334, "step": 994 }, { "epoch": 0.19, "grad_norm": 10.75, "learning_rate": 4.946415042158958e-06, "loss": 0.3433, "step": 995 }, { "epoch": 0.19, "grad_norm": 13.25, "learning_rate": 4.9462571940810915e-06, "loss": 0.2624, "step": 996 }, { "epoch": 0.19, "grad_norm": 10.4375, "learning_rate": 4.946099116380305e-06, "loss": 0.3924, "step": 997 }, { "epoch": 0.19, "grad_norm": 11.0625, "learning_rate": 4.94594080907144e-06, "loss": 0.3705, "step": 998 }, { "epoch": 0.19, "grad_norm": 8.6875, "learning_rate": 4.945782272169355e-06, "loss": 0.3117, "step": 999 }, { "epoch": 0.19, "grad_norm": 12.1875, "learning_rate": 4.9456235056889314e-06, "loss": 0.3395, "step": 1000 }, { "epoch": 0.19, "grad_norm": 11.5625, "learning_rate": 4.945464509645072e-06, "loss": 0.2964, "step": 1001 }, { "epoch": 0.19, "grad_norm": 11.5625, "learning_rate": 4.945305284052701e-06, "loss": 0.3165, "step": 1002 }, { "epoch": 0.19, "grad_norm": 12.25, "learning_rate": 4.945145828926766e-06, "loss": 0.3563, "step": 1003 }, { "epoch": 0.19, "grad_norm": 13.0, "learning_rate": 4.944986144282232e-06, "loss": 0.3912, "step": 1004 }, { "epoch": 0.19, "grad_norm": 8.5625, "learning_rate": 4.944826230134091e-06, "loss": 0.2994, "step": 1005 }, { "epoch": 0.19, "grad_norm": 11.75, "learning_rate": 4.944666086497352e-06, "loss": 0.2736, "step": 1006 }, { "epoch": 0.19, "grad_norm": 11.0625, "learning_rate": 4.944505713387048e-06, "loss": 0.3046, "step": 1007 }, { "epoch": 0.19, "grad_norm": 14.25, "learning_rate": 4.944345110818231e-06, "loss": 0.3416, "step": 1008 }, { "epoch": 0.19, "grad_norm": 11.0625, "learning_rate": 4.944184278805979e-06, "loss": 0.3065, "step": 1009 }, { "epoch": 0.19, "grad_norm": 12.625, "learning_rate": 4.944023217365386e-06, "loss": 0.4307, "step": 1010 }, { "epoch": 0.19, "grad_norm": 9.25, "learning_rate": 4.943861926511572e-06, "loss": 0.2918, "step": 1011 }, { "epoch": 0.19, "grad_norm": 25.5, "learning_rate": 4.943700406259677e-06, "loss": 0.3871, "step": 1012 }, { "epoch": 0.19, "grad_norm": 11.4375, "learning_rate": 4.943538656624862e-06, "loss": 0.3093, "step": 1013 }, { "epoch": 0.19, "grad_norm": 17.625, "learning_rate": 4.94337667762231e-06, "loss": 0.3654, "step": 1014 }, { "epoch": 0.19, "grad_norm": 12.4375, "learning_rate": 4.943214469267226e-06, "loss": 0.311, "step": 1015 }, { "epoch": 0.19, "grad_norm": 11.625, "learning_rate": 4.943052031574835e-06, "loss": 0.247, "step": 1016 }, { "epoch": 0.19, "grad_norm": 8.9375, "learning_rate": 4.942889364560386e-06, "loss": 0.28, "step": 1017 }, { "epoch": 0.19, "grad_norm": 12.1875, "learning_rate": 4.942726468239145e-06, "loss": 0.2396, "step": 1018 }, { "epoch": 0.19, "grad_norm": 9.5625, "learning_rate": 4.942563342626406e-06, "loss": 0.272, "step": 1019 }, { "epoch": 0.19, "grad_norm": 15.625, "learning_rate": 4.942399987737481e-06, "loss": 0.3407, "step": 1020 }, { "epoch": 0.19, "grad_norm": 17.125, "learning_rate": 4.942236403587701e-06, "loss": 0.3265, "step": 1021 }, { "epoch": 0.19, "grad_norm": 9.375, "learning_rate": 4.9420725901924226e-06, "loss": 0.2652, "step": 1022 }, { "epoch": 0.19, "grad_norm": 14.5, "learning_rate": 4.941908547567024e-06, "loss": 0.3828, "step": 1023 }, { "epoch": 0.19, "grad_norm": 9.3125, "learning_rate": 4.9417442757269e-06, "loss": 0.3345, "step": 1024 }, { "epoch": 0.19, "grad_norm": 11.6875, "learning_rate": 4.941579774687473e-06, "loss": 0.3163, "step": 1025 }, { "epoch": 0.19, "grad_norm": 12.4375, "learning_rate": 4.941415044464184e-06, "loss": 0.4024, "step": 1026 }, { "epoch": 0.19, "grad_norm": 11.0625, "learning_rate": 4.941250085072495e-06, "loss": 0.3971, "step": 1027 }, { "epoch": 0.19, "grad_norm": 12.0625, "learning_rate": 4.94108489652789e-06, "loss": 0.3382, "step": 1028 }, { "epoch": 0.19, "grad_norm": 12.5625, "learning_rate": 4.940919478845876e-06, "loss": 0.3579, "step": 1029 }, { "epoch": 0.19, "grad_norm": 9.9375, "learning_rate": 4.9407538320419785e-06, "loss": 0.2894, "step": 1030 }, { "epoch": 0.2, "grad_norm": 20.875, "learning_rate": 4.940587956131748e-06, "loss": 0.328, "step": 1031 }, { "epoch": 0.2, "grad_norm": 19.375, "learning_rate": 4.940421851130753e-06, "loss": 0.369, "step": 1032 }, { "epoch": 0.2, "grad_norm": 10.5, "learning_rate": 4.940255517054587e-06, "loss": 0.3495, "step": 1033 }, { "epoch": 0.2, "grad_norm": 13.375, "learning_rate": 4.9400889539188634e-06, "loss": 0.4299, "step": 1034 }, { "epoch": 0.2, "grad_norm": 14.5, "learning_rate": 4.939922161739215e-06, "loss": 0.3248, "step": 1035 }, { "epoch": 0.2, "grad_norm": 10.6875, "learning_rate": 4.9397551405313e-06, "loss": 0.3863, "step": 1036 }, { "epoch": 0.2, "grad_norm": 8.75, "learning_rate": 4.939587890310795e-06, "loss": 0.3269, "step": 1037 }, { "epoch": 0.2, "grad_norm": 11.3125, "learning_rate": 4.9394204110934e-06, "loss": 0.3121, "step": 1038 }, { "epoch": 0.2, "grad_norm": 9.5, "learning_rate": 4.939252702894835e-06, "loss": 0.3413, "step": 1039 }, { "epoch": 0.2, "grad_norm": 10.75, "learning_rate": 4.939084765730844e-06, "loss": 0.3139, "step": 1040 }, { "epoch": 0.2, "grad_norm": 9.1875, "learning_rate": 4.938916599617189e-06, "loss": 0.2843, "step": 1041 }, { "epoch": 0.2, "grad_norm": 16.375, "learning_rate": 4.938748204569656e-06, "loss": 0.3999, "step": 1042 }, { "epoch": 0.2, "grad_norm": 19.25, "learning_rate": 4.93857958060405e-06, "loss": 0.3383, "step": 1043 }, { "epoch": 0.2, "grad_norm": 12.5, "learning_rate": 4.9384107277362026e-06, "loss": 0.2772, "step": 1044 }, { "epoch": 0.2, "grad_norm": 9.75, "learning_rate": 4.938241645981962e-06, "loss": 0.3501, "step": 1045 }, { "epoch": 0.2, "grad_norm": 11.875, "learning_rate": 4.938072335357198e-06, "loss": 0.3263, "step": 1046 }, { "epoch": 0.2, "grad_norm": 7.65625, "learning_rate": 4.937902795877804e-06, "loss": 0.3601, "step": 1047 }, { "epoch": 0.2, "grad_norm": 11.5625, "learning_rate": 4.9377330275596955e-06, "loss": 0.3592, "step": 1048 }, { "epoch": 0.2, "grad_norm": 17.25, "learning_rate": 4.937563030418807e-06, "loss": 0.4364, "step": 1049 }, { "epoch": 0.2, "grad_norm": 7.75, "learning_rate": 4.937392804471096e-06, "loss": 0.279, "step": 1050 }, { "epoch": 0.2, "grad_norm": 17.375, "learning_rate": 4.937222349732541e-06, "loss": 0.3821, "step": 1051 }, { "epoch": 0.2, "grad_norm": 21.5, "learning_rate": 4.937051666219141e-06, "loss": 0.3234, "step": 1052 }, { "epoch": 0.2, "grad_norm": 16.875, "learning_rate": 4.93688075394692e-06, "loss": 0.3986, "step": 1053 }, { "epoch": 0.2, "grad_norm": 11.0625, "learning_rate": 4.9367096129319194e-06, "loss": 0.2654, "step": 1054 }, { "epoch": 0.2, "grad_norm": 12.375, "learning_rate": 4.936538243190203e-06, "loss": 0.3113, "step": 1055 }, { "epoch": 0.2, "grad_norm": 12.8125, "learning_rate": 4.9363666447378584e-06, "loss": 0.3185, "step": 1056 }, { "epoch": 0.2, "grad_norm": 12.0, "learning_rate": 4.936194817590992e-06, "loss": 0.3268, "step": 1057 }, { "epoch": 0.2, "grad_norm": 15.875, "learning_rate": 4.936022761765734e-06, "loss": 0.3891, "step": 1058 }, { "epoch": 0.2, "grad_norm": 15.8125, "learning_rate": 4.935850477278232e-06, "loss": 0.2841, "step": 1059 }, { "epoch": 0.2, "grad_norm": 21.125, "learning_rate": 4.9356779641446615e-06, "loss": 0.3186, "step": 1060 }, { "epoch": 0.2, "grad_norm": 9.4375, "learning_rate": 4.935505222381212e-06, "loss": 0.3325, "step": 1061 }, { "epoch": 0.2, "grad_norm": 12.9375, "learning_rate": 4.935332252004103e-06, "loss": 0.2998, "step": 1062 }, { "epoch": 0.2, "grad_norm": 21.5, "learning_rate": 4.935159053029566e-06, "loss": 0.3686, "step": 1063 }, { "epoch": 0.2, "grad_norm": 15.8125, "learning_rate": 4.934985625473861e-06, "loss": 0.3617, "step": 1064 }, { "epoch": 0.2, "grad_norm": 7.5625, "learning_rate": 4.9348119693532655e-06, "loss": 0.3149, "step": 1065 }, { "epoch": 0.2, "grad_norm": 10.375, "learning_rate": 4.934638084684082e-06, "loss": 0.3712, "step": 1066 }, { "epoch": 0.2, "grad_norm": 13.5, "learning_rate": 4.934463971482632e-06, "loss": 0.3359, "step": 1067 }, { "epoch": 0.2, "grad_norm": 12.0625, "learning_rate": 4.9342896297652594e-06, "loss": 0.3645, "step": 1068 }, { "epoch": 0.2, "grad_norm": 16.5, "learning_rate": 4.934115059548329e-06, "loss": 0.389, "step": 1069 }, { "epoch": 0.2, "grad_norm": 11.875, "learning_rate": 4.933940260848226e-06, "loss": 0.3493, "step": 1070 }, { "epoch": 0.2, "grad_norm": 9.25, "learning_rate": 4.933765233681359e-06, "loss": 0.2523, "step": 1071 }, { "epoch": 0.2, "grad_norm": 14.1875, "learning_rate": 4.933589978064157e-06, "loss": 0.3578, "step": 1072 }, { "epoch": 0.2, "grad_norm": 15.0625, "learning_rate": 4.933414494013071e-06, "loss": 0.3195, "step": 1073 }, { "epoch": 0.2, "grad_norm": 14.9375, "learning_rate": 4.933238781544574e-06, "loss": 0.394, "step": 1074 }, { "epoch": 0.2, "grad_norm": 15.375, "learning_rate": 4.933062840675158e-06, "loss": 0.2727, "step": 1075 }, { "epoch": 0.2, "grad_norm": 10.5, "learning_rate": 4.932886671421339e-06, "loss": 0.3007, "step": 1076 }, { "epoch": 0.2, "grad_norm": 20.5, "learning_rate": 4.932710273799654e-06, "loss": 0.3477, "step": 1077 }, { "epoch": 0.2, "grad_norm": 10.1875, "learning_rate": 4.93253364782666e-06, "loss": 0.3361, "step": 1078 }, { "epoch": 0.2, "grad_norm": 10.625, "learning_rate": 4.9323567935189354e-06, "loss": 0.3286, "step": 1079 }, { "epoch": 0.2, "grad_norm": 10.0, "learning_rate": 4.932179710893084e-06, "loss": 0.3222, "step": 1080 }, { "epoch": 0.2, "grad_norm": 26.0, "learning_rate": 4.932002399965725e-06, "loss": 0.2995, "step": 1081 }, { "epoch": 0.2, "grad_norm": 8.125, "learning_rate": 4.931824860753503e-06, "loss": 0.2952, "step": 1082 }, { "epoch": 0.2, "grad_norm": 27.5, "learning_rate": 4.931647093273084e-06, "loss": 0.2486, "step": 1083 }, { "epoch": 0.21, "grad_norm": 9.75, "learning_rate": 4.931469097541154e-06, "loss": 0.3666, "step": 1084 }, { "epoch": 0.21, "grad_norm": 12.5625, "learning_rate": 4.93129087357442e-06, "loss": 0.3786, "step": 1085 }, { "epoch": 0.21, "grad_norm": 11.6875, "learning_rate": 4.9311124213896124e-06, "loss": 0.393, "step": 1086 }, { "epoch": 0.21, "grad_norm": 11.9375, "learning_rate": 4.930933741003482e-06, "loss": 0.3157, "step": 1087 }, { "epoch": 0.21, "grad_norm": 10.8125, "learning_rate": 4.9307548324328016e-06, "loss": 0.2766, "step": 1088 }, { "epoch": 0.21, "grad_norm": 13.8125, "learning_rate": 4.930575695694363e-06, "loss": 0.2933, "step": 1089 }, { "epoch": 0.21, "grad_norm": 16.75, "learning_rate": 4.9303963308049814e-06, "loss": 0.3051, "step": 1090 }, { "epoch": 0.21, "grad_norm": 17.25, "learning_rate": 4.930216737781495e-06, "loss": 0.3425, "step": 1091 }, { "epoch": 0.21, "grad_norm": 12.25, "learning_rate": 4.93003691664076e-06, "loss": 0.402, "step": 1092 }, { "epoch": 0.21, "grad_norm": 12.1875, "learning_rate": 4.929856867399658e-06, "loss": 0.3601, "step": 1093 }, { "epoch": 0.21, "grad_norm": 11.375, "learning_rate": 4.9296765900750865e-06, "loss": 0.3367, "step": 1094 }, { "epoch": 0.21, "grad_norm": 11.875, "learning_rate": 4.9294960846839695e-06, "loss": 0.3953, "step": 1095 }, { "epoch": 0.21, "grad_norm": 12.25, "learning_rate": 4.92931535124325e-06, "loss": 0.4331, "step": 1096 }, { "epoch": 0.21, "grad_norm": 11.125, "learning_rate": 4.929134389769894e-06, "loss": 0.3059, "step": 1097 }, { "epoch": 0.21, "grad_norm": 16.125, "learning_rate": 4.928953200280886e-06, "loss": 0.437, "step": 1098 }, { "epoch": 0.21, "grad_norm": 12.5, "learning_rate": 4.928771782793235e-06, "loss": 0.3729, "step": 1099 }, { "epoch": 0.21, "grad_norm": 13.8125, "learning_rate": 4.92859013732397e-06, "loss": 0.3655, "step": 1100 }, { "epoch": 0.21, "grad_norm": 11.1875, "learning_rate": 4.928408263890139e-06, "loss": 0.3284, "step": 1101 }, { "epoch": 0.21, "grad_norm": 12.5625, "learning_rate": 4.928226162508819e-06, "loss": 0.3084, "step": 1102 }, { "epoch": 0.21, "grad_norm": 20.75, "learning_rate": 4.9280438331971e-06, "loss": 0.2609, "step": 1103 }, { "epoch": 0.21, "grad_norm": 14.3125, "learning_rate": 4.927861275972096e-06, "loss": 0.2751, "step": 1104 }, { "epoch": 0.21, "grad_norm": 10.0, "learning_rate": 4.9276784908509445e-06, "loss": 0.298, "step": 1105 }, { "epoch": 0.21, "grad_norm": 12.375, "learning_rate": 4.927495477850803e-06, "loss": 0.3525, "step": 1106 }, { "epoch": 0.21, "grad_norm": 9.25, "learning_rate": 4.927312236988851e-06, "loss": 0.2997, "step": 1107 }, { "epoch": 0.21, "grad_norm": 16.5, "learning_rate": 4.927128768282288e-06, "loss": 0.3596, "step": 1108 }, { "epoch": 0.21, "grad_norm": 7.9375, "learning_rate": 4.926945071748335e-06, "loss": 0.2697, "step": 1109 }, { "epoch": 0.21, "grad_norm": 10.5625, "learning_rate": 4.926761147404235e-06, "loss": 0.3411, "step": 1110 }, { "epoch": 0.21, "grad_norm": 10.25, "learning_rate": 4.926576995267254e-06, "loss": 0.2855, "step": 1111 }, { "epoch": 0.21, "grad_norm": 20.5, "learning_rate": 4.926392615354677e-06, "loss": 0.2781, "step": 1112 }, { "epoch": 0.21, "grad_norm": 16.375, "learning_rate": 4.926208007683811e-06, "loss": 0.3331, "step": 1113 }, { "epoch": 0.21, "grad_norm": 10.375, "learning_rate": 4.9260231722719845e-06, "loss": 0.3292, "step": 1114 }, { "epoch": 0.21, "grad_norm": 10.875, "learning_rate": 4.925838109136548e-06, "loss": 0.2301, "step": 1115 }, { "epoch": 0.21, "grad_norm": 11.1875, "learning_rate": 4.925652818294872e-06, "loss": 0.3874, "step": 1116 }, { "epoch": 0.21, "grad_norm": 13.8125, "learning_rate": 4.92546729976435e-06, "loss": 0.3034, "step": 1117 }, { "epoch": 0.21, "grad_norm": 10.9375, "learning_rate": 4.925281553562396e-06, "loss": 0.3591, "step": 1118 }, { "epoch": 0.21, "grad_norm": 12.875, "learning_rate": 4.925095579706445e-06, "loss": 0.366, "step": 1119 }, { "epoch": 0.21, "grad_norm": 12.875, "learning_rate": 4.924909378213954e-06, "loss": 0.2993, "step": 1120 }, { "epoch": 0.21, "grad_norm": 10.625, "learning_rate": 4.924722949102401e-06, "loss": 0.3824, "step": 1121 }, { "epoch": 0.21, "grad_norm": 11.25, "learning_rate": 4.924536292389286e-06, "loss": 0.3762, "step": 1122 }, { "epoch": 0.21, "grad_norm": 12.25, "learning_rate": 4.924349408092129e-06, "loss": 0.3505, "step": 1123 }, { "epoch": 0.21, "grad_norm": 7.6875, "learning_rate": 4.924162296228474e-06, "loss": 0.3163, "step": 1124 }, { "epoch": 0.21, "grad_norm": 14.3125, "learning_rate": 4.923974956815883e-06, "loss": 0.341, "step": 1125 }, { "epoch": 0.21, "grad_norm": 12.5625, "learning_rate": 4.9237873898719426e-06, "loss": 0.2675, "step": 1126 }, { "epoch": 0.21, "grad_norm": 10.9375, "learning_rate": 4.923599595414257e-06, "loss": 0.3715, "step": 1127 }, { "epoch": 0.21, "grad_norm": 11.25, "learning_rate": 4.923411573460455e-06, "loss": 0.2843, "step": 1128 }, { "epoch": 0.21, "grad_norm": 8.5625, "learning_rate": 4.923223324028187e-06, "loss": 0.2355, "step": 1129 }, { "epoch": 0.21, "grad_norm": 11.1875, "learning_rate": 4.92303484713512e-06, "loss": 0.3632, "step": 1130 }, { "epoch": 0.21, "grad_norm": 8.375, "learning_rate": 4.922846142798949e-06, "loss": 0.3275, "step": 1131 }, { "epoch": 0.21, "grad_norm": 13.1875, "learning_rate": 4.922657211037387e-06, "loss": 0.3978, "step": 1132 }, { "epoch": 0.21, "grad_norm": 13.25, "learning_rate": 4.9224680518681666e-06, "loss": 0.3067, "step": 1133 }, { "epoch": 0.21, "grad_norm": 11.0625, "learning_rate": 4.922278665309044e-06, "loss": 0.3097, "step": 1134 }, { "epoch": 0.21, "grad_norm": 9.4375, "learning_rate": 4.922089051377797e-06, "loss": 0.3878, "step": 1135 }, { "epoch": 0.21, "grad_norm": 11.875, "learning_rate": 4.921899210092225e-06, "loss": 0.2868, "step": 1136 }, { "epoch": 0.22, "grad_norm": 11.125, "learning_rate": 4.9217091414701455e-06, "loss": 0.3338, "step": 1137 }, { "epoch": 0.22, "grad_norm": 15.9375, "learning_rate": 4.921518845529401e-06, "loss": 0.3781, "step": 1138 }, { "epoch": 0.22, "grad_norm": 17.875, "learning_rate": 4.921328322287855e-06, "loss": 0.2876, "step": 1139 }, { "epoch": 0.22, "grad_norm": 10.9375, "learning_rate": 4.9211375717633895e-06, "loss": 0.3737, "step": 1140 }, { "epoch": 0.22, "grad_norm": 11.75, "learning_rate": 4.9209465939739105e-06, "loss": 0.3402, "step": 1141 }, { "epoch": 0.22, "grad_norm": 13.5625, "learning_rate": 4.920755388937345e-06, "loss": 0.3223, "step": 1142 }, { "epoch": 0.22, "grad_norm": 17.125, "learning_rate": 4.9205639566716404e-06, "loss": 0.3885, "step": 1143 }, { "epoch": 0.22, "grad_norm": 10.5, "learning_rate": 4.920372297194765e-06, "loss": 0.3592, "step": 1144 }, { "epoch": 0.22, "grad_norm": 12.125, "learning_rate": 4.920180410524711e-06, "loss": 0.3446, "step": 1145 }, { "epoch": 0.22, "grad_norm": 9.375, "learning_rate": 4.91998829667949e-06, "loss": 0.3262, "step": 1146 }, { "epoch": 0.22, "grad_norm": 15.6875, "learning_rate": 4.919795955677133e-06, "loss": 0.4057, "step": 1147 }, { "epoch": 0.22, "grad_norm": 34.0, "learning_rate": 4.919603387535697e-06, "loss": 0.3319, "step": 1148 }, { "epoch": 0.22, "grad_norm": 16.5, "learning_rate": 4.919410592273256e-06, "loss": 0.3861, "step": 1149 }, { "epoch": 0.22, "grad_norm": 18.875, "learning_rate": 4.919217569907909e-06, "loss": 0.3729, "step": 1150 }, { "epoch": 0.22, "grad_norm": 12.375, "learning_rate": 4.919024320457773e-06, "loss": 0.3654, "step": 1151 }, { "epoch": 0.22, "grad_norm": 12.25, "learning_rate": 4.918830843940988e-06, "loss": 0.291, "step": 1152 }, { "epoch": 0.22, "grad_norm": 16.125, "learning_rate": 4.918637140375716e-06, "loss": 0.3205, "step": 1153 }, { "epoch": 0.22, "grad_norm": 9.0625, "learning_rate": 4.918443209780137e-06, "loss": 0.315, "step": 1154 }, { "epoch": 0.22, "grad_norm": 32.75, "learning_rate": 4.918249052172458e-06, "loss": 0.2807, "step": 1155 }, { "epoch": 0.22, "grad_norm": 10.625, "learning_rate": 4.9180546675709014e-06, "loss": 0.2374, "step": 1156 }, { "epoch": 0.22, "grad_norm": 17.0, "learning_rate": 4.917860055993715e-06, "loss": 0.4365, "step": 1157 }, { "epoch": 0.22, "grad_norm": 12.8125, "learning_rate": 4.9176652174591646e-06, "loss": 0.2961, "step": 1158 }, { "epoch": 0.22, "grad_norm": 9.25, "learning_rate": 4.917470151985542e-06, "loss": 0.3423, "step": 1159 }, { "epoch": 0.22, "grad_norm": 11.0625, "learning_rate": 4.917274859591154e-06, "loss": 0.3528, "step": 1160 }, { "epoch": 0.22, "grad_norm": 10.25, "learning_rate": 4.917079340294335e-06, "loss": 0.3194, "step": 1161 }, { "epoch": 0.22, "grad_norm": 10.625, "learning_rate": 4.916883594113436e-06, "loss": 0.3649, "step": 1162 }, { "epoch": 0.22, "grad_norm": 12.875, "learning_rate": 4.916687621066832e-06, "loss": 0.3195, "step": 1163 }, { "epoch": 0.22, "grad_norm": 12.6875, "learning_rate": 4.916491421172919e-06, "loss": 0.3608, "step": 1164 }, { "epoch": 0.22, "grad_norm": 11.875, "learning_rate": 4.916294994450112e-06, "loss": 0.3456, "step": 1165 }, { "epoch": 0.22, "grad_norm": 22.0, "learning_rate": 4.91609834091685e-06, "loss": 0.2828, "step": 1166 }, { "epoch": 0.22, "grad_norm": 10.5, "learning_rate": 4.915901460591592e-06, "loss": 0.2716, "step": 1167 }, { "epoch": 0.22, "grad_norm": 9.125, "learning_rate": 4.915704353492819e-06, "loss": 0.3029, "step": 1168 }, { "epoch": 0.22, "grad_norm": 9.4375, "learning_rate": 4.915507019639033e-06, "loss": 0.3444, "step": 1169 }, { "epoch": 0.22, "grad_norm": 25.5, "learning_rate": 4.915309459048757e-06, "loss": 0.3337, "step": 1170 }, { "epoch": 0.22, "grad_norm": 8.1875, "learning_rate": 4.915111671740535e-06, "loss": 0.3273, "step": 1171 }, { "epoch": 0.22, "grad_norm": 13.8125, "learning_rate": 4.9149136577329326e-06, "loss": 0.3963, "step": 1172 }, { "epoch": 0.22, "grad_norm": 9.9375, "learning_rate": 4.914715417044536e-06, "loss": 0.339, "step": 1173 }, { "epoch": 0.22, "grad_norm": 13.6875, "learning_rate": 4.914516949693956e-06, "loss": 0.463, "step": 1174 }, { "epoch": 0.22, "grad_norm": 9.6875, "learning_rate": 4.914318255699821e-06, "loss": 0.2989, "step": 1175 }, { "epoch": 0.22, "grad_norm": 10.375, "learning_rate": 4.91411933508078e-06, "loss": 0.3969, "step": 1176 }, { "epoch": 0.22, "grad_norm": 12.1875, "learning_rate": 4.913920187855508e-06, "loss": 0.3743, "step": 1177 }, { "epoch": 0.22, "grad_norm": 10.0625, "learning_rate": 4.913720814042696e-06, "loss": 0.2718, "step": 1178 }, { "epoch": 0.22, "grad_norm": 9.5625, "learning_rate": 4.91352121366106e-06, "loss": 0.3813, "step": 1179 }, { "epoch": 0.22, "grad_norm": 8.5, "learning_rate": 4.913321386729336e-06, "loss": 0.3654, "step": 1180 }, { "epoch": 0.22, "grad_norm": 9.3125, "learning_rate": 4.91312133326628e-06, "loss": 0.2832, "step": 1181 }, { "epoch": 0.22, "grad_norm": 14.0625, "learning_rate": 4.912921053290671e-06, "loss": 0.3198, "step": 1182 }, { "epoch": 0.22, "grad_norm": 10.5, "learning_rate": 4.912720546821309e-06, "loss": 0.2935, "step": 1183 }, { "epoch": 0.22, "grad_norm": 10.125, "learning_rate": 4.912519813877015e-06, "loss": 0.3021, "step": 1184 }, { "epoch": 0.22, "grad_norm": 14.5625, "learning_rate": 4.912318854476631e-06, "loss": 0.3827, "step": 1185 }, { "epoch": 0.22, "grad_norm": 9.625, "learning_rate": 4.9121176686390195e-06, "loss": 0.299, "step": 1186 }, { "epoch": 0.22, "grad_norm": 10.5625, "learning_rate": 4.911916256383067e-06, "loss": 0.4385, "step": 1187 }, { "epoch": 0.22, "grad_norm": 12.75, "learning_rate": 4.911714617727678e-06, "loss": 0.3806, "step": 1188 }, { "epoch": 0.22, "grad_norm": 9.9375, "learning_rate": 4.91151275269178e-06, "loss": 0.3603, "step": 1189 }, { "epoch": 0.23, "grad_norm": 13.25, "learning_rate": 4.911310661294323e-06, "loss": 0.3035, "step": 1190 }, { "epoch": 0.23, "grad_norm": 16.375, "learning_rate": 4.911108343554274e-06, "loss": 0.3251, "step": 1191 }, { "epoch": 0.23, "grad_norm": 8.125, "learning_rate": 4.910905799490627e-06, "loss": 0.2523, "step": 1192 }, { "epoch": 0.23, "grad_norm": 13.3125, "learning_rate": 4.910703029122391e-06, "loss": 0.3436, "step": 1193 }, { "epoch": 0.23, "grad_norm": 14.0625, "learning_rate": 4.910500032468602e-06, "loss": 0.3672, "step": 1194 }, { "epoch": 0.23, "grad_norm": 9.875, "learning_rate": 4.910296809548315e-06, "loss": 0.347, "step": 1195 }, { "epoch": 0.23, "grad_norm": 12.6875, "learning_rate": 4.9100933603806035e-06, "loss": 0.2516, "step": 1196 }, { "epoch": 0.23, "grad_norm": 11.875, "learning_rate": 4.909889684984565e-06, "loss": 0.3394, "step": 1197 }, { "epoch": 0.23, "grad_norm": 14.0625, "learning_rate": 4.90968578337932e-06, "loss": 0.2699, "step": 1198 }, { "epoch": 0.23, "grad_norm": 15.9375, "learning_rate": 4.909481655584007e-06, "loss": 0.4539, "step": 1199 }, { "epoch": 0.23, "grad_norm": 14.5625, "learning_rate": 4.909277301617787e-06, "loss": 0.3903, "step": 1200 }, { "epoch": 0.23, "grad_norm": 20.625, "learning_rate": 4.9090727214998415e-06, "loss": 0.3907, "step": 1201 }, { "epoch": 0.23, "grad_norm": 14.6875, "learning_rate": 4.9088679152493744e-06, "loss": 0.3088, "step": 1202 }, { "epoch": 0.23, "grad_norm": 12.8125, "learning_rate": 4.90866288288561e-06, "loss": 0.3273, "step": 1203 }, { "epoch": 0.23, "grad_norm": 10.5, "learning_rate": 4.908457624427795e-06, "loss": 0.3258, "step": 1204 }, { "epoch": 0.23, "grad_norm": 10.5, "learning_rate": 4.908252139895195e-06, "loss": 0.3471, "step": 1205 }, { "epoch": 0.23, "grad_norm": 17.125, "learning_rate": 4.908046429307099e-06, "loss": 0.3153, "step": 1206 }, { "epoch": 0.23, "grad_norm": 13.5, "learning_rate": 4.907840492682816e-06, "loss": 0.2876, "step": 1207 }, { "epoch": 0.23, "grad_norm": 12.0, "learning_rate": 4.907634330041677e-06, "loss": 0.3362, "step": 1208 }, { "epoch": 0.23, "grad_norm": 14.6875, "learning_rate": 4.907427941403034e-06, "loss": 0.3316, "step": 1209 }, { "epoch": 0.23, "grad_norm": 13.25, "learning_rate": 4.907221326786259e-06, "loss": 0.3443, "step": 1210 }, { "epoch": 0.23, "grad_norm": 10.4375, "learning_rate": 4.907014486210749e-06, "loss": 0.2869, "step": 1211 }, { "epoch": 0.23, "grad_norm": 10.4375, "learning_rate": 4.906807419695916e-06, "loss": 0.3414, "step": 1212 }, { "epoch": 0.23, "grad_norm": 14.875, "learning_rate": 4.906600127261199e-06, "loss": 0.4043, "step": 1213 }, { "epoch": 0.23, "grad_norm": 9.4375, "learning_rate": 4.906392608926056e-06, "loss": 0.2886, "step": 1214 }, { "epoch": 0.23, "grad_norm": 14.125, "learning_rate": 4.906184864709964e-06, "loss": 0.3804, "step": 1215 }, { "epoch": 0.23, "grad_norm": 12.375, "learning_rate": 4.905976894632426e-06, "loss": 0.3551, "step": 1216 }, { "epoch": 0.23, "grad_norm": 11.8125, "learning_rate": 4.905768698712962e-06, "loss": 0.3565, "step": 1217 }, { "epoch": 0.23, "grad_norm": 9.6875, "learning_rate": 4.905560276971116e-06, "loss": 0.2621, "step": 1218 }, { "epoch": 0.23, "grad_norm": 12.3125, "learning_rate": 4.90535162942645e-06, "loss": 0.3159, "step": 1219 }, { "epoch": 0.23, "grad_norm": 14.875, "learning_rate": 4.905142756098551e-06, "loss": 0.3023, "step": 1220 }, { "epoch": 0.23, "grad_norm": 14.25, "learning_rate": 4.9049336570070234e-06, "loss": 0.3826, "step": 1221 }, { "epoch": 0.23, "grad_norm": 8.875, "learning_rate": 4.904724332171497e-06, "loss": 0.3631, "step": 1222 }, { "epoch": 0.23, "grad_norm": 13.5625, "learning_rate": 4.904514781611619e-06, "loss": 0.3652, "step": 1223 }, { "epoch": 0.23, "grad_norm": 9.0, "learning_rate": 4.904305005347059e-06, "loss": 0.2529, "step": 1224 }, { "epoch": 0.23, "grad_norm": 13.375, "learning_rate": 4.904095003397509e-06, "loss": 0.3667, "step": 1225 }, { "epoch": 0.23, "grad_norm": 18.5, "learning_rate": 4.903884775782681e-06, "loss": 0.2363, "step": 1226 }, { "epoch": 0.23, "grad_norm": 11.75, "learning_rate": 4.9036743225223084e-06, "loss": 0.3189, "step": 1227 }, { "epoch": 0.23, "grad_norm": 11.6875, "learning_rate": 4.9034636436361455e-06, "loss": 0.3413, "step": 1228 }, { "epoch": 0.23, "grad_norm": 10.5, "learning_rate": 4.903252739143969e-06, "loss": 0.3558, "step": 1229 }, { "epoch": 0.23, "grad_norm": 10.5, "learning_rate": 4.903041609065574e-06, "loss": 0.3602, "step": 1230 }, { "epoch": 0.23, "grad_norm": 9.5, "learning_rate": 4.902830253420782e-06, "loss": 0.3185, "step": 1231 }, { "epoch": 0.23, "grad_norm": 13.875, "learning_rate": 4.9026186722294285e-06, "loss": 0.3313, "step": 1232 }, { "epoch": 0.23, "grad_norm": 20.75, "learning_rate": 4.902406865511376e-06, "loss": 0.2937, "step": 1233 }, { "epoch": 0.23, "grad_norm": 11.5, "learning_rate": 4.902194833286507e-06, "loss": 0.2902, "step": 1234 }, { "epoch": 0.23, "grad_norm": 9.75, "learning_rate": 4.901982575574722e-06, "loss": 0.2801, "step": 1235 }, { "epoch": 0.23, "grad_norm": 12.8125, "learning_rate": 4.901770092395947e-06, "loss": 0.2781, "step": 1236 }, { "epoch": 0.23, "grad_norm": 11.4375, "learning_rate": 4.901557383770126e-06, "loss": 0.3336, "step": 1237 }, { "epoch": 0.23, "grad_norm": 13.3125, "learning_rate": 4.901344449717227e-06, "loss": 0.3167, "step": 1238 }, { "epoch": 0.23, "grad_norm": 12.25, "learning_rate": 4.901131290257235e-06, "loss": 0.3249, "step": 1239 }, { "epoch": 0.23, "grad_norm": 12.375, "learning_rate": 4.90091790541016e-06, "loss": 0.2923, "step": 1240 }, { "epoch": 0.23, "grad_norm": 10.5, "learning_rate": 4.900704295196032e-06, "loss": 0.344, "step": 1241 }, { "epoch": 0.23, "grad_norm": 11.0, "learning_rate": 4.900490459634901e-06, "loss": 0.3234, "step": 1242 }, { "epoch": 0.24, "grad_norm": 8.8125, "learning_rate": 4.9002763987468404e-06, "loss": 0.3113, "step": 1243 }, { "epoch": 0.24, "grad_norm": 9.1875, "learning_rate": 4.900062112551943e-06, "loss": 0.3776, "step": 1244 }, { "epoch": 0.24, "grad_norm": 13.25, "learning_rate": 4.8998476010703235e-06, "loss": 0.2606, "step": 1245 }, { "epoch": 0.24, "grad_norm": 9.375, "learning_rate": 4.899632864322116e-06, "loss": 0.267, "step": 1246 }, { "epoch": 0.24, "grad_norm": 12.875, "learning_rate": 4.899417902327478e-06, "loss": 0.3712, "step": 1247 }, { "epoch": 0.24, "grad_norm": 11.0625, "learning_rate": 4.899202715106588e-06, "loss": 0.28, "step": 1248 }, { "epoch": 0.24, "grad_norm": 8.125, "learning_rate": 4.898987302679646e-06, "loss": 0.2932, "step": 1249 }, { "epoch": 0.24, "grad_norm": 12.5625, "learning_rate": 4.898771665066868e-06, "loss": 0.3359, "step": 1250 }, { "epoch": 0.24, "grad_norm": 14.375, "learning_rate": 4.8985558022885e-06, "loss": 0.3157, "step": 1251 }, { "epoch": 0.24, "grad_norm": 9.0, "learning_rate": 4.898339714364802e-06, "loss": 0.3265, "step": 1252 }, { "epoch": 0.24, "grad_norm": 13.5, "learning_rate": 4.898123401316057e-06, "loss": 0.37, "step": 1253 }, { "epoch": 0.24, "grad_norm": 15.75, "learning_rate": 4.897906863162571e-06, "loss": 0.2851, "step": 1254 }, { "epoch": 0.24, "grad_norm": 12.875, "learning_rate": 4.897690099924669e-06, "loss": 0.2899, "step": 1255 }, { "epoch": 0.24, "grad_norm": 16.0, "learning_rate": 4.897473111622699e-06, "loss": 0.3203, "step": 1256 }, { "epoch": 0.24, "grad_norm": 12.9375, "learning_rate": 4.897255898277028e-06, "loss": 0.3734, "step": 1257 }, { "epoch": 0.24, "grad_norm": 18.125, "learning_rate": 4.897038459908045e-06, "loss": 0.408, "step": 1258 }, { "epoch": 0.24, "grad_norm": 11.5625, "learning_rate": 4.896820796536162e-06, "loss": 0.294, "step": 1259 }, { "epoch": 0.24, "grad_norm": 9.125, "learning_rate": 4.896602908181808e-06, "loss": 0.2643, "step": 1260 }, { "epoch": 0.24, "grad_norm": 13.375, "learning_rate": 4.896384794865436e-06, "loss": 0.3739, "step": 1261 }, { "epoch": 0.24, "grad_norm": 8.75, "learning_rate": 4.896166456607523e-06, "loss": 0.3449, "step": 1262 }, { "epoch": 0.24, "grad_norm": 12.1875, "learning_rate": 4.895947893428559e-06, "loss": 0.3999, "step": 1263 }, { "epoch": 0.24, "grad_norm": 20.5, "learning_rate": 4.895729105349063e-06, "loss": 0.3444, "step": 1264 }, { "epoch": 0.24, "grad_norm": 11.8125, "learning_rate": 4.89551009238957e-06, "loss": 0.3769, "step": 1265 }, { "epoch": 0.24, "grad_norm": 9.375, "learning_rate": 4.89529085457064e-06, "loss": 0.3003, "step": 1266 }, { "epoch": 0.24, "grad_norm": 10.5, "learning_rate": 4.895071391912851e-06, "loss": 0.3223, "step": 1267 }, { "epoch": 0.24, "grad_norm": 17.125, "learning_rate": 4.8948517044368035e-06, "loss": 0.3168, "step": 1268 }, { "epoch": 0.24, "grad_norm": 9.875, "learning_rate": 4.894631792163119e-06, "loss": 0.3058, "step": 1269 }, { "epoch": 0.24, "grad_norm": 10.875, "learning_rate": 4.89441165511244e-06, "loss": 0.2849, "step": 1270 }, { "epoch": 0.24, "grad_norm": 14.3125, "learning_rate": 4.89419129330543e-06, "loss": 0.2923, "step": 1271 }, { "epoch": 0.24, "grad_norm": 9.875, "learning_rate": 4.893970706762775e-06, "loss": 0.31, "step": 1272 }, { "epoch": 0.24, "grad_norm": 13.9375, "learning_rate": 4.893749895505179e-06, "loss": 0.3374, "step": 1273 }, { "epoch": 0.24, "grad_norm": 13.8125, "learning_rate": 4.89352885955337e-06, "loss": 0.269, "step": 1274 }, { "epoch": 0.24, "grad_norm": 10.875, "learning_rate": 4.893307598928094e-06, "loss": 0.3395, "step": 1275 }, { "epoch": 0.24, "grad_norm": 12.5, "learning_rate": 4.893086113650123e-06, "loss": 0.3279, "step": 1276 }, { "epoch": 0.24, "grad_norm": 11.0, "learning_rate": 4.892864403740245e-06, "loss": 0.2943, "step": 1277 }, { "epoch": 0.24, "grad_norm": 12.125, "learning_rate": 4.8926424692192724e-06, "loss": 0.3331, "step": 1278 }, { "epoch": 0.24, "grad_norm": 11.5, "learning_rate": 4.8924203101080375e-06, "loss": 0.33, "step": 1279 }, { "epoch": 0.24, "grad_norm": 10.9375, "learning_rate": 4.892197926427393e-06, "loss": 0.3809, "step": 1280 }, { "epoch": 0.24, "grad_norm": 13.125, "learning_rate": 4.8919753181982146e-06, "loss": 0.4182, "step": 1281 }, { "epoch": 0.24, "grad_norm": 14.5625, "learning_rate": 4.891752485441396e-06, "loss": 0.3188, "step": 1282 }, { "epoch": 0.24, "grad_norm": 16.875, "learning_rate": 4.891529428177856e-06, "loss": 0.3554, "step": 1283 }, { "epoch": 0.24, "grad_norm": 10.3125, "learning_rate": 4.89130614642853e-06, "loss": 0.296, "step": 1284 }, { "epoch": 0.24, "grad_norm": 10.25, "learning_rate": 4.89108264021438e-06, "loss": 0.3852, "step": 1285 }, { "epoch": 0.24, "grad_norm": 12.5625, "learning_rate": 4.890858909556382e-06, "loss": 0.3369, "step": 1286 }, { "epoch": 0.24, "grad_norm": 10.875, "learning_rate": 4.890634954475539e-06, "loss": 0.3708, "step": 1287 }, { "epoch": 0.24, "grad_norm": 12.0625, "learning_rate": 4.890410774992874e-06, "loss": 0.2712, "step": 1288 }, { "epoch": 0.24, "grad_norm": 15.6875, "learning_rate": 4.890186371129428e-06, "loss": 0.3313, "step": 1289 }, { "epoch": 0.24, "grad_norm": 19.0, "learning_rate": 4.889961742906267e-06, "loss": 0.385, "step": 1290 }, { "epoch": 0.24, "grad_norm": 9.4375, "learning_rate": 4.8897368903444745e-06, "loss": 0.293, "step": 1291 }, { "epoch": 0.24, "grad_norm": 12.25, "learning_rate": 4.889511813465157e-06, "loss": 0.363, "step": 1292 }, { "epoch": 0.24, "grad_norm": 11.375, "learning_rate": 4.889286512289443e-06, "loss": 0.3009, "step": 1293 }, { "epoch": 0.24, "grad_norm": 13.4375, "learning_rate": 4.8890609868384805e-06, "loss": 0.332, "step": 1294 }, { "epoch": 0.24, "grad_norm": 10.1875, "learning_rate": 4.888835237133438e-06, "loss": 0.2679, "step": 1295 }, { "epoch": 0.25, "grad_norm": 16.25, "learning_rate": 4.8886092631955066e-06, "loss": 0.3563, "step": 1296 }, { "epoch": 0.25, "grad_norm": 11.3125, "learning_rate": 4.888383065045898e-06, "loss": 0.3377, "step": 1297 }, { "epoch": 0.25, "grad_norm": 9.75, "learning_rate": 4.888156642705844e-06, "loss": 0.2854, "step": 1298 }, { "epoch": 0.25, "grad_norm": 9.625, "learning_rate": 4.887929996196599e-06, "loss": 0.3211, "step": 1299 }, { "epoch": 0.25, "grad_norm": 6.6875, "learning_rate": 4.887703125539437e-06, "loss": 0.2372, "step": 1300 }, { "epoch": 0.25, "grad_norm": 10.9375, "learning_rate": 4.887476030755653e-06, "loss": 0.2856, "step": 1301 }, { "epoch": 0.25, "grad_norm": 17.0, "learning_rate": 4.887248711866567e-06, "loss": 0.3152, "step": 1302 }, { "epoch": 0.25, "grad_norm": 12.3125, "learning_rate": 4.887021168893513e-06, "loss": 0.3011, "step": 1303 }, { "epoch": 0.25, "grad_norm": 19.875, "learning_rate": 4.886793401857851e-06, "loss": 0.3337, "step": 1304 }, { "epoch": 0.25, "grad_norm": 19.125, "learning_rate": 4.8865654107809614e-06, "loss": 0.3489, "step": 1305 }, { "epoch": 0.25, "grad_norm": 9.5625, "learning_rate": 4.886337195684244e-06, "loss": 0.3262, "step": 1306 }, { "epoch": 0.25, "grad_norm": 9.75, "learning_rate": 4.886108756589123e-06, "loss": 0.2975, "step": 1307 }, { "epoch": 0.25, "grad_norm": 13.3125, "learning_rate": 4.885880093517038e-06, "loss": 0.3594, "step": 1308 }, { "epoch": 0.25, "grad_norm": 17.375, "learning_rate": 4.885651206489455e-06, "loss": 0.2217, "step": 1309 }, { "epoch": 0.25, "grad_norm": 13.25, "learning_rate": 4.885422095527859e-06, "loss": 0.3196, "step": 1310 }, { "epoch": 0.25, "grad_norm": 13.125, "learning_rate": 4.8851927606537555e-06, "loss": 0.3085, "step": 1311 }, { "epoch": 0.25, "grad_norm": 11.3125, "learning_rate": 4.884963201888671e-06, "loss": 0.4044, "step": 1312 }, { "epoch": 0.25, "grad_norm": 8.625, "learning_rate": 4.884733419254154e-06, "loss": 0.3156, "step": 1313 }, { "epoch": 0.25, "grad_norm": 11.5, "learning_rate": 4.884503412771774e-06, "loss": 0.2578, "step": 1314 }, { "epoch": 0.25, "grad_norm": 8.9375, "learning_rate": 4.884273182463121e-06, "loss": 0.2941, "step": 1315 }, { "epoch": 0.25, "grad_norm": 8.9375, "learning_rate": 4.884042728349804e-06, "loss": 0.2929, "step": 1316 }, { "epoch": 0.25, "grad_norm": 13.1875, "learning_rate": 4.883812050453457e-06, "loss": 0.3914, "step": 1317 }, { "epoch": 0.25, "grad_norm": 10.5625, "learning_rate": 4.883581148795733e-06, "loss": 0.2678, "step": 1318 }, { "epoch": 0.25, "grad_norm": 8.75, "learning_rate": 4.8833500233983064e-06, "loss": 0.3602, "step": 1319 }, { "epoch": 0.25, "grad_norm": 10.3125, "learning_rate": 4.8831186742828705e-06, "loss": 0.2866, "step": 1320 }, { "epoch": 0.25, "grad_norm": 8.5, "learning_rate": 4.8828871014711434e-06, "loss": 0.2949, "step": 1321 }, { "epoch": 0.25, "grad_norm": 14.875, "learning_rate": 4.88265530498486e-06, "loss": 0.331, "step": 1322 }, { "epoch": 0.25, "grad_norm": 10.625, "learning_rate": 4.88242328484578e-06, "loss": 0.3165, "step": 1323 }, { "epoch": 0.25, "grad_norm": 12.125, "learning_rate": 4.882191041075682e-06, "loss": 0.3443, "step": 1324 }, { "epoch": 0.25, "grad_norm": 9.4375, "learning_rate": 4.8819585736963665e-06, "loss": 0.2165, "step": 1325 }, { "epoch": 0.25, "grad_norm": 9.375, "learning_rate": 4.8817258827296535e-06, "loss": 0.327, "step": 1326 }, { "epoch": 0.25, "grad_norm": 16.625, "learning_rate": 4.8814929681973854e-06, "loss": 0.4187, "step": 1327 }, { "epoch": 0.25, "grad_norm": 7.9375, "learning_rate": 4.881259830121426e-06, "loss": 0.3247, "step": 1328 }, { "epoch": 0.25, "grad_norm": 9.5, "learning_rate": 4.881026468523659e-06, "loss": 0.2672, "step": 1329 }, { "epoch": 0.25, "grad_norm": 9.4375, "learning_rate": 4.880792883425988e-06, "loss": 0.2954, "step": 1330 }, { "epoch": 0.25, "grad_norm": 24.375, "learning_rate": 4.88055907485034e-06, "loss": 0.3244, "step": 1331 }, { "epoch": 0.25, "grad_norm": 10.75, "learning_rate": 4.880325042818661e-06, "loss": 0.2888, "step": 1332 }, { "epoch": 0.25, "grad_norm": 8.5625, "learning_rate": 4.880090787352921e-06, "loss": 0.2944, "step": 1333 }, { "epoch": 0.25, "grad_norm": 13.0, "learning_rate": 4.8798563084751075e-06, "loss": 0.3438, "step": 1334 }, { "epoch": 0.25, "grad_norm": 13.375, "learning_rate": 4.87962160620723e-06, "loss": 0.3498, "step": 1335 }, { "epoch": 0.25, "grad_norm": 14.375, "learning_rate": 4.8793866805713205e-06, "loss": 0.3691, "step": 1336 }, { "epoch": 0.25, "grad_norm": 15.6875, "learning_rate": 4.8791515315894285e-06, "loss": 0.3061, "step": 1337 }, { "epoch": 0.25, "grad_norm": 20.375, "learning_rate": 4.87891615928363e-06, "loss": 0.3195, "step": 1338 }, { "epoch": 0.25, "grad_norm": 13.375, "learning_rate": 4.878680563676016e-06, "loss": 0.358, "step": 1339 }, { "epoch": 0.25, "grad_norm": 14.3125, "learning_rate": 4.878444744788703e-06, "loss": 0.4158, "step": 1340 }, { "epoch": 0.25, "grad_norm": 11.0625, "learning_rate": 4.878208702643824e-06, "loss": 0.277, "step": 1341 }, { "epoch": 0.25, "grad_norm": 10.125, "learning_rate": 4.877972437263539e-06, "loss": 0.2787, "step": 1342 }, { "epoch": 0.25, "grad_norm": 10.75, "learning_rate": 4.877735948670023e-06, "loss": 0.3254, "step": 1343 }, { "epoch": 0.25, "grad_norm": 14.6875, "learning_rate": 4.877499236885476e-06, "loss": 0.436, "step": 1344 }, { "epoch": 0.25, "grad_norm": 12.0, "learning_rate": 4.877262301932116e-06, "loss": 0.2876, "step": 1345 }, { "epoch": 0.25, "grad_norm": 10.9375, "learning_rate": 4.877025143832185e-06, "loss": 0.3287, "step": 1346 }, { "epoch": 0.25, "grad_norm": 10.25, "learning_rate": 4.8767877626079426e-06, "loss": 0.2972, "step": 1347 }, { "epoch": 0.26, "grad_norm": 8.1875, "learning_rate": 4.876550158281672e-06, "loss": 0.2477, "step": 1348 }, { "epoch": 0.26, "grad_norm": 12.1875, "learning_rate": 4.876312330875677e-06, "loss": 0.3437, "step": 1349 }, { "epoch": 0.26, "grad_norm": 10.9375, "learning_rate": 4.876074280412281e-06, "loss": 0.2802, "step": 1350 }, { "epoch": 0.26, "grad_norm": 13.625, "learning_rate": 4.875836006913829e-06, "loss": 0.2652, "step": 1351 }, { "epoch": 0.26, "grad_norm": 12.125, "learning_rate": 4.875597510402688e-06, "loss": 0.3498, "step": 1352 }, { "epoch": 0.26, "grad_norm": 11.25, "learning_rate": 4.875358790901243e-06, "loss": 0.2823, "step": 1353 }, { "epoch": 0.26, "grad_norm": 11.0625, "learning_rate": 4.875119848431903e-06, "loss": 0.3598, "step": 1354 }, { "epoch": 0.26, "grad_norm": 8.6875, "learning_rate": 4.874880683017098e-06, "loss": 0.2983, "step": 1355 }, { "epoch": 0.26, "grad_norm": 11.5, "learning_rate": 4.874641294679276e-06, "loss": 0.3745, "step": 1356 }, { "epoch": 0.26, "grad_norm": 10.9375, "learning_rate": 4.874401683440908e-06, "loss": 0.2915, "step": 1357 }, { "epoch": 0.26, "grad_norm": 13.5625, "learning_rate": 4.874161849324487e-06, "loss": 0.2751, "step": 1358 }, { "epoch": 0.26, "grad_norm": 8.375, "learning_rate": 4.8739217923525236e-06, "loss": 0.2633, "step": 1359 }, { "epoch": 0.26, "grad_norm": 14.75, "learning_rate": 4.873681512547552e-06, "loss": 0.3397, "step": 1360 }, { "epoch": 0.26, "grad_norm": 11.6875, "learning_rate": 4.873441009932127e-06, "loss": 0.3152, "step": 1361 }, { "epoch": 0.26, "grad_norm": 12.125, "learning_rate": 4.873200284528824e-06, "loss": 0.2892, "step": 1362 }, { "epoch": 0.26, "grad_norm": 11.125, "learning_rate": 4.8729593363602385e-06, "loss": 0.3963, "step": 1363 }, { "epoch": 0.26, "grad_norm": 9.75, "learning_rate": 4.872718165448987e-06, "loss": 0.3134, "step": 1364 }, { "epoch": 0.26, "grad_norm": 13.75, "learning_rate": 4.872476771817709e-06, "loss": 0.4047, "step": 1365 }, { "epoch": 0.26, "grad_norm": 13.3125, "learning_rate": 4.872235155489063e-06, "loss": 0.3771, "step": 1366 }, { "epoch": 0.26, "grad_norm": 12.75, "learning_rate": 4.871993316485728e-06, "loss": 0.3243, "step": 1367 }, { "epoch": 0.26, "grad_norm": 13.25, "learning_rate": 4.871751254830405e-06, "loss": 0.2911, "step": 1368 }, { "epoch": 0.26, "grad_norm": 11.375, "learning_rate": 4.871508970545815e-06, "loss": 0.3509, "step": 1369 }, { "epoch": 0.26, "grad_norm": 11.9375, "learning_rate": 4.871266463654703e-06, "loss": 0.3271, "step": 1370 }, { "epoch": 0.26, "grad_norm": 15.8125, "learning_rate": 4.8710237341798305e-06, "loss": 0.3395, "step": 1371 }, { "epoch": 0.26, "grad_norm": 11.5625, "learning_rate": 4.870780782143981e-06, "loss": 0.2966, "step": 1372 }, { "epoch": 0.26, "grad_norm": 13.8125, "learning_rate": 4.870537607569962e-06, "loss": 0.287, "step": 1373 }, { "epoch": 0.26, "grad_norm": 10.5, "learning_rate": 4.870294210480597e-06, "loss": 0.2851, "step": 1374 }, { "epoch": 0.26, "grad_norm": 8.875, "learning_rate": 4.870050590898735e-06, "loss": 0.3195, "step": 1375 }, { "epoch": 0.26, "grad_norm": 15.8125, "learning_rate": 4.869806748847244e-06, "loss": 0.2809, "step": 1376 }, { "epoch": 0.26, "grad_norm": 8.8125, "learning_rate": 4.8695626843490094e-06, "loss": 0.2656, "step": 1377 }, { "epoch": 0.26, "grad_norm": 9.9375, "learning_rate": 4.869318397426945e-06, "loss": 0.3654, "step": 1378 }, { "epoch": 0.26, "grad_norm": 10.0625, "learning_rate": 4.86907388810398e-06, "loss": 0.3411, "step": 1379 }, { "epoch": 0.26, "grad_norm": 13.125, "learning_rate": 4.868829156403064e-06, "loss": 0.3209, "step": 1380 }, { "epoch": 0.26, "grad_norm": 10.4375, "learning_rate": 4.86858420234717e-06, "loss": 0.3638, "step": 1381 }, { "epoch": 0.26, "grad_norm": 9.875, "learning_rate": 4.868339025959294e-06, "loss": 0.3057, "step": 1382 }, { "epoch": 0.26, "grad_norm": 18.375, "learning_rate": 4.868093627262445e-06, "loss": 0.3527, "step": 1383 }, { "epoch": 0.26, "grad_norm": 19.25, "learning_rate": 4.867848006279663e-06, "loss": 0.3851, "step": 1384 }, { "epoch": 0.26, "grad_norm": 8.3125, "learning_rate": 4.867602163034e-06, "loss": 0.2479, "step": 1385 }, { "epoch": 0.26, "grad_norm": 16.25, "learning_rate": 4.867356097548534e-06, "loss": 0.2658, "step": 1386 }, { "epoch": 0.26, "grad_norm": 14.75, "learning_rate": 4.867109809846362e-06, "loss": 0.3084, "step": 1387 }, { "epoch": 0.26, "grad_norm": 9.375, "learning_rate": 4.866863299950603e-06, "loss": 0.3176, "step": 1388 }, { "epoch": 0.26, "grad_norm": 9.125, "learning_rate": 4.8666165678843945e-06, "loss": 0.3572, "step": 1389 }, { "epoch": 0.26, "grad_norm": 8.25, "learning_rate": 4.866369613670899e-06, "loss": 0.2958, "step": 1390 }, { "epoch": 0.26, "grad_norm": 12.875, "learning_rate": 4.866122437333295e-06, "loss": 0.4413, "step": 1391 }, { "epoch": 0.26, "grad_norm": 9.625, "learning_rate": 4.865875038894786e-06, "loss": 0.278, "step": 1392 }, { "epoch": 0.26, "grad_norm": 10.25, "learning_rate": 4.865627418378593e-06, "loss": 0.3061, "step": 1393 }, { "epoch": 0.26, "grad_norm": 17.75, "learning_rate": 4.865379575807961e-06, "loss": 0.3045, "step": 1394 }, { "epoch": 0.26, "grad_norm": 12.4375, "learning_rate": 4.865131511206153e-06, "loss": 0.2779, "step": 1395 }, { "epoch": 0.26, "grad_norm": 15.5625, "learning_rate": 4.864883224596455e-06, "loss": 0.3332, "step": 1396 }, { "epoch": 0.26, "grad_norm": 16.5, "learning_rate": 4.864634716002172e-06, "loss": 0.3703, "step": 1397 }, { "epoch": 0.26, "grad_norm": 14.8125, "learning_rate": 4.864385985446631e-06, "loss": 0.3543, "step": 1398 }, { "epoch": 0.26, "grad_norm": 16.5, "learning_rate": 4.864137032953181e-06, "loss": 0.3579, "step": 1399 }, { "epoch": 0.26, "grad_norm": 17.875, "learning_rate": 4.8638878585451884e-06, "loss": 0.3757, "step": 1400 }, { "epoch": 0.27, "grad_norm": 10.5625, "learning_rate": 4.863638462246044e-06, "loss": 0.3658, "step": 1401 }, { "epoch": 0.27, "grad_norm": 10.75, "learning_rate": 4.8633888440791566e-06, "loss": 0.3521, "step": 1402 }, { "epoch": 0.27, "grad_norm": 15.125, "learning_rate": 4.863139004067958e-06, "loss": 0.2807, "step": 1403 }, { "epoch": 0.27, "grad_norm": 10.6875, "learning_rate": 4.862888942235899e-06, "loss": 0.3353, "step": 1404 }, { "epoch": 0.27, "grad_norm": 12.6875, "learning_rate": 4.862638658606454e-06, "loss": 0.2997, "step": 1405 }, { "epoch": 0.27, "grad_norm": 17.75, "learning_rate": 4.862388153203115e-06, "loss": 0.2769, "step": 1406 }, { "epoch": 0.27, "grad_norm": 15.75, "learning_rate": 4.862137426049396e-06, "loss": 0.3641, "step": 1407 }, { "epoch": 0.27, "grad_norm": 13.25, "learning_rate": 4.861886477168832e-06, "loss": 0.2624, "step": 1408 }, { "epoch": 0.27, "grad_norm": 14.6875, "learning_rate": 4.86163530658498e-06, "loss": 0.3654, "step": 1409 }, { "epoch": 0.27, "grad_norm": 13.875, "learning_rate": 4.861383914321416e-06, "loss": 0.368, "step": 1410 }, { "epoch": 0.27, "grad_norm": 13.3125, "learning_rate": 4.8611323004017365e-06, "loss": 0.3506, "step": 1411 }, { "epoch": 0.27, "grad_norm": 10.4375, "learning_rate": 4.860880464849562e-06, "loss": 0.3259, "step": 1412 }, { "epoch": 0.27, "grad_norm": 10.4375, "learning_rate": 4.860628407688529e-06, "loss": 0.3728, "step": 1413 }, { "epoch": 0.27, "grad_norm": 9.75, "learning_rate": 4.860376128942299e-06, "loss": 0.2938, "step": 1414 }, { "epoch": 0.27, "grad_norm": 15.25, "learning_rate": 4.860123628634552e-06, "loss": 0.2585, "step": 1415 }, { "epoch": 0.27, "grad_norm": 13.4375, "learning_rate": 4.859870906788991e-06, "loss": 0.3077, "step": 1416 }, { "epoch": 0.27, "grad_norm": 13.625, "learning_rate": 4.859617963429335e-06, "loss": 0.3505, "step": 1417 }, { "epoch": 0.27, "grad_norm": 15.5, "learning_rate": 4.85936479857933e-06, "loss": 0.3196, "step": 1418 }, { "epoch": 0.27, "grad_norm": 9.75, "learning_rate": 4.859111412262738e-06, "loss": 0.2624, "step": 1419 }, { "epoch": 0.27, "grad_norm": 12.25, "learning_rate": 4.858857804503345e-06, "loss": 0.3595, "step": 1420 }, { "epoch": 0.27, "grad_norm": 16.375, "learning_rate": 4.858603975324956e-06, "loss": 0.3296, "step": 1421 }, { "epoch": 0.27, "grad_norm": 14.0, "learning_rate": 4.858349924751397e-06, "loss": 0.2976, "step": 1422 }, { "epoch": 0.27, "grad_norm": 17.375, "learning_rate": 4.858095652806515e-06, "loss": 0.3278, "step": 1423 }, { "epoch": 0.27, "grad_norm": 9.125, "learning_rate": 4.857841159514177e-06, "loss": 0.328, "step": 1424 }, { "epoch": 0.27, "grad_norm": 9.3125, "learning_rate": 4.857586444898273e-06, "loss": 0.2604, "step": 1425 }, { "epoch": 0.27, "grad_norm": 15.5, "learning_rate": 4.857331508982711e-06, "loss": 0.3263, "step": 1426 }, { "epoch": 0.27, "grad_norm": 25.625, "learning_rate": 4.857076351791422e-06, "loss": 0.2459, "step": 1427 }, { "epoch": 0.27, "grad_norm": 16.875, "learning_rate": 4.856820973348356e-06, "loss": 0.3127, "step": 1428 }, { "epoch": 0.27, "grad_norm": 7.4375, "learning_rate": 4.8565653736774865e-06, "loss": 0.3038, "step": 1429 }, { "epoch": 0.27, "grad_norm": 11.3125, "learning_rate": 4.856309552802804e-06, "loss": 0.3099, "step": 1430 }, { "epoch": 0.27, "grad_norm": 13.5, "learning_rate": 4.856053510748321e-06, "loss": 0.3301, "step": 1431 }, { "epoch": 0.27, "grad_norm": 16.125, "learning_rate": 4.8557972475380735e-06, "loss": 0.4157, "step": 1432 }, { "epoch": 0.27, "grad_norm": 17.0, "learning_rate": 4.8555407631961145e-06, "loss": 0.2877, "step": 1433 }, { "epoch": 0.27, "grad_norm": 25.0, "learning_rate": 4.855284057746521e-06, "loss": 0.2984, "step": 1434 }, { "epoch": 0.27, "grad_norm": 11.3125, "learning_rate": 4.855027131213388e-06, "loss": 0.3335, "step": 1435 }, { "epoch": 0.27, "grad_norm": 8.9375, "learning_rate": 4.854769983620833e-06, "loss": 0.2742, "step": 1436 }, { "epoch": 0.27, "grad_norm": 14.5625, "learning_rate": 4.854512614992993e-06, "loss": 0.3749, "step": 1437 }, { "epoch": 0.27, "grad_norm": 11.5, "learning_rate": 4.854255025354028e-06, "loss": 0.2961, "step": 1438 }, { "epoch": 0.27, "grad_norm": 13.0625, "learning_rate": 4.853997214728115e-06, "loss": 0.3224, "step": 1439 }, { "epoch": 0.27, "grad_norm": 10.25, "learning_rate": 4.853739183139456e-06, "loss": 0.2545, "step": 1440 }, { "epoch": 0.27, "grad_norm": 10.9375, "learning_rate": 4.8534809306122696e-06, "loss": 0.3695, "step": 1441 }, { "epoch": 0.27, "grad_norm": 18.0, "learning_rate": 4.853222457170798e-06, "loss": 0.3973, "step": 1442 }, { "epoch": 0.27, "grad_norm": 10.5, "learning_rate": 4.852963762839304e-06, "loss": 0.2594, "step": 1443 }, { "epoch": 0.27, "grad_norm": 14.4375, "learning_rate": 4.852704847642071e-06, "loss": 0.2945, "step": 1444 }, { "epoch": 0.27, "grad_norm": 10.125, "learning_rate": 4.852445711603401e-06, "loss": 0.3025, "step": 1445 }, { "epoch": 0.27, "grad_norm": 12.0, "learning_rate": 4.8521863547476185e-06, "loss": 0.312, "step": 1446 }, { "epoch": 0.27, "grad_norm": 10.125, "learning_rate": 4.85192677709907e-06, "loss": 0.385, "step": 1447 }, { "epoch": 0.27, "grad_norm": 11.4375, "learning_rate": 4.851666978682119e-06, "loss": 0.4469, "step": 1448 }, { "epoch": 0.27, "grad_norm": 15.3125, "learning_rate": 4.851406959521154e-06, "loss": 0.334, "step": 1449 }, { "epoch": 0.27, "grad_norm": 16.25, "learning_rate": 4.851146719640581e-06, "loss": 0.2565, "step": 1450 }, { "epoch": 0.27, "grad_norm": 9.1875, "learning_rate": 4.85088625906483e-06, "loss": 0.3235, "step": 1451 }, { "epoch": 0.27, "grad_norm": 13.0, "learning_rate": 4.850625577818348e-06, "loss": 0.374, "step": 1452 }, { "epoch": 0.27, "grad_norm": 10.625, "learning_rate": 4.850364675925603e-06, "loss": 0.3386, "step": 1453 }, { "epoch": 0.28, "grad_norm": 10.0, "learning_rate": 4.850103553411087e-06, "loss": 0.2176, "step": 1454 }, { "epoch": 0.28, "grad_norm": 7.90625, "learning_rate": 4.849842210299311e-06, "loss": 0.3134, "step": 1455 }, { "epoch": 0.28, "grad_norm": 11.3125, "learning_rate": 4.849580646614807e-06, "loss": 0.295, "step": 1456 }, { "epoch": 0.28, "grad_norm": 11.9375, "learning_rate": 4.849318862382125e-06, "loss": 0.2975, "step": 1457 }, { "epoch": 0.28, "grad_norm": 10.875, "learning_rate": 4.849056857625839e-06, "loss": 0.3641, "step": 1458 }, { "epoch": 0.28, "grad_norm": 11.3125, "learning_rate": 4.848794632370542e-06, "loss": 0.304, "step": 1459 }, { "epoch": 0.28, "grad_norm": 10.4375, "learning_rate": 4.848532186640851e-06, "loss": 0.315, "step": 1460 }, { "epoch": 0.28, "grad_norm": 14.6875, "learning_rate": 4.848269520461397e-06, "loss": 0.346, "step": 1461 }, { "epoch": 0.28, "grad_norm": 14.5625, "learning_rate": 4.848006633856839e-06, "loss": 0.275, "step": 1462 }, { "epoch": 0.28, "grad_norm": 10.6875, "learning_rate": 4.847743526851852e-06, "loss": 0.2867, "step": 1463 }, { "epoch": 0.28, "grad_norm": 11.0, "learning_rate": 4.847480199471132e-06, "loss": 0.3251, "step": 1464 }, { "epoch": 0.28, "grad_norm": 9.375, "learning_rate": 4.847216651739399e-06, "loss": 0.3106, "step": 1465 }, { "epoch": 0.28, "grad_norm": 10.5625, "learning_rate": 4.8469528836813895e-06, "loss": 0.3467, "step": 1466 }, { "epoch": 0.28, "grad_norm": 10.6875, "learning_rate": 4.846688895321864e-06, "loss": 0.3683, "step": 1467 }, { "epoch": 0.28, "grad_norm": 12.8125, "learning_rate": 4.846424686685601e-06, "loss": 0.2844, "step": 1468 }, { "epoch": 0.28, "grad_norm": 10.3125, "learning_rate": 4.846160257797403e-06, "loss": 0.4096, "step": 1469 }, { "epoch": 0.28, "grad_norm": 10.6875, "learning_rate": 4.84589560868209e-06, "loss": 0.337, "step": 1470 }, { "epoch": 0.28, "grad_norm": 11.125, "learning_rate": 4.845630739364503e-06, "loss": 0.2659, "step": 1471 }, { "epoch": 0.28, "grad_norm": 21.75, "learning_rate": 4.845365649869505e-06, "loss": 0.4811, "step": 1472 }, { "epoch": 0.28, "grad_norm": 12.25, "learning_rate": 4.84510034022198e-06, "loss": 0.3724, "step": 1473 }, { "epoch": 0.28, "grad_norm": 10.75, "learning_rate": 4.8448348104468315e-06, "loss": 0.2854, "step": 1474 }, { "epoch": 0.28, "grad_norm": 38.75, "learning_rate": 4.8445690605689835e-06, "loss": 0.4101, "step": 1475 }, { "epoch": 0.28, "grad_norm": 8.5625, "learning_rate": 4.84430309061338e-06, "loss": 0.332, "step": 1476 }, { "epoch": 0.28, "grad_norm": 10.25, "learning_rate": 4.84403690060499e-06, "loss": 0.3537, "step": 1477 }, { "epoch": 0.28, "grad_norm": 12.9375, "learning_rate": 4.8437704905687974e-06, "loss": 0.3573, "step": 1478 }, { "epoch": 0.28, "grad_norm": 11.3125, "learning_rate": 4.84350386052981e-06, "loss": 0.3039, "step": 1479 }, { "epoch": 0.28, "grad_norm": 15.625, "learning_rate": 4.843237010513057e-06, "loss": 0.3554, "step": 1480 }, { "epoch": 0.28, "grad_norm": 8.375, "learning_rate": 4.842969940543584e-06, "loss": 0.3213, "step": 1481 }, { "epoch": 0.28, "grad_norm": 11.4375, "learning_rate": 4.842702650646462e-06, "loss": 0.3546, "step": 1482 }, { "epoch": 0.28, "grad_norm": 7.84375, "learning_rate": 4.842435140846779e-06, "loss": 0.2854, "step": 1483 }, { "epoch": 0.28, "grad_norm": 15.1875, "learning_rate": 4.8421674111696485e-06, "loss": 0.2382, "step": 1484 }, { "epoch": 0.28, "grad_norm": 23.0, "learning_rate": 4.841899461640199e-06, "loss": 0.353, "step": 1485 }, { "epoch": 0.28, "grad_norm": 10.3125, "learning_rate": 4.841631292283583e-06, "loss": 0.3516, "step": 1486 }, { "epoch": 0.28, "grad_norm": 18.25, "learning_rate": 4.841362903124972e-06, "loss": 0.2777, "step": 1487 }, { "epoch": 0.28, "grad_norm": 10.4375, "learning_rate": 4.84109429418956e-06, "loss": 0.2916, "step": 1488 }, { "epoch": 0.28, "grad_norm": 30.0, "learning_rate": 4.84082546550256e-06, "loss": 0.3325, "step": 1489 }, { "epoch": 0.28, "grad_norm": 8.8125, "learning_rate": 4.840556417089205e-06, "loss": 0.3497, "step": 1490 }, { "epoch": 0.28, "grad_norm": 10.1875, "learning_rate": 4.8402871489747516e-06, "loss": 0.2922, "step": 1491 }, { "epoch": 0.28, "grad_norm": 12.75, "learning_rate": 4.840017661184474e-06, "loss": 0.3882, "step": 1492 }, { "epoch": 0.28, "grad_norm": 8.375, "learning_rate": 4.839747953743669e-06, "loss": 0.2806, "step": 1493 }, { "epoch": 0.28, "grad_norm": 13.6875, "learning_rate": 4.839478026677653e-06, "loss": 0.3467, "step": 1494 }, { "epoch": 0.28, "grad_norm": 10.125, "learning_rate": 4.839207880011763e-06, "loss": 0.2925, "step": 1495 }, { "epoch": 0.28, "grad_norm": 13.8125, "learning_rate": 4.838937513771358e-06, "loss": 0.293, "step": 1496 }, { "epoch": 0.28, "grad_norm": 10.125, "learning_rate": 4.838666927981814e-06, "loss": 0.2868, "step": 1497 }, { "epoch": 0.28, "grad_norm": 13.375, "learning_rate": 4.838396122668532e-06, "loss": 0.255, "step": 1498 }, { "epoch": 0.28, "grad_norm": 10.5, "learning_rate": 4.8381250978569316e-06, "loss": 0.3332, "step": 1499 }, { "epoch": 0.28, "grad_norm": 10.5625, "learning_rate": 4.837853853572453e-06, "loss": 0.2802, "step": 1500 }, { "epoch": 0.28, "grad_norm": 13.0, "learning_rate": 4.837582389840556e-06, "loss": 0.4288, "step": 1501 }, { "epoch": 0.28, "grad_norm": 9.8125, "learning_rate": 4.837310706686724e-06, "loss": 0.3277, "step": 1502 }, { "epoch": 0.28, "grad_norm": 8.6875, "learning_rate": 4.837038804136458e-06, "loss": 0.3538, "step": 1503 }, { "epoch": 0.28, "grad_norm": 10.1875, "learning_rate": 4.83676668221528e-06, "loss": 0.2868, "step": 1504 }, { "epoch": 0.28, "grad_norm": 16.625, "learning_rate": 4.836494340948734e-06, "loss": 0.299, "step": 1505 }, { "epoch": 0.28, "grad_norm": 13.1875, "learning_rate": 4.836221780362385e-06, "loss": 0.3918, "step": 1506 }, { "epoch": 0.29, "grad_norm": 11.5, "learning_rate": 4.835949000481816e-06, "loss": 0.3574, "step": 1507 }, { "epoch": 0.29, "grad_norm": 14.375, "learning_rate": 4.835676001332632e-06, "loss": 0.2275, "step": 1508 }, { "epoch": 0.29, "grad_norm": 13.6875, "learning_rate": 4.8354027829404595e-06, "loss": 0.3562, "step": 1509 }, { "epoch": 0.29, "grad_norm": 13.0, "learning_rate": 4.835129345330943e-06, "loss": 0.3724, "step": 1510 }, { "epoch": 0.29, "grad_norm": 12.8125, "learning_rate": 4.834855688529752e-06, "loss": 0.294, "step": 1511 }, { "epoch": 0.29, "grad_norm": 9.3125, "learning_rate": 4.834581812562571e-06, "loss": 0.3409, "step": 1512 }, { "epoch": 0.29, "grad_norm": 8.75, "learning_rate": 4.83430771745511e-06, "loss": 0.357, "step": 1513 }, { "epoch": 0.29, "grad_norm": 11.0625, "learning_rate": 4.8340334032330975e-06, "loss": 0.3311, "step": 1514 }, { "epoch": 0.29, "grad_norm": 9.5, "learning_rate": 4.833758869922282e-06, "loss": 0.3253, "step": 1515 }, { "epoch": 0.29, "grad_norm": 10.0, "learning_rate": 4.833484117548431e-06, "loss": 0.3115, "step": 1516 }, { "epoch": 0.29, "grad_norm": 11.0, "learning_rate": 4.8332091461373384e-06, "loss": 0.3725, "step": 1517 }, { "epoch": 0.29, "grad_norm": 11.3125, "learning_rate": 4.832933955714812e-06, "loss": 0.323, "step": 1518 }, { "epoch": 0.29, "grad_norm": 9.6875, "learning_rate": 4.832658546306686e-06, "loss": 0.2491, "step": 1519 }, { "epoch": 0.29, "grad_norm": 16.125, "learning_rate": 4.832382917938809e-06, "loss": 0.353, "step": 1520 }, { "epoch": 0.29, "grad_norm": 9.875, "learning_rate": 4.8321070706370555e-06, "loss": 0.2483, "step": 1521 }, { "epoch": 0.29, "grad_norm": 12.5625, "learning_rate": 4.831831004427318e-06, "loss": 0.368, "step": 1522 }, { "epoch": 0.29, "grad_norm": 8.9375, "learning_rate": 4.83155471933551e-06, "loss": 0.241, "step": 1523 }, { "epoch": 0.29, "grad_norm": 7.96875, "learning_rate": 4.831278215387566e-06, "loss": 0.2913, "step": 1524 }, { "epoch": 0.29, "grad_norm": 13.375, "learning_rate": 4.831001492609439e-06, "loss": 0.3498, "step": 1525 }, { "epoch": 0.29, "grad_norm": 11.3125, "learning_rate": 4.830724551027107e-06, "loss": 0.3243, "step": 1526 }, { "epoch": 0.29, "grad_norm": 10.25, "learning_rate": 4.830447390666563e-06, "loss": 0.2891, "step": 1527 }, { "epoch": 0.29, "grad_norm": 12.8125, "learning_rate": 4.8301700115538245e-06, "loss": 0.3151, "step": 1528 }, { "epoch": 0.29, "grad_norm": 13.0625, "learning_rate": 4.829892413714928e-06, "loss": 0.301, "step": 1529 }, { "epoch": 0.29, "grad_norm": 11.1875, "learning_rate": 4.82961459717593e-06, "loss": 0.304, "step": 1530 }, { "epoch": 0.29, "grad_norm": 7.46875, "learning_rate": 4.829336561962909e-06, "loss": 0.275, "step": 1531 }, { "epoch": 0.29, "grad_norm": 17.75, "learning_rate": 4.829058308101965e-06, "loss": 0.3358, "step": 1532 }, { "epoch": 0.29, "grad_norm": 10.8125, "learning_rate": 4.828779835619214e-06, "loss": 0.2759, "step": 1533 }, { "epoch": 0.29, "grad_norm": 19.75, "learning_rate": 4.828501144540797e-06, "loss": 0.331, "step": 1534 }, { "epoch": 0.29, "grad_norm": 12.375, "learning_rate": 4.828222234892874e-06, "loss": 0.4063, "step": 1535 }, { "epoch": 0.29, "grad_norm": 13.0625, "learning_rate": 4.827943106701625e-06, "loss": 0.3329, "step": 1536 }, { "epoch": 0.29, "grad_norm": 10.1875, "learning_rate": 4.82766375999325e-06, "loss": 0.3324, "step": 1537 }, { "epoch": 0.29, "grad_norm": 16.375, "learning_rate": 4.827384194793973e-06, "loss": 0.4254, "step": 1538 }, { "epoch": 0.29, "grad_norm": 11.6875, "learning_rate": 4.827104411130032e-06, "loss": 0.3172, "step": 1539 }, { "epoch": 0.29, "grad_norm": 9.1875, "learning_rate": 4.826824409027694e-06, "loss": 0.2392, "step": 1540 }, { "epoch": 0.29, "grad_norm": 12.6875, "learning_rate": 4.8265441885132385e-06, "loss": 0.3195, "step": 1541 }, { "epoch": 0.29, "grad_norm": 6.40625, "learning_rate": 4.82626374961297e-06, "loss": 0.2705, "step": 1542 }, { "epoch": 0.29, "grad_norm": 10.3125, "learning_rate": 4.8259830923532146e-06, "loss": 0.3867, "step": 1543 }, { "epoch": 0.29, "grad_norm": 11.5, "learning_rate": 4.8257022167603135e-06, "loss": 0.3006, "step": 1544 }, { "epoch": 0.29, "grad_norm": 15.25, "learning_rate": 4.825421122860633e-06, "loss": 0.3325, "step": 1545 }, { "epoch": 0.29, "grad_norm": 11.25, "learning_rate": 4.825139810680559e-06, "loss": 0.3986, "step": 1546 }, { "epoch": 0.29, "grad_norm": 20.75, "learning_rate": 4.824858280246497e-06, "loss": 0.4261, "step": 1547 }, { "epoch": 0.29, "grad_norm": 10.0625, "learning_rate": 4.824576531584874e-06, "loss": 0.2618, "step": 1548 }, { "epoch": 0.29, "grad_norm": 15.875, "learning_rate": 4.824294564722135e-06, "loss": 0.378, "step": 1549 }, { "epoch": 0.29, "grad_norm": 13.3125, "learning_rate": 4.824012379684749e-06, "loss": 0.3853, "step": 1550 }, { "epoch": 0.29, "grad_norm": 13.5, "learning_rate": 4.823729976499204e-06, "loss": 0.2729, "step": 1551 }, { "epoch": 0.29, "grad_norm": 8.6875, "learning_rate": 4.823447355192009e-06, "loss": 0.3157, "step": 1552 }, { "epoch": 0.29, "grad_norm": 10.75, "learning_rate": 4.823164515789691e-06, "loss": 0.3512, "step": 1553 }, { "epoch": 0.29, "grad_norm": 11.5, "learning_rate": 4.8228814583188e-06, "loss": 0.3102, "step": 1554 }, { "epoch": 0.29, "grad_norm": 8.375, "learning_rate": 4.822598182805906e-06, "loss": 0.3114, "step": 1555 }, { "epoch": 0.29, "grad_norm": 9.5, "learning_rate": 4.822314689277599e-06, "loss": 0.3443, "step": 1556 }, { "epoch": 0.29, "grad_norm": 11.8125, "learning_rate": 4.8220309777604905e-06, "loss": 0.3642, "step": 1557 }, { "epoch": 0.29, "grad_norm": 10.875, "learning_rate": 4.82174704828121e-06, "loss": 0.3716, "step": 1558 }, { "epoch": 0.29, "grad_norm": 10.1875, "learning_rate": 4.8214629008664105e-06, "loss": 0.341, "step": 1559 }, { "epoch": 0.3, "grad_norm": 12.9375, "learning_rate": 4.821178535542764e-06, "loss": 0.3212, "step": 1560 }, { "epoch": 0.3, "grad_norm": 12.5625, "learning_rate": 4.8208939523369615e-06, "loss": 0.3499, "step": 1561 }, { "epoch": 0.3, "grad_norm": 10.6875, "learning_rate": 4.8206091512757184e-06, "loss": 0.3241, "step": 1562 }, { "epoch": 0.3, "grad_norm": 14.0, "learning_rate": 4.820324132385766e-06, "loss": 0.318, "step": 1563 }, { "epoch": 0.3, "grad_norm": 12.1875, "learning_rate": 4.82003889569386e-06, "loss": 0.2494, "step": 1564 }, { "epoch": 0.3, "grad_norm": 9.3125, "learning_rate": 4.819753441226774e-06, "loss": 0.2834, "step": 1565 }, { "epoch": 0.3, "grad_norm": 11.5625, "learning_rate": 4.8194677690113024e-06, "loss": 0.3096, "step": 1566 }, { "epoch": 0.3, "grad_norm": 10.625, "learning_rate": 4.81918187907426e-06, "loss": 0.3073, "step": 1567 }, { "epoch": 0.3, "grad_norm": 11.0625, "learning_rate": 4.818895771442483e-06, "loss": 0.2938, "step": 1568 }, { "epoch": 0.3, "grad_norm": 10.25, "learning_rate": 4.818609446142828e-06, "loss": 0.3246, "step": 1569 }, { "epoch": 0.3, "grad_norm": 16.0, "learning_rate": 4.818322903202171e-06, "loss": 0.3965, "step": 1570 }, { "epoch": 0.3, "grad_norm": 12.75, "learning_rate": 4.818036142647409e-06, "loss": 0.3532, "step": 1571 }, { "epoch": 0.3, "grad_norm": 9.5, "learning_rate": 4.817749164505459e-06, "loss": 0.3584, "step": 1572 }, { "epoch": 0.3, "grad_norm": 14.875, "learning_rate": 4.817461968803259e-06, "loss": 0.2565, "step": 1573 }, { "epoch": 0.3, "grad_norm": 12.0625, "learning_rate": 4.817174555567768e-06, "loss": 0.288, "step": 1574 }, { "epoch": 0.3, "grad_norm": 10.75, "learning_rate": 4.816886924825963e-06, "loss": 0.3012, "step": 1575 }, { "epoch": 0.3, "grad_norm": 18.25, "learning_rate": 4.8165990766048444e-06, "loss": 0.308, "step": 1576 }, { "epoch": 0.3, "grad_norm": 18.5, "learning_rate": 4.816311010931431e-06, "loss": 0.2944, "step": 1577 }, { "epoch": 0.3, "grad_norm": 9.0, "learning_rate": 4.816022727832764e-06, "loss": 0.2536, "step": 1578 }, { "epoch": 0.3, "grad_norm": 13.5, "learning_rate": 4.815734227335902e-06, "loss": 0.326, "step": 1579 }, { "epoch": 0.3, "grad_norm": 10.8125, "learning_rate": 4.815445509467927e-06, "loss": 0.2823, "step": 1580 }, { "epoch": 0.3, "grad_norm": 16.25, "learning_rate": 4.815156574255938e-06, "loss": 0.3271, "step": 1581 }, { "epoch": 0.3, "grad_norm": 9.5, "learning_rate": 4.814867421727059e-06, "loss": 0.3752, "step": 1582 }, { "epoch": 0.3, "grad_norm": 9.8125, "learning_rate": 4.81457805190843e-06, "loss": 0.3208, "step": 1583 }, { "epoch": 0.3, "grad_norm": 11.0625, "learning_rate": 4.8142884648272135e-06, "loss": 0.3788, "step": 1584 }, { "epoch": 0.3, "grad_norm": 9.8125, "learning_rate": 4.813998660510594e-06, "loss": 0.3214, "step": 1585 }, { "epoch": 0.3, "grad_norm": 12.3125, "learning_rate": 4.813708638985772e-06, "loss": 0.3317, "step": 1586 }, { "epoch": 0.3, "grad_norm": 34.0, "learning_rate": 4.813418400279974e-06, "loss": 0.3462, "step": 1587 }, { "epoch": 0.3, "grad_norm": 10.75, "learning_rate": 4.813127944420441e-06, "loss": 0.2791, "step": 1588 }, { "epoch": 0.3, "grad_norm": 8.6875, "learning_rate": 4.812837271434438e-06, "loss": 0.2921, "step": 1589 }, { "epoch": 0.3, "grad_norm": 10.625, "learning_rate": 4.812546381349251e-06, "loss": 0.3458, "step": 1590 }, { "epoch": 0.3, "grad_norm": 17.0, "learning_rate": 4.812255274192183e-06, "loss": 0.3903, "step": 1591 }, { "epoch": 0.3, "grad_norm": 11.875, "learning_rate": 4.811963949990561e-06, "loss": 0.3496, "step": 1592 }, { "epoch": 0.3, "grad_norm": 12.75, "learning_rate": 4.811672408771731e-06, "loss": 0.344, "step": 1593 }, { "epoch": 0.3, "grad_norm": 13.625, "learning_rate": 4.811380650563057e-06, "loss": 0.3818, "step": 1594 }, { "epoch": 0.3, "grad_norm": 10.3125, "learning_rate": 4.8110886753919265e-06, "loss": 0.2778, "step": 1595 }, { "epoch": 0.3, "grad_norm": 11.25, "learning_rate": 4.810796483285746e-06, "loss": 0.2894, "step": 1596 }, { "epoch": 0.3, "grad_norm": 9.8125, "learning_rate": 4.810504074271944e-06, "loss": 0.2923, "step": 1597 }, { "epoch": 0.3, "grad_norm": 12.4375, "learning_rate": 4.8102114483779674e-06, "loss": 0.3117, "step": 1598 }, { "epoch": 0.3, "grad_norm": 8.75, "learning_rate": 4.809918605631284e-06, "loss": 0.2722, "step": 1599 }, { "epoch": 0.3, "grad_norm": 13.1875, "learning_rate": 4.809625546059382e-06, "loss": 0.3271, "step": 1600 }, { "epoch": 0.3, "grad_norm": 13.8125, "learning_rate": 4.809332269689769e-06, "loss": 0.3177, "step": 1601 }, { "epoch": 0.3, "grad_norm": 7.96875, "learning_rate": 4.809038776549977e-06, "loss": 0.3067, "step": 1602 }, { "epoch": 0.3, "grad_norm": 14.75, "learning_rate": 4.808745066667552e-06, "loss": 0.3453, "step": 1603 }, { "epoch": 0.3, "grad_norm": 10.75, "learning_rate": 4.808451140070066e-06, "loss": 0.3122, "step": 1604 }, { "epoch": 0.3, "grad_norm": 9.5625, "learning_rate": 4.8081569967851075e-06, "loss": 0.305, "step": 1605 }, { "epoch": 0.3, "grad_norm": 9.125, "learning_rate": 4.807862636840288e-06, "loss": 0.2353, "step": 1606 }, { "epoch": 0.3, "grad_norm": 13.0, "learning_rate": 4.807568060263237e-06, "loss": 0.3798, "step": 1607 }, { "epoch": 0.3, "grad_norm": 14.25, "learning_rate": 4.807273267081607e-06, "loss": 0.4736, "step": 1608 }, { "epoch": 0.3, "grad_norm": 11.25, "learning_rate": 4.806978257323068e-06, "loss": 0.405, "step": 1609 }, { "epoch": 0.3, "grad_norm": 9.5, "learning_rate": 4.806683031015313e-06, "loss": 0.2826, "step": 1610 }, { "epoch": 0.3, "grad_norm": 11.6875, "learning_rate": 4.806387588186052e-06, "loss": 0.2784, "step": 1611 }, { "epoch": 0.3, "grad_norm": 12.125, "learning_rate": 4.806091928863019e-06, "loss": 0.3573, "step": 1612 }, { "epoch": 0.31, "grad_norm": 7.6875, "learning_rate": 4.805796053073967e-06, "loss": 0.2532, "step": 1613 }, { "epoch": 0.31, "grad_norm": 15.5, "learning_rate": 4.805499960846668e-06, "loss": 0.3437, "step": 1614 }, { "epoch": 0.31, "grad_norm": 10.875, "learning_rate": 4.805203652208915e-06, "loss": 0.3656, "step": 1615 }, { "epoch": 0.31, "grad_norm": 8.3125, "learning_rate": 4.804907127188523e-06, "loss": 0.3271, "step": 1616 }, { "epoch": 0.31, "grad_norm": 11.5, "learning_rate": 4.804610385813325e-06, "loss": 0.3611, "step": 1617 }, { "epoch": 0.31, "grad_norm": 15.125, "learning_rate": 4.804313428111175e-06, "loss": 0.4206, "step": 1618 }, { "epoch": 0.31, "grad_norm": 8.375, "learning_rate": 4.804016254109947e-06, "loss": 0.3165, "step": 1619 }, { "epoch": 0.31, "grad_norm": 10.375, "learning_rate": 4.803718863837538e-06, "loss": 0.3766, "step": 1620 }, { "epoch": 0.31, "grad_norm": 10.5625, "learning_rate": 4.803421257321862e-06, "loss": 0.3408, "step": 1621 }, { "epoch": 0.31, "grad_norm": 9.5, "learning_rate": 4.8031234345908545e-06, "loss": 0.3158, "step": 1622 }, { "epoch": 0.31, "grad_norm": 10.75, "learning_rate": 4.80282539567247e-06, "loss": 0.2601, "step": 1623 }, { "epoch": 0.31, "grad_norm": 10.8125, "learning_rate": 4.802527140594687e-06, "loss": 0.2427, "step": 1624 }, { "epoch": 0.31, "grad_norm": 21.75, "learning_rate": 4.802228669385499e-06, "loss": 0.2875, "step": 1625 }, { "epoch": 0.31, "grad_norm": 17.0, "learning_rate": 4.801929982072925e-06, "loss": 0.3956, "step": 1626 }, { "epoch": 0.31, "grad_norm": 12.875, "learning_rate": 4.801631078685001e-06, "loss": 0.3888, "step": 1627 }, { "epoch": 0.31, "grad_norm": 7.71875, "learning_rate": 4.801331959249783e-06, "loss": 0.3017, "step": 1628 }, { "epoch": 0.31, "grad_norm": 12.9375, "learning_rate": 4.80103262379535e-06, "loss": 0.2783, "step": 1629 }, { "epoch": 0.31, "grad_norm": 11.8125, "learning_rate": 4.800733072349799e-06, "loss": 0.3349, "step": 1630 }, { "epoch": 0.31, "grad_norm": 12.3125, "learning_rate": 4.80043330494125e-06, "loss": 0.4224, "step": 1631 }, { "epoch": 0.31, "grad_norm": 12.4375, "learning_rate": 4.800133321597838e-06, "loss": 0.287, "step": 1632 }, { "epoch": 0.31, "grad_norm": 12.1875, "learning_rate": 4.799833122347723e-06, "loss": 0.2686, "step": 1633 }, { "epoch": 0.31, "grad_norm": 12.5625, "learning_rate": 4.799532707219085e-06, "loss": 0.3239, "step": 1634 }, { "epoch": 0.31, "grad_norm": 16.75, "learning_rate": 4.799232076240121e-06, "loss": 0.3463, "step": 1635 }, { "epoch": 0.31, "grad_norm": 11.25, "learning_rate": 4.7989312294390515e-06, "loss": 0.399, "step": 1636 }, { "epoch": 0.31, "grad_norm": 12.0, "learning_rate": 4.798630166844116e-06, "loss": 0.3543, "step": 1637 }, { "epoch": 0.31, "grad_norm": 13.5, "learning_rate": 4.7983288884835745e-06, "loss": 0.2725, "step": 1638 }, { "epoch": 0.31, "grad_norm": 9.3125, "learning_rate": 4.798027394385706e-06, "loss": 0.2735, "step": 1639 }, { "epoch": 0.31, "grad_norm": 15.1875, "learning_rate": 4.7977256845788134e-06, "loss": 0.4311, "step": 1640 }, { "epoch": 0.31, "grad_norm": 11.5625, "learning_rate": 4.797423759091214e-06, "loss": 0.2921, "step": 1641 }, { "epoch": 0.31, "grad_norm": 11.6875, "learning_rate": 4.797121617951251e-06, "loss": 0.346, "step": 1642 }, { "epoch": 0.31, "grad_norm": 10.25, "learning_rate": 4.796819261187286e-06, "loss": 0.3338, "step": 1643 }, { "epoch": 0.31, "grad_norm": 11.3125, "learning_rate": 4.796516688827698e-06, "loss": 0.3123, "step": 1644 }, { "epoch": 0.31, "grad_norm": 9.75, "learning_rate": 4.79621390090089e-06, "loss": 0.3064, "step": 1645 }, { "epoch": 0.31, "grad_norm": 12.3125, "learning_rate": 4.795910897435283e-06, "loss": 0.2568, "step": 1646 }, { "epoch": 0.31, "grad_norm": 26.625, "learning_rate": 4.79560767845932e-06, "loss": 0.3558, "step": 1647 }, { "epoch": 0.31, "grad_norm": 13.9375, "learning_rate": 4.795304244001462e-06, "loss": 0.4237, "step": 1648 }, { "epoch": 0.31, "grad_norm": 18.625, "learning_rate": 4.795000594090193e-06, "loss": 0.4244, "step": 1649 }, { "epoch": 0.31, "grad_norm": 19.25, "learning_rate": 4.794696728754016e-06, "loss": 0.2985, "step": 1650 }, { "epoch": 0.31, "grad_norm": 10.4375, "learning_rate": 4.794392648021452e-06, "loss": 0.3029, "step": 1651 }, { "epoch": 0.31, "grad_norm": 17.375, "learning_rate": 4.794088351921044e-06, "loss": 0.2557, "step": 1652 }, { "epoch": 0.31, "grad_norm": 8.5625, "learning_rate": 4.793783840481358e-06, "loss": 0.3022, "step": 1653 }, { "epoch": 0.31, "grad_norm": 9.8125, "learning_rate": 4.793479113730977e-06, "loss": 0.3025, "step": 1654 }, { "epoch": 0.31, "grad_norm": 12.6875, "learning_rate": 4.793174171698503e-06, "loss": 0.2872, "step": 1655 }, { "epoch": 0.31, "grad_norm": 14.125, "learning_rate": 4.79286901441256e-06, "loss": 0.3221, "step": 1656 }, { "epoch": 0.31, "grad_norm": 14.3125, "learning_rate": 4.792563641901794e-06, "loss": 0.4114, "step": 1657 }, { "epoch": 0.31, "grad_norm": 13.5625, "learning_rate": 4.792258054194868e-06, "loss": 0.414, "step": 1658 }, { "epoch": 0.31, "grad_norm": 13.6875, "learning_rate": 4.791952251320468e-06, "loss": 0.273, "step": 1659 }, { "epoch": 0.31, "grad_norm": 13.125, "learning_rate": 4.791646233307296e-06, "loss": 0.3424, "step": 1660 }, { "epoch": 0.31, "grad_norm": 9.5625, "learning_rate": 4.791340000184081e-06, "loss": 0.3141, "step": 1661 }, { "epoch": 0.31, "grad_norm": 8.6875, "learning_rate": 4.7910335519795654e-06, "loss": 0.2749, "step": 1662 }, { "epoch": 0.31, "grad_norm": 22.0, "learning_rate": 4.7907268887225156e-06, "loss": 0.3621, "step": 1663 }, { "epoch": 0.31, "grad_norm": 7.84375, "learning_rate": 4.790420010441716e-06, "loss": 0.2539, "step": 1664 }, { "epoch": 0.31, "grad_norm": 7.28125, "learning_rate": 4.790112917165975e-06, "loss": 0.3234, "step": 1665 }, { "epoch": 0.32, "grad_norm": 9.0625, "learning_rate": 4.789805608924115e-06, "loss": 0.333, "step": 1666 }, { "epoch": 0.32, "grad_norm": 7.71875, "learning_rate": 4.789498085744984e-06, "loss": 0.3196, "step": 1667 }, { "epoch": 0.32, "grad_norm": 11.375, "learning_rate": 4.78919034765745e-06, "loss": 0.2761, "step": 1668 }, { "epoch": 0.32, "grad_norm": 11.1875, "learning_rate": 4.7888823946903965e-06, "loss": 0.3165, "step": 1669 }, { "epoch": 0.32, "grad_norm": 13.125, "learning_rate": 4.788574226872731e-06, "loss": 0.3799, "step": 1670 }, { "epoch": 0.32, "grad_norm": 16.875, "learning_rate": 4.788265844233381e-06, "loss": 0.3533, "step": 1671 }, { "epoch": 0.32, "grad_norm": 9.375, "learning_rate": 4.787957246801294e-06, "loss": 0.3231, "step": 1672 }, { "epoch": 0.32, "grad_norm": 12.75, "learning_rate": 4.7876484346054344e-06, "loss": 0.3543, "step": 1673 }, { "epoch": 0.32, "grad_norm": 11.4375, "learning_rate": 4.787339407674792e-06, "loss": 0.2906, "step": 1674 }, { "epoch": 0.32, "grad_norm": 11.375, "learning_rate": 4.787030166038374e-06, "loss": 0.3434, "step": 1675 }, { "epoch": 0.32, "grad_norm": 12.875, "learning_rate": 4.786720709725207e-06, "loss": 0.4326, "step": 1676 }, { "epoch": 0.32, "grad_norm": 12.375, "learning_rate": 4.786411038764339e-06, "loss": 0.3462, "step": 1677 }, { "epoch": 0.32, "grad_norm": 12.3125, "learning_rate": 4.786101153184839e-06, "loss": 0.3304, "step": 1678 }, { "epoch": 0.32, "grad_norm": 9.5625, "learning_rate": 4.785791053015793e-06, "loss": 0.2917, "step": 1679 }, { "epoch": 0.32, "grad_norm": 13.4375, "learning_rate": 4.785480738286312e-06, "loss": 0.2467, "step": 1680 }, { "epoch": 0.32, "grad_norm": 9.9375, "learning_rate": 4.785170209025521e-06, "loss": 0.2957, "step": 1681 }, { "epoch": 0.32, "grad_norm": 23.375, "learning_rate": 4.784859465262572e-06, "loss": 0.3718, "step": 1682 }, { "epoch": 0.32, "grad_norm": 8.0, "learning_rate": 4.78454850702663e-06, "loss": 0.2855, "step": 1683 }, { "epoch": 0.32, "grad_norm": 19.25, "learning_rate": 4.784237334346887e-06, "loss": 0.3409, "step": 1684 }, { "epoch": 0.32, "grad_norm": 35.75, "learning_rate": 4.78392594725255e-06, "loss": 0.3537, "step": 1685 }, { "epoch": 0.32, "grad_norm": 17.375, "learning_rate": 4.783614345772848e-06, "loss": 0.3099, "step": 1686 }, { "epoch": 0.32, "grad_norm": 12.625, "learning_rate": 4.78330252993703e-06, "loss": 0.2873, "step": 1687 }, { "epoch": 0.32, "grad_norm": 18.875, "learning_rate": 4.782990499774366e-06, "loss": 0.3164, "step": 1688 }, { "epoch": 0.32, "grad_norm": 15.875, "learning_rate": 4.7826782553141454e-06, "loss": 0.3803, "step": 1689 }, { "epoch": 0.32, "grad_norm": 9.75, "learning_rate": 4.782365796585678e-06, "loss": 0.2534, "step": 1690 }, { "epoch": 0.32, "grad_norm": 15.625, "learning_rate": 4.782053123618291e-06, "loss": 0.3867, "step": 1691 }, { "epoch": 0.32, "grad_norm": 9.5, "learning_rate": 4.7817402364413365e-06, "loss": 0.307, "step": 1692 }, { "epoch": 0.32, "grad_norm": 11.375, "learning_rate": 4.781427135084184e-06, "loss": 0.2422, "step": 1693 }, { "epoch": 0.32, "grad_norm": 9.875, "learning_rate": 4.781113819576222e-06, "loss": 0.2611, "step": 1694 }, { "epoch": 0.32, "grad_norm": 11.4375, "learning_rate": 4.780800289946861e-06, "loss": 0.2631, "step": 1695 }, { "epoch": 0.32, "grad_norm": 13.75, "learning_rate": 4.780486546225533e-06, "loss": 0.3301, "step": 1696 }, { "epoch": 0.32, "grad_norm": 9.9375, "learning_rate": 4.780172588441686e-06, "loss": 0.3104, "step": 1697 }, { "epoch": 0.32, "grad_norm": 9.6875, "learning_rate": 4.7798584166247915e-06, "loss": 0.3625, "step": 1698 }, { "epoch": 0.32, "grad_norm": 12.0625, "learning_rate": 4.779544030804339e-06, "loss": 0.2847, "step": 1699 }, { "epoch": 0.32, "grad_norm": 13.3125, "learning_rate": 4.77922943100984e-06, "loss": 0.3131, "step": 1700 }, { "epoch": 0.32, "grad_norm": 16.375, "learning_rate": 4.778914617270824e-06, "loss": 0.3626, "step": 1701 }, { "epoch": 0.32, "grad_norm": 17.0, "learning_rate": 4.778599589616841e-06, "loss": 0.3195, "step": 1702 }, { "epoch": 0.32, "grad_norm": 15.6875, "learning_rate": 4.778284348077464e-06, "loss": 0.3527, "step": 1703 }, { "epoch": 0.32, "grad_norm": 9.25, "learning_rate": 4.777968892682282e-06, "loss": 0.2972, "step": 1704 }, { "epoch": 0.32, "grad_norm": 17.25, "learning_rate": 4.7776532234609065e-06, "loss": 0.4042, "step": 1705 }, { "epoch": 0.32, "grad_norm": 8.75, "learning_rate": 4.777337340442968e-06, "loss": 0.3422, "step": 1706 }, { "epoch": 0.32, "grad_norm": 12.5625, "learning_rate": 4.777021243658119e-06, "loss": 0.2631, "step": 1707 }, { "epoch": 0.32, "grad_norm": 11.875, "learning_rate": 4.776704933136028e-06, "loss": 0.2982, "step": 1708 }, { "epoch": 0.32, "grad_norm": 9.5, "learning_rate": 4.776388408906389e-06, "loss": 0.2822, "step": 1709 }, { "epoch": 0.32, "grad_norm": 10.5, "learning_rate": 4.776071670998911e-06, "loss": 0.3164, "step": 1710 }, { "epoch": 0.32, "grad_norm": 9.9375, "learning_rate": 4.775754719443326e-06, "loss": 0.3046, "step": 1711 }, { "epoch": 0.32, "grad_norm": 8.375, "learning_rate": 4.775437554269385e-06, "loss": 0.251, "step": 1712 }, { "epoch": 0.32, "grad_norm": 8.5, "learning_rate": 4.77512017550686e-06, "loss": 0.3304, "step": 1713 }, { "epoch": 0.32, "grad_norm": 12.5, "learning_rate": 4.774802583185542e-06, "loss": 0.2871, "step": 1714 }, { "epoch": 0.32, "grad_norm": 13.875, "learning_rate": 4.774484777335243e-06, "loss": 0.3305, "step": 1715 }, { "epoch": 0.32, "grad_norm": 8.1875, "learning_rate": 4.7741667579857934e-06, "loss": 0.3067, "step": 1716 }, { "epoch": 0.32, "grad_norm": 10.125, "learning_rate": 4.7738485251670456e-06, "loss": 0.3316, "step": 1717 }, { "epoch": 0.32, "grad_norm": 12.0625, "learning_rate": 4.773530078908871e-06, "loss": 0.2596, "step": 1718 }, { "epoch": 0.33, "grad_norm": 9.375, "learning_rate": 4.773211419241161e-06, "loss": 0.3145, "step": 1719 }, { "epoch": 0.33, "grad_norm": 19.25, "learning_rate": 4.772892546193827e-06, "loss": 0.3319, "step": 1720 }, { "epoch": 0.33, "grad_norm": 11.5, "learning_rate": 4.772573459796801e-06, "loss": 0.375, "step": 1721 }, { "epoch": 0.33, "grad_norm": 7.96875, "learning_rate": 4.772254160080035e-06, "loss": 0.2901, "step": 1722 }, { "epoch": 0.33, "grad_norm": 8.375, "learning_rate": 4.7719346470735004e-06, "loss": 0.3249, "step": 1723 }, { "epoch": 0.33, "grad_norm": 9.9375, "learning_rate": 4.7716149208071895e-06, "loss": 0.3362, "step": 1724 }, { "epoch": 0.33, "grad_norm": 12.5625, "learning_rate": 4.7712949813111125e-06, "loss": 0.3435, "step": 1725 }, { "epoch": 0.33, "grad_norm": 15.5, "learning_rate": 4.770974828615302e-06, "loss": 0.4642, "step": 1726 }, { "epoch": 0.33, "grad_norm": 11.875, "learning_rate": 4.7706544627498115e-06, "loss": 0.3934, "step": 1727 }, { "epoch": 0.33, "grad_norm": 12.375, "learning_rate": 4.770333883744711e-06, "loss": 0.3139, "step": 1728 }, { "epoch": 0.33, "grad_norm": 14.0, "learning_rate": 4.770013091630091e-06, "loss": 0.3092, "step": 1729 }, { "epoch": 0.33, "grad_norm": 13.8125, "learning_rate": 4.769692086436067e-06, "loss": 0.2971, "step": 1730 }, { "epoch": 0.33, "grad_norm": 11.625, "learning_rate": 4.769370868192767e-06, "loss": 0.3957, "step": 1731 }, { "epoch": 0.33, "grad_norm": 10.625, "learning_rate": 4.769049436930345e-06, "loss": 0.3358, "step": 1732 }, { "epoch": 0.33, "grad_norm": 12.0, "learning_rate": 4.768727792678973e-06, "loss": 0.2143, "step": 1733 }, { "epoch": 0.33, "grad_norm": 10.75, "learning_rate": 4.768405935468842e-06, "loss": 0.3389, "step": 1734 }, { "epoch": 0.33, "grad_norm": 10.3125, "learning_rate": 4.768083865330163e-06, "loss": 0.3873, "step": 1735 }, { "epoch": 0.33, "grad_norm": 11.125, "learning_rate": 4.767761582293169e-06, "loss": 0.2778, "step": 1736 }, { "epoch": 0.33, "grad_norm": 10.75, "learning_rate": 4.767439086388112e-06, "loss": 0.2783, "step": 1737 }, { "epoch": 0.33, "grad_norm": 8.5625, "learning_rate": 4.767116377645263e-06, "loss": 0.2553, "step": 1738 }, { "epoch": 0.33, "grad_norm": 11.375, "learning_rate": 4.766793456094913e-06, "loss": 0.3474, "step": 1739 }, { "epoch": 0.33, "grad_norm": 10.0625, "learning_rate": 4.766470321767376e-06, "loss": 0.3357, "step": 1740 }, { "epoch": 0.33, "grad_norm": 9.0, "learning_rate": 4.766146974692981e-06, "loss": 0.2624, "step": 1741 }, { "epoch": 0.33, "grad_norm": 10.75, "learning_rate": 4.765823414902081e-06, "loss": 0.241, "step": 1742 }, { "epoch": 0.33, "grad_norm": 11.375, "learning_rate": 4.765499642425048e-06, "loss": 0.3116, "step": 1743 }, { "epoch": 0.33, "grad_norm": 11.5, "learning_rate": 4.765175657292273e-06, "loss": 0.2618, "step": 1744 }, { "epoch": 0.33, "grad_norm": 9.875, "learning_rate": 4.764851459534167e-06, "loss": 0.283, "step": 1745 }, { "epoch": 0.33, "grad_norm": 10.3125, "learning_rate": 4.764527049181161e-06, "loss": 0.3169, "step": 1746 }, { "epoch": 0.33, "grad_norm": 12.0625, "learning_rate": 4.7642024262637095e-06, "loss": 0.3209, "step": 1747 }, { "epoch": 0.33, "grad_norm": 15.5625, "learning_rate": 4.7638775908122805e-06, "loss": 0.4156, "step": 1748 }, { "epoch": 0.33, "grad_norm": 9.75, "learning_rate": 4.763552542857367e-06, "loss": 0.2467, "step": 1749 }, { "epoch": 0.33, "grad_norm": 9.6875, "learning_rate": 4.7632272824294805e-06, "loss": 0.3153, "step": 1750 }, { "epoch": 0.33, "grad_norm": 13.5625, "learning_rate": 4.762901809559152e-06, "loss": 0.2965, "step": 1751 }, { "epoch": 0.33, "grad_norm": 10.375, "learning_rate": 4.762576124276931e-06, "loss": 0.345, "step": 1752 }, { "epoch": 0.33, "grad_norm": 10.5625, "learning_rate": 4.7622502266133905e-06, "loss": 0.406, "step": 1753 }, { "epoch": 0.33, "grad_norm": 20.75, "learning_rate": 4.7619241165991216e-06, "loss": 0.3343, "step": 1754 }, { "epoch": 0.33, "grad_norm": 9.8125, "learning_rate": 4.761597794264734e-06, "loss": 0.329, "step": 1755 }, { "epoch": 0.33, "grad_norm": 15.25, "learning_rate": 4.76127125964086e-06, "loss": 0.3615, "step": 1756 }, { "epoch": 0.33, "grad_norm": 14.25, "learning_rate": 4.76094451275815e-06, "loss": 0.3991, "step": 1757 }, { "epoch": 0.33, "grad_norm": 16.125, "learning_rate": 4.760617553647274e-06, "loss": 0.3653, "step": 1758 }, { "epoch": 0.33, "grad_norm": 32.75, "learning_rate": 4.760290382338923e-06, "loss": 0.3422, "step": 1759 }, { "epoch": 0.33, "grad_norm": 9.875, "learning_rate": 4.7599629988638076e-06, "loss": 0.2497, "step": 1760 }, { "epoch": 0.33, "grad_norm": 13.125, "learning_rate": 4.75963540325266e-06, "loss": 0.3078, "step": 1761 }, { "epoch": 0.33, "grad_norm": 9.25, "learning_rate": 4.759307595536228e-06, "loss": 0.295, "step": 1762 }, { "epoch": 0.33, "grad_norm": 9.5, "learning_rate": 4.758979575745283e-06, "loss": 0.3133, "step": 1763 }, { "epoch": 0.33, "grad_norm": 11.125, "learning_rate": 4.7586513439106155e-06, "loss": 0.3488, "step": 1764 }, { "epoch": 0.33, "grad_norm": 7.65625, "learning_rate": 4.758322900063036e-06, "loss": 0.2633, "step": 1765 }, { "epoch": 0.33, "grad_norm": 15.0, "learning_rate": 4.7579942442333736e-06, "loss": 0.3769, "step": 1766 }, { "epoch": 0.33, "grad_norm": 23.75, "learning_rate": 4.757665376452478e-06, "loss": 0.3342, "step": 1767 }, { "epoch": 0.33, "grad_norm": 11.5625, "learning_rate": 4.757336296751221e-06, "loss": 0.3193, "step": 1768 }, { "epoch": 0.33, "grad_norm": 16.625, "learning_rate": 4.757007005160491e-06, "loss": 0.3085, "step": 1769 }, { "epoch": 0.33, "grad_norm": 11.125, "learning_rate": 4.756677501711198e-06, "loss": 0.2923, "step": 1770 }, { "epoch": 0.34, "grad_norm": 13.25, "learning_rate": 4.7563477864342696e-06, "loss": 0.2804, "step": 1771 }, { "epoch": 0.34, "grad_norm": 13.5625, "learning_rate": 4.7560178593606585e-06, "loss": 0.3657, "step": 1772 }, { "epoch": 0.34, "grad_norm": 13.375, "learning_rate": 4.755687720521331e-06, "loss": 0.3537, "step": 1773 }, { "epoch": 0.34, "grad_norm": 9.9375, "learning_rate": 4.7553573699472785e-06, "loss": 0.3612, "step": 1774 }, { "epoch": 0.34, "grad_norm": 12.625, "learning_rate": 4.755026807669508e-06, "loss": 0.2211, "step": 1775 }, { "epoch": 0.34, "grad_norm": 10.0625, "learning_rate": 4.75469603371905e-06, "loss": 0.306, "step": 1776 }, { "epoch": 0.34, "grad_norm": 13.3125, "learning_rate": 4.754365048126954e-06, "loss": 0.4141, "step": 1777 }, { "epoch": 0.34, "grad_norm": 16.875, "learning_rate": 4.754033850924287e-06, "loss": 0.4019, "step": 1778 }, { "epoch": 0.34, "grad_norm": 15.0625, "learning_rate": 4.7537024421421364e-06, "loss": 0.3507, "step": 1779 }, { "epoch": 0.34, "grad_norm": 7.5, "learning_rate": 4.7533708218116134e-06, "loss": 0.2296, "step": 1780 }, { "epoch": 0.34, "grad_norm": 40.0, "learning_rate": 4.753038989963844e-06, "loss": 0.455, "step": 1781 }, { "epoch": 0.34, "grad_norm": 10.625, "learning_rate": 4.752706946629978e-06, "loss": 0.2986, "step": 1782 }, { "epoch": 0.34, "grad_norm": 10.375, "learning_rate": 4.752374691841182e-06, "loss": 0.2859, "step": 1783 }, { "epoch": 0.34, "grad_norm": 8.5, "learning_rate": 4.752042225628644e-06, "loss": 0.3091, "step": 1784 }, { "epoch": 0.34, "grad_norm": 10.0625, "learning_rate": 4.751709548023572e-06, "loss": 0.3114, "step": 1785 }, { "epoch": 0.34, "grad_norm": 10.1875, "learning_rate": 4.751376659057194e-06, "loss": 0.3501, "step": 1786 }, { "epoch": 0.34, "grad_norm": 8.0625, "learning_rate": 4.7510435587607564e-06, "loss": 0.38, "step": 1787 }, { "epoch": 0.34, "grad_norm": 7.53125, "learning_rate": 4.750710247165526e-06, "loss": 0.2624, "step": 1788 }, { "epoch": 0.34, "grad_norm": 11.375, "learning_rate": 4.750376724302791e-06, "loss": 0.2898, "step": 1789 }, { "epoch": 0.34, "grad_norm": 11.5, "learning_rate": 4.750042990203857e-06, "loss": 0.3499, "step": 1790 }, { "epoch": 0.34, "grad_norm": 13.5625, "learning_rate": 4.749709044900051e-06, "loss": 0.3139, "step": 1791 }, { "epoch": 0.34, "grad_norm": 12.6875, "learning_rate": 4.74937488842272e-06, "loss": 0.355, "step": 1792 }, { "epoch": 0.34, "grad_norm": 13.125, "learning_rate": 4.74904052080323e-06, "loss": 0.3217, "step": 1793 }, { "epoch": 0.34, "grad_norm": 9.8125, "learning_rate": 4.7487059420729665e-06, "loss": 0.2916, "step": 1794 }, { "epoch": 0.34, "grad_norm": 11.1875, "learning_rate": 4.748371152263337e-06, "loss": 0.3243, "step": 1795 }, { "epoch": 0.34, "grad_norm": 12.375, "learning_rate": 4.748036151405765e-06, "loss": 0.2645, "step": 1796 }, { "epoch": 0.34, "grad_norm": 10.5, "learning_rate": 4.747700939531697e-06, "loss": 0.3104, "step": 1797 }, { "epoch": 0.34, "grad_norm": 13.4375, "learning_rate": 4.7473655166725996e-06, "loss": 0.2988, "step": 1798 }, { "epoch": 0.34, "grad_norm": 9.9375, "learning_rate": 4.747029882859957e-06, "loss": 0.2798, "step": 1799 }, { "epoch": 0.34, "grad_norm": 12.5, "learning_rate": 4.746694038125273e-06, "loss": 0.3317, "step": 1800 }, { "epoch": 0.34, "grad_norm": 13.125, "learning_rate": 4.746357982500074e-06, "loss": 0.312, "step": 1801 }, { "epoch": 0.34, "grad_norm": 13.0625, "learning_rate": 4.746021716015904e-06, "loss": 0.2903, "step": 1802 }, { "epoch": 0.34, "grad_norm": 9.0, "learning_rate": 4.7456852387043276e-06, "loss": 0.2313, "step": 1803 }, { "epoch": 0.34, "grad_norm": 8.6875, "learning_rate": 4.745348550596927e-06, "loss": 0.3025, "step": 1804 }, { "epoch": 0.34, "grad_norm": 11.8125, "learning_rate": 4.7450116517253096e-06, "loss": 0.3227, "step": 1805 }, { "epoch": 0.34, "grad_norm": 11.1875, "learning_rate": 4.744674542121096e-06, "loss": 0.3793, "step": 1806 }, { "epoch": 0.34, "grad_norm": 10.8125, "learning_rate": 4.7443372218159314e-06, "loss": 0.3001, "step": 1807 }, { "epoch": 0.34, "grad_norm": 11.125, "learning_rate": 4.743999690841479e-06, "loss": 0.3684, "step": 1808 }, { "epoch": 0.34, "grad_norm": 10.75, "learning_rate": 4.7436619492294205e-06, "loss": 0.3484, "step": 1809 }, { "epoch": 0.34, "grad_norm": 17.375, "learning_rate": 4.74332399701146e-06, "loss": 0.2681, "step": 1810 }, { "epoch": 0.34, "grad_norm": 9.375, "learning_rate": 4.74298583421932e-06, "loss": 0.2768, "step": 1811 }, { "epoch": 0.34, "grad_norm": 12.4375, "learning_rate": 4.742647460884742e-06, "loss": 0.3543, "step": 1812 }, { "epoch": 0.34, "grad_norm": 13.625, "learning_rate": 4.742308877039489e-06, "loss": 0.3356, "step": 1813 }, { "epoch": 0.34, "grad_norm": 15.9375, "learning_rate": 4.7419700827153415e-06, "loss": 0.3521, "step": 1814 }, { "epoch": 0.34, "grad_norm": 12.25, "learning_rate": 4.741631077944103e-06, "loss": 0.3481, "step": 1815 }, { "epoch": 0.34, "grad_norm": 9.0, "learning_rate": 4.741291862757594e-06, "loss": 0.2703, "step": 1816 }, { "epoch": 0.34, "grad_norm": 13.625, "learning_rate": 4.740952437187655e-06, "loss": 0.2598, "step": 1817 }, { "epoch": 0.34, "grad_norm": 10.6875, "learning_rate": 4.740612801266148e-06, "loss": 0.3409, "step": 1818 }, { "epoch": 0.34, "grad_norm": 10.0, "learning_rate": 4.740272955024952e-06, "loss": 0.3734, "step": 1819 }, { "epoch": 0.34, "grad_norm": 11.8125, "learning_rate": 4.739932898495971e-06, "loss": 0.2823, "step": 1820 }, { "epoch": 0.34, "grad_norm": 11.625, "learning_rate": 4.73959263171112e-06, "loss": 0.381, "step": 1821 }, { "epoch": 0.34, "grad_norm": 19.625, "learning_rate": 4.739252154702342e-06, "loss": 0.27, "step": 1822 }, { "epoch": 0.34, "grad_norm": 21.875, "learning_rate": 4.738911467501597e-06, "loss": 0.386, "step": 1823 }, { "epoch": 0.35, "grad_norm": 13.0, "learning_rate": 4.738570570140862e-06, "loss": 0.3233, "step": 1824 }, { "epoch": 0.35, "grad_norm": 7.21875, "learning_rate": 4.738229462652138e-06, "loss": 0.2847, "step": 1825 }, { "epoch": 0.35, "grad_norm": 19.125, "learning_rate": 4.737888145067442e-06, "loss": 0.3658, "step": 1826 }, { "epoch": 0.35, "grad_norm": 9.0, "learning_rate": 4.7375466174188146e-06, "loss": 0.3186, "step": 1827 }, { "epoch": 0.35, "grad_norm": 7.84375, "learning_rate": 4.737204879738313e-06, "loss": 0.2845, "step": 1828 }, { "epoch": 0.35, "grad_norm": 10.625, "learning_rate": 4.736862932058014e-06, "loss": 0.2824, "step": 1829 }, { "epoch": 0.35, "grad_norm": 14.0, "learning_rate": 4.736520774410017e-06, "loss": 0.2479, "step": 1830 }, { "epoch": 0.35, "grad_norm": 11.5625, "learning_rate": 4.7361784068264385e-06, "loss": 0.2496, "step": 1831 }, { "epoch": 0.35, "grad_norm": 9.125, "learning_rate": 4.7358358293394145e-06, "loss": 0.3142, "step": 1832 }, { "epoch": 0.35, "grad_norm": 12.375, "learning_rate": 4.735493041981104e-06, "loss": 0.3662, "step": 1833 }, { "epoch": 0.35, "grad_norm": 15.25, "learning_rate": 4.735150044783682e-06, "loss": 0.3542, "step": 1834 }, { "epoch": 0.35, "grad_norm": 12.0, "learning_rate": 4.734806837779344e-06, "loss": 0.3226, "step": 1835 }, { "epoch": 0.35, "grad_norm": 14.25, "learning_rate": 4.734463421000307e-06, "loss": 0.336, "step": 1836 }, { "epoch": 0.35, "grad_norm": 11.625, "learning_rate": 4.734119794478807e-06, "loss": 0.2641, "step": 1837 }, { "epoch": 0.35, "grad_norm": 15.75, "learning_rate": 4.733775958247098e-06, "loss": 0.296, "step": 1838 }, { "epoch": 0.35, "grad_norm": 10.375, "learning_rate": 4.733431912337455e-06, "loss": 0.2732, "step": 1839 }, { "epoch": 0.35, "grad_norm": 11.0, "learning_rate": 4.733087656782173e-06, "loss": 0.2721, "step": 1840 }, { "epoch": 0.35, "grad_norm": 8.75, "learning_rate": 4.732743191613565e-06, "loss": 0.3477, "step": 1841 }, { "epoch": 0.35, "grad_norm": 14.75, "learning_rate": 4.7323985168639684e-06, "loss": 0.3273, "step": 1842 }, { "epoch": 0.35, "grad_norm": 9.1875, "learning_rate": 4.732053632565733e-06, "loss": 0.298, "step": 1843 }, { "epoch": 0.35, "grad_norm": 8.875, "learning_rate": 4.731708538751234e-06, "loss": 0.3177, "step": 1844 }, { "epoch": 0.35, "grad_norm": 10.9375, "learning_rate": 4.7313632354528635e-06, "loss": 0.2829, "step": 1845 }, { "epoch": 0.35, "grad_norm": 15.6875, "learning_rate": 4.731017722703034e-06, "loss": 0.4004, "step": 1846 }, { "epoch": 0.35, "grad_norm": 21.125, "learning_rate": 4.730672000534179e-06, "loss": 0.3299, "step": 1847 }, { "epoch": 0.35, "grad_norm": 16.375, "learning_rate": 4.730326068978749e-06, "loss": 0.3552, "step": 1848 }, { "epoch": 0.35, "grad_norm": 14.75, "learning_rate": 4.7299799280692175e-06, "loss": 0.3134, "step": 1849 }, { "epoch": 0.35, "grad_norm": 15.5, "learning_rate": 4.729633577838073e-06, "loss": 0.3653, "step": 1850 }, { "epoch": 0.35, "grad_norm": 9.5, "learning_rate": 4.729287018317828e-06, "loss": 0.3091, "step": 1851 }, { "epoch": 0.35, "grad_norm": 17.0, "learning_rate": 4.728940249541012e-06, "loss": 0.3327, "step": 1852 }, { "epoch": 0.35, "grad_norm": 11.375, "learning_rate": 4.728593271540177e-06, "loss": 0.2629, "step": 1853 }, { "epoch": 0.35, "grad_norm": 11.6875, "learning_rate": 4.7282460843478915e-06, "loss": 0.3717, "step": 1854 }, { "epoch": 0.35, "grad_norm": 10.8125, "learning_rate": 4.727898687996746e-06, "loss": 0.2707, "step": 1855 }, { "epoch": 0.35, "grad_norm": 25.125, "learning_rate": 4.727551082519348e-06, "loss": 0.3303, "step": 1856 }, { "epoch": 0.35, "grad_norm": 14.8125, "learning_rate": 4.7272032679483266e-06, "loss": 0.3184, "step": 1857 }, { "epoch": 0.35, "grad_norm": 12.3125, "learning_rate": 4.726855244316331e-06, "loss": 0.3354, "step": 1858 }, { "epoch": 0.35, "grad_norm": 14.1875, "learning_rate": 4.726507011656028e-06, "loss": 0.2872, "step": 1859 }, { "epoch": 0.35, "grad_norm": 13.125, "learning_rate": 4.726158570000107e-06, "loss": 0.2739, "step": 1860 }, { "epoch": 0.35, "grad_norm": 9.5, "learning_rate": 4.725809919381272e-06, "loss": 0.2103, "step": 1861 }, { "epoch": 0.35, "grad_norm": 8.6875, "learning_rate": 4.725461059832253e-06, "loss": 0.279, "step": 1862 }, { "epoch": 0.35, "grad_norm": 10.4375, "learning_rate": 4.725111991385795e-06, "loss": 0.2883, "step": 1863 }, { "epoch": 0.35, "grad_norm": 9.1875, "learning_rate": 4.724762714074664e-06, "loss": 0.3404, "step": 1864 }, { "epoch": 0.35, "grad_norm": 26.125, "learning_rate": 4.724413227931647e-06, "loss": 0.38, "step": 1865 }, { "epoch": 0.35, "grad_norm": 9.875, "learning_rate": 4.724063532989547e-06, "loss": 0.2735, "step": 1866 }, { "epoch": 0.35, "grad_norm": 11.125, "learning_rate": 4.72371362928119e-06, "loss": 0.2813, "step": 1867 }, { "epoch": 0.35, "grad_norm": 8.5625, "learning_rate": 4.7233635168394196e-06, "loss": 0.2786, "step": 1868 }, { "epoch": 0.35, "grad_norm": 8.0, "learning_rate": 4.723013195697101e-06, "loss": 0.2342, "step": 1869 }, { "epoch": 0.35, "grad_norm": 8.9375, "learning_rate": 4.722662665887119e-06, "loss": 0.2703, "step": 1870 }, { "epoch": 0.35, "grad_norm": 10.1875, "learning_rate": 4.722311927442373e-06, "loss": 0.3017, "step": 1871 }, { "epoch": 0.35, "grad_norm": 14.6875, "learning_rate": 4.721960980395789e-06, "loss": 0.2307, "step": 1872 }, { "epoch": 0.35, "grad_norm": 13.4375, "learning_rate": 4.721609824780308e-06, "loss": 0.3054, "step": 1873 }, { "epoch": 0.35, "grad_norm": 10.625, "learning_rate": 4.721258460628891e-06, "loss": 0.3535, "step": 1874 }, { "epoch": 0.35, "grad_norm": 10.875, "learning_rate": 4.7209068879745225e-06, "loss": 0.2964, "step": 1875 }, { "epoch": 0.35, "grad_norm": 13.25, "learning_rate": 4.720555106850201e-06, "loss": 0.2963, "step": 1876 }, { "epoch": 0.36, "grad_norm": 13.25, "learning_rate": 4.720203117288949e-06, "loss": 0.3923, "step": 1877 }, { "epoch": 0.36, "grad_norm": 13.1875, "learning_rate": 4.719850919323804e-06, "loss": 0.2753, "step": 1878 }, { "epoch": 0.36, "grad_norm": 12.6875, "learning_rate": 4.719498512987828e-06, "loss": 0.3359, "step": 1879 }, { "epoch": 0.36, "grad_norm": 10.8125, "learning_rate": 4.719145898314101e-06, "loss": 0.294, "step": 1880 }, { "epoch": 0.36, "grad_norm": 12.5, "learning_rate": 4.718793075335719e-06, "loss": 0.3055, "step": 1881 }, { "epoch": 0.36, "grad_norm": 9.5, "learning_rate": 4.718440044085802e-06, "loss": 0.3575, "step": 1882 }, { "epoch": 0.36, "grad_norm": 10.1875, "learning_rate": 4.718086804597489e-06, "loss": 0.2893, "step": 1883 }, { "epoch": 0.36, "grad_norm": 12.375, "learning_rate": 4.7177333569039375e-06, "loss": 0.3727, "step": 1884 }, { "epoch": 0.36, "grad_norm": 8.5625, "learning_rate": 4.717379701038323e-06, "loss": 0.2933, "step": 1885 }, { "epoch": 0.36, "grad_norm": 13.9375, "learning_rate": 4.717025837033843e-06, "loss": 0.3109, "step": 1886 }, { "epoch": 0.36, "grad_norm": 12.0625, "learning_rate": 4.716671764923713e-06, "loss": 0.3543, "step": 1887 }, { "epoch": 0.36, "grad_norm": 16.625, "learning_rate": 4.7163174847411695e-06, "loss": 0.3218, "step": 1888 }, { "epoch": 0.36, "grad_norm": 14.5625, "learning_rate": 4.715962996519468e-06, "loss": 0.3505, "step": 1889 }, { "epoch": 0.36, "grad_norm": 15.25, "learning_rate": 4.715608300291882e-06, "loss": 0.2995, "step": 1890 }, { "epoch": 0.36, "grad_norm": 17.375, "learning_rate": 4.715253396091707e-06, "loss": 0.3568, "step": 1891 }, { "epoch": 0.36, "grad_norm": 6.65625, "learning_rate": 4.714898283952256e-06, "loss": 0.2633, "step": 1892 }, { "epoch": 0.36, "grad_norm": 18.625, "learning_rate": 4.714542963906863e-06, "loss": 0.4045, "step": 1893 }, { "epoch": 0.36, "grad_norm": 10.125, "learning_rate": 4.71418743598888e-06, "loss": 0.4021, "step": 1894 }, { "epoch": 0.36, "grad_norm": 8.875, "learning_rate": 4.71383170023168e-06, "loss": 0.3394, "step": 1895 }, { "epoch": 0.36, "grad_norm": 8.6875, "learning_rate": 4.713475756668655e-06, "loss": 0.314, "step": 1896 }, { "epoch": 0.36, "grad_norm": 10.5, "learning_rate": 4.7131196053332165e-06, "loss": 0.2981, "step": 1897 }, { "epoch": 0.36, "grad_norm": 9.5625, "learning_rate": 4.712763246258794e-06, "loss": 0.3759, "step": 1898 }, { "epoch": 0.36, "grad_norm": 8.375, "learning_rate": 4.712406679478839e-06, "loss": 0.2858, "step": 1899 }, { "epoch": 0.36, "grad_norm": 10.75, "learning_rate": 4.712049905026822e-06, "loss": 0.2899, "step": 1900 }, { "epoch": 0.36, "grad_norm": 14.1875, "learning_rate": 4.711692922936231e-06, "loss": 0.3813, "step": 1901 }, { "epoch": 0.36, "grad_norm": 13.5625, "learning_rate": 4.711335733240575e-06, "loss": 0.2671, "step": 1902 }, { "epoch": 0.36, "grad_norm": 12.0625, "learning_rate": 4.710978335973384e-06, "loss": 0.3292, "step": 1903 }, { "epoch": 0.36, "grad_norm": 8.125, "learning_rate": 4.710620731168204e-06, "loss": 0.2877, "step": 1904 }, { "epoch": 0.36, "grad_norm": 8.4375, "learning_rate": 4.7102629188586026e-06, "loss": 0.2478, "step": 1905 }, { "epoch": 0.36, "grad_norm": 15.3125, "learning_rate": 4.709904899078166e-06, "loss": 0.3255, "step": 1906 }, { "epoch": 0.36, "grad_norm": 12.0, "learning_rate": 4.709546671860503e-06, "loss": 0.292, "step": 1907 }, { "epoch": 0.36, "grad_norm": 12.8125, "learning_rate": 4.709188237239237e-06, "loss": 0.3576, "step": 1908 }, { "epoch": 0.36, "grad_norm": 14.875, "learning_rate": 4.708829595248013e-06, "loss": 0.3437, "step": 1909 }, { "epoch": 0.36, "grad_norm": 12.875, "learning_rate": 4.708470745920498e-06, "loss": 0.3394, "step": 1910 }, { "epoch": 0.36, "grad_norm": 8.8125, "learning_rate": 4.708111689290373e-06, "loss": 0.3933, "step": 1911 }, { "epoch": 0.36, "grad_norm": 14.9375, "learning_rate": 4.707752425391345e-06, "loss": 0.3694, "step": 1912 }, { "epoch": 0.36, "grad_norm": 12.5, "learning_rate": 4.707392954257134e-06, "loss": 0.3142, "step": 1913 }, { "epoch": 0.36, "grad_norm": 9.875, "learning_rate": 4.707033275921484e-06, "loss": 0.2793, "step": 1914 }, { "epoch": 0.36, "grad_norm": 11.5625, "learning_rate": 4.706673390418157e-06, "loss": 0.3275, "step": 1915 }, { "epoch": 0.36, "grad_norm": 11.9375, "learning_rate": 4.706313297780934e-06, "loss": 0.288, "step": 1916 }, { "epoch": 0.36, "grad_norm": 11.125, "learning_rate": 4.705952998043615e-06, "loss": 0.3588, "step": 1917 }, { "epoch": 0.36, "grad_norm": 15.8125, "learning_rate": 4.705592491240022e-06, "loss": 0.2658, "step": 1918 }, { "epoch": 0.36, "grad_norm": 14.5625, "learning_rate": 4.705231777403994e-06, "loss": 0.3106, "step": 1919 }, { "epoch": 0.36, "grad_norm": 9.5, "learning_rate": 4.70487085656939e-06, "loss": 0.2943, "step": 1920 }, { "epoch": 0.36, "grad_norm": 11.8125, "learning_rate": 4.704509728770088e-06, "loss": 0.3688, "step": 1921 }, { "epoch": 0.36, "grad_norm": 9.3125, "learning_rate": 4.704148394039987e-06, "loss": 0.3194, "step": 1922 }, { "epoch": 0.36, "grad_norm": 9.6875, "learning_rate": 4.703786852413004e-06, "loss": 0.3254, "step": 1923 }, { "epoch": 0.36, "grad_norm": 14.75, "learning_rate": 4.703425103923075e-06, "loss": 0.279, "step": 1924 }, { "epoch": 0.36, "grad_norm": 12.3125, "learning_rate": 4.703063148604159e-06, "loss": 0.2689, "step": 1925 }, { "epoch": 0.36, "grad_norm": 10.1875, "learning_rate": 4.702700986490229e-06, "loss": 0.3118, "step": 1926 }, { "epoch": 0.36, "grad_norm": 8.625, "learning_rate": 4.702338617615279e-06, "loss": 0.3318, "step": 1927 }, { "epoch": 0.36, "grad_norm": 9.3125, "learning_rate": 4.701976042013327e-06, "loss": 0.2814, "step": 1928 }, { "epoch": 0.36, "grad_norm": 9.9375, "learning_rate": 4.701613259718405e-06, "loss": 0.3296, "step": 1929 }, { "epoch": 0.37, "grad_norm": 11.1875, "learning_rate": 4.701250270764567e-06, "loss": 0.4168, "step": 1930 }, { "epoch": 0.37, "grad_norm": 10.3125, "learning_rate": 4.700887075185885e-06, "loss": 0.3833, "step": 1931 }, { "epoch": 0.37, "grad_norm": 12.8125, "learning_rate": 4.70052367301645e-06, "loss": 0.3528, "step": 1932 }, { "epoch": 0.37, "grad_norm": 10.5, "learning_rate": 4.700160064290377e-06, "loss": 0.2707, "step": 1933 }, { "epoch": 0.37, "grad_norm": 10.125, "learning_rate": 4.699796249041794e-06, "loss": 0.32, "step": 1934 }, { "epoch": 0.37, "grad_norm": 11.375, "learning_rate": 4.699432227304851e-06, "loss": 0.3582, "step": 1935 }, { "epoch": 0.37, "grad_norm": 11.75, "learning_rate": 4.699067999113719e-06, "loss": 0.2866, "step": 1936 }, { "epoch": 0.37, "grad_norm": 11.125, "learning_rate": 4.698703564502587e-06, "loss": 0.2695, "step": 1937 }, { "epoch": 0.37, "grad_norm": 11.5, "learning_rate": 4.698338923505663e-06, "loss": 0.3407, "step": 1938 }, { "epoch": 0.37, "grad_norm": 12.5625, "learning_rate": 4.697974076157175e-06, "loss": 0.297, "step": 1939 }, { "epoch": 0.37, "grad_norm": 11.5, "learning_rate": 4.697609022491369e-06, "loss": 0.3192, "step": 1940 }, { "epoch": 0.37, "grad_norm": 11.125, "learning_rate": 4.697243762542513e-06, "loss": 0.2764, "step": 1941 }, { "epoch": 0.37, "grad_norm": 12.0625, "learning_rate": 4.696878296344893e-06, "loss": 0.3395, "step": 1942 }, { "epoch": 0.37, "grad_norm": 14.9375, "learning_rate": 4.696512623932813e-06, "loss": 0.3571, "step": 1943 }, { "epoch": 0.37, "grad_norm": 17.375, "learning_rate": 4.696146745340597e-06, "loss": 0.2901, "step": 1944 }, { "epoch": 0.37, "grad_norm": 25.125, "learning_rate": 4.695780660602592e-06, "loss": 0.3182, "step": 1945 }, { "epoch": 0.37, "grad_norm": 22.0, "learning_rate": 4.695414369753159e-06, "loss": 0.2823, "step": 1946 }, { "epoch": 0.37, "grad_norm": 19.0, "learning_rate": 4.6950478728266795e-06, "loss": 0.3058, "step": 1947 }, { "epoch": 0.37, "grad_norm": 9.6875, "learning_rate": 4.694681169857558e-06, "loss": 0.2461, "step": 1948 }, { "epoch": 0.37, "grad_norm": 13.3125, "learning_rate": 4.694314260880215e-06, "loss": 0.4028, "step": 1949 }, { "epoch": 0.37, "grad_norm": 13.8125, "learning_rate": 4.69394714592909e-06, "loss": 0.3371, "step": 1950 }, { "epoch": 0.37, "grad_norm": 11.5, "learning_rate": 4.693579825038644e-06, "loss": 0.3434, "step": 1951 }, { "epoch": 0.37, "grad_norm": 10.6875, "learning_rate": 4.693212298243357e-06, "loss": 0.2885, "step": 1952 }, { "epoch": 0.37, "grad_norm": 13.3125, "learning_rate": 4.692844565577728e-06, "loss": 0.2557, "step": 1953 }, { "epoch": 0.37, "grad_norm": 11.8125, "learning_rate": 4.692476627076272e-06, "loss": 0.32, "step": 1954 }, { "epoch": 0.37, "grad_norm": 11.25, "learning_rate": 4.692108482773529e-06, "loss": 0.3804, "step": 1955 }, { "epoch": 0.37, "grad_norm": 13.0625, "learning_rate": 4.6917401327040544e-06, "loss": 0.3095, "step": 1956 }, { "epoch": 0.37, "grad_norm": 20.0, "learning_rate": 4.6913715769024246e-06, "loss": 0.2834, "step": 1957 }, { "epoch": 0.37, "grad_norm": 8.5, "learning_rate": 4.691002815403235e-06, "loss": 0.3126, "step": 1958 }, { "epoch": 0.37, "grad_norm": 9.3125, "learning_rate": 4.6906338482411e-06, "loss": 0.2928, "step": 1959 }, { "epoch": 0.37, "grad_norm": 9.8125, "learning_rate": 4.690264675450653e-06, "loss": 0.2966, "step": 1960 }, { "epoch": 0.37, "grad_norm": 25.25, "learning_rate": 4.689895297066548e-06, "loss": 0.1831, "step": 1961 }, { "epoch": 0.37, "grad_norm": 13.1875, "learning_rate": 4.689525713123457e-06, "loss": 0.3381, "step": 1962 }, { "epoch": 0.37, "grad_norm": 17.875, "learning_rate": 4.689155923656071e-06, "loss": 0.2621, "step": 1963 }, { "epoch": 0.37, "grad_norm": 10.5, "learning_rate": 4.6887859286991026e-06, "loss": 0.2698, "step": 1964 }, { "epoch": 0.37, "grad_norm": 10.0, "learning_rate": 4.688415728287281e-06, "loss": 0.292, "step": 1965 }, { "epoch": 0.37, "grad_norm": 11.1875, "learning_rate": 4.688045322455357e-06, "loss": 0.2442, "step": 1966 }, { "epoch": 0.37, "grad_norm": 8.25, "learning_rate": 4.687674711238097e-06, "loss": 0.3006, "step": 1967 }, { "epoch": 0.37, "grad_norm": 16.875, "learning_rate": 4.687303894670292e-06, "loss": 0.2459, "step": 1968 }, { "epoch": 0.37, "grad_norm": 10.25, "learning_rate": 4.686932872786748e-06, "loss": 0.3103, "step": 1969 }, { "epoch": 0.37, "grad_norm": 10.125, "learning_rate": 4.686561645622292e-06, "loss": 0.2694, "step": 1970 }, { "epoch": 0.37, "grad_norm": 8.3125, "learning_rate": 4.68619021321177e-06, "loss": 0.3051, "step": 1971 }, { "epoch": 0.37, "grad_norm": 16.625, "learning_rate": 4.685818575590047e-06, "loss": 0.3544, "step": 1972 }, { "epoch": 0.37, "grad_norm": 11.5, "learning_rate": 4.685446732792008e-06, "loss": 0.3068, "step": 1973 }, { "epoch": 0.37, "grad_norm": 9.4375, "learning_rate": 4.685074684852556e-06, "loss": 0.2964, "step": 1974 }, { "epoch": 0.37, "grad_norm": 17.125, "learning_rate": 4.684702431806615e-06, "loss": 0.344, "step": 1975 }, { "epoch": 0.37, "grad_norm": 12.25, "learning_rate": 4.684329973689127e-06, "loss": 0.2948, "step": 1976 }, { "epoch": 0.37, "grad_norm": 10.3125, "learning_rate": 4.683957310535054e-06, "loss": 0.2826, "step": 1977 }, { "epoch": 0.37, "grad_norm": 14.625, "learning_rate": 4.6835844423793755e-06, "loss": 0.4012, "step": 1978 }, { "epoch": 0.37, "grad_norm": 7.75, "learning_rate": 4.683211369257092e-06, "loss": 0.2699, "step": 1979 }, { "epoch": 0.37, "grad_norm": 9.4375, "learning_rate": 4.682838091203223e-06, "loss": 0.3118, "step": 1980 }, { "epoch": 0.37, "grad_norm": 10.125, "learning_rate": 4.682464608252807e-06, "loss": 0.3146, "step": 1981 }, { "epoch": 0.37, "grad_norm": 10.1875, "learning_rate": 4.682090920440902e-06, "loss": 0.3091, "step": 1982 }, { "epoch": 0.38, "grad_norm": 12.25, "learning_rate": 4.681717027802585e-06, "loss": 0.3163, "step": 1983 }, { "epoch": 0.38, "grad_norm": 14.4375, "learning_rate": 4.681342930372952e-06, "loss": 0.3742, "step": 1984 }, { "epoch": 0.38, "grad_norm": 12.1875, "learning_rate": 4.680968628187118e-06, "loss": 0.3027, "step": 1985 }, { "epoch": 0.38, "grad_norm": 14.4375, "learning_rate": 4.680594121280218e-06, "loss": 0.2605, "step": 1986 }, { "epoch": 0.38, "grad_norm": 9.875, "learning_rate": 4.6802194096874054e-06, "loss": 0.3831, "step": 1987 }, { "epoch": 0.38, "grad_norm": 10.6875, "learning_rate": 4.679844493443855e-06, "loss": 0.3165, "step": 1988 }, { "epoch": 0.38, "grad_norm": 11.1875, "learning_rate": 4.679469372584756e-06, "loss": 0.278, "step": 1989 }, { "epoch": 0.38, "grad_norm": 8.0, "learning_rate": 4.679094047145323e-06, "loss": 0.229, "step": 1990 }, { "epoch": 0.38, "grad_norm": 8.8125, "learning_rate": 4.6787185171607846e-06, "loss": 0.2694, "step": 1991 }, { "epoch": 0.38, "grad_norm": 11.6875, "learning_rate": 4.6783427826663915e-06, "loss": 0.3018, "step": 1992 }, { "epoch": 0.38, "grad_norm": 14.125, "learning_rate": 4.677966843697413e-06, "loss": 0.3003, "step": 1993 }, { "epoch": 0.38, "grad_norm": 9.125, "learning_rate": 4.677590700289138e-06, "loss": 0.2827, "step": 1994 }, { "epoch": 0.38, "grad_norm": 12.625, "learning_rate": 4.677214352476871e-06, "loss": 0.2859, "step": 1995 }, { "epoch": 0.38, "grad_norm": 14.9375, "learning_rate": 4.676837800295942e-06, "loss": 0.3247, "step": 1996 }, { "epoch": 0.38, "grad_norm": 7.84375, "learning_rate": 4.676461043781695e-06, "loss": 0.2793, "step": 1997 }, { "epoch": 0.38, "grad_norm": 14.5, "learning_rate": 4.676084082969496e-06, "loss": 0.3274, "step": 1998 }, { "epoch": 0.38, "grad_norm": 8.0625, "learning_rate": 4.675706917894728e-06, "loss": 0.4, "step": 1999 }, { "epoch": 0.38, "grad_norm": 9.6875, "learning_rate": 4.675329548592796e-06, "loss": 0.2477, "step": 2000 }, { "epoch": 0.38, "grad_norm": 9.6875, "learning_rate": 4.674951975099122e-06, "loss": 0.2627, "step": 2001 }, { "epoch": 0.38, "grad_norm": 29.625, "learning_rate": 4.674574197449146e-06, "loss": 0.3942, "step": 2002 }, { "epoch": 0.38, "grad_norm": 10.0, "learning_rate": 4.6741962156783315e-06, "loss": 0.2894, "step": 2003 }, { "epoch": 0.38, "grad_norm": 10.0625, "learning_rate": 4.6738180298221566e-06, "loss": 0.3422, "step": 2004 }, { "epoch": 0.38, "grad_norm": 10.0, "learning_rate": 4.6734396399161215e-06, "loss": 0.316, "step": 2005 }, { "epoch": 0.38, "grad_norm": 6.90625, "learning_rate": 4.673061045995743e-06, "loss": 0.2782, "step": 2006 }, { "epoch": 0.38, "grad_norm": 12.0625, "learning_rate": 4.672682248096562e-06, "loss": 0.2561, "step": 2007 }, { "epoch": 0.38, "grad_norm": 10.9375, "learning_rate": 4.672303246254132e-06, "loss": 0.3116, "step": 2008 }, { "epoch": 0.38, "grad_norm": 10.4375, "learning_rate": 4.671924040504029e-06, "loss": 0.3064, "step": 2009 }, { "epoch": 0.38, "grad_norm": 9.0, "learning_rate": 4.671544630881849e-06, "loss": 0.275, "step": 2010 }, { "epoch": 0.38, "grad_norm": 12.75, "learning_rate": 4.671165017423206e-06, "loss": 0.3566, "step": 2011 }, { "epoch": 0.38, "grad_norm": 15.5, "learning_rate": 4.670785200163732e-06, "loss": 0.32, "step": 2012 }, { "epoch": 0.38, "grad_norm": 13.625, "learning_rate": 4.670405179139082e-06, "loss": 0.2918, "step": 2013 }, { "epoch": 0.38, "grad_norm": 15.1875, "learning_rate": 4.670024954384924e-06, "loss": 0.3184, "step": 2014 }, { "epoch": 0.38, "grad_norm": 14.6875, "learning_rate": 4.66964452593695e-06, "loss": 0.3533, "step": 2015 }, { "epoch": 0.38, "grad_norm": 14.6875, "learning_rate": 4.669263893830871e-06, "loss": 0.346, "step": 2016 }, { "epoch": 0.38, "grad_norm": 10.875, "learning_rate": 4.668883058102414e-06, "loss": 0.2623, "step": 2017 }, { "epoch": 0.38, "grad_norm": 8.1875, "learning_rate": 4.668502018787329e-06, "loss": 0.3213, "step": 2018 }, { "epoch": 0.38, "grad_norm": 16.75, "learning_rate": 4.6681207759213795e-06, "loss": 0.3568, "step": 2019 }, { "epoch": 0.38, "grad_norm": 13.0625, "learning_rate": 4.667739329540355e-06, "loss": 0.354, "step": 2020 }, { "epoch": 0.38, "grad_norm": 9.4375, "learning_rate": 4.66735767968006e-06, "loss": 0.2905, "step": 2021 }, { "epoch": 0.38, "grad_norm": 13.3125, "learning_rate": 4.666975826376318e-06, "loss": 0.2876, "step": 2022 }, { "epoch": 0.38, "grad_norm": 15.25, "learning_rate": 4.666593769664973e-06, "loss": 0.3288, "step": 2023 }, { "epoch": 0.38, "grad_norm": 15.1875, "learning_rate": 4.6662115095818875e-06, "loss": 0.3604, "step": 2024 }, { "epoch": 0.38, "grad_norm": 11.0625, "learning_rate": 4.665829046162942e-06, "loss": 0.3026, "step": 2025 }, { "epoch": 0.38, "grad_norm": 11.625, "learning_rate": 4.6654463794440384e-06, "loss": 0.2781, "step": 2026 }, { "epoch": 0.38, "grad_norm": 17.25, "learning_rate": 4.665063509461098e-06, "loss": 0.3125, "step": 2027 }, { "epoch": 0.38, "grad_norm": 10.1875, "learning_rate": 4.664680436250055e-06, "loss": 0.3212, "step": 2028 }, { "epoch": 0.38, "grad_norm": 10.75, "learning_rate": 4.664297159846872e-06, "loss": 0.3778, "step": 2029 }, { "epoch": 0.38, "grad_norm": 9.5625, "learning_rate": 4.663913680287525e-06, "loss": 0.2971, "step": 2030 }, { "epoch": 0.38, "grad_norm": 15.9375, "learning_rate": 4.6635299976080085e-06, "loss": 0.3874, "step": 2031 }, { "epoch": 0.38, "grad_norm": 10.375, "learning_rate": 4.663146111844339e-06, "loss": 0.2518, "step": 2032 }, { "epoch": 0.38, "grad_norm": 8.1875, "learning_rate": 4.66276202303255e-06, "loss": 0.3238, "step": 2033 }, { "epoch": 0.38, "grad_norm": 11.125, "learning_rate": 4.662377731208695e-06, "loss": 0.2678, "step": 2034 }, { "epoch": 0.38, "grad_norm": 15.9375, "learning_rate": 4.661993236408846e-06, "loss": 0.2892, "step": 2035 }, { "epoch": 0.39, "grad_norm": 11.875, "learning_rate": 4.661608538669096e-06, "loss": 0.3695, "step": 2036 }, { "epoch": 0.39, "grad_norm": 14.4375, "learning_rate": 4.661223638025553e-06, "loss": 0.385, "step": 2037 }, { "epoch": 0.39, "grad_norm": 11.1875, "learning_rate": 4.660838534514349e-06, "loss": 0.3255, "step": 2038 }, { "epoch": 0.39, "grad_norm": 15.4375, "learning_rate": 4.66045322817163e-06, "loss": 0.3554, "step": 2039 }, { "epoch": 0.39, "grad_norm": 11.25, "learning_rate": 4.660067719033565e-06, "loss": 0.3267, "step": 2040 }, { "epoch": 0.39, "grad_norm": 13.0625, "learning_rate": 4.659682007136341e-06, "loss": 0.2593, "step": 2041 }, { "epoch": 0.39, "grad_norm": 14.25, "learning_rate": 4.659296092516163e-06, "loss": 0.2681, "step": 2042 }, { "epoch": 0.39, "grad_norm": 9.0625, "learning_rate": 4.658909975209255e-06, "loss": 0.2936, "step": 2043 }, { "epoch": 0.39, "grad_norm": 15.0625, "learning_rate": 4.658523655251862e-06, "loss": 0.3146, "step": 2044 }, { "epoch": 0.39, "grad_norm": 8.9375, "learning_rate": 4.658137132680246e-06, "loss": 0.3024, "step": 2045 }, { "epoch": 0.39, "grad_norm": 11.625, "learning_rate": 4.657750407530689e-06, "loss": 0.271, "step": 2046 }, { "epoch": 0.39, "grad_norm": 10.25, "learning_rate": 4.657363479839492e-06, "loss": 0.3441, "step": 2047 }, { "epoch": 0.39, "grad_norm": 10.875, "learning_rate": 4.6569763496429735e-06, "loss": 0.3024, "step": 2048 }, { "epoch": 0.39, "grad_norm": 9.25, "learning_rate": 4.656589016977473e-06, "loss": 0.2889, "step": 2049 }, { "epoch": 0.39, "grad_norm": 11.875, "learning_rate": 4.656201481879349e-06, "loss": 0.267, "step": 2050 }, { "epoch": 0.39, "grad_norm": 10.5625, "learning_rate": 4.655813744384977e-06, "loss": 0.3466, "step": 2051 }, { "epoch": 0.39, "grad_norm": 16.0, "learning_rate": 4.655425804530754e-06, "loss": 0.2687, "step": 2052 }, { "epoch": 0.39, "grad_norm": 9.4375, "learning_rate": 4.655037662353093e-06, "loss": 0.295, "step": 2053 }, { "epoch": 0.39, "grad_norm": 13.125, "learning_rate": 4.65464931788843e-06, "loss": 0.3051, "step": 2054 }, { "epoch": 0.39, "grad_norm": 10.6875, "learning_rate": 4.654260771173217e-06, "loss": 0.3108, "step": 2055 }, { "epoch": 0.39, "grad_norm": 10.5625, "learning_rate": 4.653872022243925e-06, "loss": 0.338, "step": 2056 }, { "epoch": 0.39, "grad_norm": 18.75, "learning_rate": 4.6534830711370445e-06, "loss": 0.3192, "step": 2057 }, { "epoch": 0.39, "grad_norm": 12.25, "learning_rate": 4.653093917889087e-06, "loss": 0.3168, "step": 2058 }, { "epoch": 0.39, "grad_norm": 11.375, "learning_rate": 4.65270456253658e-06, "loss": 0.3194, "step": 2059 }, { "epoch": 0.39, "grad_norm": 10.4375, "learning_rate": 4.65231500511607e-06, "loss": 0.3235, "step": 2060 }, { "epoch": 0.39, "grad_norm": 8.5, "learning_rate": 4.651925245664125e-06, "loss": 0.2781, "step": 2061 }, { "epoch": 0.39, "grad_norm": 15.375, "learning_rate": 4.651535284217331e-06, "loss": 0.3483, "step": 2062 }, { "epoch": 0.39, "grad_norm": 12.8125, "learning_rate": 4.651145120812291e-06, "loss": 0.3445, "step": 2063 }, { "epoch": 0.39, "grad_norm": 9.0625, "learning_rate": 4.650754755485629e-06, "loss": 0.3043, "step": 2064 }, { "epoch": 0.39, "grad_norm": 9.4375, "learning_rate": 4.650364188273988e-06, "loss": 0.3269, "step": 2065 }, { "epoch": 0.39, "grad_norm": 19.5, "learning_rate": 4.64997341921403e-06, "loss": 0.3698, "step": 2066 }, { "epoch": 0.39, "grad_norm": 10.625, "learning_rate": 4.649582448342434e-06, "loss": 0.3172, "step": 2067 }, { "epoch": 0.39, "grad_norm": 14.0, "learning_rate": 4.649191275695899e-06, "loss": 0.251, "step": 2068 }, { "epoch": 0.39, "grad_norm": 15.75, "learning_rate": 4.648799901311144e-06, "loss": 0.3303, "step": 2069 }, { "epoch": 0.39, "grad_norm": 10.25, "learning_rate": 4.648408325224907e-06, "loss": 0.2904, "step": 2070 }, { "epoch": 0.39, "grad_norm": 19.25, "learning_rate": 4.648016547473942e-06, "loss": 0.3186, "step": 2071 }, { "epoch": 0.39, "grad_norm": 11.375, "learning_rate": 4.6476245680950265e-06, "loss": 0.3525, "step": 2072 }, { "epoch": 0.39, "grad_norm": 10.9375, "learning_rate": 4.647232387124953e-06, "loss": 0.3443, "step": 2073 }, { "epoch": 0.39, "grad_norm": 10.625, "learning_rate": 4.646840004600534e-06, "loss": 0.2509, "step": 2074 }, { "epoch": 0.39, "grad_norm": 7.6875, "learning_rate": 4.646447420558601e-06, "loss": 0.3016, "step": 2075 }, { "epoch": 0.39, "grad_norm": 9.0625, "learning_rate": 4.646054635036007e-06, "loss": 0.27, "step": 2076 }, { "epoch": 0.39, "grad_norm": 14.125, "learning_rate": 4.6456616480696185e-06, "loss": 0.2842, "step": 2077 }, { "epoch": 0.39, "grad_norm": 14.5, "learning_rate": 4.645268459696327e-06, "loss": 0.2117, "step": 2078 }, { "epoch": 0.39, "grad_norm": 10.6875, "learning_rate": 4.644875069953038e-06, "loss": 0.2752, "step": 2079 }, { "epoch": 0.39, "grad_norm": 11.75, "learning_rate": 4.6444814788766775e-06, "loss": 0.3012, "step": 2080 }, { "epoch": 0.39, "grad_norm": 15.1875, "learning_rate": 4.6440876865041935e-06, "loss": 0.3264, "step": 2081 }, { "epoch": 0.39, "grad_norm": 11.8125, "learning_rate": 4.643693692872547e-06, "loss": 0.2485, "step": 2082 }, { "epoch": 0.39, "grad_norm": 12.3125, "learning_rate": 4.6432994980187226e-06, "loss": 0.3864, "step": 2083 }, { "epoch": 0.39, "grad_norm": 9.625, "learning_rate": 4.642905101979722e-06, "loss": 0.2768, "step": 2084 }, { "epoch": 0.39, "grad_norm": 15.1875, "learning_rate": 4.642510504792565e-06, "loss": 0.3349, "step": 2085 }, { "epoch": 0.39, "grad_norm": 14.5625, "learning_rate": 4.642115706494294e-06, "loss": 0.3284, "step": 2086 }, { "epoch": 0.39, "grad_norm": 10.75, "learning_rate": 4.641720707121964e-06, "loss": 0.2799, "step": 2087 }, { "epoch": 0.39, "grad_norm": 15.5625, "learning_rate": 4.6413255067126545e-06, "loss": 0.2757, "step": 2088 }, { "epoch": 0.4, "grad_norm": 8.8125, "learning_rate": 4.6409301053034616e-06, "loss": 0.2745, "step": 2089 }, { "epoch": 0.4, "grad_norm": 13.3125, "learning_rate": 4.6405345029315e-06, "loss": 0.2683, "step": 2090 }, { "epoch": 0.4, "grad_norm": 15.1875, "learning_rate": 4.640138699633904e-06, "loss": 0.4196, "step": 2091 }, { "epoch": 0.4, "grad_norm": 10.625, "learning_rate": 4.6397426954478265e-06, "loss": 0.3052, "step": 2092 }, { "epoch": 0.4, "grad_norm": 12.0, "learning_rate": 4.639346490410439e-06, "loss": 0.3528, "step": 2093 }, { "epoch": 0.4, "grad_norm": 9.6875, "learning_rate": 4.638950084558932e-06, "loss": 0.2865, "step": 2094 }, { "epoch": 0.4, "grad_norm": 9.5, "learning_rate": 4.638553477930516e-06, "loss": 0.3631, "step": 2095 }, { "epoch": 0.4, "grad_norm": 9.5, "learning_rate": 4.6381566705624185e-06, "loss": 0.2565, "step": 2096 }, { "epoch": 0.4, "grad_norm": 9.3125, "learning_rate": 4.637759662491885e-06, "loss": 0.2313, "step": 2097 }, { "epoch": 0.4, "grad_norm": 12.5625, "learning_rate": 4.637362453756184e-06, "loss": 0.3121, "step": 2098 }, { "epoch": 0.4, "grad_norm": 15.8125, "learning_rate": 4.6369650443926e-06, "loss": 0.3512, "step": 2099 }, { "epoch": 0.4, "grad_norm": 10.4375, "learning_rate": 4.636567434438436e-06, "loss": 0.2923, "step": 2100 }, { "epoch": 0.4, "grad_norm": 13.625, "learning_rate": 4.636169623931015e-06, "loss": 0.3131, "step": 2101 }, { "epoch": 0.4, "grad_norm": 10.5, "learning_rate": 4.635771612907677e-06, "loss": 0.3363, "step": 2102 }, { "epoch": 0.4, "grad_norm": 9.125, "learning_rate": 4.635373401405783e-06, "loss": 0.2658, "step": 2103 }, { "epoch": 0.4, "grad_norm": 14.4375, "learning_rate": 4.634974989462712e-06, "loss": 0.3396, "step": 2104 }, { "epoch": 0.4, "grad_norm": 12.5625, "learning_rate": 4.634576377115862e-06, "loss": 0.3128, "step": 2105 }, { "epoch": 0.4, "grad_norm": 12.875, "learning_rate": 4.634177564402649e-06, "loss": 0.2389, "step": 2106 }, { "epoch": 0.4, "grad_norm": 14.1875, "learning_rate": 4.6337785513605085e-06, "loss": 0.3633, "step": 2107 }, { "epoch": 0.4, "grad_norm": 8.5625, "learning_rate": 4.633379338026894e-06, "loss": 0.3399, "step": 2108 }, { "epoch": 0.4, "grad_norm": 10.625, "learning_rate": 4.632979924439279e-06, "loss": 0.3095, "step": 2109 }, { "epoch": 0.4, "grad_norm": 12.625, "learning_rate": 4.632580310635157e-06, "loss": 0.2698, "step": 2110 }, { "epoch": 0.4, "grad_norm": 10.625, "learning_rate": 4.6321804966520365e-06, "loss": 0.3115, "step": 2111 }, { "epoch": 0.4, "grad_norm": 10.4375, "learning_rate": 4.631780482527447e-06, "loss": 0.3155, "step": 2112 }, { "epoch": 0.4, "grad_norm": 9.875, "learning_rate": 4.631380268298938e-06, "loss": 0.3096, "step": 2113 }, { "epoch": 0.4, "grad_norm": 10.4375, "learning_rate": 4.6309798540040736e-06, "loss": 0.3286, "step": 2114 }, { "epoch": 0.4, "grad_norm": 13.125, "learning_rate": 4.630579239680442e-06, "loss": 0.2614, "step": 2115 }, { "epoch": 0.4, "grad_norm": 11.9375, "learning_rate": 4.630178425365648e-06, "loss": 0.2649, "step": 2116 }, { "epoch": 0.4, "grad_norm": 7.625, "learning_rate": 4.6297774110973135e-06, "loss": 0.2644, "step": 2117 }, { "epoch": 0.4, "grad_norm": 11.375, "learning_rate": 4.62937619691308e-06, "loss": 0.2595, "step": 2118 }, { "epoch": 0.4, "grad_norm": 8.25, "learning_rate": 4.62897478285061e-06, "loss": 0.2608, "step": 2119 }, { "epoch": 0.4, "grad_norm": 11.4375, "learning_rate": 4.628573168947582e-06, "loss": 0.3547, "step": 2120 }, { "epoch": 0.4, "grad_norm": 9.5625, "learning_rate": 4.628171355241695e-06, "loss": 0.2669, "step": 2121 }, { "epoch": 0.4, "grad_norm": 18.25, "learning_rate": 4.627769341770665e-06, "loss": 0.3333, "step": 2122 }, { "epoch": 0.4, "grad_norm": 15.125, "learning_rate": 4.6273671285722285e-06, "loss": 0.3528, "step": 2123 }, { "epoch": 0.4, "grad_norm": 11.75, "learning_rate": 4.626964715684141e-06, "loss": 0.3386, "step": 2124 }, { "epoch": 0.4, "grad_norm": 9.0, "learning_rate": 4.626562103144174e-06, "loss": 0.2979, "step": 2125 }, { "epoch": 0.4, "grad_norm": 8.9375, "learning_rate": 4.62615929099012e-06, "loss": 0.3283, "step": 2126 }, { "epoch": 0.4, "grad_norm": 16.625, "learning_rate": 4.625756279259791e-06, "loss": 0.3655, "step": 2127 }, { "epoch": 0.4, "grad_norm": 12.9375, "learning_rate": 4.625353067991015e-06, "loss": 0.325, "step": 2128 }, { "epoch": 0.4, "grad_norm": 10.8125, "learning_rate": 4.624949657221641e-06, "loss": 0.2747, "step": 2129 }, { "epoch": 0.4, "grad_norm": 8.5, "learning_rate": 4.624546046989535e-06, "loss": 0.2371, "step": 2130 }, { "epoch": 0.4, "grad_norm": 13.6875, "learning_rate": 4.624142237332585e-06, "loss": 0.3422, "step": 2131 }, { "epoch": 0.4, "grad_norm": 11.875, "learning_rate": 4.623738228288693e-06, "loss": 0.2908, "step": 2132 }, { "epoch": 0.4, "grad_norm": 11.3125, "learning_rate": 4.623334019895784e-06, "loss": 0.289, "step": 2133 }, { "epoch": 0.4, "grad_norm": 8.3125, "learning_rate": 4.622929612191798e-06, "loss": 0.2575, "step": 2134 }, { "epoch": 0.4, "grad_norm": 55.0, "learning_rate": 4.622525005214698e-06, "loss": 0.3302, "step": 2135 }, { "epoch": 0.4, "grad_norm": 10.4375, "learning_rate": 4.622120199002461e-06, "loss": 0.2535, "step": 2136 }, { "epoch": 0.4, "grad_norm": 10.3125, "learning_rate": 4.6217151935930845e-06, "loss": 0.2878, "step": 2137 }, { "epoch": 0.4, "grad_norm": 12.8125, "learning_rate": 4.621309989024588e-06, "loss": 0.3231, "step": 2138 }, { "epoch": 0.4, "grad_norm": 12.0, "learning_rate": 4.620904585335004e-06, "loss": 0.2301, "step": 2139 }, { "epoch": 0.4, "grad_norm": 11.625, "learning_rate": 4.620498982562389e-06, "loss": 0.3103, "step": 2140 }, { "epoch": 0.41, "grad_norm": 9.25, "learning_rate": 4.620093180744814e-06, "loss": 0.3008, "step": 2141 }, { "epoch": 0.41, "grad_norm": 12.5, "learning_rate": 4.619687179920371e-06, "loss": 0.3096, "step": 2142 }, { "epoch": 0.41, "grad_norm": 12.75, "learning_rate": 4.61928098012717e-06, "loss": 0.3491, "step": 2143 }, { "epoch": 0.41, "grad_norm": 10.1875, "learning_rate": 4.61887458140334e-06, "loss": 0.3382, "step": 2144 }, { "epoch": 0.41, "grad_norm": 8.5, "learning_rate": 4.618467983787027e-06, "loss": 0.3104, "step": 2145 }, { "epoch": 0.41, "grad_norm": 16.25, "learning_rate": 4.6180611873164e-06, "loss": 0.3657, "step": 2146 }, { "epoch": 0.41, "grad_norm": 13.125, "learning_rate": 4.617654192029642e-06, "loss": 0.3411, "step": 2147 }, { "epoch": 0.41, "grad_norm": 12.9375, "learning_rate": 4.617246997964956e-06, "loss": 0.2692, "step": 2148 }, { "epoch": 0.41, "grad_norm": 12.8125, "learning_rate": 4.616839605160564e-06, "loss": 0.3121, "step": 2149 }, { "epoch": 0.41, "grad_norm": 13.0, "learning_rate": 4.616432013654708e-06, "loss": 0.2852, "step": 2150 }, { "epoch": 0.41, "grad_norm": 13.8125, "learning_rate": 4.616024223485648e-06, "loss": 0.3535, "step": 2151 }, { "epoch": 0.41, "grad_norm": 7.625, "learning_rate": 4.61561623469166e-06, "loss": 0.2993, "step": 2152 }, { "epoch": 0.41, "grad_norm": 20.0, "learning_rate": 4.61520804731104e-06, "loss": 0.4162, "step": 2153 }, { "epoch": 0.41, "grad_norm": 22.0, "learning_rate": 4.614799661382107e-06, "loss": 0.4195, "step": 2154 }, { "epoch": 0.41, "grad_norm": 12.625, "learning_rate": 4.614391076943193e-06, "loss": 0.321, "step": 2155 }, { "epoch": 0.41, "grad_norm": 11.375, "learning_rate": 4.6139822940326495e-06, "loss": 0.2587, "step": 2156 }, { "epoch": 0.41, "grad_norm": 8.625, "learning_rate": 4.6135733126888495e-06, "loss": 0.2818, "step": 2157 }, { "epoch": 0.41, "grad_norm": 9.125, "learning_rate": 4.613164132950181e-06, "loss": 0.2763, "step": 2158 }, { "epoch": 0.41, "grad_norm": 13.75, "learning_rate": 4.6127547548550545e-06, "loss": 0.3275, "step": 2159 }, { "epoch": 0.41, "grad_norm": 10.0625, "learning_rate": 4.612345178441896e-06, "loss": 0.2924, "step": 2160 }, { "epoch": 0.41, "grad_norm": 10.5, "learning_rate": 4.611935403749152e-06, "loss": 0.2778, "step": 2161 }, { "epoch": 0.41, "grad_norm": 10.9375, "learning_rate": 4.611525430815286e-06, "loss": 0.3066, "step": 2162 }, { "epoch": 0.41, "grad_norm": 14.9375, "learning_rate": 4.611115259678781e-06, "loss": 0.4171, "step": 2163 }, { "epoch": 0.41, "grad_norm": 9.875, "learning_rate": 4.6107048903781396e-06, "loss": 0.3366, "step": 2164 }, { "epoch": 0.41, "grad_norm": 9.375, "learning_rate": 4.61029432295188e-06, "loss": 0.298, "step": 2165 }, { "epoch": 0.41, "grad_norm": 14.625, "learning_rate": 4.609883557438543e-06, "loss": 0.2913, "step": 2166 }, { "epoch": 0.41, "grad_norm": 10.0, "learning_rate": 4.609472593876685e-06, "loss": 0.2791, "step": 2167 }, { "epoch": 0.41, "grad_norm": 11.625, "learning_rate": 4.609061432304881e-06, "loss": 0.2957, "step": 2168 }, { "epoch": 0.41, "grad_norm": 9.25, "learning_rate": 4.608650072761726e-06, "loss": 0.3022, "step": 2169 }, { "epoch": 0.41, "grad_norm": 13.5, "learning_rate": 4.608238515285836e-06, "loss": 0.4214, "step": 2170 }, { "epoch": 0.41, "grad_norm": 10.8125, "learning_rate": 4.607826759915839e-06, "loss": 0.2769, "step": 2171 }, { "epoch": 0.41, "grad_norm": 9.0, "learning_rate": 4.607414806690386e-06, "loss": 0.3541, "step": 2172 }, { "epoch": 0.41, "grad_norm": 10.8125, "learning_rate": 4.607002655648147e-06, "loss": 0.3478, "step": 2173 }, { "epoch": 0.41, "grad_norm": 11.875, "learning_rate": 4.606590306827808e-06, "loss": 0.3711, "step": 2174 }, { "epoch": 0.41, "grad_norm": 13.375, "learning_rate": 4.6061777602680764e-06, "loss": 0.2754, "step": 2175 }, { "epoch": 0.41, "grad_norm": 17.0, "learning_rate": 4.605765016007676e-06, "loss": 0.4055, "step": 2176 }, { "epoch": 0.41, "grad_norm": 11.25, "learning_rate": 4.60535207408535e-06, "loss": 0.2926, "step": 2177 }, { "epoch": 0.41, "grad_norm": 9.75, "learning_rate": 4.6049389345398596e-06, "loss": 0.3166, "step": 2178 }, { "epoch": 0.41, "grad_norm": 9.0, "learning_rate": 4.604525597409985e-06, "loss": 0.2354, "step": 2179 }, { "epoch": 0.41, "grad_norm": 11.1875, "learning_rate": 4.604112062734526e-06, "loss": 0.3555, "step": 2180 }, { "epoch": 0.41, "grad_norm": 11.8125, "learning_rate": 4.6036983305522995e-06, "loss": 0.2875, "step": 2181 }, { "epoch": 0.41, "grad_norm": 13.375, "learning_rate": 4.603284400902141e-06, "loss": 0.3342, "step": 2182 }, { "epoch": 0.41, "grad_norm": 12.875, "learning_rate": 4.602870273822904e-06, "loss": 0.2884, "step": 2183 }, { "epoch": 0.41, "grad_norm": 10.5625, "learning_rate": 4.602455949353462e-06, "loss": 0.2619, "step": 2184 }, { "epoch": 0.41, "grad_norm": 10.3125, "learning_rate": 4.602041427532707e-06, "loss": 0.3284, "step": 2185 }, { "epoch": 0.41, "grad_norm": 8.75, "learning_rate": 4.601626708399549e-06, "loss": 0.2977, "step": 2186 }, { "epoch": 0.41, "grad_norm": 11.3125, "learning_rate": 4.601211791992916e-06, "loss": 0.2606, "step": 2187 }, { "epoch": 0.41, "grad_norm": 9.5, "learning_rate": 4.600796678351754e-06, "loss": 0.2873, "step": 2188 }, { "epoch": 0.41, "grad_norm": 12.25, "learning_rate": 4.60038136751503e-06, "loss": 0.2896, "step": 2189 }, { "epoch": 0.41, "grad_norm": 7.90625, "learning_rate": 4.599965859521728e-06, "loss": 0.254, "step": 2190 }, { "epoch": 0.41, "grad_norm": 9.125, "learning_rate": 4.599550154410849e-06, "loss": 0.2644, "step": 2191 }, { "epoch": 0.41, "grad_norm": 8.4375, "learning_rate": 4.599134252221415e-06, "loss": 0.2657, "step": 2192 }, { "epoch": 0.41, "grad_norm": 12.75, "learning_rate": 4.598718152992466e-06, "loss": 0.313, "step": 2193 }, { "epoch": 0.42, "grad_norm": 10.0, "learning_rate": 4.59830185676306e-06, "loss": 0.2431, "step": 2194 }, { "epoch": 0.42, "grad_norm": 19.25, "learning_rate": 4.597885363572272e-06, "loss": 0.3842, "step": 2195 }, { "epoch": 0.42, "grad_norm": 10.9375, "learning_rate": 4.597468673459199e-06, "loss": 0.3101, "step": 2196 }, { "epoch": 0.42, "grad_norm": 9.9375, "learning_rate": 4.5970517864629524e-06, "loss": 0.3102, "step": 2197 }, { "epoch": 0.42, "grad_norm": 11.4375, "learning_rate": 4.596634702622666e-06, "loss": 0.3923, "step": 2198 }, { "epoch": 0.42, "grad_norm": 32.75, "learning_rate": 4.5962174219774884e-06, "loss": 0.2324, "step": 2199 }, { "epoch": 0.42, "grad_norm": 9.3125, "learning_rate": 4.59579994456659e-06, "loss": 0.2209, "step": 2200 }, { "epoch": 0.42, "grad_norm": 8.8125, "learning_rate": 4.595382270429158e-06, "loss": 0.2862, "step": 2201 }, { "epoch": 0.42, "grad_norm": 12.625, "learning_rate": 4.594964399604397e-06, "loss": 0.3243, "step": 2202 }, { "epoch": 0.42, "grad_norm": 11.6875, "learning_rate": 4.594546332131533e-06, "loss": 0.2682, "step": 2203 }, { "epoch": 0.42, "grad_norm": 11.3125, "learning_rate": 4.594128068049808e-06, "loss": 0.34, "step": 2204 }, { "epoch": 0.42, "grad_norm": 10.9375, "learning_rate": 4.593709607398484e-06, "loss": 0.319, "step": 2205 }, { "epoch": 0.42, "grad_norm": 17.5, "learning_rate": 4.593290950216838e-06, "loss": 0.3416, "step": 2206 }, { "epoch": 0.42, "grad_norm": 9.5, "learning_rate": 4.5928720965441705e-06, "loss": 0.2979, "step": 2207 }, { "epoch": 0.42, "grad_norm": 10.1875, "learning_rate": 4.592453046419798e-06, "loss": 0.3504, "step": 2208 }, { "epoch": 0.42, "grad_norm": 17.125, "learning_rate": 4.5920337998830544e-06, "loss": 0.3297, "step": 2209 }, { "epoch": 0.42, "grad_norm": 28.0, "learning_rate": 4.591614356973294e-06, "loss": 0.2924, "step": 2210 }, { "epoch": 0.42, "grad_norm": 11.25, "learning_rate": 4.591194717729888e-06, "loss": 0.2983, "step": 2211 }, { "epoch": 0.42, "grad_norm": 9.5, "learning_rate": 4.590774882192228e-06, "loss": 0.2812, "step": 2212 }, { "epoch": 0.42, "grad_norm": 13.125, "learning_rate": 4.590354850399721e-06, "loss": 0.3302, "step": 2213 }, { "epoch": 0.42, "grad_norm": 11.875, "learning_rate": 4.589934622391795e-06, "loss": 0.32, "step": 2214 }, { "epoch": 0.42, "grad_norm": 11.875, "learning_rate": 4.589514198207896e-06, "loss": 0.3117, "step": 2215 }, { "epoch": 0.42, "grad_norm": 11.0625, "learning_rate": 4.589093577887487e-06, "loss": 0.2719, "step": 2216 }, { "epoch": 0.42, "grad_norm": 15.125, "learning_rate": 4.588672761470051e-06, "loss": 0.3295, "step": 2217 }, { "epoch": 0.42, "grad_norm": 7.21875, "learning_rate": 4.588251748995088e-06, "loss": 0.2581, "step": 2218 }, { "epoch": 0.42, "grad_norm": 15.9375, "learning_rate": 4.587830540502118e-06, "loss": 0.2889, "step": 2219 }, { "epoch": 0.42, "grad_norm": 12.3125, "learning_rate": 4.587409136030679e-06, "loss": 0.3129, "step": 2220 }, { "epoch": 0.42, "grad_norm": 9.75, "learning_rate": 4.586987535620327e-06, "loss": 0.3218, "step": 2221 }, { "epoch": 0.42, "grad_norm": 11.875, "learning_rate": 4.586565739310635e-06, "loss": 0.2961, "step": 2222 }, { "epoch": 0.42, "grad_norm": 13.125, "learning_rate": 4.586143747141197e-06, "loss": 0.3109, "step": 2223 }, { "epoch": 0.42, "grad_norm": 11.0, "learning_rate": 4.585721559151622e-06, "loss": 0.3023, "step": 2224 }, { "epoch": 0.42, "grad_norm": 15.3125, "learning_rate": 4.585299175381543e-06, "loss": 0.3356, "step": 2225 }, { "epoch": 0.42, "grad_norm": 12.875, "learning_rate": 4.584876595870606e-06, "loss": 0.2482, "step": 2226 }, { "epoch": 0.42, "grad_norm": 15.9375, "learning_rate": 4.584453820658476e-06, "loss": 0.3102, "step": 2227 }, { "epoch": 0.42, "grad_norm": 11.1875, "learning_rate": 4.584030849784841e-06, "loss": 0.2938, "step": 2228 }, { "epoch": 0.42, "grad_norm": 7.1875, "learning_rate": 4.583607683289402e-06, "loss": 0.2439, "step": 2229 }, { "epoch": 0.42, "grad_norm": 11.1875, "learning_rate": 4.58318432121188e-06, "loss": 0.3032, "step": 2230 }, { "epoch": 0.42, "grad_norm": 7.5, "learning_rate": 4.582760763592015e-06, "loss": 0.2987, "step": 2231 }, { "epoch": 0.42, "grad_norm": 13.4375, "learning_rate": 4.582337010469565e-06, "loss": 0.3057, "step": 2232 }, { "epoch": 0.42, "grad_norm": 11.9375, "learning_rate": 4.5819130618843074e-06, "loss": 0.3057, "step": 2233 }, { "epoch": 0.42, "grad_norm": 8.25, "learning_rate": 4.581488917876037e-06, "loss": 0.2952, "step": 2234 }, { "epoch": 0.42, "grad_norm": 12.25, "learning_rate": 4.581064578484566e-06, "loss": 0.3044, "step": 2235 }, { "epoch": 0.42, "grad_norm": 11.875, "learning_rate": 4.580640043749725e-06, "loss": 0.2948, "step": 2236 }, { "epoch": 0.42, "grad_norm": 12.0625, "learning_rate": 4.580215313711367e-06, "loss": 0.3831, "step": 2237 }, { "epoch": 0.42, "grad_norm": 9.625, "learning_rate": 4.5797903884093575e-06, "loss": 0.2955, "step": 2238 }, { "epoch": 0.42, "grad_norm": 13.3125, "learning_rate": 4.579365267883583e-06, "loss": 0.3312, "step": 2239 }, { "epoch": 0.42, "grad_norm": 11.4375, "learning_rate": 4.57893995217395e-06, "loss": 0.2493, "step": 2240 }, { "epoch": 0.42, "grad_norm": 12.3125, "learning_rate": 4.57851444132038e-06, "loss": 0.3152, "step": 2241 }, { "epoch": 0.42, "grad_norm": 12.0, "learning_rate": 4.578088735362815e-06, "loss": 0.2812, "step": 2242 }, { "epoch": 0.42, "grad_norm": 8.75, "learning_rate": 4.577662834341214e-06, "loss": 0.3339, "step": 2243 }, { "epoch": 0.42, "grad_norm": 32.0, "learning_rate": 4.577236738295557e-06, "loss": 0.3093, "step": 2244 }, { "epoch": 0.42, "grad_norm": 11.75, "learning_rate": 4.57681044726584e-06, "loss": 0.3775, "step": 2245 }, { "epoch": 0.42, "grad_norm": 13.1875, "learning_rate": 4.576383961292076e-06, "loss": 0.243, "step": 2246 }, { "epoch": 0.43, "grad_norm": 9.5625, "learning_rate": 4.575957280414299e-06, "loss": 0.2591, "step": 2247 }, { "epoch": 0.43, "grad_norm": 16.625, "learning_rate": 4.5755304046725605e-06, "loss": 0.3653, "step": 2248 }, { "epoch": 0.43, "grad_norm": 10.875, "learning_rate": 4.575103334106929e-06, "loss": 0.2869, "step": 2249 }, { "epoch": 0.43, "grad_norm": 6.78125, "learning_rate": 4.5746760687574935e-06, "loss": 0.2467, "step": 2250 }, { "epoch": 0.43, "grad_norm": 8.3125, "learning_rate": 4.5742486086643605e-06, "loss": 0.2935, "step": 2251 }, { "epoch": 0.43, "grad_norm": 10.375, "learning_rate": 4.573820953867652e-06, "loss": 0.2691, "step": 2252 }, { "epoch": 0.43, "grad_norm": 9.0, "learning_rate": 4.573393104407513e-06, "loss": 0.2812, "step": 2253 }, { "epoch": 0.43, "grad_norm": 9.5625, "learning_rate": 4.5729650603241044e-06, "loss": 0.3007, "step": 2254 }, { "epoch": 0.43, "grad_norm": 19.25, "learning_rate": 4.5725368216576045e-06, "loss": 0.3287, "step": 2255 }, { "epoch": 0.43, "grad_norm": 8.125, "learning_rate": 4.572108388448211e-06, "loss": 0.2857, "step": 2256 }, { "epoch": 0.43, "grad_norm": 9.8125, "learning_rate": 4.5716797607361394e-06, "loss": 0.2092, "step": 2257 }, { "epoch": 0.43, "grad_norm": 9.9375, "learning_rate": 4.571250938561625e-06, "loss": 0.287, "step": 2258 }, { "epoch": 0.43, "grad_norm": 9.6875, "learning_rate": 4.5708219219649185e-06, "loss": 0.323, "step": 2259 }, { "epoch": 0.43, "grad_norm": 10.8125, "learning_rate": 4.5703927109862914e-06, "loss": 0.2945, "step": 2260 }, { "epoch": 0.43, "grad_norm": 12.0, "learning_rate": 4.569963305666032e-06, "loss": 0.311, "step": 2261 }, { "epoch": 0.43, "grad_norm": 14.25, "learning_rate": 4.569533706044447e-06, "loss": 0.3481, "step": 2262 }, { "epoch": 0.43, "grad_norm": 11.75, "learning_rate": 4.569103912161863e-06, "loss": 0.3405, "step": 2263 }, { "epoch": 0.43, "grad_norm": 8.25, "learning_rate": 4.568673924058622e-06, "loss": 0.3067, "step": 2264 }, { "epoch": 0.43, "grad_norm": 9.625, "learning_rate": 4.568243741775086e-06, "loss": 0.2818, "step": 2265 }, { "epoch": 0.43, "grad_norm": 9.25, "learning_rate": 4.567813365351635e-06, "loss": 0.2855, "step": 2266 }, { "epoch": 0.43, "grad_norm": 14.0, "learning_rate": 4.567382794828669e-06, "loss": 0.3694, "step": 2267 }, { "epoch": 0.43, "grad_norm": 13.0, "learning_rate": 4.5669520302466e-06, "loss": 0.3691, "step": 2268 }, { "epoch": 0.43, "grad_norm": 14.6875, "learning_rate": 4.566521071645869e-06, "loss": 0.2503, "step": 2269 }, { "epoch": 0.43, "grad_norm": 14.5625, "learning_rate": 4.566089919066922e-06, "loss": 0.2853, "step": 2270 }, { "epoch": 0.43, "grad_norm": 16.25, "learning_rate": 4.5656585725502345e-06, "loss": 0.2654, "step": 2271 }, { "epoch": 0.43, "grad_norm": 10.8125, "learning_rate": 4.565227032136295e-06, "loss": 0.3584, "step": 2272 }, { "epoch": 0.43, "grad_norm": 14.0625, "learning_rate": 4.564795297865609e-06, "loss": 0.3227, "step": 2273 }, { "epoch": 0.43, "grad_norm": 8.125, "learning_rate": 4.564363369778704e-06, "loss": 0.2969, "step": 2274 }, { "epoch": 0.43, "grad_norm": 9.125, "learning_rate": 4.563931247916124e-06, "loss": 0.2749, "step": 2275 }, { "epoch": 0.43, "grad_norm": 8.3125, "learning_rate": 4.563498932318429e-06, "loss": 0.2744, "step": 2276 }, { "epoch": 0.43, "grad_norm": 18.0, "learning_rate": 4.563066423026201e-06, "loss": 0.3235, "step": 2277 }, { "epoch": 0.43, "grad_norm": 10.3125, "learning_rate": 4.5626337200800375e-06, "loss": 0.2685, "step": 2278 }, { "epoch": 0.43, "grad_norm": 9.125, "learning_rate": 4.562200823520555e-06, "loss": 0.2514, "step": 2279 }, { "epoch": 0.43, "grad_norm": 8.9375, "learning_rate": 4.561767733388389e-06, "loss": 0.2552, "step": 2280 }, { "epoch": 0.43, "grad_norm": 11.0, "learning_rate": 4.561334449724193e-06, "loss": 0.2911, "step": 2281 }, { "epoch": 0.43, "grad_norm": 10.625, "learning_rate": 4.560900972568635e-06, "loss": 0.3116, "step": 2282 }, { "epoch": 0.43, "grad_norm": 13.1875, "learning_rate": 4.560467301962408e-06, "loss": 0.3179, "step": 2283 }, { "epoch": 0.43, "grad_norm": 20.875, "learning_rate": 4.560033437946217e-06, "loss": 0.2573, "step": 2284 }, { "epoch": 0.43, "grad_norm": 13.75, "learning_rate": 4.559599380560788e-06, "loss": 0.2973, "step": 2285 }, { "epoch": 0.43, "grad_norm": 11.625, "learning_rate": 4.559165129846864e-06, "loss": 0.3588, "step": 2286 }, { "epoch": 0.43, "grad_norm": 9.9375, "learning_rate": 4.558730685845208e-06, "loss": 0.3313, "step": 2287 }, { "epoch": 0.43, "grad_norm": 9.5625, "learning_rate": 4.5582960485966e-06, "loss": 0.3074, "step": 2288 }, { "epoch": 0.43, "grad_norm": 11.25, "learning_rate": 4.557861218141838e-06, "loss": 0.359, "step": 2289 }, { "epoch": 0.43, "grad_norm": 9.5, "learning_rate": 4.5574261945217364e-06, "loss": 0.2765, "step": 2290 }, { "epoch": 0.43, "grad_norm": 20.625, "learning_rate": 4.556990977777133e-06, "loss": 0.3083, "step": 2291 }, { "epoch": 0.43, "grad_norm": 14.25, "learning_rate": 4.5565555679488785e-06, "loss": 0.3248, "step": 2292 }, { "epoch": 0.43, "grad_norm": 9.8125, "learning_rate": 4.556119965077842e-06, "loss": 0.2117, "step": 2293 }, { "epoch": 0.43, "grad_norm": 16.25, "learning_rate": 4.555684169204915e-06, "loss": 0.3148, "step": 2294 }, { "epoch": 0.43, "grad_norm": 12.1875, "learning_rate": 4.555248180371003e-06, "loss": 0.3442, "step": 2295 }, { "epoch": 0.43, "grad_norm": 26.75, "learning_rate": 4.554811998617029e-06, "loss": 0.3489, "step": 2296 }, { "epoch": 0.43, "grad_norm": 12.25, "learning_rate": 4.554375623983941e-06, "loss": 0.2507, "step": 2297 }, { "epoch": 0.43, "grad_norm": 13.25, "learning_rate": 4.553939056512696e-06, "loss": 0.2417, "step": 2298 }, { "epoch": 0.43, "grad_norm": 12.125, "learning_rate": 4.5535022962442745e-06, "loss": 0.3237, "step": 2299 }, { "epoch": 0.44, "grad_norm": 9.75, "learning_rate": 4.5530653432196744e-06, "loss": 0.3199, "step": 2300 }, { "epoch": 0.44, "grad_norm": 12.4375, "learning_rate": 4.552628197479911e-06, "loss": 0.2715, "step": 2301 }, { "epoch": 0.44, "grad_norm": 15.25, "learning_rate": 4.552190859066017e-06, "loss": 0.4008, "step": 2302 }, { "epoch": 0.44, "grad_norm": 8.9375, "learning_rate": 4.551753328019046e-06, "loss": 0.2944, "step": 2303 }, { "epoch": 0.44, "grad_norm": 13.5, "learning_rate": 4.551315604380066e-06, "loss": 0.402, "step": 2304 }, { "epoch": 0.44, "grad_norm": 10.75, "learning_rate": 4.550877688190164e-06, "loss": 0.3417, "step": 2305 }, { "epoch": 0.44, "grad_norm": 9.875, "learning_rate": 4.550439579490448e-06, "loss": 0.2934, "step": 2306 }, { "epoch": 0.44, "grad_norm": 9.6875, "learning_rate": 4.550001278322043e-06, "loss": 0.3218, "step": 2307 }, { "epoch": 0.44, "grad_norm": 14.1875, "learning_rate": 4.549562784726087e-06, "loss": 0.3755, "step": 2308 }, { "epoch": 0.44, "grad_norm": 15.375, "learning_rate": 4.549124098743743e-06, "loss": 0.2692, "step": 2309 }, { "epoch": 0.44, "grad_norm": 11.5625, "learning_rate": 4.548685220416189e-06, "loss": 0.3955, "step": 2310 }, { "epoch": 0.44, "grad_norm": 10.4375, "learning_rate": 4.54824614978462e-06, "loss": 0.2625, "step": 2311 }, { "epoch": 0.44, "grad_norm": 14.4375, "learning_rate": 4.547806886890252e-06, "loss": 0.2853, "step": 2312 }, { "epoch": 0.44, "grad_norm": 8.8125, "learning_rate": 4.547367431774314e-06, "loss": 0.2847, "step": 2313 }, { "epoch": 0.44, "grad_norm": 12.125, "learning_rate": 4.546927784478061e-06, "loss": 0.248, "step": 2314 }, { "epoch": 0.44, "grad_norm": 8.875, "learning_rate": 4.546487945042758e-06, "loss": 0.2595, "step": 2315 }, { "epoch": 0.44, "grad_norm": 10.9375, "learning_rate": 4.5460479135096926e-06, "loss": 0.2766, "step": 2316 }, { "epoch": 0.44, "grad_norm": 10.25, "learning_rate": 4.545607689920169e-06, "loss": 0.3375, "step": 2317 }, { "epoch": 0.44, "grad_norm": 17.125, "learning_rate": 4.545167274315509e-06, "loss": 0.3229, "step": 2318 }, { "epoch": 0.44, "grad_norm": 12.125, "learning_rate": 4.544726666737055e-06, "loss": 0.3121, "step": 2319 }, { "epoch": 0.44, "grad_norm": 10.75, "learning_rate": 4.544285867226164e-06, "loss": 0.3131, "step": 2320 }, { "epoch": 0.44, "grad_norm": 10.3125, "learning_rate": 4.543844875824213e-06, "loss": 0.3419, "step": 2321 }, { "epoch": 0.44, "grad_norm": 15.3125, "learning_rate": 4.543403692572597e-06, "loss": 0.2436, "step": 2322 }, { "epoch": 0.44, "grad_norm": 14.3125, "learning_rate": 4.542962317512727e-06, "loss": 0.2813, "step": 2323 }, { "epoch": 0.44, "grad_norm": 11.4375, "learning_rate": 4.542520750686035e-06, "loss": 0.2832, "step": 2324 }, { "epoch": 0.44, "grad_norm": 18.25, "learning_rate": 4.542078992133969e-06, "loss": 0.2939, "step": 2325 }, { "epoch": 0.44, "grad_norm": 14.875, "learning_rate": 4.541637041897996e-06, "loss": 0.2385, "step": 2326 }, { "epoch": 0.44, "grad_norm": 8.3125, "learning_rate": 4.541194900019601e-06, "loss": 0.2596, "step": 2327 }, { "epoch": 0.44, "grad_norm": 12.875, "learning_rate": 4.540752566540285e-06, "loss": 0.3492, "step": 2328 }, { "epoch": 0.44, "grad_norm": 12.3125, "learning_rate": 4.5403100415015696e-06, "loss": 0.2659, "step": 2329 }, { "epoch": 0.44, "grad_norm": 8.3125, "learning_rate": 4.539867324944994e-06, "loss": 0.2963, "step": 2330 }, { "epoch": 0.44, "grad_norm": 12.375, "learning_rate": 4.539424416912113e-06, "loss": 0.2485, "step": 2331 }, { "epoch": 0.44, "grad_norm": 8.875, "learning_rate": 4.538981317444502e-06, "loss": 0.314, "step": 2332 }, { "epoch": 0.44, "grad_norm": 10.6875, "learning_rate": 4.538538026583753e-06, "loss": 0.3325, "step": 2333 }, { "epoch": 0.44, "grad_norm": 13.125, "learning_rate": 4.538094544371478e-06, "loss": 0.3408, "step": 2334 }, { "epoch": 0.44, "grad_norm": 30.375, "learning_rate": 4.537650870849303e-06, "loss": 0.2617, "step": 2335 }, { "epoch": 0.44, "grad_norm": 10.375, "learning_rate": 4.5372070060588745e-06, "loss": 0.358, "step": 2336 }, { "epoch": 0.44, "grad_norm": 9.0, "learning_rate": 4.536762950041859e-06, "loss": 0.2732, "step": 2337 }, { "epoch": 0.44, "grad_norm": 11.6875, "learning_rate": 4.536318702839937e-06, "loss": 0.3334, "step": 2338 }, { "epoch": 0.44, "grad_norm": 14.8125, "learning_rate": 4.535874264494809e-06, "loss": 0.3379, "step": 2339 }, { "epoch": 0.44, "grad_norm": 10.5625, "learning_rate": 4.535429635048193e-06, "loss": 0.2554, "step": 2340 }, { "epoch": 0.44, "grad_norm": 10.8125, "learning_rate": 4.534984814541827e-06, "loss": 0.3196, "step": 2341 }, { "epoch": 0.44, "grad_norm": 10.6875, "learning_rate": 4.534539803017461e-06, "loss": 0.2747, "step": 2342 }, { "epoch": 0.44, "grad_norm": 11.375, "learning_rate": 4.53409460051687e-06, "loss": 0.3104, "step": 2343 }, { "epoch": 0.44, "grad_norm": 9.3125, "learning_rate": 4.533649207081842e-06, "loss": 0.3106, "step": 2344 }, { "epoch": 0.44, "grad_norm": 10.5, "learning_rate": 4.533203622754187e-06, "loss": 0.2604, "step": 2345 }, { "epoch": 0.44, "grad_norm": 11.0625, "learning_rate": 4.532757847575729e-06, "loss": 0.298, "step": 2346 }, { "epoch": 0.44, "grad_norm": 9.6875, "learning_rate": 4.532311881588312e-06, "loss": 0.3073, "step": 2347 }, { "epoch": 0.44, "grad_norm": 11.375, "learning_rate": 4.5318657248337976e-06, "loss": 0.289, "step": 2348 }, { "epoch": 0.44, "grad_norm": 7.90625, "learning_rate": 4.531419377354065e-06, "loss": 0.3207, "step": 2349 }, { "epoch": 0.44, "grad_norm": 9.3125, "learning_rate": 4.530972839191012e-06, "loss": 0.3095, "step": 2350 }, { "epoch": 0.44, "grad_norm": 12.625, "learning_rate": 4.530526110386553e-06, "loss": 0.2948, "step": 2351 }, { "epoch": 0.44, "grad_norm": 7.15625, "learning_rate": 4.530079190982621e-06, "loss": 0.2861, "step": 2352 }, { "epoch": 0.45, "grad_norm": 8.625, "learning_rate": 4.5296320810211695e-06, "loss": 0.2872, "step": 2353 }, { "epoch": 0.45, "grad_norm": 12.9375, "learning_rate": 4.529184780544164e-06, "loss": 0.3124, "step": 2354 }, { "epoch": 0.45, "grad_norm": 7.96875, "learning_rate": 4.528737289593592e-06, "loss": 0.2376, "step": 2355 }, { "epoch": 0.45, "grad_norm": 10.75, "learning_rate": 4.52828960821146e-06, "loss": 0.3086, "step": 2356 }, { "epoch": 0.45, "grad_norm": 12.25, "learning_rate": 4.5278417364397895e-06, "loss": 0.3353, "step": 2357 }, { "epoch": 0.45, "grad_norm": 9.75, "learning_rate": 4.52739367432062e-06, "loss": 0.2986, "step": 2358 }, { "epoch": 0.45, "grad_norm": 13.5625, "learning_rate": 4.526945421896012e-06, "loss": 0.2551, "step": 2359 }, { "epoch": 0.45, "grad_norm": 9.25, "learning_rate": 4.52649697920804e-06, "loss": 0.3057, "step": 2360 }, { "epoch": 0.45, "grad_norm": 8.4375, "learning_rate": 4.526048346298797e-06, "loss": 0.2972, "step": 2361 }, { "epoch": 0.45, "grad_norm": 12.6875, "learning_rate": 4.525599523210396e-06, "loss": 0.2727, "step": 2362 }, { "epoch": 0.45, "grad_norm": 9.4375, "learning_rate": 4.5251505099849674e-06, "loss": 0.311, "step": 2363 }, { "epoch": 0.45, "grad_norm": 12.75, "learning_rate": 4.524701306664658e-06, "loss": 0.283, "step": 2364 }, { "epoch": 0.45, "grad_norm": 9.0625, "learning_rate": 4.524251913291633e-06, "loss": 0.2044, "step": 2365 }, { "epoch": 0.45, "grad_norm": 11.3125, "learning_rate": 4.523802329908076e-06, "loss": 0.2587, "step": 2366 }, { "epoch": 0.45, "grad_norm": 15.3125, "learning_rate": 4.523352556556187e-06, "loss": 0.3346, "step": 2367 }, { "epoch": 0.45, "grad_norm": 10.0, "learning_rate": 4.5229025932781876e-06, "loss": 0.2816, "step": 2368 }, { "epoch": 0.45, "grad_norm": 10.8125, "learning_rate": 4.522452440116312e-06, "loss": 0.2937, "step": 2369 }, { "epoch": 0.45, "grad_norm": 10.75, "learning_rate": 4.522002097112815e-06, "loss": 0.263, "step": 2370 }, { "epoch": 0.45, "grad_norm": 10.8125, "learning_rate": 4.52155156430997e-06, "loss": 0.3664, "step": 2371 }, { "epoch": 0.45, "grad_norm": 10.625, "learning_rate": 4.521100841750066e-06, "loss": 0.3375, "step": 2372 }, { "epoch": 0.45, "grad_norm": 15.5625, "learning_rate": 4.520649929475412e-06, "loss": 0.2929, "step": 2373 }, { "epoch": 0.45, "grad_norm": 12.4375, "learning_rate": 4.520198827528334e-06, "loss": 0.3232, "step": 2374 }, { "epoch": 0.45, "grad_norm": 15.875, "learning_rate": 4.519747535951174e-06, "loss": 0.3027, "step": 2375 }, { "epoch": 0.45, "grad_norm": 10.8125, "learning_rate": 4.519296054786295e-06, "loss": 0.353, "step": 2376 }, { "epoch": 0.45, "grad_norm": 11.0, "learning_rate": 4.518844384076075e-06, "loss": 0.3091, "step": 2377 }, { "epoch": 0.45, "grad_norm": 13.625, "learning_rate": 4.518392523862913e-06, "loss": 0.3981, "step": 2378 }, { "epoch": 0.45, "grad_norm": 13.25, "learning_rate": 4.517940474189222e-06, "loss": 0.3135, "step": 2379 }, { "epoch": 0.45, "grad_norm": 13.5625, "learning_rate": 4.5174882350974335e-06, "loss": 0.3072, "step": 2380 }, { "epoch": 0.45, "grad_norm": 8.8125, "learning_rate": 4.517035806630001e-06, "loss": 0.3065, "step": 2381 }, { "epoch": 0.45, "grad_norm": 8.75, "learning_rate": 4.51658318882939e-06, "loss": 0.2496, "step": 2382 }, { "epoch": 0.45, "grad_norm": 9.375, "learning_rate": 4.516130381738087e-06, "loss": 0.3201, "step": 2383 }, { "epoch": 0.45, "grad_norm": 8.875, "learning_rate": 4.515677385398597e-06, "loss": 0.2542, "step": 2384 }, { "epoch": 0.45, "grad_norm": 9.9375, "learning_rate": 4.5152241998534405e-06, "loss": 0.3373, "step": 2385 }, { "epoch": 0.45, "grad_norm": 8.875, "learning_rate": 4.514770825145156e-06, "loss": 0.2919, "step": 2386 }, { "epoch": 0.45, "grad_norm": 9.25, "learning_rate": 4.514317261316301e-06, "loss": 0.275, "step": 2387 }, { "epoch": 0.45, "grad_norm": 11.875, "learning_rate": 4.51386350840945e-06, "loss": 0.3761, "step": 2388 }, { "epoch": 0.45, "grad_norm": 9.375, "learning_rate": 4.513409566467196e-06, "loss": 0.2737, "step": 2389 }, { "epoch": 0.45, "grad_norm": 17.75, "learning_rate": 4.512955435532149e-06, "loss": 0.2751, "step": 2390 }, { "epoch": 0.45, "grad_norm": 8.125, "learning_rate": 4.512501115646937e-06, "loss": 0.2399, "step": 2391 }, { "epoch": 0.45, "grad_norm": 23.0, "learning_rate": 4.512046606854205e-06, "loss": 0.3092, "step": 2392 }, { "epoch": 0.45, "grad_norm": 9.625, "learning_rate": 4.511591909196616e-06, "loss": 0.3448, "step": 2393 }, { "epoch": 0.45, "grad_norm": 10.8125, "learning_rate": 4.511137022716853e-06, "loss": 0.2789, "step": 2394 }, { "epoch": 0.45, "grad_norm": 9.375, "learning_rate": 4.510681947457614e-06, "loss": 0.2751, "step": 2395 }, { "epoch": 0.45, "grad_norm": 9.0625, "learning_rate": 4.510226683461615e-06, "loss": 0.3111, "step": 2396 }, { "epoch": 0.45, "grad_norm": 8.625, "learning_rate": 4.509771230771591e-06, "loss": 0.3165, "step": 2397 }, { "epoch": 0.45, "grad_norm": 12.5, "learning_rate": 4.5093155894302935e-06, "loss": 0.2968, "step": 2398 }, { "epoch": 0.45, "grad_norm": 11.375, "learning_rate": 4.508859759480492e-06, "loss": 0.3531, "step": 2399 }, { "epoch": 0.45, "grad_norm": 10.3125, "learning_rate": 4.508403740964974e-06, "loss": 0.3076, "step": 2400 }, { "epoch": 0.45, "grad_norm": 15.625, "learning_rate": 4.507947533926546e-06, "loss": 0.313, "step": 2401 }, { "epoch": 0.45, "grad_norm": 13.875, "learning_rate": 4.507491138408028e-06, "loss": 0.2843, "step": 2402 }, { "epoch": 0.45, "grad_norm": 14.8125, "learning_rate": 4.507034554452263e-06, "loss": 0.3469, "step": 2403 }, { "epoch": 0.45, "grad_norm": 13.6875, "learning_rate": 4.506577782102109e-06, "loss": 0.4005, "step": 2404 }, { "epoch": 0.45, "grad_norm": 7.0, "learning_rate": 4.506120821400441e-06, "loss": 0.2547, "step": 2405 }, { "epoch": 0.46, "grad_norm": 15.3125, "learning_rate": 4.505663672390153e-06, "loss": 0.297, "step": 2406 }, { "epoch": 0.46, "grad_norm": 8.75, "learning_rate": 4.505206335114155e-06, "loss": 0.316, "step": 2407 }, { "epoch": 0.46, "grad_norm": 9.3125, "learning_rate": 4.504748809615377e-06, "loss": 0.3249, "step": 2408 }, { "epoch": 0.46, "grad_norm": 12.5, "learning_rate": 4.504291095936766e-06, "loss": 0.3298, "step": 2409 }, { "epoch": 0.46, "grad_norm": 10.625, "learning_rate": 4.503833194121286e-06, "loss": 0.3418, "step": 2410 }, { "epoch": 0.46, "grad_norm": 12.375, "learning_rate": 4.503375104211917e-06, "loss": 0.2785, "step": 2411 }, { "epoch": 0.46, "grad_norm": 9.8125, "learning_rate": 4.5029168262516615e-06, "loss": 0.2772, "step": 2412 }, { "epoch": 0.46, "grad_norm": 11.0, "learning_rate": 4.502458360283535e-06, "loss": 0.3124, "step": 2413 }, { "epoch": 0.46, "grad_norm": 10.6875, "learning_rate": 4.501999706350572e-06, "loss": 0.3032, "step": 2414 }, { "epoch": 0.46, "grad_norm": 13.5, "learning_rate": 4.501540864495827e-06, "loss": 0.3342, "step": 2415 }, { "epoch": 0.46, "grad_norm": 13.25, "learning_rate": 4.501081834762368e-06, "loss": 0.2887, "step": 2416 }, { "epoch": 0.46, "grad_norm": 8.4375, "learning_rate": 4.5006226171932835e-06, "loss": 0.3518, "step": 2417 }, { "epoch": 0.46, "grad_norm": 8.5, "learning_rate": 4.500163211831678e-06, "loss": 0.2669, "step": 2418 }, { "epoch": 0.46, "grad_norm": 8.75, "learning_rate": 4.499703618720677e-06, "loss": 0.3097, "step": 2419 }, { "epoch": 0.46, "grad_norm": 11.5625, "learning_rate": 4.499243837903419e-06, "loss": 0.3294, "step": 2420 }, { "epoch": 0.46, "grad_norm": 11.375, "learning_rate": 4.498783869423063e-06, "loss": 0.3107, "step": 2421 }, { "epoch": 0.46, "grad_norm": 12.125, "learning_rate": 4.4983237133227845e-06, "loss": 0.328, "step": 2422 }, { "epoch": 0.46, "grad_norm": 11.1875, "learning_rate": 4.497863369645777e-06, "loss": 0.3001, "step": 2423 }, { "epoch": 0.46, "grad_norm": 9.25, "learning_rate": 4.497402838435252e-06, "loss": 0.3036, "step": 2424 }, { "epoch": 0.46, "grad_norm": 9.875, "learning_rate": 4.496942119734438e-06, "loss": 0.2504, "step": 2425 }, { "epoch": 0.46, "grad_norm": 8.0, "learning_rate": 4.496481213586581e-06, "loss": 0.2818, "step": 2426 }, { "epoch": 0.46, "grad_norm": 21.125, "learning_rate": 4.496020120034946e-06, "loss": 0.4093, "step": 2427 }, { "epoch": 0.46, "grad_norm": 8.875, "learning_rate": 4.495558839122813e-06, "loss": 0.403, "step": 2428 }, { "epoch": 0.46, "grad_norm": 9.8125, "learning_rate": 4.495097370893482e-06, "loss": 0.2439, "step": 2429 }, { "epoch": 0.46, "grad_norm": 18.25, "learning_rate": 4.49463571539027e-06, "loss": 0.3591, "step": 2430 }, { "epoch": 0.46, "grad_norm": 15.0625, "learning_rate": 4.49417387265651e-06, "loss": 0.319, "step": 2431 }, { "epoch": 0.46, "grad_norm": 11.375, "learning_rate": 4.493711842735554e-06, "loss": 0.2414, "step": 2432 }, { "epoch": 0.46, "grad_norm": 13.625, "learning_rate": 4.4932496256707725e-06, "loss": 0.301, "step": 2433 }, { "epoch": 0.46, "grad_norm": 12.0625, "learning_rate": 4.492787221505551e-06, "loss": 0.3333, "step": 2434 }, { "epoch": 0.46, "grad_norm": 10.4375, "learning_rate": 4.492324630283296e-06, "loss": 0.3456, "step": 2435 }, { "epoch": 0.46, "grad_norm": 7.96875, "learning_rate": 4.491861852047428e-06, "loss": 0.2044, "step": 2436 }, { "epoch": 0.46, "grad_norm": 11.875, "learning_rate": 4.491398886841386e-06, "loss": 0.2798, "step": 2437 }, { "epoch": 0.46, "grad_norm": 8.4375, "learning_rate": 4.490935734708629e-06, "loss": 0.2868, "step": 2438 }, { "epoch": 0.46, "grad_norm": 10.5, "learning_rate": 4.490472395692631e-06, "loss": 0.2727, "step": 2439 }, { "epoch": 0.46, "grad_norm": 9.4375, "learning_rate": 4.490008869836884e-06, "loss": 0.2707, "step": 2440 }, { "epoch": 0.46, "grad_norm": 8.4375, "learning_rate": 4.489545157184898e-06, "loss": 0.3627, "step": 2441 }, { "epoch": 0.46, "grad_norm": 8.4375, "learning_rate": 4.4890812577802e-06, "loss": 0.3075, "step": 2442 }, { "epoch": 0.46, "grad_norm": 7.375, "learning_rate": 4.488617171666336e-06, "loss": 0.2427, "step": 2443 }, { "epoch": 0.46, "grad_norm": 15.0625, "learning_rate": 4.488152898886866e-06, "loss": 0.3143, "step": 2444 }, { "epoch": 0.46, "grad_norm": 11.875, "learning_rate": 4.487688439485372e-06, "loss": 0.2548, "step": 2445 }, { "epoch": 0.46, "grad_norm": 11.6875, "learning_rate": 4.487223793505452e-06, "loss": 0.3084, "step": 2446 }, { "epoch": 0.46, "grad_norm": 11.625, "learning_rate": 4.4867589609907196e-06, "loss": 0.2465, "step": 2447 }, { "epoch": 0.46, "grad_norm": 8.875, "learning_rate": 4.486293941984806e-06, "loss": 0.3044, "step": 2448 }, { "epoch": 0.46, "grad_norm": 14.3125, "learning_rate": 4.485828736531364e-06, "loss": 0.3344, "step": 2449 }, { "epoch": 0.46, "grad_norm": 9.9375, "learning_rate": 4.485363344674059e-06, "loss": 0.2815, "step": 2450 }, { "epoch": 0.46, "grad_norm": 10.0625, "learning_rate": 4.484897766456576e-06, "loss": 0.2764, "step": 2451 }, { "epoch": 0.46, "grad_norm": 13.25, "learning_rate": 4.48443200192262e-06, "loss": 0.3268, "step": 2452 }, { "epoch": 0.46, "grad_norm": 10.6875, "learning_rate": 4.4839660511159075e-06, "loss": 0.3309, "step": 2453 }, { "epoch": 0.46, "grad_norm": 12.0, "learning_rate": 4.4834999140801774e-06, "loss": 0.2867, "step": 2454 }, { "epoch": 0.46, "grad_norm": 9.0625, "learning_rate": 4.4830335908591855e-06, "loss": 0.3063, "step": 2455 }, { "epoch": 0.46, "grad_norm": 9.5625, "learning_rate": 4.482567081496702e-06, "loss": 0.2844, "step": 2456 }, { "epoch": 0.46, "grad_norm": 9.75, "learning_rate": 4.48210038603652e-06, "loss": 0.2464, "step": 2457 }, { "epoch": 0.46, "grad_norm": 13.9375, "learning_rate": 4.481633504522443e-06, "loss": 0.2616, "step": 2458 }, { "epoch": 0.47, "grad_norm": 11.3125, "learning_rate": 4.481166436998299e-06, "loss": 0.2687, "step": 2459 }, { "epoch": 0.47, "grad_norm": 8.5625, "learning_rate": 4.480699183507929e-06, "loss": 0.2711, "step": 2460 }, { "epoch": 0.47, "grad_norm": 10.125, "learning_rate": 4.480231744095191e-06, "loss": 0.3309, "step": 2461 }, { "epoch": 0.47, "grad_norm": 8.6875, "learning_rate": 4.479764118803965e-06, "loss": 0.2019, "step": 2462 }, { "epoch": 0.47, "grad_norm": 10.25, "learning_rate": 4.479296307678145e-06, "loss": 0.3107, "step": 2463 }, { "epoch": 0.47, "grad_norm": 9.4375, "learning_rate": 4.478828310761642e-06, "loss": 0.2777, "step": 2464 }, { "epoch": 0.47, "grad_norm": 10.4375, "learning_rate": 4.478360128098385e-06, "loss": 0.3117, "step": 2465 }, { "epoch": 0.47, "grad_norm": 10.25, "learning_rate": 4.477891759732322e-06, "loss": 0.3408, "step": 2466 }, { "epoch": 0.47, "grad_norm": 12.4375, "learning_rate": 4.477423205707417e-06, "loss": 0.274, "step": 2467 }, { "epoch": 0.47, "grad_norm": 12.0, "learning_rate": 4.476954466067654e-06, "loss": 0.355, "step": 2468 }, { "epoch": 0.47, "grad_norm": 19.75, "learning_rate": 4.476485540857028e-06, "loss": 0.2601, "step": 2469 }, { "epoch": 0.47, "grad_norm": 15.375, "learning_rate": 4.476016430119558e-06, "loss": 0.3347, "step": 2470 }, { "epoch": 0.47, "grad_norm": 10.875, "learning_rate": 4.475547133899279e-06, "loss": 0.3064, "step": 2471 }, { "epoch": 0.47, "grad_norm": 12.75, "learning_rate": 4.475077652240241e-06, "loss": 0.3169, "step": 2472 }, { "epoch": 0.47, "grad_norm": 8.75, "learning_rate": 4.474607985186512e-06, "loss": 0.2611, "step": 2473 }, { "epoch": 0.47, "grad_norm": 10.1875, "learning_rate": 4.474138132782181e-06, "loss": 0.299, "step": 2474 }, { "epoch": 0.47, "grad_norm": 12.5625, "learning_rate": 4.473668095071348e-06, "loss": 0.3729, "step": 2475 }, { "epoch": 0.47, "grad_norm": 10.0625, "learning_rate": 4.473197872098137e-06, "loss": 0.2328, "step": 2476 }, { "epoch": 0.47, "grad_norm": 8.5625, "learning_rate": 4.472727463906687e-06, "loss": 0.3351, "step": 2477 }, { "epoch": 0.47, "grad_norm": 9.625, "learning_rate": 4.47225687054115e-06, "loss": 0.2221, "step": 2478 }, { "epoch": 0.47, "grad_norm": 10.3125, "learning_rate": 4.4717860920457026e-06, "loss": 0.3712, "step": 2479 }, { "epoch": 0.47, "grad_norm": 9.125, "learning_rate": 4.4713151284645346e-06, "loss": 0.3369, "step": 2480 }, { "epoch": 0.47, "grad_norm": 11.75, "learning_rate": 4.470843979841853e-06, "loss": 0.2648, "step": 2481 }, { "epoch": 0.47, "grad_norm": 9.625, "learning_rate": 4.470372646221884e-06, "loss": 0.3084, "step": 2482 }, { "epoch": 0.47, "grad_norm": 13.375, "learning_rate": 4.469901127648871e-06, "loss": 0.4029, "step": 2483 }, { "epoch": 0.47, "grad_norm": 9.3125, "learning_rate": 4.469429424167072e-06, "loss": 0.3147, "step": 2484 }, { "epoch": 0.47, "grad_norm": 11.375, "learning_rate": 4.468957535820766e-06, "loss": 0.3928, "step": 2485 }, { "epoch": 0.47, "grad_norm": 8.125, "learning_rate": 4.468485462654247e-06, "loss": 0.2661, "step": 2486 }, { "epoch": 0.47, "grad_norm": 8.875, "learning_rate": 4.468013204711828e-06, "loss": 0.3162, "step": 2487 }, { "epoch": 0.47, "grad_norm": 10.9375, "learning_rate": 4.467540762037837e-06, "loss": 0.3624, "step": 2488 }, { "epoch": 0.47, "grad_norm": 8.375, "learning_rate": 4.467068134676623e-06, "loss": 0.2911, "step": 2489 }, { "epoch": 0.47, "grad_norm": 16.0, "learning_rate": 4.466595322672548e-06, "loss": 0.4002, "step": 2490 }, { "epoch": 0.47, "grad_norm": 8.875, "learning_rate": 4.466122326069995e-06, "loss": 0.305, "step": 2491 }, { "epoch": 0.47, "grad_norm": 14.4375, "learning_rate": 4.465649144913361e-06, "loss": 0.3376, "step": 2492 }, { "epoch": 0.47, "grad_norm": 8.5625, "learning_rate": 4.465175779247064e-06, "loss": 0.2512, "step": 2493 }, { "epoch": 0.47, "grad_norm": 9.875, "learning_rate": 4.464702229115536e-06, "loss": 0.3209, "step": 2494 }, { "epoch": 0.47, "grad_norm": 11.0, "learning_rate": 4.464228494563228e-06, "loss": 0.3197, "step": 2495 }, { "epoch": 0.47, "grad_norm": 16.375, "learning_rate": 4.46375457563461e-06, "loss": 0.3933, "step": 2496 }, { "epoch": 0.47, "grad_norm": 9.25, "learning_rate": 4.4632804723741645e-06, "loss": 0.293, "step": 2497 }, { "epoch": 0.47, "grad_norm": 10.875, "learning_rate": 4.462806184826396e-06, "loss": 0.3436, "step": 2498 }, { "epoch": 0.47, "grad_norm": 8.8125, "learning_rate": 4.462331713035823e-06, "loss": 0.3042, "step": 2499 }, { "epoch": 0.47, "grad_norm": 11.8125, "learning_rate": 4.461857057046985e-06, "loss": 0.3266, "step": 2500 }, { "epoch": 0.47, "grad_norm": 7.71875, "learning_rate": 4.461382216904434e-06, "loss": 0.2223, "step": 2501 }, { "epoch": 0.47, "grad_norm": 13.5625, "learning_rate": 4.4609071926527445e-06, "loss": 0.3314, "step": 2502 }, { "epoch": 0.47, "grad_norm": 11.25, "learning_rate": 4.460431984336503e-06, "loss": 0.3009, "step": 2503 }, { "epoch": 0.47, "grad_norm": 27.125, "learning_rate": 4.459956592000317e-06, "loss": 0.2319, "step": 2504 }, { "epoch": 0.47, "grad_norm": 13.9375, "learning_rate": 4.459481015688811e-06, "loss": 0.2826, "step": 2505 }, { "epoch": 0.47, "grad_norm": 10.5, "learning_rate": 4.459005255446625e-06, "loss": 0.3346, "step": 2506 }, { "epoch": 0.47, "grad_norm": 9.0, "learning_rate": 4.458529311318417e-06, "loss": 0.237, "step": 2507 }, { "epoch": 0.47, "grad_norm": 11.0, "learning_rate": 4.458053183348864e-06, "loss": 0.3318, "step": 2508 }, { "epoch": 0.47, "grad_norm": 9.3125, "learning_rate": 4.457576871582656e-06, "loss": 0.3782, "step": 2509 }, { "epoch": 0.47, "grad_norm": 13.1875, "learning_rate": 4.457100376064506e-06, "loss": 0.2171, "step": 2510 }, { "epoch": 0.48, "grad_norm": 11.625, "learning_rate": 4.456623696839139e-06, "loss": 0.2788, "step": 2511 }, { "epoch": 0.48, "grad_norm": 8.5, "learning_rate": 4.4561468339513005e-06, "loss": 0.2551, "step": 2512 }, { "epoch": 0.48, "grad_norm": 10.125, "learning_rate": 4.455669787445752e-06, "loss": 0.3243, "step": 2513 }, { "epoch": 0.48, "grad_norm": 7.0625, "learning_rate": 4.455192557367272e-06, "loss": 0.2212, "step": 2514 }, { "epoch": 0.48, "grad_norm": 11.8125, "learning_rate": 4.454715143760657e-06, "loss": 0.4141, "step": 2515 }, { "epoch": 0.48, "grad_norm": 8.6875, "learning_rate": 4.4542375466707214e-06, "loss": 0.2845, "step": 2516 }, { "epoch": 0.48, "grad_norm": 10.5625, "learning_rate": 4.453759766142294e-06, "loss": 0.4211, "step": 2517 }, { "epoch": 0.48, "grad_norm": 10.9375, "learning_rate": 4.453281802220224e-06, "loss": 0.2707, "step": 2518 }, { "epoch": 0.48, "grad_norm": 13.3125, "learning_rate": 4.452803654949376e-06, "loss": 0.2531, "step": 2519 }, { "epoch": 0.48, "grad_norm": 12.25, "learning_rate": 4.452325324374632e-06, "loss": 0.3341, "step": 2520 }, { "epoch": 0.48, "grad_norm": 11.625, "learning_rate": 4.451846810540892e-06, "loss": 0.2952, "step": 2521 }, { "epoch": 0.48, "grad_norm": 11.3125, "learning_rate": 4.451368113493073e-06, "loss": 0.2943, "step": 2522 }, { "epoch": 0.48, "grad_norm": 12.375, "learning_rate": 4.450889233276108e-06, "loss": 0.2908, "step": 2523 }, { "epoch": 0.48, "grad_norm": 12.625, "learning_rate": 4.450410169934948e-06, "loss": 0.3789, "step": 2524 }, { "epoch": 0.48, "grad_norm": 13.375, "learning_rate": 4.4499309235145615e-06, "loss": 0.3616, "step": 2525 }, { "epoch": 0.48, "grad_norm": 11.0625, "learning_rate": 4.449451494059934e-06, "loss": 0.3032, "step": 2526 }, { "epoch": 0.48, "grad_norm": 12.0625, "learning_rate": 4.448971881616069e-06, "loss": 0.2568, "step": 2527 }, { "epoch": 0.48, "grad_norm": 11.25, "learning_rate": 4.448492086227984e-06, "loss": 0.3361, "step": 2528 }, { "epoch": 0.48, "grad_norm": 10.125, "learning_rate": 4.4480121079407194e-06, "loss": 0.26, "step": 2529 }, { "epoch": 0.48, "grad_norm": 9.625, "learning_rate": 4.447531946799326e-06, "loss": 0.2923, "step": 2530 }, { "epoch": 0.48, "grad_norm": 12.25, "learning_rate": 4.447051602848877e-06, "loss": 0.3253, "step": 2531 }, { "epoch": 0.48, "grad_norm": 11.8125, "learning_rate": 4.44657107613446e-06, "loss": 0.348, "step": 2532 }, { "epoch": 0.48, "grad_norm": 13.75, "learning_rate": 4.446090366701181e-06, "loss": 0.4045, "step": 2533 }, { "epoch": 0.48, "grad_norm": 11.3125, "learning_rate": 4.445609474594164e-06, "loss": 0.2782, "step": 2534 }, { "epoch": 0.48, "grad_norm": 12.8125, "learning_rate": 4.445128399858546e-06, "loss": 0.2708, "step": 2535 }, { "epoch": 0.48, "grad_norm": 14.8125, "learning_rate": 4.444647142539486e-06, "loss": 0.3227, "step": 2536 }, { "epoch": 0.48, "grad_norm": 9.9375, "learning_rate": 4.44416570268216e-06, "loss": 0.2369, "step": 2537 }, { "epoch": 0.48, "grad_norm": 10.4375, "learning_rate": 4.443684080331755e-06, "loss": 0.2595, "step": 2538 }, { "epoch": 0.48, "grad_norm": 8.6875, "learning_rate": 4.4432022755334825e-06, "loss": 0.3045, "step": 2539 }, { "epoch": 0.48, "grad_norm": 9.3125, "learning_rate": 4.442720288332567e-06, "loss": 0.2618, "step": 2540 }, { "epoch": 0.48, "grad_norm": 14.4375, "learning_rate": 4.442238118774252e-06, "loss": 0.3404, "step": 2541 }, { "epoch": 0.48, "grad_norm": 12.6875, "learning_rate": 4.441755766903797e-06, "loss": 0.3223, "step": 2542 }, { "epoch": 0.48, "grad_norm": 25.625, "learning_rate": 4.441273232766478e-06, "loss": 0.274, "step": 2543 }, { "epoch": 0.48, "grad_norm": 14.5625, "learning_rate": 4.44079051640759e-06, "loss": 0.4341, "step": 2544 }, { "epoch": 0.48, "grad_norm": 12.4375, "learning_rate": 4.4403076178724444e-06, "loss": 0.2977, "step": 2545 }, { "epoch": 0.48, "grad_norm": 10.0, "learning_rate": 4.43982453720637e-06, "loss": 0.2484, "step": 2546 }, { "epoch": 0.48, "grad_norm": 11.1875, "learning_rate": 4.43934127445471e-06, "loss": 0.2898, "step": 2547 }, { "epoch": 0.48, "grad_norm": 13.6875, "learning_rate": 4.438857829662828e-06, "loss": 0.3269, "step": 2548 }, { "epoch": 0.48, "grad_norm": 12.125, "learning_rate": 4.438374202876104e-06, "loss": 0.3123, "step": 2549 }, { "epoch": 0.48, "grad_norm": 13.6875, "learning_rate": 4.437890394139934e-06, "loss": 0.3814, "step": 2550 }, { "epoch": 0.48, "grad_norm": 21.25, "learning_rate": 4.437406403499732e-06, "loss": 0.3733, "step": 2551 }, { "epoch": 0.48, "grad_norm": 8.375, "learning_rate": 4.436922231000929e-06, "loss": 0.3195, "step": 2552 }, { "epoch": 0.48, "grad_norm": 11.375, "learning_rate": 4.436437876688972e-06, "loss": 0.2957, "step": 2553 }, { "epoch": 0.48, "grad_norm": 15.1875, "learning_rate": 4.435953340609327e-06, "loss": 0.2363, "step": 2554 }, { "epoch": 0.48, "grad_norm": 11.9375, "learning_rate": 4.435468622807475e-06, "loss": 0.3154, "step": 2555 }, { "epoch": 0.48, "grad_norm": 10.5625, "learning_rate": 4.434983723328915e-06, "loss": 0.3339, "step": 2556 }, { "epoch": 0.48, "grad_norm": 11.75, "learning_rate": 4.4344986422191634e-06, "loss": 0.312, "step": 2557 }, { "epoch": 0.48, "grad_norm": 10.1875, "learning_rate": 4.434013379523755e-06, "loss": 0.3333, "step": 2558 }, { "epoch": 0.48, "grad_norm": 13.5625, "learning_rate": 4.433527935288237e-06, "loss": 0.2867, "step": 2559 }, { "epoch": 0.48, "grad_norm": 16.625, "learning_rate": 4.433042309558178e-06, "loss": 0.3173, "step": 2560 }, { "epoch": 0.48, "grad_norm": 18.25, "learning_rate": 4.432556502379163e-06, "loss": 0.2911, "step": 2561 }, { "epoch": 0.48, "grad_norm": 13.5625, "learning_rate": 4.432070513796792e-06, "loss": 0.2743, "step": 2562 }, { "epoch": 0.48, "grad_norm": 14.4375, "learning_rate": 4.431584343856684e-06, "loss": 0.2898, "step": 2563 }, { "epoch": 0.49, "grad_norm": 10.75, "learning_rate": 4.431097992604474e-06, "loss": 0.2682, "step": 2564 }, { "epoch": 0.49, "grad_norm": 15.3125, "learning_rate": 4.430611460085815e-06, "loss": 0.354, "step": 2565 }, { "epoch": 0.49, "grad_norm": 9.6875, "learning_rate": 4.4301247463463746e-06, "loss": 0.3147, "step": 2566 }, { "epoch": 0.49, "grad_norm": 10.1875, "learning_rate": 4.429637851431842e-06, "loss": 0.3496, "step": 2567 }, { "epoch": 0.49, "grad_norm": 10.375, "learning_rate": 4.429150775387918e-06, "loss": 0.2869, "step": 2568 }, { "epoch": 0.49, "grad_norm": 11.375, "learning_rate": 4.428663518260324e-06, "loss": 0.3143, "step": 2569 }, { "epoch": 0.49, "grad_norm": 8.6875, "learning_rate": 4.428176080094798e-06, "loss": 0.285, "step": 2570 }, { "epoch": 0.49, "grad_norm": 10.875, "learning_rate": 4.427688460937094e-06, "loss": 0.3823, "step": 2571 }, { "epoch": 0.49, "grad_norm": 7.78125, "learning_rate": 4.427200660832982e-06, "loss": 0.274, "step": 2572 }, { "epoch": 0.49, "grad_norm": 15.875, "learning_rate": 4.426712679828254e-06, "loss": 0.3401, "step": 2573 }, { "epoch": 0.49, "grad_norm": 13.0, "learning_rate": 4.42622451796871e-06, "loss": 0.2766, "step": 2574 }, { "epoch": 0.49, "grad_norm": 9.5, "learning_rate": 4.425736175300177e-06, "loss": 0.3047, "step": 2575 }, { "epoch": 0.49, "grad_norm": 10.8125, "learning_rate": 4.425247651868491e-06, "loss": 0.2697, "step": 2576 }, { "epoch": 0.49, "grad_norm": 11.0, "learning_rate": 4.424758947719511e-06, "loss": 0.2269, "step": 2577 }, { "epoch": 0.49, "grad_norm": 14.375, "learning_rate": 4.424270062899108e-06, "loss": 0.31, "step": 2578 }, { "epoch": 0.49, "grad_norm": 10.6875, "learning_rate": 4.423780997453173e-06, "loss": 0.324, "step": 2579 }, { "epoch": 0.49, "grad_norm": 10.75, "learning_rate": 4.423291751427614e-06, "loss": 0.3004, "step": 2580 }, { "epoch": 0.49, "grad_norm": 10.75, "learning_rate": 4.422802324868353e-06, "loss": 0.2903, "step": 2581 }, { "epoch": 0.49, "grad_norm": 33.75, "learning_rate": 4.422312717821333e-06, "loss": 0.2955, "step": 2582 }, { "epoch": 0.49, "grad_norm": 17.875, "learning_rate": 4.421822930332512e-06, "loss": 0.2819, "step": 2583 }, { "epoch": 0.49, "grad_norm": 10.1875, "learning_rate": 4.421332962447864e-06, "loss": 0.3104, "step": 2584 }, { "epoch": 0.49, "grad_norm": 10.75, "learning_rate": 4.42084281421338e-06, "loss": 0.3041, "step": 2585 }, { "epoch": 0.49, "grad_norm": 21.875, "learning_rate": 4.42035248567507e-06, "loss": 0.2899, "step": 2586 }, { "epoch": 0.49, "grad_norm": 14.0625, "learning_rate": 4.419861976878959e-06, "loss": 0.2316, "step": 2587 }, { "epoch": 0.49, "grad_norm": 9.1875, "learning_rate": 4.419371287871091e-06, "loss": 0.3384, "step": 2588 }, { "epoch": 0.49, "grad_norm": 13.4375, "learning_rate": 4.418880418697525e-06, "loss": 0.2851, "step": 2589 }, { "epoch": 0.49, "grad_norm": 17.875, "learning_rate": 4.418389369404335e-06, "loss": 0.3377, "step": 2590 }, { "epoch": 0.49, "grad_norm": 10.875, "learning_rate": 4.4178981400376185e-06, "loss": 0.313, "step": 2591 }, { "epoch": 0.49, "grad_norm": 9.0625, "learning_rate": 4.417406730643482e-06, "loss": 0.2734, "step": 2592 }, { "epoch": 0.49, "grad_norm": 8.6875, "learning_rate": 4.416915141268055e-06, "loss": 0.2708, "step": 2593 }, { "epoch": 0.49, "grad_norm": 8.1875, "learning_rate": 4.416423371957481e-06, "loss": 0.3232, "step": 2594 }, { "epoch": 0.49, "grad_norm": 11.375, "learning_rate": 4.415931422757921e-06, "loss": 0.266, "step": 2595 }, { "epoch": 0.49, "grad_norm": 7.78125, "learning_rate": 4.415439293715551e-06, "loss": 0.2619, "step": 2596 }, { "epoch": 0.49, "grad_norm": 10.5, "learning_rate": 4.414946984876568e-06, "loss": 0.3054, "step": 2597 }, { "epoch": 0.49, "grad_norm": 7.875, "learning_rate": 4.414454496287183e-06, "loss": 0.2426, "step": 2598 }, { "epoch": 0.49, "grad_norm": 10.875, "learning_rate": 4.413961827993625e-06, "loss": 0.3196, "step": 2599 }, { "epoch": 0.49, "grad_norm": 13.6875, "learning_rate": 4.413468980042137e-06, "loss": 0.3139, "step": 2600 }, { "epoch": 0.49, "grad_norm": 8.875, "learning_rate": 4.4129759524789835e-06, "loss": 0.3481, "step": 2601 }, { "epoch": 0.49, "grad_norm": 10.4375, "learning_rate": 4.412482745350443e-06, "loss": 0.2981, "step": 2602 }, { "epoch": 0.49, "grad_norm": 10.75, "learning_rate": 4.41198935870281e-06, "loss": 0.2549, "step": 2603 }, { "epoch": 0.49, "grad_norm": 12.4375, "learning_rate": 4.4114957925823995e-06, "loss": 0.2725, "step": 2604 }, { "epoch": 0.49, "grad_norm": 11.0625, "learning_rate": 4.41100204703554e-06, "loss": 0.3172, "step": 2605 }, { "epoch": 0.49, "grad_norm": 10.8125, "learning_rate": 4.410508122108577e-06, "loss": 0.3049, "step": 2606 }, { "epoch": 0.49, "grad_norm": 13.0, "learning_rate": 4.410014017847875e-06, "loss": 0.3187, "step": 2607 }, { "epoch": 0.49, "grad_norm": 11.3125, "learning_rate": 4.409519734299814e-06, "loss": 0.1759, "step": 2608 }, { "epoch": 0.49, "grad_norm": 13.0, "learning_rate": 4.409025271510791e-06, "loss": 0.3007, "step": 2609 }, { "epoch": 0.49, "grad_norm": 15.0, "learning_rate": 4.40853062952722e-06, "loss": 0.2698, "step": 2610 }, { "epoch": 0.49, "grad_norm": 15.6875, "learning_rate": 4.40803580839553e-06, "loss": 0.2946, "step": 2611 }, { "epoch": 0.49, "grad_norm": 11.4375, "learning_rate": 4.40754080816217e-06, "loss": 0.3112, "step": 2612 }, { "epoch": 0.49, "grad_norm": 9.6875, "learning_rate": 4.407045628873603e-06, "loss": 0.3732, "step": 2613 }, { "epoch": 0.49, "grad_norm": 9.1875, "learning_rate": 4.406550270576311e-06, "loss": 0.2542, "step": 2614 }, { "epoch": 0.49, "grad_norm": 11.75, "learning_rate": 4.406054733316791e-06, "loss": 0.2606, "step": 2615 }, { "epoch": 0.49, "grad_norm": 12.9375, "learning_rate": 4.405559017141559e-06, "loss": 0.3267, "step": 2616 }, { "epoch": 0.5, "grad_norm": 11.0625, "learning_rate": 4.405063122097145e-06, "loss": 0.3538, "step": 2617 }, { "epoch": 0.5, "grad_norm": 8.75, "learning_rate": 4.404567048230098e-06, "loss": 0.2794, "step": 2618 }, { "epoch": 0.5, "grad_norm": 12.375, "learning_rate": 4.404070795586983e-06, "loss": 0.2668, "step": 2619 }, { "epoch": 0.5, "grad_norm": 13.25, "learning_rate": 4.403574364214381e-06, "loss": 0.4047, "step": 2620 }, { "epoch": 0.5, "grad_norm": 17.375, "learning_rate": 4.403077754158891e-06, "loss": 0.289, "step": 2621 }, { "epoch": 0.5, "grad_norm": 9.375, "learning_rate": 4.402580965467128e-06, "loss": 0.2972, "step": 2622 }, { "epoch": 0.5, "grad_norm": 14.5625, "learning_rate": 4.402083998185725e-06, "loss": 0.4264, "step": 2623 }, { "epoch": 0.5, "grad_norm": 8.9375, "learning_rate": 4.4015868523613295e-06, "loss": 0.3005, "step": 2624 }, { "epoch": 0.5, "grad_norm": 10.5, "learning_rate": 4.401089528040609e-06, "loss": 0.2426, "step": 2625 }, { "epoch": 0.5, "grad_norm": 10.0625, "learning_rate": 4.400592025270244e-06, "loss": 0.2762, "step": 2626 }, { "epoch": 0.5, "grad_norm": 18.625, "learning_rate": 4.400094344096935e-06, "loss": 0.3308, "step": 2627 }, { "epoch": 0.5, "grad_norm": 11.25, "learning_rate": 4.399596484567396e-06, "loss": 0.2622, "step": 2628 }, { "epoch": 0.5, "grad_norm": 17.0, "learning_rate": 4.3990984467283615e-06, "loss": 0.2894, "step": 2629 }, { "epoch": 0.5, "grad_norm": 8.375, "learning_rate": 4.39860023062658e-06, "loss": 0.2343, "step": 2630 }, { "epoch": 0.5, "grad_norm": 10.75, "learning_rate": 4.398101836308817e-06, "loss": 0.2157, "step": 2631 }, { "epoch": 0.5, "grad_norm": 11.8125, "learning_rate": 4.397603263821857e-06, "loss": 0.3359, "step": 2632 }, { "epoch": 0.5, "grad_norm": 9.3125, "learning_rate": 4.397104513212498e-06, "loss": 0.2681, "step": 2633 }, { "epoch": 0.5, "grad_norm": 10.1875, "learning_rate": 4.396605584527557e-06, "loss": 0.3683, "step": 2634 }, { "epoch": 0.5, "grad_norm": 10.0625, "learning_rate": 4.396106477813866e-06, "loss": 0.2965, "step": 2635 }, { "epoch": 0.5, "grad_norm": 12.5, "learning_rate": 4.395607193118276e-06, "loss": 0.3235, "step": 2636 }, { "epoch": 0.5, "grad_norm": 9.3125, "learning_rate": 4.395107730487653e-06, "loss": 0.2645, "step": 2637 }, { "epoch": 0.5, "grad_norm": 8.0, "learning_rate": 4.394608089968881e-06, "loss": 0.273, "step": 2638 }, { "epoch": 0.5, "grad_norm": 11.375, "learning_rate": 4.394108271608857e-06, "loss": 0.2878, "step": 2639 }, { "epoch": 0.5, "grad_norm": 16.25, "learning_rate": 4.393608275454499e-06, "loss": 0.2454, "step": 2640 }, { "epoch": 0.5, "grad_norm": 9.8125, "learning_rate": 4.39310810155274e-06, "loss": 0.3061, "step": 2641 }, { "epoch": 0.5, "grad_norm": 6.9375, "learning_rate": 4.392607749950531e-06, "loss": 0.2916, "step": 2642 }, { "epoch": 0.5, "grad_norm": 9.5, "learning_rate": 4.392107220694837e-06, "loss": 0.3149, "step": 2643 }, { "epoch": 0.5, "grad_norm": 28.0, "learning_rate": 4.3916065138326425e-06, "loss": 0.2628, "step": 2644 }, { "epoch": 0.5, "grad_norm": 15.5, "learning_rate": 4.391105629410946e-06, "loss": 0.3425, "step": 2645 }, { "epoch": 0.5, "grad_norm": 10.5625, "learning_rate": 4.390604567476765e-06, "loss": 0.2828, "step": 2646 }, { "epoch": 0.5, "grad_norm": 11.6875, "learning_rate": 4.390103328077132e-06, "loss": 0.2888, "step": 2647 }, { "epoch": 0.5, "grad_norm": 15.125, "learning_rate": 4.389601911259098e-06, "loss": 0.2806, "step": 2648 }, { "epoch": 0.5, "grad_norm": 37.25, "learning_rate": 4.389100317069728e-06, "loss": 0.2624, "step": 2649 }, { "epoch": 0.5, "grad_norm": 14.8125, "learning_rate": 4.388598545556107e-06, "loss": 0.3013, "step": 2650 }, { "epoch": 0.5, "grad_norm": 10.3125, "learning_rate": 4.388096596765333e-06, "loss": 0.2969, "step": 2651 }, { "epoch": 0.5, "grad_norm": 8.375, "learning_rate": 4.387594470744523e-06, "loss": 0.3206, "step": 2652 }, { "epoch": 0.5, "grad_norm": 21.25, "learning_rate": 4.387092167540811e-06, "loss": 0.3663, "step": 2653 }, { "epoch": 0.5, "grad_norm": 12.6875, "learning_rate": 4.386589687201345e-06, "loss": 0.3721, "step": 2654 }, { "epoch": 0.5, "grad_norm": 14.5, "learning_rate": 4.386087029773293e-06, "loss": 0.2868, "step": 2655 }, { "epoch": 0.5, "grad_norm": 10.375, "learning_rate": 4.385584195303838e-06, "loss": 0.276, "step": 2656 }, { "epoch": 0.5, "grad_norm": 11.0, "learning_rate": 4.385081183840178e-06, "loss": 0.3376, "step": 2657 }, { "epoch": 0.5, "grad_norm": 15.4375, "learning_rate": 4.38457799542953e-06, "loss": 0.3205, "step": 2658 }, { "epoch": 0.5, "grad_norm": 12.1875, "learning_rate": 4.3840746301191265e-06, "loss": 0.2874, "step": 2659 }, { "epoch": 0.5, "grad_norm": 12.5, "learning_rate": 4.383571087956218e-06, "loss": 0.3293, "step": 2660 }, { "epoch": 0.5, "grad_norm": 8.875, "learning_rate": 4.38306736898807e-06, "loss": 0.312, "step": 2661 }, { "epoch": 0.5, "grad_norm": 8.3125, "learning_rate": 4.382563473261964e-06, "loss": 0.3725, "step": 2662 }, { "epoch": 0.5, "grad_norm": 12.3125, "learning_rate": 4.382059400825201e-06, "loss": 0.3271, "step": 2663 }, { "epoch": 0.5, "grad_norm": 27.75, "learning_rate": 4.381555151725095e-06, "loss": 0.3118, "step": 2664 }, { "epoch": 0.5, "grad_norm": 9.9375, "learning_rate": 4.38105072600898e-06, "loss": 0.2961, "step": 2665 }, { "epoch": 0.5, "grad_norm": 11.625, "learning_rate": 4.380546123724204e-06, "loss": 0.2919, "step": 2666 }, { "epoch": 0.5, "grad_norm": 8.6875, "learning_rate": 4.380041344918133e-06, "loss": 0.239, "step": 2667 }, { "epoch": 0.5, "grad_norm": 18.875, "learning_rate": 4.379536389638149e-06, "loss": 0.2335, "step": 2668 }, { "epoch": 0.5, "grad_norm": 11.1875, "learning_rate": 4.37903125793165e-06, "loss": 0.2358, "step": 2669 }, { "epoch": 0.51, "grad_norm": 12.375, "learning_rate": 4.378525949846052e-06, "loss": 0.272, "step": 2670 }, { "epoch": 0.51, "grad_norm": 8.0625, "learning_rate": 4.378020465428786e-06, "loss": 0.2255, "step": 2671 }, { "epoch": 0.51, "grad_norm": 9.125, "learning_rate": 4.3775148047273005e-06, "loss": 0.2548, "step": 2672 }, { "epoch": 0.51, "grad_norm": 12.375, "learning_rate": 4.377008967789062e-06, "loss": 0.344, "step": 2673 }, { "epoch": 0.51, "grad_norm": 10.5625, "learning_rate": 4.3765029546615495e-06, "loss": 0.3294, "step": 2674 }, { "epoch": 0.51, "grad_norm": 9.8125, "learning_rate": 4.375996765392263e-06, "loss": 0.3207, "step": 2675 }, { "epoch": 0.51, "grad_norm": 9.5, "learning_rate": 4.375490400028715e-06, "loss": 0.291, "step": 2676 }, { "epoch": 0.51, "grad_norm": 9.5, "learning_rate": 4.3749838586184375e-06, "loss": 0.2395, "step": 2677 }, { "epoch": 0.51, "grad_norm": 8.375, "learning_rate": 4.374477141208979e-06, "loss": 0.3068, "step": 2678 }, { "epoch": 0.51, "grad_norm": 7.8125, "learning_rate": 4.373970247847902e-06, "loss": 0.2463, "step": 2679 }, { "epoch": 0.51, "grad_norm": 10.5625, "learning_rate": 4.373463178582787e-06, "loss": 0.3414, "step": 2680 }, { "epoch": 0.51, "grad_norm": 6.65625, "learning_rate": 4.372955933461234e-06, "loss": 0.2848, "step": 2681 }, { "epoch": 0.51, "grad_norm": 14.5, "learning_rate": 4.372448512530852e-06, "loss": 0.3264, "step": 2682 }, { "epoch": 0.51, "grad_norm": 16.5, "learning_rate": 4.371940915839276e-06, "loss": 0.3144, "step": 2683 }, { "epoch": 0.51, "grad_norm": 7.75, "learning_rate": 4.371433143434148e-06, "loss": 0.2081, "step": 2684 }, { "epoch": 0.51, "grad_norm": 13.9375, "learning_rate": 4.370925195363134e-06, "loss": 0.3718, "step": 2685 }, { "epoch": 0.51, "grad_norm": 12.625, "learning_rate": 4.370417071673913e-06, "loss": 0.2497, "step": 2686 }, { "epoch": 0.51, "grad_norm": 7.5625, "learning_rate": 4.369908772414181e-06, "loss": 0.3114, "step": 2687 }, { "epoch": 0.51, "grad_norm": 9.625, "learning_rate": 4.36940029763165e-06, "loss": 0.3094, "step": 2688 }, { "epoch": 0.51, "grad_norm": 17.375, "learning_rate": 4.368891647374049e-06, "loss": 0.3867, "step": 2689 }, { "epoch": 0.51, "grad_norm": 11.0625, "learning_rate": 4.3683828216891255e-06, "loss": 0.3512, "step": 2690 }, { "epoch": 0.51, "grad_norm": 12.3125, "learning_rate": 4.367873820624638e-06, "loss": 0.3222, "step": 2691 }, { "epoch": 0.51, "grad_norm": 12.0, "learning_rate": 4.367364644228368e-06, "loss": 0.2776, "step": 2692 }, { "epoch": 0.51, "grad_norm": 11.0625, "learning_rate": 4.366855292548109e-06, "loss": 0.3098, "step": 2693 }, { "epoch": 0.51, "grad_norm": 8.25, "learning_rate": 4.366345765631672e-06, "loss": 0.2722, "step": 2694 }, { "epoch": 0.51, "grad_norm": 10.6875, "learning_rate": 4.365836063526886e-06, "loss": 0.3539, "step": 2695 }, { "epoch": 0.51, "grad_norm": 8.8125, "learning_rate": 4.365326186281594e-06, "loss": 0.2806, "step": 2696 }, { "epoch": 0.51, "grad_norm": 11.25, "learning_rate": 4.364816133943657e-06, "loss": 0.2959, "step": 2697 }, { "epoch": 0.51, "grad_norm": 12.75, "learning_rate": 4.3643059065609526e-06, "loss": 0.3016, "step": 2698 }, { "epoch": 0.51, "grad_norm": 8.25, "learning_rate": 4.363795504181374e-06, "loss": 0.3151, "step": 2699 }, { "epoch": 0.51, "grad_norm": 6.28125, "learning_rate": 4.363284926852831e-06, "loss": 0.2138, "step": 2700 }, { "epoch": 0.51, "grad_norm": 8.6875, "learning_rate": 4.362774174623251e-06, "loss": 0.2858, "step": 2701 }, { "epoch": 0.51, "grad_norm": 14.9375, "learning_rate": 4.362263247540575e-06, "loss": 0.2996, "step": 2702 }, { "epoch": 0.51, "grad_norm": 14.875, "learning_rate": 4.361752145652763e-06, "loss": 0.281, "step": 2703 }, { "epoch": 0.51, "grad_norm": 11.0, "learning_rate": 4.361240869007792e-06, "loss": 0.2686, "step": 2704 }, { "epoch": 0.51, "grad_norm": 10.0, "learning_rate": 4.360729417653652e-06, "loss": 0.3422, "step": 2705 }, { "epoch": 0.51, "grad_norm": 8.6875, "learning_rate": 4.360217791638353e-06, "loss": 0.2436, "step": 2706 }, { "epoch": 0.51, "grad_norm": 9.1875, "learning_rate": 4.359705991009918e-06, "loss": 0.2554, "step": 2707 }, { "epoch": 0.51, "grad_norm": 9.75, "learning_rate": 4.3591940158163895e-06, "loss": 0.275, "step": 2708 }, { "epoch": 0.51, "grad_norm": 12.5, "learning_rate": 4.358681866105825e-06, "loss": 0.233, "step": 2709 }, { "epoch": 0.51, "grad_norm": 12.5625, "learning_rate": 4.3581695419262996e-06, "loss": 0.2966, "step": 2710 }, { "epoch": 0.51, "grad_norm": 15.9375, "learning_rate": 4.3576570433259e-06, "loss": 0.3392, "step": 2711 }, { "epoch": 0.51, "grad_norm": 10.4375, "learning_rate": 4.3571443703527375e-06, "loss": 0.3191, "step": 2712 }, { "epoch": 0.51, "grad_norm": 11.0, "learning_rate": 4.356631523054932e-06, "loss": 0.2597, "step": 2713 }, { "epoch": 0.51, "grad_norm": 13.3125, "learning_rate": 4.356118501480624e-06, "loss": 0.3449, "step": 2714 }, { "epoch": 0.51, "grad_norm": 7.46875, "learning_rate": 4.355605305677969e-06, "loss": 0.2304, "step": 2715 }, { "epoch": 0.51, "grad_norm": 12.25, "learning_rate": 4.3550919356951406e-06, "loss": 0.2959, "step": 2716 }, { "epoch": 0.51, "grad_norm": 8.125, "learning_rate": 4.354578391580325e-06, "loss": 0.2265, "step": 2717 }, { "epoch": 0.51, "grad_norm": 12.4375, "learning_rate": 4.3540646733817286e-06, "loss": 0.285, "step": 2718 }, { "epoch": 0.51, "grad_norm": 10.125, "learning_rate": 4.353550781147573e-06, "loss": 0.2695, "step": 2719 }, { "epoch": 0.51, "grad_norm": 9.5625, "learning_rate": 4.353036714926094e-06, "loss": 0.3003, "step": 2720 }, { "epoch": 0.51, "grad_norm": 7.375, "learning_rate": 4.352522474765546e-06, "loss": 0.2315, "step": 2721 }, { "epoch": 0.51, "grad_norm": 11.6875, "learning_rate": 4.352008060714201e-06, "loss": 0.2571, "step": 2722 }, { "epoch": 0.52, "grad_norm": 12.9375, "learning_rate": 4.351493472820344e-06, "loss": 0.283, "step": 2723 }, { "epoch": 0.52, "grad_norm": 10.6875, "learning_rate": 4.350978711132278e-06, "loss": 0.278, "step": 2724 }, { "epoch": 0.52, "grad_norm": 9.6875, "learning_rate": 4.350463775698322e-06, "loss": 0.3274, "step": 2725 }, { "epoch": 0.52, "grad_norm": 10.0, "learning_rate": 4.349948666566811e-06, "loss": 0.3643, "step": 2726 }, { "epoch": 0.52, "grad_norm": 9.875, "learning_rate": 4.349433383786098e-06, "loss": 0.3956, "step": 2727 }, { "epoch": 0.52, "grad_norm": 14.0, "learning_rate": 4.34891792740455e-06, "loss": 0.3071, "step": 2728 }, { "epoch": 0.52, "grad_norm": 10.375, "learning_rate": 4.348402297470553e-06, "loss": 0.2541, "step": 2729 }, { "epoch": 0.52, "grad_norm": 10.5625, "learning_rate": 4.3478864940325056e-06, "loss": 0.3154, "step": 2730 }, { "epoch": 0.52, "grad_norm": 9.625, "learning_rate": 4.347370517138826e-06, "loss": 0.329, "step": 2731 }, { "epoch": 0.52, "grad_norm": 16.125, "learning_rate": 4.346854366837947e-06, "loss": 0.2221, "step": 2732 }, { "epoch": 0.52, "grad_norm": 14.0625, "learning_rate": 4.3463380431783185e-06, "loss": 0.2977, "step": 2733 }, { "epoch": 0.52, "grad_norm": 8.125, "learning_rate": 4.345821546208405e-06, "loss": 0.226, "step": 2734 }, { "epoch": 0.52, "grad_norm": 13.9375, "learning_rate": 4.3453048759766905e-06, "loss": 0.3668, "step": 2735 }, { "epoch": 0.52, "grad_norm": 13.5625, "learning_rate": 4.344788032531672e-06, "loss": 0.2996, "step": 2736 }, { "epoch": 0.52, "grad_norm": 36.75, "learning_rate": 4.344271015921864e-06, "loss": 0.2504, "step": 2737 }, { "epoch": 0.52, "grad_norm": 8.5625, "learning_rate": 4.343753826195798e-06, "loss": 0.2727, "step": 2738 }, { "epoch": 0.52, "grad_norm": 8.0625, "learning_rate": 4.3432364634020205e-06, "loss": 0.2918, "step": 2739 }, { "epoch": 0.52, "grad_norm": 11.125, "learning_rate": 4.342718927589096e-06, "loss": 0.3022, "step": 2740 }, { "epoch": 0.52, "grad_norm": 13.375, "learning_rate": 4.342201218805602e-06, "loss": 0.3369, "step": 2741 }, { "epoch": 0.52, "grad_norm": 14.3125, "learning_rate": 4.341683337100136e-06, "loss": 0.3098, "step": 2742 }, { "epoch": 0.52, "grad_norm": 13.875, "learning_rate": 4.341165282521308e-06, "loss": 0.3152, "step": 2743 }, { "epoch": 0.52, "grad_norm": 11.25, "learning_rate": 4.340647055117749e-06, "loss": 0.2379, "step": 2744 }, { "epoch": 0.52, "grad_norm": 10.6875, "learning_rate": 4.340128654938103e-06, "loss": 0.3262, "step": 2745 }, { "epoch": 0.52, "grad_norm": 9.375, "learning_rate": 4.339610082031028e-06, "loss": 0.302, "step": 2746 }, { "epoch": 0.52, "grad_norm": 8.3125, "learning_rate": 4.3390913364452035e-06, "loss": 0.2853, "step": 2747 }, { "epoch": 0.52, "grad_norm": 14.25, "learning_rate": 4.338572418229322e-06, "loss": 0.3711, "step": 2748 }, { "epoch": 0.52, "grad_norm": 8.8125, "learning_rate": 4.3380533274320925e-06, "loss": 0.325, "step": 2749 }, { "epoch": 0.52, "grad_norm": 14.5, "learning_rate": 4.337534064102241e-06, "loss": 0.3165, "step": 2750 }, { "epoch": 0.52, "grad_norm": 8.8125, "learning_rate": 4.337014628288508e-06, "loss": 0.2233, "step": 2751 }, { "epoch": 0.52, "grad_norm": 11.25, "learning_rate": 4.336495020039654e-06, "loss": 0.3251, "step": 2752 }, { "epoch": 0.52, "grad_norm": 10.625, "learning_rate": 4.33597523940445e-06, "loss": 0.2618, "step": 2753 }, { "epoch": 0.52, "grad_norm": 10.375, "learning_rate": 4.335455286431688e-06, "loss": 0.3129, "step": 2754 }, { "epoch": 0.52, "grad_norm": 9.25, "learning_rate": 4.334935161170174e-06, "loss": 0.2689, "step": 2755 }, { "epoch": 0.52, "grad_norm": 15.0, "learning_rate": 4.3344148636687314e-06, "loss": 0.2728, "step": 2756 }, { "epoch": 0.52, "grad_norm": 10.75, "learning_rate": 4.3338943939761974e-06, "loss": 0.2527, "step": 2757 }, { "epoch": 0.52, "grad_norm": 9.0, "learning_rate": 4.333373752141428e-06, "loss": 0.2686, "step": 2758 }, { "epoch": 0.52, "grad_norm": 9.125, "learning_rate": 4.3328529382132946e-06, "loss": 0.2806, "step": 2759 }, { "epoch": 0.52, "grad_norm": 14.9375, "learning_rate": 4.3323319522406834e-06, "loss": 0.2557, "step": 2760 }, { "epoch": 0.52, "grad_norm": 20.625, "learning_rate": 4.331810794272499e-06, "loss": 0.3129, "step": 2761 }, { "epoch": 0.52, "grad_norm": 10.5625, "learning_rate": 4.33128946435766e-06, "loss": 0.225, "step": 2762 }, { "epoch": 0.52, "grad_norm": 8.1875, "learning_rate": 4.330767962545103e-06, "loss": 0.2712, "step": 2763 }, { "epoch": 0.52, "grad_norm": 9.6875, "learning_rate": 4.330246288883778e-06, "loss": 0.28, "step": 2764 }, { "epoch": 0.52, "grad_norm": 24.125, "learning_rate": 4.329724443422655e-06, "loss": 0.2703, "step": 2765 }, { "epoch": 0.52, "grad_norm": 10.9375, "learning_rate": 4.329202426210718e-06, "loss": 0.2303, "step": 2766 }, { "epoch": 0.52, "grad_norm": 11.125, "learning_rate": 4.328680237296966e-06, "loss": 0.3525, "step": 2767 }, { "epoch": 0.52, "grad_norm": 7.84375, "learning_rate": 4.328157876730416e-06, "loss": 0.2389, "step": 2768 }, { "epoch": 0.52, "grad_norm": 8.9375, "learning_rate": 4.327635344560101e-06, "loss": 0.2524, "step": 2769 }, { "epoch": 0.52, "grad_norm": 12.3125, "learning_rate": 4.3271126408350675e-06, "loss": 0.3255, "step": 2770 }, { "epoch": 0.52, "grad_norm": 13.9375, "learning_rate": 4.326589765604382e-06, "loss": 0.2973, "step": 2771 }, { "epoch": 0.52, "grad_norm": 14.0625, "learning_rate": 4.3260667189171255e-06, "loss": 0.2665, "step": 2772 }, { "epoch": 0.52, "grad_norm": 14.5, "learning_rate": 4.325543500822394e-06, "loss": 0.3574, "step": 2773 }, { "epoch": 0.52, "grad_norm": 9.5, "learning_rate": 4.3250201113692995e-06, "loss": 0.2861, "step": 2774 }, { "epoch": 0.52, "grad_norm": 10.75, "learning_rate": 4.324496550606972e-06, "loss": 0.3232, "step": 2775 }, { "epoch": 0.53, "grad_norm": 8.4375, "learning_rate": 4.3239728185845585e-06, "loss": 0.2505, "step": 2776 }, { "epoch": 0.53, "grad_norm": 7.875, "learning_rate": 4.323448915351217e-06, "loss": 0.2572, "step": 2777 }, { "epoch": 0.53, "grad_norm": 8.125, "learning_rate": 4.322924840956127e-06, "loss": 0.2467, "step": 2778 }, { "epoch": 0.53, "grad_norm": 10.9375, "learning_rate": 4.3224005954484794e-06, "loss": 0.238, "step": 2779 }, { "epoch": 0.53, "grad_norm": 8.5625, "learning_rate": 4.321876178877486e-06, "loss": 0.2567, "step": 2780 }, { "epoch": 0.53, "grad_norm": 8.6875, "learning_rate": 4.321351591292371e-06, "loss": 0.3177, "step": 2781 }, { "epoch": 0.53, "grad_norm": 11.25, "learning_rate": 4.320826832742376e-06, "loss": 0.4028, "step": 2782 }, { "epoch": 0.53, "grad_norm": 11.9375, "learning_rate": 4.32030190327676e-06, "loss": 0.239, "step": 2783 }, { "epoch": 0.53, "grad_norm": 7.875, "learning_rate": 4.319776802944794e-06, "loss": 0.2546, "step": 2784 }, { "epoch": 0.53, "grad_norm": 11.125, "learning_rate": 4.31925153179577e-06, "loss": 0.2966, "step": 2785 }, { "epoch": 0.53, "grad_norm": 16.375, "learning_rate": 4.318726089878991e-06, "loss": 0.3192, "step": 2786 }, { "epoch": 0.53, "grad_norm": 15.5625, "learning_rate": 4.3182004772437815e-06, "loss": 0.3166, "step": 2787 }, { "epoch": 0.53, "grad_norm": 10.5625, "learning_rate": 4.317674693939478e-06, "loss": 0.3272, "step": 2788 }, { "epoch": 0.53, "grad_norm": 10.125, "learning_rate": 4.317148740015433e-06, "loss": 0.2621, "step": 2789 }, { "epoch": 0.53, "grad_norm": 12.875, "learning_rate": 4.316622615521019e-06, "loss": 0.2975, "step": 2790 }, { "epoch": 0.53, "grad_norm": 18.625, "learning_rate": 4.316096320505619e-06, "loss": 0.2825, "step": 2791 }, { "epoch": 0.53, "grad_norm": 7.0625, "learning_rate": 4.315569855018636e-06, "loss": 0.2831, "step": 2792 }, { "epoch": 0.53, "grad_norm": 11.5, "learning_rate": 4.315043219109488e-06, "loss": 0.3182, "step": 2793 }, { "epoch": 0.53, "grad_norm": 11.6875, "learning_rate": 4.314516412827608e-06, "loss": 0.2752, "step": 2794 }, { "epoch": 0.53, "grad_norm": 12.1875, "learning_rate": 4.313989436222447e-06, "loss": 0.3241, "step": 2795 }, { "epoch": 0.53, "grad_norm": 12.6875, "learning_rate": 4.313462289343469e-06, "loss": 0.2687, "step": 2796 }, { "epoch": 0.53, "grad_norm": 13.5, "learning_rate": 4.312934972240157e-06, "loss": 0.3313, "step": 2797 }, { "epoch": 0.53, "grad_norm": 8.25, "learning_rate": 4.3124074849620075e-06, "loss": 0.2394, "step": 2798 }, { "epoch": 0.53, "grad_norm": 12.9375, "learning_rate": 4.311879827558536e-06, "loss": 0.2374, "step": 2799 }, { "epoch": 0.53, "grad_norm": 10.375, "learning_rate": 4.31135200007927e-06, "loss": 0.2901, "step": 2800 }, { "epoch": 0.53, "grad_norm": 13.8125, "learning_rate": 4.310824002573757e-06, "loss": 0.3101, "step": 2801 }, { "epoch": 0.53, "grad_norm": 14.625, "learning_rate": 4.310295835091558e-06, "loss": 0.2803, "step": 2802 }, { "epoch": 0.53, "grad_norm": 10.25, "learning_rate": 4.309767497682249e-06, "loss": 0.2879, "step": 2803 }, { "epoch": 0.53, "grad_norm": 12.9375, "learning_rate": 4.309238990395426e-06, "loss": 0.2732, "step": 2804 }, { "epoch": 0.53, "grad_norm": 10.5625, "learning_rate": 4.308710313280697e-06, "loss": 0.3019, "step": 2805 }, { "epoch": 0.53, "grad_norm": 12.75, "learning_rate": 4.308181466387688e-06, "loss": 0.3141, "step": 2806 }, { "epoch": 0.53, "grad_norm": 12.8125, "learning_rate": 4.307652449766039e-06, "loss": 0.359, "step": 2807 }, { "epoch": 0.53, "grad_norm": 9.125, "learning_rate": 4.307123263465408e-06, "loss": 0.2369, "step": 2808 }, { "epoch": 0.53, "grad_norm": 23.625, "learning_rate": 4.3065939075354676e-06, "loss": 0.3071, "step": 2809 }, { "epoch": 0.53, "grad_norm": 11.75, "learning_rate": 4.306064382025909e-06, "loss": 0.2932, "step": 2810 }, { "epoch": 0.53, "grad_norm": 9.875, "learning_rate": 4.305534686986435e-06, "loss": 0.274, "step": 2811 }, { "epoch": 0.53, "grad_norm": 10.0, "learning_rate": 4.305004822466767e-06, "loss": 0.2996, "step": 2812 }, { "epoch": 0.53, "grad_norm": 11.9375, "learning_rate": 4.3044747885166415e-06, "loss": 0.2109, "step": 2813 }, { "epoch": 0.53, "grad_norm": 8.625, "learning_rate": 4.3039445851858135e-06, "loss": 0.2674, "step": 2814 }, { "epoch": 0.53, "grad_norm": 10.125, "learning_rate": 4.303414212524047e-06, "loss": 0.3481, "step": 2815 }, { "epoch": 0.53, "grad_norm": 11.0, "learning_rate": 4.302883670581132e-06, "loss": 0.2254, "step": 2816 }, { "epoch": 0.53, "grad_norm": 7.84375, "learning_rate": 4.3023529594068645e-06, "loss": 0.2531, "step": 2817 }, { "epoch": 0.53, "grad_norm": 8.875, "learning_rate": 4.301822079051063e-06, "loss": 0.2868, "step": 2818 }, { "epoch": 0.53, "grad_norm": 16.125, "learning_rate": 4.301291029563559e-06, "loss": 0.2696, "step": 2819 }, { "epoch": 0.53, "grad_norm": 10.4375, "learning_rate": 4.300759810994201e-06, "loss": 0.306, "step": 2820 }, { "epoch": 0.53, "grad_norm": 12.1875, "learning_rate": 4.300228423392852e-06, "loss": 0.2801, "step": 2821 }, { "epoch": 0.53, "grad_norm": 14.5625, "learning_rate": 4.299696866809393e-06, "loss": 0.2991, "step": 2822 }, { "epoch": 0.53, "grad_norm": 10.3125, "learning_rate": 4.29916514129372e-06, "loss": 0.2729, "step": 2823 }, { "epoch": 0.53, "grad_norm": 9.8125, "learning_rate": 4.298633246895742e-06, "loss": 0.3211, "step": 2824 }, { "epoch": 0.53, "grad_norm": 11.0625, "learning_rate": 4.2981011836653886e-06, "loss": 0.3254, "step": 2825 }, { "epoch": 0.53, "grad_norm": 11.5625, "learning_rate": 4.297568951652602e-06, "loss": 0.2679, "step": 2826 }, { "epoch": 0.53, "grad_norm": 9.875, "learning_rate": 4.297036550907341e-06, "loss": 0.2593, "step": 2827 }, { "epoch": 0.53, "grad_norm": 10.3125, "learning_rate": 4.296503981479582e-06, "loss": 0.2627, "step": 2828 }, { "epoch": 0.54, "grad_norm": 13.25, "learning_rate": 4.295971243419314e-06, "loss": 0.3889, "step": 2829 }, { "epoch": 0.54, "grad_norm": 10.125, "learning_rate": 4.295438336776545e-06, "loss": 0.3447, "step": 2830 }, { "epoch": 0.54, "grad_norm": 10.75, "learning_rate": 4.2949052616012965e-06, "loss": 0.2, "step": 2831 }, { "epoch": 0.54, "grad_norm": 14.125, "learning_rate": 4.294372017943607e-06, "loss": 0.2981, "step": 2832 }, { "epoch": 0.54, "grad_norm": 10.125, "learning_rate": 4.29383860585353e-06, "loss": 0.2707, "step": 2833 }, { "epoch": 0.54, "grad_norm": 14.75, "learning_rate": 4.293305025381135e-06, "loss": 0.3548, "step": 2834 }, { "epoch": 0.54, "grad_norm": 7.84375, "learning_rate": 4.292771276576509e-06, "loss": 0.2683, "step": 2835 }, { "epoch": 0.54, "grad_norm": 9.0, "learning_rate": 4.292237359489753e-06, "loss": 0.3025, "step": 2836 }, { "epoch": 0.54, "grad_norm": 11.6875, "learning_rate": 4.291703274170983e-06, "loss": 0.3289, "step": 2837 }, { "epoch": 0.54, "grad_norm": 9.5, "learning_rate": 4.291169020670334e-06, "loss": 0.2121, "step": 2838 }, { "epoch": 0.54, "grad_norm": 9.875, "learning_rate": 4.290634599037953e-06, "loss": 0.2684, "step": 2839 }, { "epoch": 0.54, "grad_norm": 11.5625, "learning_rate": 4.2901000093240055e-06, "loss": 0.3706, "step": 2840 }, { "epoch": 0.54, "grad_norm": 18.75, "learning_rate": 4.2895652515786715e-06, "loss": 0.2904, "step": 2841 }, { "epoch": 0.54, "grad_norm": 17.625, "learning_rate": 4.289030325852147e-06, "loss": 0.3093, "step": 2842 }, { "epoch": 0.54, "grad_norm": 9.75, "learning_rate": 4.288495232194644e-06, "loss": 0.2766, "step": 2843 }, { "epoch": 0.54, "grad_norm": 9.875, "learning_rate": 4.287959970656391e-06, "loss": 0.2893, "step": 2844 }, { "epoch": 0.54, "grad_norm": 10.5, "learning_rate": 4.28742454128763e-06, "loss": 0.3503, "step": 2845 }, { "epoch": 0.54, "grad_norm": 10.0625, "learning_rate": 4.286888944138622e-06, "loss": 0.3451, "step": 2846 }, { "epoch": 0.54, "grad_norm": 7.90625, "learning_rate": 4.28635317925964e-06, "loss": 0.26, "step": 2847 }, { "epoch": 0.54, "grad_norm": 11.0625, "learning_rate": 4.2858172467009755e-06, "loss": 0.3182, "step": 2848 }, { "epoch": 0.54, "grad_norm": 8.1875, "learning_rate": 4.2852811465129355e-06, "loss": 0.3101, "step": 2849 }, { "epoch": 0.54, "grad_norm": 18.5, "learning_rate": 4.284744878745841e-06, "loss": 0.379, "step": 2850 }, { "epoch": 0.54, "grad_norm": 9.8125, "learning_rate": 4.28420844345003e-06, "loss": 0.344, "step": 2851 }, { "epoch": 0.54, "grad_norm": 13.9375, "learning_rate": 4.283671840675857e-06, "loss": 0.3273, "step": 2852 }, { "epoch": 0.54, "grad_norm": 10.5625, "learning_rate": 4.2831350704736905e-06, "loss": 0.2545, "step": 2853 }, { "epoch": 0.54, "grad_norm": 9.5, "learning_rate": 4.282598132893916e-06, "loss": 0.2943, "step": 2854 }, { "epoch": 0.54, "grad_norm": 8.8125, "learning_rate": 4.282061027986934e-06, "loss": 0.2387, "step": 2855 }, { "epoch": 0.54, "grad_norm": 9.9375, "learning_rate": 4.281523755803161e-06, "loss": 0.3363, "step": 2856 }, { "epoch": 0.54, "grad_norm": 13.0, "learning_rate": 4.28098631639303e-06, "loss": 0.2847, "step": 2857 }, { "epoch": 0.54, "grad_norm": 11.6875, "learning_rate": 4.280448709806987e-06, "loss": 0.278, "step": 2858 }, { "epoch": 0.54, "grad_norm": 18.125, "learning_rate": 4.279910936095496e-06, "loss": 0.3624, "step": 2859 }, { "epoch": 0.54, "grad_norm": 9.875, "learning_rate": 4.279372995309038e-06, "loss": 0.2733, "step": 2860 }, { "epoch": 0.54, "grad_norm": 9.5, "learning_rate": 4.278834887498106e-06, "loss": 0.233, "step": 2861 }, { "epoch": 0.54, "grad_norm": 20.125, "learning_rate": 4.278296612713212e-06, "loss": 0.2823, "step": 2862 }, { "epoch": 0.54, "grad_norm": 9.875, "learning_rate": 4.277758171004881e-06, "loss": 0.3304, "step": 2863 }, { "epoch": 0.54, "grad_norm": 9.75, "learning_rate": 4.277219562423655e-06, "loss": 0.2909, "step": 2864 }, { "epoch": 0.54, "grad_norm": 9.125, "learning_rate": 4.276680787020092e-06, "loss": 0.3234, "step": 2865 }, { "epoch": 0.54, "grad_norm": 9.8125, "learning_rate": 4.2761418448447665e-06, "loss": 0.2663, "step": 2866 }, { "epoch": 0.54, "grad_norm": 10.9375, "learning_rate": 4.275602735948266e-06, "loss": 0.3014, "step": 2867 }, { "epoch": 0.54, "grad_norm": 14.25, "learning_rate": 4.2750634603811944e-06, "loss": 0.2871, "step": 2868 }, { "epoch": 0.54, "grad_norm": 17.25, "learning_rate": 4.274524018194173e-06, "loss": 0.321, "step": 2869 }, { "epoch": 0.54, "grad_norm": 13.25, "learning_rate": 4.273984409437837e-06, "loss": 0.2594, "step": 2870 }, { "epoch": 0.54, "grad_norm": 9.0625, "learning_rate": 4.273444634162839e-06, "loss": 0.237, "step": 2871 }, { "epoch": 0.54, "grad_norm": 10.125, "learning_rate": 4.2729046924198444e-06, "loss": 0.3053, "step": 2872 }, { "epoch": 0.54, "grad_norm": 5.3125, "learning_rate": 4.272364584259538e-06, "loss": 0.1454, "step": 2873 }, { "epoch": 0.54, "grad_norm": 13.9375, "learning_rate": 4.271824309732616e-06, "loss": 0.3222, "step": 2874 }, { "epoch": 0.54, "grad_norm": 8.1875, "learning_rate": 4.271283868889795e-06, "loss": 0.3197, "step": 2875 }, { "epoch": 0.54, "grad_norm": 10.0625, "learning_rate": 4.270743261781802e-06, "loss": 0.2968, "step": 2876 }, { "epoch": 0.54, "grad_norm": 13.5625, "learning_rate": 4.270202488459383e-06, "loss": 0.2101, "step": 2877 }, { "epoch": 0.54, "grad_norm": 17.375, "learning_rate": 4.269661548973299e-06, "loss": 0.241, "step": 2878 }, { "epoch": 0.54, "grad_norm": 9.625, "learning_rate": 4.269120443374327e-06, "loss": 0.1933, "step": 2879 }, { "epoch": 0.54, "grad_norm": 12.4375, "learning_rate": 4.268579171713258e-06, "loss": 0.4075, "step": 2880 }, { "epoch": 0.55, "grad_norm": 9.75, "learning_rate": 4.268037734040899e-06, "loss": 0.2926, "step": 2881 }, { "epoch": 0.55, "grad_norm": 11.875, "learning_rate": 4.267496130408076e-06, "loss": 0.3424, "step": 2882 }, { "epoch": 0.55, "grad_norm": 18.0, "learning_rate": 4.266954360865625e-06, "loss": 0.2562, "step": 2883 }, { "epoch": 0.55, "grad_norm": 8.1875, "learning_rate": 4.266412425464401e-06, "loss": 0.238, "step": 2884 }, { "epoch": 0.55, "grad_norm": 12.75, "learning_rate": 4.265870324255273e-06, "loss": 0.2786, "step": 2885 }, { "epoch": 0.55, "grad_norm": 11.4375, "learning_rate": 4.265328057289129e-06, "loss": 0.3871, "step": 2886 }, { "epoch": 0.55, "grad_norm": 9.6875, "learning_rate": 4.264785624616868e-06, "loss": 0.329, "step": 2887 }, { "epoch": 0.55, "grad_norm": 10.5, "learning_rate": 4.264243026289406e-06, "loss": 0.2818, "step": 2888 }, { "epoch": 0.55, "grad_norm": 9.5625, "learning_rate": 4.263700262357677e-06, "loss": 0.2895, "step": 2889 }, { "epoch": 0.55, "grad_norm": 12.0625, "learning_rate": 4.263157332872627e-06, "loss": 0.2245, "step": 2890 }, { "epoch": 0.55, "grad_norm": 20.25, "learning_rate": 4.262614237885221e-06, "loss": 0.2791, "step": 2891 }, { "epoch": 0.55, "grad_norm": 14.4375, "learning_rate": 4.262070977446436e-06, "loss": 0.3065, "step": 2892 }, { "epoch": 0.55, "grad_norm": 10.5, "learning_rate": 4.261527551607267e-06, "loss": 0.3004, "step": 2893 }, { "epoch": 0.55, "grad_norm": 12.875, "learning_rate": 4.260983960418723e-06, "loss": 0.3011, "step": 2894 }, { "epoch": 0.55, "grad_norm": 10.375, "learning_rate": 4.26044020393183e-06, "loss": 0.2323, "step": 2895 }, { "epoch": 0.55, "grad_norm": 13.8125, "learning_rate": 4.259896282197629e-06, "loss": 0.2957, "step": 2896 }, { "epoch": 0.55, "grad_norm": 10.1875, "learning_rate": 4.259352195267177e-06, "loss": 0.4119, "step": 2897 }, { "epoch": 0.55, "grad_norm": 8.75, "learning_rate": 4.258807943191543e-06, "loss": 0.2795, "step": 2898 }, { "epoch": 0.55, "grad_norm": 10.8125, "learning_rate": 4.2582635260218174e-06, "loss": 0.3132, "step": 2899 }, { "epoch": 0.55, "grad_norm": 10.125, "learning_rate": 4.257718943809102e-06, "loss": 0.2909, "step": 2900 }, { "epoch": 0.55, "grad_norm": 12.75, "learning_rate": 4.257174196604514e-06, "loss": 0.3201, "step": 2901 }, { "epoch": 0.55, "grad_norm": 15.3125, "learning_rate": 4.256629284459189e-06, "loss": 0.33, "step": 2902 }, { "epoch": 0.55, "grad_norm": 10.1875, "learning_rate": 4.256084207424275e-06, "loss": 0.3465, "step": 2903 }, { "epoch": 0.55, "grad_norm": 8.6875, "learning_rate": 4.255538965550937e-06, "loss": 0.2899, "step": 2904 }, { "epoch": 0.55, "grad_norm": 18.0, "learning_rate": 4.254993558890357e-06, "loss": 0.2815, "step": 2905 }, { "epoch": 0.55, "grad_norm": 12.8125, "learning_rate": 4.254447987493727e-06, "loss": 0.3018, "step": 2906 }, { "epoch": 0.55, "grad_norm": 10.0, "learning_rate": 4.253902251412262e-06, "loss": 0.3125, "step": 2907 }, { "epoch": 0.55, "grad_norm": 9.25, "learning_rate": 4.253356350697186e-06, "loss": 0.2748, "step": 2908 }, { "epoch": 0.55, "grad_norm": 8.125, "learning_rate": 4.252810285399742e-06, "loss": 0.221, "step": 2909 }, { "epoch": 0.55, "grad_norm": 10.5, "learning_rate": 4.252264055571188e-06, "loss": 0.2416, "step": 2910 }, { "epoch": 0.55, "grad_norm": 9.8125, "learning_rate": 4.251717661262797e-06, "loss": 0.296, "step": 2911 }, { "epoch": 0.55, "grad_norm": 16.0, "learning_rate": 4.251171102525856e-06, "loss": 0.3195, "step": 2912 }, { "epoch": 0.55, "grad_norm": 9.5, "learning_rate": 4.250624379411671e-06, "loss": 0.2404, "step": 2913 }, { "epoch": 0.55, "grad_norm": 13.625, "learning_rate": 4.2500774919715595e-06, "loss": 0.2057, "step": 2914 }, { "epoch": 0.55, "grad_norm": 10.9375, "learning_rate": 4.249530440256857e-06, "loss": 0.3208, "step": 2915 }, { "epoch": 0.55, "grad_norm": 9.6875, "learning_rate": 4.248983224318913e-06, "loss": 0.2798, "step": 2916 }, { "epoch": 0.55, "grad_norm": 13.9375, "learning_rate": 4.248435844209093e-06, "loss": 0.2824, "step": 2917 }, { "epoch": 0.55, "grad_norm": 12.9375, "learning_rate": 4.247888299978779e-06, "loss": 0.2943, "step": 2918 }, { "epoch": 0.55, "grad_norm": 8.1875, "learning_rate": 4.247340591679367e-06, "loss": 0.3149, "step": 2919 }, { "epoch": 0.55, "grad_norm": 14.5, "learning_rate": 4.2467927193622685e-06, "loss": 0.3177, "step": 2920 }, { "epoch": 0.55, "grad_norm": 9.875, "learning_rate": 4.246244683078911e-06, "loss": 0.2115, "step": 2921 }, { "epoch": 0.55, "grad_norm": 23.0, "learning_rate": 4.245696482880736e-06, "loss": 0.321, "step": 2922 }, { "epoch": 0.55, "grad_norm": 9.9375, "learning_rate": 4.2451481188192025e-06, "loss": 0.2973, "step": 2923 }, { "epoch": 0.55, "grad_norm": 9.1875, "learning_rate": 4.244599590945783e-06, "loss": 0.2973, "step": 2924 }, { "epoch": 0.55, "grad_norm": 15.9375, "learning_rate": 4.244050899311967e-06, "loss": 0.2527, "step": 2925 }, { "epoch": 0.55, "grad_norm": 16.75, "learning_rate": 4.243502043969259e-06, "loss": 0.2905, "step": 2926 }, { "epoch": 0.55, "grad_norm": 58.5, "learning_rate": 4.2429530249691765e-06, "loss": 0.3572, "step": 2927 }, { "epoch": 0.55, "grad_norm": 12.5625, "learning_rate": 4.2424038423632564e-06, "loss": 0.364, "step": 2928 }, { "epoch": 0.55, "grad_norm": 9.3125, "learning_rate": 4.241854496203047e-06, "loss": 0.2905, "step": 2929 }, { "epoch": 0.55, "grad_norm": 19.375, "learning_rate": 4.241304986540114e-06, "loss": 0.2875, "step": 2930 }, { "epoch": 0.55, "grad_norm": 13.0, "learning_rate": 4.240755313426041e-06, "loss": 0.3484, "step": 2931 }, { "epoch": 0.55, "grad_norm": 9.75, "learning_rate": 4.24020547691242e-06, "loss": 0.3618, "step": 2932 }, { "epoch": 0.55, "grad_norm": 20.625, "learning_rate": 4.239655477050866e-06, "loss": 0.3455, "step": 2933 }, { "epoch": 0.56, "grad_norm": 10.4375, "learning_rate": 4.239105313893004e-06, "loss": 0.2667, "step": 2934 }, { "epoch": 0.56, "grad_norm": 15.25, "learning_rate": 4.238554987490477e-06, "loss": 0.2833, "step": 2935 }, { "epoch": 0.56, "grad_norm": 12.8125, "learning_rate": 4.238004497894942e-06, "loss": 0.2382, "step": 2936 }, { "epoch": 0.56, "grad_norm": 10.6875, "learning_rate": 4.2374538451580714e-06, "loss": 0.2774, "step": 2937 }, { "epoch": 0.56, "grad_norm": 14.625, "learning_rate": 4.236903029331555e-06, "loss": 0.3137, "step": 2938 }, { "epoch": 0.56, "grad_norm": 10.375, "learning_rate": 4.236352050467096e-06, "loss": 0.2473, "step": 2939 }, { "epoch": 0.56, "grad_norm": 8.8125, "learning_rate": 4.23580090861641e-06, "loss": 0.2976, "step": 2940 }, { "epoch": 0.56, "grad_norm": 11.375, "learning_rate": 4.235249603831234e-06, "loss": 0.2879, "step": 2941 }, { "epoch": 0.56, "grad_norm": 17.125, "learning_rate": 4.234698136163318e-06, "loss": 0.417, "step": 2942 }, { "epoch": 0.56, "grad_norm": 9.5625, "learning_rate": 4.234146505664424e-06, "loss": 0.3295, "step": 2943 }, { "epoch": 0.56, "grad_norm": 8.6875, "learning_rate": 4.233594712386334e-06, "loss": 0.2478, "step": 2944 }, { "epoch": 0.56, "grad_norm": 15.3125, "learning_rate": 4.233042756380843e-06, "loss": 0.3085, "step": 2945 }, { "epoch": 0.56, "grad_norm": 9.8125, "learning_rate": 4.23249063769976e-06, "loss": 0.2862, "step": 2946 }, { "epoch": 0.56, "grad_norm": 6.96875, "learning_rate": 4.231938356394912e-06, "loss": 0.2214, "step": 2947 }, { "epoch": 0.56, "grad_norm": 13.25, "learning_rate": 4.23138591251814e-06, "loss": 0.3319, "step": 2948 }, { "epoch": 0.56, "grad_norm": 13.0625, "learning_rate": 4.2308333061213e-06, "loss": 0.3185, "step": 2949 }, { "epoch": 0.56, "grad_norm": 12.9375, "learning_rate": 4.230280537256264e-06, "loss": 0.331, "step": 2950 }, { "epoch": 0.56, "grad_norm": 11.875, "learning_rate": 4.229727605974917e-06, "loss": 0.2588, "step": 2951 }, { "epoch": 0.56, "grad_norm": 11.375, "learning_rate": 4.229174512329163e-06, "loss": 0.3436, "step": 2952 }, { "epoch": 0.56, "grad_norm": 10.5625, "learning_rate": 4.228621256370919e-06, "loss": 0.2599, "step": 2953 }, { "epoch": 0.56, "grad_norm": 31.125, "learning_rate": 4.228067838152117e-06, "loss": 0.2749, "step": 2954 }, { "epoch": 0.56, "grad_norm": 13.0625, "learning_rate": 4.227514257724704e-06, "loss": 0.3492, "step": 2955 }, { "epoch": 0.56, "grad_norm": 15.375, "learning_rate": 4.226960515140644e-06, "loss": 0.2263, "step": 2956 }, { "epoch": 0.56, "grad_norm": 24.875, "learning_rate": 4.226406610451917e-06, "loss": 0.2829, "step": 2957 }, { "epoch": 0.56, "grad_norm": 11.9375, "learning_rate": 4.2258525437105115e-06, "loss": 0.2786, "step": 2958 }, { "epoch": 0.56, "grad_norm": 10.6875, "learning_rate": 4.225298314968441e-06, "loss": 0.2944, "step": 2959 }, { "epoch": 0.56, "grad_norm": 11.375, "learning_rate": 4.224743924277726e-06, "loss": 0.3643, "step": 2960 }, { "epoch": 0.56, "grad_norm": 10.375, "learning_rate": 4.224189371690408e-06, "loss": 0.2844, "step": 2961 }, { "epoch": 0.56, "grad_norm": 11.125, "learning_rate": 4.2236346572585385e-06, "loss": 0.3393, "step": 2962 }, { "epoch": 0.56, "grad_norm": 8.5625, "learning_rate": 4.223079781034189e-06, "loss": 0.3204, "step": 2963 }, { "epoch": 0.56, "grad_norm": 12.1875, "learning_rate": 4.2225247430694435e-06, "loss": 0.2995, "step": 2964 }, { "epoch": 0.56, "grad_norm": 14.5625, "learning_rate": 4.221969543416403e-06, "loss": 0.2748, "step": 2965 }, { "epoch": 0.56, "grad_norm": 11.0, "learning_rate": 4.221414182127179e-06, "loss": 0.346, "step": 2966 }, { "epoch": 0.56, "grad_norm": 10.6875, "learning_rate": 4.2208586592539055e-06, "loss": 0.2649, "step": 2967 }, { "epoch": 0.56, "grad_norm": 9.4375, "learning_rate": 4.220302974848725e-06, "loss": 0.251, "step": 2968 }, { "epoch": 0.56, "grad_norm": 12.125, "learning_rate": 4.2197471289638e-06, "loss": 0.2512, "step": 2969 }, { "epoch": 0.56, "grad_norm": 10.5, "learning_rate": 4.2191911216513045e-06, "loss": 0.2562, "step": 2970 }, { "epoch": 0.56, "grad_norm": 9.5, "learning_rate": 4.21863495296343e-06, "loss": 0.3114, "step": 2971 }, { "epoch": 0.56, "grad_norm": 9.75, "learning_rate": 4.218078622952383e-06, "loss": 0.2231, "step": 2972 }, { "epoch": 0.56, "grad_norm": 13.8125, "learning_rate": 4.217522131670383e-06, "loss": 0.276, "step": 2973 }, { "epoch": 0.56, "grad_norm": 11.5, "learning_rate": 4.216965479169668e-06, "loss": 0.2623, "step": 2974 }, { "epoch": 0.56, "grad_norm": 8.875, "learning_rate": 4.2164086655024874e-06, "loss": 0.3178, "step": 2975 }, { "epoch": 0.56, "grad_norm": 9.125, "learning_rate": 4.215851690721109e-06, "loss": 0.2662, "step": 2976 }, { "epoch": 0.56, "grad_norm": 8.9375, "learning_rate": 4.215294554877814e-06, "loss": 0.2926, "step": 2977 }, { "epoch": 0.56, "grad_norm": 12.4375, "learning_rate": 4.214737258024899e-06, "loss": 0.2917, "step": 2978 }, { "epoch": 0.56, "grad_norm": 9.8125, "learning_rate": 4.214179800214676e-06, "loss": 0.2457, "step": 2979 }, { "epoch": 0.56, "grad_norm": 8.9375, "learning_rate": 4.213622181499471e-06, "loss": 0.2169, "step": 2980 }, { "epoch": 0.56, "grad_norm": 9.1875, "learning_rate": 4.213064401931629e-06, "loss": 0.3159, "step": 2981 }, { "epoch": 0.56, "grad_norm": 8.8125, "learning_rate": 4.2125064615635034e-06, "loss": 0.2855, "step": 2982 }, { "epoch": 0.56, "grad_norm": 7.65625, "learning_rate": 4.2119483604474675e-06, "loss": 0.2158, "step": 2983 }, { "epoch": 0.56, "grad_norm": 10.1875, "learning_rate": 4.21139009863591e-06, "loss": 0.2753, "step": 2984 }, { "epoch": 0.56, "grad_norm": 17.0, "learning_rate": 4.2108316761812326e-06, "loss": 0.2616, "step": 2985 }, { "epoch": 0.56, "grad_norm": 11.0625, "learning_rate": 4.210273093135851e-06, "loss": 0.3147, "step": 2986 }, { "epoch": 0.57, "grad_norm": 14.5625, "learning_rate": 4.209714349552201e-06, "loss": 0.2767, "step": 2987 }, { "epoch": 0.57, "grad_norm": 12.0625, "learning_rate": 4.209155445482728e-06, "loss": 0.3272, "step": 2988 }, { "epoch": 0.57, "grad_norm": 10.4375, "learning_rate": 4.208596380979895e-06, "loss": 0.2534, "step": 2989 }, { "epoch": 0.57, "grad_norm": 10.9375, "learning_rate": 4.208037156096179e-06, "loss": 0.3011, "step": 2990 }, { "epoch": 0.57, "grad_norm": 6.75, "learning_rate": 4.207477770884074e-06, "loss": 0.2359, "step": 2991 }, { "epoch": 0.57, "grad_norm": 10.625, "learning_rate": 4.2069182253960885e-06, "loss": 0.3433, "step": 2992 }, { "epoch": 0.57, "grad_norm": 15.875, "learning_rate": 4.206358519684743e-06, "loss": 0.3114, "step": 2993 }, { "epoch": 0.57, "grad_norm": 14.875, "learning_rate": 4.205798653802578e-06, "loss": 0.2667, "step": 2994 }, { "epoch": 0.57, "grad_norm": 9.3125, "learning_rate": 4.205238627802144e-06, "loss": 0.2498, "step": 2995 }, { "epoch": 0.57, "grad_norm": 27.75, "learning_rate": 4.204678441736011e-06, "loss": 0.2709, "step": 2996 }, { "epoch": 0.57, "grad_norm": 17.125, "learning_rate": 4.204118095656762e-06, "loss": 0.3126, "step": 2997 }, { "epoch": 0.57, "grad_norm": 12.1875, "learning_rate": 4.2035575896169924e-06, "loss": 0.2644, "step": 2998 }, { "epoch": 0.57, "grad_norm": 8.9375, "learning_rate": 4.202996923669318e-06, "loss": 0.278, "step": 2999 }, { "epoch": 0.57, "grad_norm": 12.0, "learning_rate": 4.202436097866365e-06, "loss": 0.3262, "step": 3000 }, { "epoch": 0.57, "grad_norm": 9.0, "learning_rate": 4.2018751122607785e-06, "loss": 0.222, "step": 3001 }, { "epoch": 0.57, "grad_norm": 8.0, "learning_rate": 4.201313966905215e-06, "loss": 0.2922, "step": 3002 }, { "epoch": 0.57, "grad_norm": 15.1875, "learning_rate": 4.200752661852349e-06, "loss": 0.319, "step": 3003 }, { "epoch": 0.57, "grad_norm": 8.0625, "learning_rate": 4.200191197154866e-06, "loss": 0.2516, "step": 3004 }, { "epoch": 0.57, "grad_norm": 10.125, "learning_rate": 4.199629572865471e-06, "loss": 0.3419, "step": 3005 }, { "epoch": 0.57, "grad_norm": 10.25, "learning_rate": 4.199067789036882e-06, "loss": 0.2732, "step": 3006 }, { "epoch": 0.57, "grad_norm": 9.1875, "learning_rate": 4.198505845721832e-06, "loss": 0.3198, "step": 3007 }, { "epoch": 0.57, "grad_norm": 14.1875, "learning_rate": 4.197943742973068e-06, "loss": 0.3156, "step": 3008 }, { "epoch": 0.57, "grad_norm": 7.53125, "learning_rate": 4.197381480843353e-06, "loss": 0.2685, "step": 3009 }, { "epoch": 0.57, "grad_norm": 15.125, "learning_rate": 4.196819059385465e-06, "loss": 0.3256, "step": 3010 }, { "epoch": 0.57, "grad_norm": 25.375, "learning_rate": 4.196256478652197e-06, "loss": 0.3045, "step": 3011 }, { "epoch": 0.57, "grad_norm": 11.25, "learning_rate": 4.195693738696359e-06, "loss": 0.282, "step": 3012 }, { "epoch": 0.57, "grad_norm": 8.0625, "learning_rate": 4.19513083957077e-06, "loss": 0.2602, "step": 3013 }, { "epoch": 0.57, "grad_norm": 10.5625, "learning_rate": 4.19456778132827e-06, "loss": 0.3903, "step": 3014 }, { "epoch": 0.57, "grad_norm": 16.5, "learning_rate": 4.19400456402171e-06, "loss": 0.4406, "step": 3015 }, { "epoch": 0.57, "grad_norm": 6.90625, "learning_rate": 4.19344118770396e-06, "loss": 0.2538, "step": 3016 }, { "epoch": 0.57, "grad_norm": 9.3125, "learning_rate": 4.192877652427899e-06, "loss": 0.2894, "step": 3017 }, { "epoch": 0.57, "grad_norm": 12.75, "learning_rate": 4.192313958246428e-06, "loss": 0.2747, "step": 3018 }, { "epoch": 0.57, "grad_norm": 13.5625, "learning_rate": 4.191750105212457e-06, "loss": 0.2734, "step": 3019 }, { "epoch": 0.57, "grad_norm": 11.3125, "learning_rate": 4.191186093378914e-06, "loss": 0.3098, "step": 3020 }, { "epoch": 0.57, "grad_norm": 9.9375, "learning_rate": 4.190621922798741e-06, "loss": 0.2812, "step": 3021 }, { "epoch": 0.57, "grad_norm": 12.5, "learning_rate": 4.190057593524894e-06, "loss": 0.233, "step": 3022 }, { "epoch": 0.57, "grad_norm": 10.125, "learning_rate": 4.1894931056103475e-06, "loss": 0.2447, "step": 3023 }, { "epoch": 0.57, "grad_norm": 13.8125, "learning_rate": 4.1889284591080855e-06, "loss": 0.2809, "step": 3024 }, { "epoch": 0.57, "grad_norm": 14.5625, "learning_rate": 4.188363654071113e-06, "loss": 0.2887, "step": 3025 }, { "epoch": 0.57, "grad_norm": 13.4375, "learning_rate": 4.187798690552442e-06, "loss": 0.2524, "step": 3026 }, { "epoch": 0.57, "grad_norm": 9.625, "learning_rate": 4.187233568605108e-06, "loss": 0.258, "step": 3027 }, { "epoch": 0.57, "grad_norm": 9.75, "learning_rate": 4.1866682882821545e-06, "loss": 0.2246, "step": 3028 }, { "epoch": 0.57, "grad_norm": 6.9375, "learning_rate": 4.186102849636645e-06, "loss": 0.2667, "step": 3029 }, { "epoch": 0.57, "grad_norm": 8.5, "learning_rate": 4.185537252721654e-06, "loss": 0.2373, "step": 3030 }, { "epoch": 0.57, "grad_norm": 15.1875, "learning_rate": 4.1849714975902735e-06, "loss": 0.3402, "step": 3031 }, { "epoch": 0.57, "grad_norm": 14.75, "learning_rate": 4.184405584295609e-06, "loss": 0.3124, "step": 3032 }, { "epoch": 0.57, "grad_norm": 10.375, "learning_rate": 4.1838395128907795e-06, "loss": 0.332, "step": 3033 }, { "epoch": 0.57, "grad_norm": 10.5, "learning_rate": 4.183273283428922e-06, "loss": 0.2574, "step": 3034 }, { "epoch": 0.57, "grad_norm": 13.3125, "learning_rate": 4.1827068959631865e-06, "loss": 0.2723, "step": 3035 }, { "epoch": 0.57, "grad_norm": 10.5, "learning_rate": 4.18214035054674e-06, "loss": 0.2394, "step": 3036 }, { "epoch": 0.57, "grad_norm": 6.90625, "learning_rate": 4.1815736472327584e-06, "loss": 0.2558, "step": 3037 }, { "epoch": 0.57, "grad_norm": 17.75, "learning_rate": 4.18100678607444e-06, "loss": 0.2943, "step": 3038 }, { "epoch": 0.57, "grad_norm": 8.875, "learning_rate": 4.180439767124992e-06, "loss": 0.2617, "step": 3039 }, { "epoch": 0.58, "grad_norm": 12.5625, "learning_rate": 4.17987259043764e-06, "loss": 0.2765, "step": 3040 }, { "epoch": 0.58, "grad_norm": 12.5, "learning_rate": 4.1793052560656226e-06, "loss": 0.29, "step": 3041 }, { "epoch": 0.58, "grad_norm": 11.8125, "learning_rate": 4.178737764062195e-06, "loss": 0.2551, "step": 3042 }, { "epoch": 0.58, "grad_norm": 20.875, "learning_rate": 4.178170114480625e-06, "loss": 0.3256, "step": 3043 }, { "epoch": 0.58, "grad_norm": 12.5625, "learning_rate": 4.177602307374197e-06, "loss": 0.3573, "step": 3044 }, { "epoch": 0.58, "grad_norm": 13.25, "learning_rate": 4.177034342796207e-06, "loss": 0.35, "step": 3045 }, { "epoch": 0.58, "grad_norm": 11.0, "learning_rate": 4.1764662207999705e-06, "loss": 0.2768, "step": 3046 }, { "epoch": 0.58, "grad_norm": 11.125, "learning_rate": 4.175897941438816e-06, "loss": 0.3579, "step": 3047 }, { "epoch": 0.58, "grad_norm": 9.1875, "learning_rate": 4.1753295047660845e-06, "loss": 0.2688, "step": 3048 }, { "epoch": 0.58, "grad_norm": 8.625, "learning_rate": 4.174760910835134e-06, "loss": 0.2441, "step": 3049 }, { "epoch": 0.58, "grad_norm": 8.9375, "learning_rate": 4.1741921596993365e-06, "loss": 0.2779, "step": 3050 }, { "epoch": 0.58, "grad_norm": 10.25, "learning_rate": 4.173623251412079e-06, "loss": 0.3479, "step": 3051 }, { "epoch": 0.58, "grad_norm": 10.125, "learning_rate": 4.173054186026764e-06, "loss": 0.2672, "step": 3052 }, { "epoch": 0.58, "grad_norm": 9.1875, "learning_rate": 4.172484963596807e-06, "loss": 0.2402, "step": 3053 }, { "epoch": 0.58, "grad_norm": 7.0, "learning_rate": 4.1719155841756395e-06, "loss": 0.2611, "step": 3054 }, { "epoch": 0.58, "grad_norm": 18.625, "learning_rate": 4.1713460478167075e-06, "loss": 0.2912, "step": 3055 }, { "epoch": 0.58, "grad_norm": 12.25, "learning_rate": 4.170776354573472e-06, "loss": 0.2401, "step": 3056 }, { "epoch": 0.58, "grad_norm": 10.3125, "learning_rate": 4.170206504499409e-06, "loss": 0.3303, "step": 3057 }, { "epoch": 0.58, "grad_norm": 11.5, "learning_rate": 4.169636497648008e-06, "loss": 0.3403, "step": 3058 }, { "epoch": 0.58, "grad_norm": 14.0, "learning_rate": 4.1690663340727725e-06, "loss": 0.3306, "step": 3059 }, { "epoch": 0.58, "grad_norm": 25.75, "learning_rate": 4.168496013827223e-06, "loss": 0.3023, "step": 3060 }, { "epoch": 0.58, "grad_norm": 17.75, "learning_rate": 4.167925536964895e-06, "loss": 0.3195, "step": 3061 }, { "epoch": 0.58, "grad_norm": 12.5, "learning_rate": 4.167354903539337e-06, "loss": 0.312, "step": 3062 }, { "epoch": 0.58, "grad_norm": 15.375, "learning_rate": 4.166784113604111e-06, "loss": 0.2667, "step": 3063 }, { "epoch": 0.58, "grad_norm": 10.0, "learning_rate": 4.166213167212797e-06, "loss": 0.2889, "step": 3064 }, { "epoch": 0.58, "grad_norm": 8.8125, "learning_rate": 4.165642064418988e-06, "loss": 0.2858, "step": 3065 }, { "epoch": 0.58, "grad_norm": 10.625, "learning_rate": 4.165070805276291e-06, "loss": 0.3079, "step": 3066 }, { "epoch": 0.58, "grad_norm": 10.125, "learning_rate": 4.164499389838329e-06, "loss": 0.3366, "step": 3067 }, { "epoch": 0.58, "grad_norm": 13.8125, "learning_rate": 4.163927818158738e-06, "loss": 0.3006, "step": 3068 }, { "epoch": 0.58, "grad_norm": 16.75, "learning_rate": 4.163356090291171e-06, "loss": 0.3101, "step": 3069 }, { "epoch": 0.58, "grad_norm": 9.875, "learning_rate": 4.162784206289295e-06, "loss": 0.2883, "step": 3070 }, { "epoch": 0.58, "grad_norm": 8.75, "learning_rate": 4.162212166206788e-06, "loss": 0.3168, "step": 3071 }, { "epoch": 0.58, "grad_norm": 11.375, "learning_rate": 4.161639970097349e-06, "loss": 0.3169, "step": 3072 }, { "epoch": 0.58, "grad_norm": 10.1875, "learning_rate": 4.161067618014687e-06, "loss": 0.2994, "step": 3073 }, { "epoch": 0.58, "grad_norm": 8.4375, "learning_rate": 4.160495110012526e-06, "loss": 0.2985, "step": 3074 }, { "epoch": 0.58, "grad_norm": 14.875, "learning_rate": 4.159922446144609e-06, "loss": 0.3851, "step": 3075 }, { "epoch": 0.58, "grad_norm": 10.25, "learning_rate": 4.159349626464686e-06, "loss": 0.2204, "step": 3076 }, { "epoch": 0.58, "grad_norm": 9.3125, "learning_rate": 4.158776651026529e-06, "loss": 0.2482, "step": 3077 }, { "epoch": 0.58, "grad_norm": 8.625, "learning_rate": 4.1582035198839195e-06, "loss": 0.2625, "step": 3078 }, { "epoch": 0.58, "grad_norm": 9.6875, "learning_rate": 4.157630233090657e-06, "loss": 0.3106, "step": 3079 }, { "epoch": 0.58, "grad_norm": 13.9375, "learning_rate": 4.1570567907005536e-06, "loss": 0.3625, "step": 3080 }, { "epoch": 0.58, "grad_norm": 11.3125, "learning_rate": 4.156483192767436e-06, "loss": 0.2749, "step": 3081 }, { "epoch": 0.58, "grad_norm": 8.5625, "learning_rate": 4.155909439345148e-06, "loss": 0.2868, "step": 3082 }, { "epoch": 0.58, "grad_norm": 11.0625, "learning_rate": 4.155335530487544e-06, "loss": 0.2917, "step": 3083 }, { "epoch": 0.58, "grad_norm": 9.8125, "learning_rate": 4.154761466248497e-06, "loss": 0.3315, "step": 3084 }, { "epoch": 0.58, "grad_norm": 14.0625, "learning_rate": 4.154187246681891e-06, "loss": 0.3549, "step": 3085 }, { "epoch": 0.58, "grad_norm": 7.59375, "learning_rate": 4.153612871841628e-06, "loss": 0.2068, "step": 3086 }, { "epoch": 0.58, "grad_norm": 9.75, "learning_rate": 4.1530383417816225e-06, "loss": 0.2645, "step": 3087 }, { "epoch": 0.58, "grad_norm": 11.6875, "learning_rate": 4.152463656555802e-06, "loss": 0.3416, "step": 3088 }, { "epoch": 0.58, "grad_norm": 29.125, "learning_rate": 4.151888816218113e-06, "loss": 0.2151, "step": 3089 }, { "epoch": 0.58, "grad_norm": 9.125, "learning_rate": 4.151313820822512e-06, "loss": 0.266, "step": 3090 }, { "epoch": 0.58, "grad_norm": 8.9375, "learning_rate": 4.150738670422974e-06, "loss": 0.2622, "step": 3091 }, { "epoch": 0.58, "grad_norm": 7.78125, "learning_rate": 4.150163365073485e-06, "loss": 0.2369, "step": 3092 }, { "epoch": 0.59, "grad_norm": 14.5625, "learning_rate": 4.1495879048280495e-06, "loss": 0.2343, "step": 3093 }, { "epoch": 0.59, "grad_norm": 9.0625, "learning_rate": 4.149012289740682e-06, "loss": 0.288, "step": 3094 }, { "epoch": 0.59, "grad_norm": 15.375, "learning_rate": 4.1484365198654135e-06, "loss": 0.3193, "step": 3095 }, { "epoch": 0.59, "grad_norm": 9.1875, "learning_rate": 4.147860595256292e-06, "loss": 0.319, "step": 3096 }, { "epoch": 0.59, "grad_norm": 9.4375, "learning_rate": 4.147284515967376e-06, "loss": 0.3192, "step": 3097 }, { "epoch": 0.59, "grad_norm": 10.3125, "learning_rate": 4.146708282052741e-06, "loss": 0.2699, "step": 3098 }, { "epoch": 0.59, "grad_norm": 7.71875, "learning_rate": 4.146131893566477e-06, "loss": 0.264, "step": 3099 }, { "epoch": 0.59, "grad_norm": 9.125, "learning_rate": 4.145555350562688e-06, "loss": 0.3281, "step": 3100 }, { "epoch": 0.59, "grad_norm": 16.0, "learning_rate": 4.14497865309549e-06, "loss": 0.3222, "step": 3101 }, { "epoch": 0.59, "grad_norm": 11.875, "learning_rate": 4.144401801219019e-06, "loss": 0.2921, "step": 3102 }, { "epoch": 0.59, "grad_norm": 11.25, "learning_rate": 4.143824794987419e-06, "loss": 0.2818, "step": 3103 }, { "epoch": 0.59, "grad_norm": 11.0625, "learning_rate": 4.143247634454855e-06, "loss": 0.3062, "step": 3104 }, { "epoch": 0.59, "grad_norm": 10.5625, "learning_rate": 4.142670319675503e-06, "loss": 0.2927, "step": 3105 }, { "epoch": 0.59, "grad_norm": 8.5625, "learning_rate": 4.142092850703552e-06, "loss": 0.2433, "step": 3106 }, { "epoch": 0.59, "grad_norm": 54.75, "learning_rate": 4.141515227593208e-06, "loss": 0.2947, "step": 3107 }, { "epoch": 0.59, "grad_norm": 9.6875, "learning_rate": 4.140937450398692e-06, "loss": 0.3523, "step": 3108 }, { "epoch": 0.59, "grad_norm": 37.75, "learning_rate": 4.140359519174236e-06, "loss": 0.2796, "step": 3109 }, { "epoch": 0.59, "grad_norm": 15.9375, "learning_rate": 4.139781433974092e-06, "loss": 0.328, "step": 3110 }, { "epoch": 0.59, "grad_norm": 10.5625, "learning_rate": 4.139203194852519e-06, "loss": 0.2664, "step": 3111 }, { "epoch": 0.59, "grad_norm": 8.6875, "learning_rate": 4.1386248018637976e-06, "loss": 0.2607, "step": 3112 }, { "epoch": 0.59, "grad_norm": 12.75, "learning_rate": 4.1380462550622194e-06, "loss": 0.2907, "step": 3113 }, { "epoch": 0.59, "grad_norm": 14.125, "learning_rate": 4.137467554502089e-06, "loss": 0.3686, "step": 3114 }, { "epoch": 0.59, "grad_norm": 15.0, "learning_rate": 4.13688870023773e-06, "loss": 0.2614, "step": 3115 }, { "epoch": 0.59, "grad_norm": 11.0625, "learning_rate": 4.1363096923234765e-06, "loss": 0.3348, "step": 3116 }, { "epoch": 0.59, "grad_norm": 9.375, "learning_rate": 4.135730530813678e-06, "loss": 0.2215, "step": 3117 }, { "epoch": 0.59, "grad_norm": 12.0, "learning_rate": 4.1351512157626985e-06, "loss": 0.2794, "step": 3118 }, { "epoch": 0.59, "grad_norm": 7.71875, "learning_rate": 4.134571747224918e-06, "loss": 0.294, "step": 3119 }, { "epoch": 0.59, "grad_norm": 10.9375, "learning_rate": 4.133992125254728e-06, "loss": 0.2764, "step": 3120 }, { "epoch": 0.59, "grad_norm": 10.9375, "learning_rate": 4.133412349906536e-06, "loss": 0.2463, "step": 3121 }, { "epoch": 0.59, "grad_norm": 9.9375, "learning_rate": 4.132832421234764e-06, "loss": 0.2314, "step": 3122 }, { "epoch": 0.59, "grad_norm": 10.625, "learning_rate": 4.132252339293849e-06, "loss": 0.446, "step": 3123 }, { "epoch": 0.59, "grad_norm": 10.625, "learning_rate": 4.131672104138241e-06, "loss": 0.2415, "step": 3124 }, { "epoch": 0.59, "grad_norm": 7.78125, "learning_rate": 4.131091715822404e-06, "loss": 0.2277, "step": 3125 }, { "epoch": 0.59, "grad_norm": 7.84375, "learning_rate": 4.130511174400819e-06, "loss": 0.2451, "step": 3126 }, { "epoch": 0.59, "grad_norm": 9.1875, "learning_rate": 4.129930479927978e-06, "loss": 0.2211, "step": 3127 }, { "epoch": 0.59, "grad_norm": 11.125, "learning_rate": 4.129349632458391e-06, "loss": 0.429, "step": 3128 }, { "epoch": 0.59, "grad_norm": 13.625, "learning_rate": 4.128768632046579e-06, "loss": 0.3619, "step": 3129 }, { "epoch": 0.59, "grad_norm": 11.0625, "learning_rate": 4.12818747874708e-06, "loss": 0.2993, "step": 3130 }, { "epoch": 0.59, "grad_norm": 8.4375, "learning_rate": 4.127606172614444e-06, "loss": 0.2698, "step": 3131 }, { "epoch": 0.59, "grad_norm": 9.75, "learning_rate": 4.127024713703236e-06, "loss": 0.2469, "step": 3132 }, { "epoch": 0.59, "grad_norm": 12.75, "learning_rate": 4.126443102068037e-06, "loss": 0.3489, "step": 3133 }, { "epoch": 0.59, "grad_norm": 10.0, "learning_rate": 4.125861337763441e-06, "loss": 0.2686, "step": 3134 }, { "epoch": 0.59, "grad_norm": 13.9375, "learning_rate": 4.125279420844057e-06, "loss": 0.3069, "step": 3135 }, { "epoch": 0.59, "grad_norm": 20.25, "learning_rate": 4.124697351364507e-06, "loss": 0.3765, "step": 3136 }, { "epoch": 0.59, "grad_norm": 17.875, "learning_rate": 4.124115129379428e-06, "loss": 0.3537, "step": 3137 }, { "epoch": 0.59, "grad_norm": 11.1875, "learning_rate": 4.1235327549434715e-06, "loss": 0.1881, "step": 3138 }, { "epoch": 0.59, "grad_norm": 8.375, "learning_rate": 4.122950228111304e-06, "loss": 0.3056, "step": 3139 }, { "epoch": 0.59, "grad_norm": 7.5625, "learning_rate": 4.122367548937605e-06, "loss": 0.2752, "step": 3140 }, { "epoch": 0.59, "grad_norm": 13.0, "learning_rate": 4.121784717477068e-06, "loss": 0.3238, "step": 3141 }, { "epoch": 0.59, "grad_norm": 14.1875, "learning_rate": 4.121201733784403e-06, "loss": 0.341, "step": 3142 }, { "epoch": 0.59, "grad_norm": 12.0, "learning_rate": 4.1206185979143335e-06, "loss": 0.3466, "step": 3143 }, { "epoch": 0.59, "grad_norm": 11.375, "learning_rate": 4.120035309921595e-06, "loss": 0.3497, "step": 3144 }, { "epoch": 0.59, "grad_norm": 20.125, "learning_rate": 4.1194518698609394e-06, "loss": 0.2791, "step": 3145 }, { "epoch": 0.6, "grad_norm": 8.6875, "learning_rate": 4.118868277787134e-06, "loss": 0.3051, "step": 3146 }, { "epoch": 0.6, "grad_norm": 11.0625, "learning_rate": 4.118284533754957e-06, "loss": 0.285, "step": 3147 }, { "epoch": 0.6, "grad_norm": 14.8125, "learning_rate": 4.117700637819204e-06, "loss": 0.3342, "step": 3148 }, { "epoch": 0.6, "grad_norm": 12.125, "learning_rate": 4.117116590034682e-06, "loss": 0.2913, "step": 3149 }, { "epoch": 0.6, "grad_norm": 11.3125, "learning_rate": 4.116532390456216e-06, "loss": 0.4037, "step": 3150 }, { "epoch": 0.6, "grad_norm": 8.75, "learning_rate": 4.115948039138642e-06, "loss": 0.2673, "step": 3151 }, { "epoch": 0.6, "grad_norm": 9.3125, "learning_rate": 4.1153635361368105e-06, "loss": 0.2642, "step": 3152 }, { "epoch": 0.6, "grad_norm": 11.4375, "learning_rate": 4.114778881505588e-06, "loss": 0.438, "step": 3153 }, { "epoch": 0.6, "grad_norm": 9.0625, "learning_rate": 4.114194075299854e-06, "loss": 0.2465, "step": 3154 }, { "epoch": 0.6, "grad_norm": 10.0625, "learning_rate": 4.113609117574504e-06, "loss": 0.3154, "step": 3155 }, { "epoch": 0.6, "grad_norm": 10.5, "learning_rate": 4.113024008384444e-06, "loss": 0.3041, "step": 3156 }, { "epoch": 0.6, "grad_norm": 10.8125, "learning_rate": 4.112438747784597e-06, "loss": 0.2771, "step": 3157 }, { "epoch": 0.6, "grad_norm": 10.8125, "learning_rate": 4.111853335829901e-06, "loss": 0.2653, "step": 3158 }, { "epoch": 0.6, "grad_norm": 17.75, "learning_rate": 4.111267772575306e-06, "loss": 0.2803, "step": 3159 }, { "epoch": 0.6, "grad_norm": 9.3125, "learning_rate": 4.110682058075777e-06, "loss": 0.3235, "step": 3160 }, { "epoch": 0.6, "grad_norm": 9.625, "learning_rate": 4.110096192386292e-06, "loss": 0.2645, "step": 3161 }, { "epoch": 0.6, "grad_norm": 9.625, "learning_rate": 4.1095101755618475e-06, "loss": 0.2687, "step": 3162 }, { "epoch": 0.6, "grad_norm": 13.5625, "learning_rate": 4.1089240076574485e-06, "loss": 0.3159, "step": 3163 }, { "epoch": 0.6, "grad_norm": 11.125, "learning_rate": 4.1083376887281185e-06, "loss": 0.2301, "step": 3164 }, { "epoch": 0.6, "grad_norm": 11.5, "learning_rate": 4.107751218828893e-06, "loss": 0.2964, "step": 3165 }, { "epoch": 0.6, "grad_norm": 9.125, "learning_rate": 4.107164598014821e-06, "loss": 0.3348, "step": 3166 }, { "epoch": 0.6, "grad_norm": 15.5, "learning_rate": 4.106577826340969e-06, "loss": 0.3282, "step": 3167 }, { "epoch": 0.6, "grad_norm": 8.8125, "learning_rate": 4.105990903862414e-06, "loss": 0.3214, "step": 3168 }, { "epoch": 0.6, "grad_norm": 9.125, "learning_rate": 4.105403830634249e-06, "loss": 0.2673, "step": 3169 }, { "epoch": 0.6, "grad_norm": 12.75, "learning_rate": 4.104816606711582e-06, "loss": 0.3264, "step": 3170 }, { "epoch": 0.6, "grad_norm": 8.0625, "learning_rate": 4.104229232149531e-06, "loss": 0.2892, "step": 3171 }, { "epoch": 0.6, "grad_norm": 10.1875, "learning_rate": 4.103641707003235e-06, "loss": 0.2892, "step": 3172 }, { "epoch": 0.6, "grad_norm": 8.9375, "learning_rate": 4.1030540313278385e-06, "loss": 0.2857, "step": 3173 }, { "epoch": 0.6, "grad_norm": 8.375, "learning_rate": 4.10246620517851e-06, "loss": 0.2967, "step": 3174 }, { "epoch": 0.6, "grad_norm": 9.75, "learning_rate": 4.101878228610424e-06, "loss": 0.2936, "step": 3175 }, { "epoch": 0.6, "grad_norm": 11.3125, "learning_rate": 4.101290101678773e-06, "loss": 0.3047, "step": 3176 }, { "epoch": 0.6, "grad_norm": 14.5, "learning_rate": 4.100701824438762e-06, "loss": 0.4193, "step": 3177 }, { "epoch": 0.6, "grad_norm": 10.0, "learning_rate": 4.100113396945612e-06, "loss": 0.2974, "step": 3178 }, { "epoch": 0.6, "grad_norm": 12.3125, "learning_rate": 4.099524819254555e-06, "loss": 0.2944, "step": 3179 }, { "epoch": 0.6, "grad_norm": 9.125, "learning_rate": 4.098936091420842e-06, "loss": 0.2675, "step": 3180 }, { "epoch": 0.6, "grad_norm": 10.75, "learning_rate": 4.098347213499733e-06, "loss": 0.328, "step": 3181 }, { "epoch": 0.6, "grad_norm": 8.625, "learning_rate": 4.097758185546504e-06, "loss": 0.2806, "step": 3182 }, { "epoch": 0.6, "grad_norm": 12.9375, "learning_rate": 4.097169007616446e-06, "loss": 0.2676, "step": 3183 }, { "epoch": 0.6, "grad_norm": 7.8125, "learning_rate": 4.096579679764864e-06, "loss": 0.2024, "step": 3184 }, { "epoch": 0.6, "grad_norm": 13.875, "learning_rate": 4.095990202047076e-06, "loss": 0.2819, "step": 3185 }, { "epoch": 0.6, "grad_norm": 9.9375, "learning_rate": 4.095400574518414e-06, "loss": 0.2272, "step": 3186 }, { "epoch": 0.6, "grad_norm": 13.9375, "learning_rate": 4.094810797234225e-06, "loss": 0.2713, "step": 3187 }, { "epoch": 0.6, "grad_norm": 13.8125, "learning_rate": 4.094220870249869e-06, "loss": 0.274, "step": 3188 }, { "epoch": 0.6, "grad_norm": 11.625, "learning_rate": 4.093630793620722e-06, "loss": 0.2972, "step": 3189 }, { "epoch": 0.6, "grad_norm": 13.1875, "learning_rate": 4.093040567402172e-06, "loss": 0.3367, "step": 3190 }, { "epoch": 0.6, "grad_norm": 9.5625, "learning_rate": 4.092450191649621e-06, "loss": 0.2533, "step": 3191 }, { "epoch": 0.6, "grad_norm": 10.125, "learning_rate": 4.091859666418488e-06, "loss": 0.2621, "step": 3192 }, { "epoch": 0.6, "grad_norm": 14.0, "learning_rate": 4.091268991764202e-06, "loss": 0.3004, "step": 3193 }, { "epoch": 0.6, "grad_norm": 13.4375, "learning_rate": 4.090678167742209e-06, "loss": 0.3823, "step": 3194 }, { "epoch": 0.6, "grad_norm": 9.6875, "learning_rate": 4.090087194407967e-06, "loss": 0.3106, "step": 3195 }, { "epoch": 0.6, "grad_norm": 8.625, "learning_rate": 4.089496071816948e-06, "loss": 0.2943, "step": 3196 }, { "epoch": 0.6, "grad_norm": 13.6875, "learning_rate": 4.0889048000246414e-06, "loss": 0.3658, "step": 3197 }, { "epoch": 0.6, "grad_norm": 11.625, "learning_rate": 4.088313379086546e-06, "loss": 0.3098, "step": 3198 }, { "epoch": 0.61, "grad_norm": 10.25, "learning_rate": 4.087721809058178e-06, "loss": 0.2728, "step": 3199 }, { "epoch": 0.61, "grad_norm": 7.8125, "learning_rate": 4.087130089995066e-06, "loss": 0.2838, "step": 3200 }, { "epoch": 0.61, "grad_norm": 12.75, "learning_rate": 4.086538221952753e-06, "loss": 0.2687, "step": 3201 }, { "epoch": 0.61, "grad_norm": 13.3125, "learning_rate": 4.085946204986796e-06, "loss": 0.3025, "step": 3202 }, { "epoch": 0.61, "grad_norm": 7.5625, "learning_rate": 4.085354039152765e-06, "loss": 0.2204, "step": 3203 }, { "epoch": 0.61, "grad_norm": 9.125, "learning_rate": 4.084761724506245e-06, "loss": 0.2297, "step": 3204 }, { "epoch": 0.61, "grad_norm": 15.125, "learning_rate": 4.084169261102837e-06, "loss": 0.3668, "step": 3205 }, { "epoch": 0.61, "grad_norm": 11.375, "learning_rate": 4.0835766489981506e-06, "loss": 0.2664, "step": 3206 }, { "epoch": 0.61, "grad_norm": 8.875, "learning_rate": 4.082983888247815e-06, "loss": 0.2873, "step": 3207 }, { "epoch": 0.61, "grad_norm": 17.125, "learning_rate": 4.08239097890747e-06, "loss": 0.2505, "step": 3208 }, { "epoch": 0.61, "grad_norm": 13.4375, "learning_rate": 4.08179792103277e-06, "loss": 0.2593, "step": 3209 }, { "epoch": 0.61, "grad_norm": 13.5625, "learning_rate": 4.081204714679382e-06, "loss": 0.2679, "step": 3210 }, { "epoch": 0.61, "grad_norm": 12.5, "learning_rate": 4.080611359902993e-06, "loss": 0.2776, "step": 3211 }, { "epoch": 0.61, "grad_norm": 10.0, "learning_rate": 4.080017856759295e-06, "loss": 0.3482, "step": 3212 }, { "epoch": 0.61, "grad_norm": 10.5, "learning_rate": 4.079424205304e-06, "loss": 0.2957, "step": 3213 }, { "epoch": 0.61, "grad_norm": 13.5, "learning_rate": 4.078830405592833e-06, "loss": 0.2965, "step": 3214 }, { "epoch": 0.61, "grad_norm": 9.125, "learning_rate": 4.078236457681532e-06, "loss": 0.2635, "step": 3215 }, { "epoch": 0.61, "grad_norm": 12.125, "learning_rate": 4.0776423616258485e-06, "loss": 0.2523, "step": 3216 }, { "epoch": 0.61, "grad_norm": 9.9375, "learning_rate": 4.077048117481549e-06, "loss": 0.3373, "step": 3217 }, { "epoch": 0.61, "grad_norm": 15.625, "learning_rate": 4.076453725304413e-06, "loss": 0.2899, "step": 3218 }, { "epoch": 0.61, "grad_norm": 9.1875, "learning_rate": 4.075859185150235e-06, "loss": 0.2131, "step": 3219 }, { "epoch": 0.61, "grad_norm": 9.8125, "learning_rate": 4.075264497074821e-06, "loss": 0.2922, "step": 3220 }, { "epoch": 0.61, "grad_norm": 13.625, "learning_rate": 4.074669661133996e-06, "loss": 0.3247, "step": 3221 }, { "epoch": 0.61, "grad_norm": 16.75, "learning_rate": 4.074074677383592e-06, "loss": 0.2654, "step": 3222 }, { "epoch": 0.61, "grad_norm": 12.3125, "learning_rate": 4.07347954587946e-06, "loss": 0.3622, "step": 3223 }, { "epoch": 0.61, "grad_norm": 13.0, "learning_rate": 4.072884266677463e-06, "loss": 0.3509, "step": 3224 }, { "epoch": 0.61, "grad_norm": 8.0, "learning_rate": 4.072288839833478e-06, "loss": 0.3207, "step": 3225 }, { "epoch": 0.61, "grad_norm": 10.8125, "learning_rate": 4.071693265403396e-06, "loss": 0.3252, "step": 3226 }, { "epoch": 0.61, "grad_norm": 8.0625, "learning_rate": 4.071097543443122e-06, "loss": 0.2692, "step": 3227 }, { "epoch": 0.61, "grad_norm": 17.625, "learning_rate": 4.070501674008575e-06, "loss": 0.2777, "step": 3228 }, { "epoch": 0.61, "grad_norm": 12.125, "learning_rate": 4.069905657155686e-06, "loss": 0.2801, "step": 3229 }, { "epoch": 0.61, "grad_norm": 10.9375, "learning_rate": 4.0693094929404025e-06, "loss": 0.308, "step": 3230 }, { "epoch": 0.61, "grad_norm": 11.3125, "learning_rate": 4.068713181418685e-06, "loss": 0.3509, "step": 3231 }, { "epoch": 0.61, "grad_norm": 12.5625, "learning_rate": 4.068116722646507e-06, "loss": 0.2637, "step": 3232 }, { "epoch": 0.61, "grad_norm": 11.25, "learning_rate": 4.067520116679856e-06, "loss": 0.3186, "step": 3233 }, { "epoch": 0.61, "grad_norm": 11.0, "learning_rate": 4.066923363574735e-06, "loss": 0.2379, "step": 3234 }, { "epoch": 0.61, "grad_norm": 17.375, "learning_rate": 4.066326463387157e-06, "loss": 0.3128, "step": 3235 }, { "epoch": 0.61, "grad_norm": 10.9375, "learning_rate": 4.065729416173154e-06, "loss": 0.2583, "step": 3236 }, { "epoch": 0.61, "grad_norm": 13.6875, "learning_rate": 4.065132221988767e-06, "loss": 0.3401, "step": 3237 }, { "epoch": 0.61, "grad_norm": 10.9375, "learning_rate": 4.064534880890053e-06, "loss": 0.2535, "step": 3238 }, { "epoch": 0.61, "grad_norm": 20.875, "learning_rate": 4.063937392933084e-06, "loss": 0.2882, "step": 3239 }, { "epoch": 0.61, "grad_norm": 10.5625, "learning_rate": 4.0633397581739434e-06, "loss": 0.3094, "step": 3240 }, { "epoch": 0.61, "grad_norm": 11.8125, "learning_rate": 4.062741976668729e-06, "loss": 0.2341, "step": 3241 }, { "epoch": 0.61, "grad_norm": 11.8125, "learning_rate": 4.062144048473555e-06, "loss": 0.2614, "step": 3242 }, { "epoch": 0.61, "grad_norm": 9.625, "learning_rate": 4.0615459736445444e-06, "loss": 0.2765, "step": 3243 }, { "epoch": 0.61, "grad_norm": 13.25, "learning_rate": 4.060947752237838e-06, "loss": 0.2994, "step": 3244 }, { "epoch": 0.61, "grad_norm": 9.875, "learning_rate": 4.06034938430959e-06, "loss": 0.2643, "step": 3245 }, { "epoch": 0.61, "grad_norm": 10.375, "learning_rate": 4.0597508699159645e-06, "loss": 0.2669, "step": 3246 }, { "epoch": 0.61, "grad_norm": 11.125, "learning_rate": 4.059152209113146e-06, "loss": 0.3078, "step": 3247 }, { "epoch": 0.61, "grad_norm": 10.25, "learning_rate": 4.058553401957326e-06, "loss": 0.2339, "step": 3248 }, { "epoch": 0.61, "grad_norm": 11.125, "learning_rate": 4.057954448504715e-06, "loss": 0.2476, "step": 3249 }, { "epoch": 0.61, "grad_norm": 9.8125, "learning_rate": 4.057355348811533e-06, "loss": 0.3453, "step": 3250 }, { "epoch": 0.61, "grad_norm": 12.8125, "learning_rate": 4.0567561029340165e-06, "loss": 0.3051, "step": 3251 }, { "epoch": 0.62, "grad_norm": 12.5, "learning_rate": 4.056156710928415e-06, "loss": 0.3371, "step": 3252 }, { "epoch": 0.62, "grad_norm": 11.5, "learning_rate": 4.055557172850993e-06, "loss": 0.2483, "step": 3253 }, { "epoch": 0.62, "grad_norm": 13.3125, "learning_rate": 4.054957488758024e-06, "loss": 0.3032, "step": 3254 }, { "epoch": 0.62, "grad_norm": 10.9375, "learning_rate": 4.054357658705803e-06, "loss": 0.2457, "step": 3255 }, { "epoch": 0.62, "grad_norm": 10.9375, "learning_rate": 4.05375768275063e-06, "loss": 0.2954, "step": 3256 }, { "epoch": 0.62, "grad_norm": 10.375, "learning_rate": 4.053157560948826e-06, "loss": 0.2187, "step": 3257 }, { "epoch": 0.62, "grad_norm": 9.375, "learning_rate": 4.052557293356721e-06, "loss": 0.2913, "step": 3258 }, { "epoch": 0.62, "grad_norm": 11.3125, "learning_rate": 4.051956880030661e-06, "loss": 0.2696, "step": 3259 }, { "epoch": 0.62, "grad_norm": 10.6875, "learning_rate": 4.051356321027005e-06, "loss": 0.2995, "step": 3260 }, { "epoch": 0.62, "grad_norm": 11.0625, "learning_rate": 4.050755616402126e-06, "loss": 0.267, "step": 3261 }, { "epoch": 0.62, "grad_norm": 10.0625, "learning_rate": 4.050154766212409e-06, "loss": 0.2387, "step": 3262 }, { "epoch": 0.62, "grad_norm": 10.625, "learning_rate": 4.049553770514255e-06, "loss": 0.2731, "step": 3263 }, { "epoch": 0.62, "grad_norm": 12.9375, "learning_rate": 4.048952629364079e-06, "loss": 0.2181, "step": 3264 }, { "epoch": 0.62, "grad_norm": 7.90625, "learning_rate": 4.048351342818307e-06, "loss": 0.2465, "step": 3265 }, { "epoch": 0.62, "grad_norm": 12.125, "learning_rate": 4.047749910933378e-06, "loss": 0.2786, "step": 3266 }, { "epoch": 0.62, "grad_norm": 9.3125, "learning_rate": 4.0471483337657505e-06, "loss": 0.252, "step": 3267 }, { "epoch": 0.62, "grad_norm": 13.8125, "learning_rate": 4.04654661137189e-06, "loss": 0.3777, "step": 3268 }, { "epoch": 0.62, "grad_norm": 6.75, "learning_rate": 4.045944743808279e-06, "loss": 0.2257, "step": 3269 }, { "epoch": 0.62, "grad_norm": 7.84375, "learning_rate": 4.045342731131414e-06, "loss": 0.2743, "step": 3270 }, { "epoch": 0.62, "grad_norm": 17.125, "learning_rate": 4.044740573397803e-06, "loss": 0.3115, "step": 3271 }, { "epoch": 0.62, "grad_norm": 11.6875, "learning_rate": 4.04413827066397e-06, "loss": 0.2534, "step": 3272 }, { "epoch": 0.62, "grad_norm": 10.75, "learning_rate": 4.04353582298645e-06, "loss": 0.2997, "step": 3273 }, { "epoch": 0.62, "grad_norm": 11.3125, "learning_rate": 4.042933230421794e-06, "loss": 0.2884, "step": 3274 }, { "epoch": 0.62, "grad_norm": 10.625, "learning_rate": 4.042330493026565e-06, "loss": 0.2964, "step": 3275 }, { "epoch": 0.62, "grad_norm": 10.25, "learning_rate": 4.04172761085734e-06, "loss": 0.2853, "step": 3276 }, { "epoch": 0.62, "grad_norm": 14.5625, "learning_rate": 4.041124583970711e-06, "loss": 0.3115, "step": 3277 }, { "epoch": 0.62, "grad_norm": 10.5, "learning_rate": 4.040521412423281e-06, "loss": 0.3082, "step": 3278 }, { "epoch": 0.62, "grad_norm": 12.5, "learning_rate": 4.039918096271668e-06, "loss": 0.3184, "step": 3279 }, { "epoch": 0.62, "grad_norm": 19.25, "learning_rate": 4.039314635572504e-06, "loss": 0.3486, "step": 3280 }, { "epoch": 0.62, "grad_norm": 14.0625, "learning_rate": 4.038711030382434e-06, "loss": 0.3555, "step": 3281 }, { "epoch": 0.62, "grad_norm": 10.3125, "learning_rate": 4.038107280758117e-06, "loss": 0.3199, "step": 3282 }, { "epoch": 0.62, "grad_norm": 11.0625, "learning_rate": 4.037503386756224e-06, "loss": 0.2768, "step": 3283 }, { "epoch": 0.62, "grad_norm": 12.25, "learning_rate": 4.036899348433442e-06, "loss": 0.3465, "step": 3284 }, { "epoch": 0.62, "grad_norm": 19.0, "learning_rate": 4.03629516584647e-06, "loss": 0.3797, "step": 3285 }, { "epoch": 0.62, "grad_norm": 11.25, "learning_rate": 4.03569083905202e-06, "loss": 0.3504, "step": 3286 }, { "epoch": 0.62, "grad_norm": 11.4375, "learning_rate": 4.0350863681068196e-06, "loss": 0.2545, "step": 3287 }, { "epoch": 0.62, "grad_norm": 12.0, "learning_rate": 4.034481753067607e-06, "loss": 0.3609, "step": 3288 }, { "epoch": 0.62, "grad_norm": 10.875, "learning_rate": 4.033876993991138e-06, "loss": 0.2695, "step": 3289 }, { "epoch": 0.62, "grad_norm": 7.875, "learning_rate": 4.033272090934178e-06, "loss": 0.2397, "step": 3290 }, { "epoch": 0.62, "grad_norm": 12.5625, "learning_rate": 4.032667043953508e-06, "loss": 0.2936, "step": 3291 }, { "epoch": 0.62, "grad_norm": 8.0625, "learning_rate": 4.032061853105921e-06, "loss": 0.3168, "step": 3292 }, { "epoch": 0.62, "grad_norm": 9.5625, "learning_rate": 4.031456518448225e-06, "loss": 0.2399, "step": 3293 }, { "epoch": 0.62, "grad_norm": 12.625, "learning_rate": 4.030851040037242e-06, "loss": 0.3436, "step": 3294 }, { "epoch": 0.62, "grad_norm": 11.0625, "learning_rate": 4.030245417929805e-06, "loss": 0.3368, "step": 3295 }, { "epoch": 0.62, "grad_norm": 11.8125, "learning_rate": 4.029639652182762e-06, "loss": 0.2363, "step": 3296 }, { "epoch": 0.62, "grad_norm": 11.1875, "learning_rate": 4.029033742852975e-06, "loss": 0.2637, "step": 3297 }, { "epoch": 0.62, "grad_norm": 7.875, "learning_rate": 4.0284276899973205e-06, "loss": 0.2248, "step": 3298 }, { "epoch": 0.62, "grad_norm": 7.34375, "learning_rate": 4.0278214936726845e-06, "loss": 0.2679, "step": 3299 }, { "epoch": 0.62, "grad_norm": 13.4375, "learning_rate": 4.027215153935969e-06, "loss": 0.3087, "step": 3300 }, { "epoch": 0.62, "grad_norm": 12.1875, "learning_rate": 4.026608670844091e-06, "loss": 0.3114, "step": 3301 }, { "epoch": 0.62, "grad_norm": 11.375, "learning_rate": 4.026002044453977e-06, "loss": 0.2598, "step": 3302 }, { "epoch": 0.62, "grad_norm": 11.375, "learning_rate": 4.025395274822571e-06, "loss": 0.2811, "step": 3303 }, { "epoch": 0.63, "grad_norm": 10.1875, "learning_rate": 4.024788362006829e-06, "loss": 0.3381, "step": 3304 }, { "epoch": 0.63, "grad_norm": 10.1875, "learning_rate": 4.024181306063718e-06, "loss": 0.2953, "step": 3305 }, { "epoch": 0.63, "grad_norm": 18.25, "learning_rate": 4.023574107050222e-06, "loss": 0.2622, "step": 3306 }, { "epoch": 0.63, "grad_norm": 9.625, "learning_rate": 4.022966765023337e-06, "loss": 0.2389, "step": 3307 }, { "epoch": 0.63, "grad_norm": 14.3125, "learning_rate": 4.022359280040073e-06, "loss": 0.3059, "step": 3308 }, { "epoch": 0.63, "grad_norm": 17.375, "learning_rate": 4.0217516521574504e-06, "loss": 0.3083, "step": 3309 }, { "epoch": 0.63, "grad_norm": 10.3125, "learning_rate": 4.021143881432508e-06, "loss": 0.2935, "step": 3310 }, { "epoch": 0.63, "grad_norm": 8.625, "learning_rate": 4.020535967922295e-06, "loss": 0.2652, "step": 3311 }, { "epoch": 0.63, "grad_norm": 9.8125, "learning_rate": 4.019927911683872e-06, "loss": 0.3028, "step": 3312 }, { "epoch": 0.63, "grad_norm": 11.0, "learning_rate": 4.019319712774319e-06, "loss": 0.3775, "step": 3313 }, { "epoch": 0.63, "grad_norm": 10.875, "learning_rate": 4.0187113712507235e-06, "loss": 0.2913, "step": 3314 }, { "epoch": 0.63, "grad_norm": 7.78125, "learning_rate": 4.018102887170189e-06, "loss": 0.2574, "step": 3315 }, { "epoch": 0.63, "grad_norm": 20.75, "learning_rate": 4.017494260589832e-06, "loss": 0.2756, "step": 3316 }, { "epoch": 0.63, "grad_norm": 12.125, "learning_rate": 4.016885491566784e-06, "loss": 0.316, "step": 3317 }, { "epoch": 0.63, "grad_norm": 12.3125, "learning_rate": 4.016276580158187e-06, "loss": 0.2993, "step": 3318 }, { "epoch": 0.63, "grad_norm": 12.125, "learning_rate": 4.015667526421198e-06, "loss": 0.2657, "step": 3319 }, { "epoch": 0.63, "grad_norm": 9.875, "learning_rate": 4.015058330412986e-06, "loss": 0.2926, "step": 3320 }, { "epoch": 0.63, "grad_norm": 10.5625, "learning_rate": 4.014448992190737e-06, "loss": 0.3149, "step": 3321 }, { "epoch": 0.63, "grad_norm": 47.75, "learning_rate": 4.013839511811646e-06, "loss": 0.3349, "step": 3322 }, { "epoch": 0.63, "grad_norm": 9.125, "learning_rate": 4.013229889332923e-06, "loss": 0.3099, "step": 3323 }, { "epoch": 0.63, "grad_norm": 9.5625, "learning_rate": 4.012620124811791e-06, "loss": 0.3331, "step": 3324 }, { "epoch": 0.63, "grad_norm": 16.25, "learning_rate": 4.012010218305489e-06, "loss": 0.3617, "step": 3325 }, { "epoch": 0.63, "grad_norm": 10.0625, "learning_rate": 4.011400169871266e-06, "loss": 0.2476, "step": 3326 }, { "epoch": 0.63, "grad_norm": 8.3125, "learning_rate": 4.010789979566384e-06, "loss": 0.3013, "step": 3327 }, { "epoch": 0.63, "grad_norm": 9.875, "learning_rate": 4.010179647448122e-06, "loss": 0.2298, "step": 3328 }, { "epoch": 0.63, "grad_norm": 12.5625, "learning_rate": 4.009569173573767e-06, "loss": 0.2703, "step": 3329 }, { "epoch": 0.63, "grad_norm": 12.625, "learning_rate": 4.008958558000627e-06, "loss": 0.2505, "step": 3330 }, { "epoch": 0.63, "grad_norm": 10.25, "learning_rate": 4.0083478007860135e-06, "loss": 0.3087, "step": 3331 }, { "epoch": 0.63, "grad_norm": 10.375, "learning_rate": 4.007736901987261e-06, "loss": 0.2609, "step": 3332 }, { "epoch": 0.63, "grad_norm": 7.5625, "learning_rate": 4.00712586166171e-06, "loss": 0.2674, "step": 3333 }, { "epoch": 0.63, "grad_norm": 7.78125, "learning_rate": 4.006514679866717e-06, "loss": 0.2668, "step": 3334 }, { "epoch": 0.63, "grad_norm": 9.625, "learning_rate": 4.005903356659654e-06, "loss": 0.273, "step": 3335 }, { "epoch": 0.63, "grad_norm": 13.5625, "learning_rate": 4.005291892097902e-06, "loss": 0.3664, "step": 3336 }, { "epoch": 0.63, "grad_norm": 9.0625, "learning_rate": 4.004680286238859e-06, "loss": 0.2654, "step": 3337 }, { "epoch": 0.63, "grad_norm": 15.0, "learning_rate": 4.004068539139933e-06, "loss": 0.3215, "step": 3338 }, { "epoch": 0.63, "grad_norm": 13.1875, "learning_rate": 4.003456650858548e-06, "loss": 0.3123, "step": 3339 }, { "epoch": 0.63, "grad_norm": 15.25, "learning_rate": 4.002844621452139e-06, "loss": 0.3405, "step": 3340 }, { "epoch": 0.63, "grad_norm": 9.875, "learning_rate": 4.002232450978157e-06, "loss": 0.271, "step": 3341 }, { "epoch": 0.63, "grad_norm": 11.3125, "learning_rate": 4.0016201394940636e-06, "loss": 0.1976, "step": 3342 }, { "epoch": 0.63, "grad_norm": 14.9375, "learning_rate": 4.001007687057336e-06, "loss": 0.237, "step": 3343 }, { "epoch": 0.63, "grad_norm": 9.3125, "learning_rate": 4.00039509372546e-06, "loss": 0.2938, "step": 3344 }, { "epoch": 0.63, "grad_norm": 10.0, "learning_rate": 3.999782359555942e-06, "loss": 0.2885, "step": 3345 }, { "epoch": 0.63, "grad_norm": 10.625, "learning_rate": 3.999169484606293e-06, "loss": 0.2918, "step": 3346 }, { "epoch": 0.63, "grad_norm": 9.5, "learning_rate": 3.998556468934047e-06, "loss": 0.2718, "step": 3347 }, { "epoch": 0.63, "grad_norm": 10.25, "learning_rate": 3.997943312596742e-06, "loss": 0.3038, "step": 3348 }, { "epoch": 0.63, "grad_norm": 11.8125, "learning_rate": 3.997330015651935e-06, "loss": 0.333, "step": 3349 }, { "epoch": 0.63, "grad_norm": 10.375, "learning_rate": 3.996716578157194e-06, "loss": 0.2527, "step": 3350 }, { "epoch": 0.63, "grad_norm": 13.1875, "learning_rate": 3.9961030001701e-06, "loss": 0.3386, "step": 3351 }, { "epoch": 0.63, "grad_norm": 13.75, "learning_rate": 3.995489281748248e-06, "loss": 0.3125, "step": 3352 }, { "epoch": 0.63, "grad_norm": 6.9375, "learning_rate": 3.994875422949247e-06, "loss": 0.2361, "step": 3353 }, { "epoch": 0.63, "grad_norm": 10.875, "learning_rate": 3.994261423830716e-06, "loss": 0.2834, "step": 3354 }, { "epoch": 0.63, "grad_norm": 11.25, "learning_rate": 3.993647284450291e-06, "loss": 0.3065, "step": 3355 }, { "epoch": 0.63, "grad_norm": 11.5, "learning_rate": 3.99303300486562e-06, "loss": 0.3282, "step": 3356 }, { "epoch": 0.64, "grad_norm": 13.9375, "learning_rate": 3.992418585134362e-06, "loss": 0.2709, "step": 3357 }, { "epoch": 0.64, "grad_norm": 12.75, "learning_rate": 3.991804025314191e-06, "loss": 0.3005, "step": 3358 }, { "epoch": 0.64, "grad_norm": 8.1875, "learning_rate": 3.991189325462794e-06, "loss": 0.2807, "step": 3359 }, { "epoch": 0.64, "grad_norm": 10.1875, "learning_rate": 3.990574485637872e-06, "loss": 0.2653, "step": 3360 }, { "epoch": 0.64, "grad_norm": 10.0, "learning_rate": 3.989959505897137e-06, "loss": 0.2583, "step": 3361 }, { "epoch": 0.64, "grad_norm": 10.8125, "learning_rate": 3.9893443862983165e-06, "loss": 0.2685, "step": 3362 }, { "epoch": 0.64, "grad_norm": 8.4375, "learning_rate": 3.988729126899148e-06, "loss": 0.2901, "step": 3363 }, { "epoch": 0.64, "grad_norm": 8.5625, "learning_rate": 3.988113727757386e-06, "loss": 0.2591, "step": 3364 }, { "epoch": 0.64, "grad_norm": 7.8125, "learning_rate": 3.987498188930795e-06, "loss": 0.2715, "step": 3365 }, { "epoch": 0.64, "grad_norm": 10.4375, "learning_rate": 3.986882510477156e-06, "loss": 0.2788, "step": 3366 }, { "epoch": 0.64, "grad_norm": 8.3125, "learning_rate": 3.986266692454258e-06, "loss": 0.3114, "step": 3367 }, { "epoch": 0.64, "grad_norm": 11.25, "learning_rate": 3.985650734919907e-06, "loss": 0.255, "step": 3368 }, { "epoch": 0.64, "grad_norm": 8.9375, "learning_rate": 3.985034637931922e-06, "loss": 0.3242, "step": 3369 }, { "epoch": 0.64, "grad_norm": 10.125, "learning_rate": 3.984418401548133e-06, "loss": 0.3607, "step": 3370 }, { "epoch": 0.64, "grad_norm": 10.9375, "learning_rate": 3.983802025826385e-06, "loss": 0.2974, "step": 3371 }, { "epoch": 0.64, "grad_norm": 13.125, "learning_rate": 3.983185510824535e-06, "loss": 0.2802, "step": 3372 }, { "epoch": 0.64, "grad_norm": 9.4375, "learning_rate": 3.982568856600454e-06, "loss": 0.2322, "step": 3373 }, { "epoch": 0.64, "grad_norm": 7.875, "learning_rate": 3.9819520632120255e-06, "loss": 0.2364, "step": 3374 }, { "epoch": 0.64, "grad_norm": 13.0, "learning_rate": 3.981335130717145e-06, "loss": 0.2962, "step": 3375 }, { "epoch": 0.64, "grad_norm": 10.375, "learning_rate": 3.980718059173724e-06, "loss": 0.2682, "step": 3376 }, { "epoch": 0.64, "grad_norm": 10.625, "learning_rate": 3.9801008486396834e-06, "loss": 0.3169, "step": 3377 }, { "epoch": 0.64, "grad_norm": 9.25, "learning_rate": 3.97948349917296e-06, "loss": 0.2249, "step": 3378 }, { "epoch": 0.64, "grad_norm": 16.375, "learning_rate": 3.978866010831501e-06, "loss": 0.3393, "step": 3379 }, { "epoch": 0.64, "grad_norm": 10.5625, "learning_rate": 3.97824838367327e-06, "loss": 0.2865, "step": 3380 }, { "epoch": 0.64, "grad_norm": 9.1875, "learning_rate": 3.9776306177562414e-06, "loss": 0.2785, "step": 3381 }, { "epoch": 0.64, "grad_norm": 8.3125, "learning_rate": 3.977012713138403e-06, "loss": 0.2395, "step": 3382 }, { "epoch": 0.64, "grad_norm": 12.125, "learning_rate": 3.976394669877755e-06, "loss": 0.301, "step": 3383 }, { "epoch": 0.64, "grad_norm": 9.0, "learning_rate": 3.975776488032311e-06, "loss": 0.2954, "step": 3384 }, { "epoch": 0.64, "grad_norm": 12.5, "learning_rate": 3.9751581676601004e-06, "loss": 0.343, "step": 3385 }, { "epoch": 0.64, "grad_norm": 16.0, "learning_rate": 3.97453970881916e-06, "loss": 0.2823, "step": 3386 }, { "epoch": 0.64, "grad_norm": 9.0625, "learning_rate": 3.9739211115675445e-06, "loss": 0.2302, "step": 3387 }, { "epoch": 0.64, "grad_norm": 14.0625, "learning_rate": 3.973302375963319e-06, "loss": 0.3161, "step": 3388 }, { "epoch": 0.64, "grad_norm": 22.375, "learning_rate": 3.972683502064564e-06, "loss": 0.3442, "step": 3389 }, { "epoch": 0.64, "grad_norm": 8.1875, "learning_rate": 3.9720644899293695e-06, "loss": 0.2774, "step": 3390 }, { "epoch": 0.64, "grad_norm": 10.9375, "learning_rate": 3.97144533961584e-06, "loss": 0.2712, "step": 3391 }, { "epoch": 0.64, "grad_norm": 16.0, "learning_rate": 3.970826051182096e-06, "loss": 0.3121, "step": 3392 }, { "epoch": 0.64, "grad_norm": 12.3125, "learning_rate": 3.970206624686264e-06, "loss": 0.2614, "step": 3393 }, { "epoch": 0.64, "grad_norm": 10.0625, "learning_rate": 3.969587060186493e-06, "loss": 0.2693, "step": 3394 }, { "epoch": 0.64, "grad_norm": 9.4375, "learning_rate": 3.968967357740936e-06, "loss": 0.2504, "step": 3395 }, { "epoch": 0.64, "grad_norm": 12.625, "learning_rate": 3.9683475174077634e-06, "loss": 0.3011, "step": 3396 }, { "epoch": 0.64, "grad_norm": 11.6875, "learning_rate": 3.967727539245158e-06, "loss": 0.2052, "step": 3397 }, { "epoch": 0.64, "grad_norm": 13.1875, "learning_rate": 3.967107423311315e-06, "loss": 0.2955, "step": 3398 }, { "epoch": 0.64, "grad_norm": 8.4375, "learning_rate": 3.9664871696644436e-06, "loss": 0.2483, "step": 3399 }, { "epoch": 0.64, "grad_norm": 8.0625, "learning_rate": 3.965866778362765e-06, "loss": 0.3133, "step": 3400 }, { "epoch": 0.64, "grad_norm": 8.5, "learning_rate": 3.965246249464511e-06, "loss": 0.2264, "step": 3401 }, { "epoch": 0.64, "grad_norm": 15.4375, "learning_rate": 3.964625583027934e-06, "loss": 0.3259, "step": 3402 }, { "epoch": 0.64, "grad_norm": 8.9375, "learning_rate": 3.964004779111289e-06, "loss": 0.3107, "step": 3403 }, { "epoch": 0.64, "grad_norm": 10.625, "learning_rate": 3.963383837772852e-06, "loss": 0.2523, "step": 3404 }, { "epoch": 0.64, "grad_norm": 8.375, "learning_rate": 3.962762759070908e-06, "loss": 0.2876, "step": 3405 }, { "epoch": 0.64, "grad_norm": 9.9375, "learning_rate": 3.9621415430637555e-06, "loss": 0.2424, "step": 3406 }, { "epoch": 0.64, "grad_norm": 20.625, "learning_rate": 3.961520189809708e-06, "loss": 0.3332, "step": 3407 }, { "epoch": 0.64, "grad_norm": 20.75, "learning_rate": 3.960898699367087e-06, "loss": 0.3077, "step": 3408 }, { "epoch": 0.64, "grad_norm": 13.0, "learning_rate": 3.960277071794233e-06, "loss": 0.2591, "step": 3409 }, { "epoch": 0.65, "grad_norm": 10.25, "learning_rate": 3.959655307149494e-06, "loss": 0.3365, "step": 3410 }, { "epoch": 0.65, "grad_norm": 12.0625, "learning_rate": 3.959033405491234e-06, "loss": 0.2695, "step": 3411 }, { "epoch": 0.65, "grad_norm": 14.3125, "learning_rate": 3.958411366877829e-06, "loss": 0.2792, "step": 3412 }, { "epoch": 0.65, "grad_norm": 9.5625, "learning_rate": 3.95778919136767e-06, "loss": 0.3062, "step": 3413 }, { "epoch": 0.65, "grad_norm": 7.6875, "learning_rate": 3.9571668790191555e-06, "loss": 0.1937, "step": 3414 }, { "epoch": 0.65, "grad_norm": 25.375, "learning_rate": 3.956544429890701e-06, "loss": 0.3111, "step": 3415 }, { "epoch": 0.65, "grad_norm": 10.75, "learning_rate": 3.955921844040736e-06, "loss": 0.2882, "step": 3416 }, { "epoch": 0.65, "grad_norm": 12.8125, "learning_rate": 3.955299121527698e-06, "loss": 0.2782, "step": 3417 }, { "epoch": 0.65, "grad_norm": 10.9375, "learning_rate": 3.954676262410043e-06, "loss": 0.2964, "step": 3418 }, { "epoch": 0.65, "grad_norm": 9.5, "learning_rate": 3.954053266746234e-06, "loss": 0.3313, "step": 3419 }, { "epoch": 0.65, "grad_norm": 16.25, "learning_rate": 3.953430134594752e-06, "loss": 0.2523, "step": 3420 }, { "epoch": 0.65, "grad_norm": 8.5625, "learning_rate": 3.952806866014088e-06, "loss": 0.2585, "step": 3421 }, { "epoch": 0.65, "grad_norm": 10.0625, "learning_rate": 3.952183461062745e-06, "loss": 0.3155, "step": 3422 }, { "epoch": 0.65, "grad_norm": 10.3125, "learning_rate": 3.951559919799241e-06, "loss": 0.2374, "step": 3423 }, { "epoch": 0.65, "grad_norm": 8.625, "learning_rate": 3.950936242282107e-06, "loss": 0.2308, "step": 3424 }, { "epoch": 0.65, "grad_norm": 10.3125, "learning_rate": 3.950312428569885e-06, "loss": 0.2744, "step": 3425 }, { "epoch": 0.65, "grad_norm": 12.6875, "learning_rate": 3.949688478721131e-06, "loss": 0.2702, "step": 3426 }, { "epoch": 0.65, "grad_norm": 9.625, "learning_rate": 3.949064392794413e-06, "loss": 0.2781, "step": 3427 }, { "epoch": 0.65, "grad_norm": 12.1875, "learning_rate": 3.94844017084831e-06, "loss": 0.2988, "step": 3428 }, { "epoch": 0.65, "grad_norm": 8.4375, "learning_rate": 3.94781581294142e-06, "loss": 0.286, "step": 3429 }, { "epoch": 0.65, "grad_norm": 12.4375, "learning_rate": 3.947191319132346e-06, "loss": 0.2884, "step": 3430 }, { "epoch": 0.65, "grad_norm": 12.6875, "learning_rate": 3.94656668947971e-06, "loss": 0.2119, "step": 3431 }, { "epoch": 0.65, "grad_norm": 9.75, "learning_rate": 3.945941924042141e-06, "loss": 0.2285, "step": 3432 }, { "epoch": 0.65, "grad_norm": 17.25, "learning_rate": 3.945317022878289e-06, "loss": 0.2966, "step": 3433 }, { "epoch": 0.65, "grad_norm": 10.9375, "learning_rate": 3.944691986046805e-06, "loss": 0.2785, "step": 3434 }, { "epoch": 0.65, "grad_norm": 10.8125, "learning_rate": 3.944066813606365e-06, "loss": 0.2037, "step": 3435 }, { "epoch": 0.65, "grad_norm": 8.75, "learning_rate": 3.943441505615649e-06, "loss": 0.2713, "step": 3436 }, { "epoch": 0.65, "grad_norm": 14.125, "learning_rate": 3.942816062133354e-06, "loss": 0.2499, "step": 3437 }, { "epoch": 0.65, "grad_norm": 13.0625, "learning_rate": 3.942190483218189e-06, "loss": 0.2978, "step": 3438 }, { "epoch": 0.65, "grad_norm": 11.6875, "learning_rate": 3.941564768928874e-06, "loss": 0.3297, "step": 3439 }, { "epoch": 0.65, "grad_norm": 15.5625, "learning_rate": 3.9409389193241435e-06, "loss": 0.2479, "step": 3440 }, { "epoch": 0.65, "grad_norm": 9.125, "learning_rate": 3.940312934462743e-06, "loss": 0.2259, "step": 3441 }, { "epoch": 0.65, "grad_norm": 10.375, "learning_rate": 3.939686814403435e-06, "loss": 0.3083, "step": 3442 }, { "epoch": 0.65, "grad_norm": 12.5, "learning_rate": 3.939060559204988e-06, "loss": 0.2016, "step": 3443 }, { "epoch": 0.65, "grad_norm": 9.5625, "learning_rate": 3.938434168926189e-06, "loss": 0.2659, "step": 3444 }, { "epoch": 0.65, "grad_norm": 12.875, "learning_rate": 3.937807643625834e-06, "loss": 0.3409, "step": 3445 }, { "epoch": 0.65, "grad_norm": 13.6875, "learning_rate": 3.937180983362734e-06, "loss": 0.2651, "step": 3446 }, { "epoch": 0.65, "grad_norm": 10.8125, "learning_rate": 3.93655418819571e-06, "loss": 0.2667, "step": 3447 }, { "epoch": 0.65, "grad_norm": 12.0625, "learning_rate": 3.935927258183601e-06, "loss": 0.2996, "step": 3448 }, { "epoch": 0.65, "grad_norm": 9.0, "learning_rate": 3.935300193385251e-06, "loss": 0.2463, "step": 3449 }, { "epoch": 0.65, "grad_norm": 9.0625, "learning_rate": 3.934672993859524e-06, "loss": 0.2572, "step": 3450 }, { "epoch": 0.65, "grad_norm": 8.625, "learning_rate": 3.934045659665291e-06, "loss": 0.1825, "step": 3451 }, { "epoch": 0.65, "grad_norm": 10.5625, "learning_rate": 3.93341819086144e-06, "loss": 0.2981, "step": 3452 }, { "epoch": 0.65, "grad_norm": 9.3125, "learning_rate": 3.932790587506868e-06, "loss": 0.2642, "step": 3453 }, { "epoch": 0.65, "grad_norm": 9.875, "learning_rate": 3.932162849660485e-06, "loss": 0.2921, "step": 3454 }, { "epoch": 0.65, "grad_norm": 18.25, "learning_rate": 3.93153497738122e-06, "loss": 0.2903, "step": 3455 }, { "epoch": 0.65, "grad_norm": 8.5625, "learning_rate": 3.930906970728005e-06, "loss": 0.2834, "step": 3456 }, { "epoch": 0.65, "grad_norm": 8.6875, "learning_rate": 3.93027882975979e-06, "loss": 0.2712, "step": 3457 }, { "epoch": 0.65, "grad_norm": 11.375, "learning_rate": 3.929650554535536e-06, "loss": 0.2942, "step": 3458 }, { "epoch": 0.65, "grad_norm": 11.9375, "learning_rate": 3.929022145114222e-06, "loss": 0.2499, "step": 3459 }, { "epoch": 0.65, "grad_norm": 14.375, "learning_rate": 3.928393601554828e-06, "loss": 0.241, "step": 3460 }, { "epoch": 0.65, "grad_norm": 11.75, "learning_rate": 3.927764923916358e-06, "loss": 0.271, "step": 3461 }, { "epoch": 0.65, "grad_norm": 9.9375, "learning_rate": 3.927136112257823e-06, "loss": 0.3591, "step": 3462 }, { "epoch": 0.66, "grad_norm": 8.1875, "learning_rate": 3.926507166638248e-06, "loss": 0.3191, "step": 3463 }, { "epoch": 0.66, "grad_norm": 14.3125, "learning_rate": 3.92587808711667e-06, "loss": 0.3094, "step": 3464 }, { "epoch": 0.66, "grad_norm": 9.625, "learning_rate": 3.925248873752139e-06, "loss": 0.2299, "step": 3465 }, { "epoch": 0.66, "grad_norm": 18.125, "learning_rate": 3.924619526603717e-06, "loss": 0.2319, "step": 3466 }, { "epoch": 0.66, "grad_norm": 11.0625, "learning_rate": 3.923990045730478e-06, "loss": 0.2431, "step": 3467 }, { "epoch": 0.66, "grad_norm": 14.0625, "learning_rate": 3.923360431191512e-06, "loss": 0.3868, "step": 3468 }, { "epoch": 0.66, "grad_norm": 8.3125, "learning_rate": 3.922730683045918e-06, "loss": 0.2592, "step": 3469 }, { "epoch": 0.66, "grad_norm": 15.625, "learning_rate": 3.922100801352807e-06, "loss": 0.3281, "step": 3470 }, { "epoch": 0.66, "grad_norm": 16.25, "learning_rate": 3.921470786171307e-06, "loss": 0.2515, "step": 3471 }, { "epoch": 0.66, "grad_norm": 27.625, "learning_rate": 3.9208406375605526e-06, "loss": 0.2832, "step": 3472 }, { "epoch": 0.66, "grad_norm": 11.375, "learning_rate": 3.9202103555796955e-06, "loss": 0.28, "step": 3473 }, { "epoch": 0.66, "grad_norm": 11.8125, "learning_rate": 3.919579940287899e-06, "loss": 0.3732, "step": 3474 }, { "epoch": 0.66, "grad_norm": 11.75, "learning_rate": 3.918949391744338e-06, "loss": 0.2962, "step": 3475 }, { "epoch": 0.66, "grad_norm": 13.0, "learning_rate": 3.918318710008199e-06, "loss": 0.2815, "step": 3476 }, { "epoch": 0.66, "grad_norm": 14.625, "learning_rate": 3.9176878951386835e-06, "loss": 0.2941, "step": 3477 }, { "epoch": 0.66, "grad_norm": 10.6875, "learning_rate": 3.917056947195004e-06, "loss": 0.2746, "step": 3478 }, { "epoch": 0.66, "grad_norm": 12.75, "learning_rate": 3.9164258662363865e-06, "loss": 0.2773, "step": 3479 }, { "epoch": 0.66, "grad_norm": 12.4375, "learning_rate": 3.915794652322066e-06, "loss": 0.2977, "step": 3480 }, { "epoch": 0.66, "grad_norm": 15.75, "learning_rate": 3.915163305511296e-06, "loss": 0.3498, "step": 3481 }, { "epoch": 0.66, "grad_norm": 10.5625, "learning_rate": 3.914531825863337e-06, "loss": 0.277, "step": 3482 }, { "epoch": 0.66, "grad_norm": 10.6875, "learning_rate": 3.9139002134374635e-06, "loss": 0.2483, "step": 3483 }, { "epoch": 0.66, "grad_norm": 8.1875, "learning_rate": 3.9132684682929665e-06, "loss": 0.2291, "step": 3484 }, { "epoch": 0.66, "grad_norm": 11.3125, "learning_rate": 3.9126365904891425e-06, "loss": 0.315, "step": 3485 }, { "epoch": 0.66, "grad_norm": 8.5, "learning_rate": 3.912004580085306e-06, "loss": 0.2815, "step": 3486 }, { "epoch": 0.66, "grad_norm": 12.9375, "learning_rate": 3.911372437140781e-06, "loss": 0.3373, "step": 3487 }, { "epoch": 0.66, "grad_norm": 10.625, "learning_rate": 3.910740161714905e-06, "loss": 0.2323, "step": 3488 }, { "epoch": 0.66, "grad_norm": 10.5, "learning_rate": 3.9101077538670285e-06, "loss": 0.2776, "step": 3489 }, { "epoch": 0.66, "grad_norm": 16.25, "learning_rate": 3.909475213656513e-06, "loss": 0.2941, "step": 3490 }, { "epoch": 0.66, "grad_norm": 12.6875, "learning_rate": 3.908842541142734e-06, "loss": 0.2868, "step": 3491 }, { "epoch": 0.66, "grad_norm": 13.125, "learning_rate": 3.908209736385078e-06, "loss": 0.2292, "step": 3492 }, { "epoch": 0.66, "grad_norm": 8.1875, "learning_rate": 3.907576799442944e-06, "loss": 0.3148, "step": 3493 }, { "epoch": 0.66, "grad_norm": 14.9375, "learning_rate": 3.906943730375744e-06, "loss": 0.273, "step": 3494 }, { "epoch": 0.66, "grad_norm": 21.5, "learning_rate": 3.906310529242903e-06, "loss": 0.2047, "step": 3495 }, { "epoch": 0.66, "grad_norm": 15.125, "learning_rate": 3.9056771961038585e-06, "loss": 0.2475, "step": 3496 }, { "epoch": 0.66, "grad_norm": 12.375, "learning_rate": 3.905043731018059e-06, "loss": 0.3369, "step": 3497 }, { "epoch": 0.66, "grad_norm": 16.75, "learning_rate": 3.904410134044964e-06, "loss": 0.362, "step": 3498 }, { "epoch": 0.66, "grad_norm": 11.25, "learning_rate": 3.903776405244049e-06, "loss": 0.3033, "step": 3499 }, { "epoch": 0.66, "grad_norm": 9.6875, "learning_rate": 3.903142544674802e-06, "loss": 0.2498, "step": 3500 }, { "epoch": 0.66, "grad_norm": 12.1875, "learning_rate": 3.902508552396719e-06, "loss": 0.2523, "step": 3501 }, { "epoch": 0.66, "grad_norm": 13.0, "learning_rate": 3.901874428469311e-06, "loss": 0.3068, "step": 3502 }, { "epoch": 0.66, "grad_norm": 8.6875, "learning_rate": 3.901240172952104e-06, "loss": 0.2916, "step": 3503 }, { "epoch": 0.66, "grad_norm": 19.5, "learning_rate": 3.900605785904629e-06, "loss": 0.2091, "step": 3504 }, { "epoch": 0.66, "grad_norm": 9.0, "learning_rate": 3.8999712673864384e-06, "loss": 0.1882, "step": 3505 }, { "epoch": 0.66, "grad_norm": 13.875, "learning_rate": 3.899336617457091e-06, "loss": 0.2981, "step": 3506 }, { "epoch": 0.66, "grad_norm": 11.5, "learning_rate": 3.89870183617616e-06, "loss": 0.2946, "step": 3507 }, { "epoch": 0.66, "grad_norm": 13.4375, "learning_rate": 3.89806692360323e-06, "loss": 0.2494, "step": 3508 }, { "epoch": 0.66, "grad_norm": 10.8125, "learning_rate": 3.897431879797897e-06, "loss": 0.2228, "step": 3509 }, { "epoch": 0.66, "grad_norm": 8.4375, "learning_rate": 3.896796704819773e-06, "loss": 0.266, "step": 3510 }, { "epoch": 0.66, "grad_norm": 16.5, "learning_rate": 3.89616139872848e-06, "loss": 0.4287, "step": 3511 }, { "epoch": 0.66, "grad_norm": 10.375, "learning_rate": 3.89552596158365e-06, "loss": 0.2734, "step": 3512 }, { "epoch": 0.66, "grad_norm": 10.375, "learning_rate": 3.894890393444931e-06, "loss": 0.3671, "step": 3513 }, { "epoch": 0.66, "grad_norm": 8.6875, "learning_rate": 3.8942546943719825e-06, "loss": 0.2623, "step": 3514 }, { "epoch": 0.66, "grad_norm": 12.625, "learning_rate": 3.893618864424475e-06, "loss": 0.2523, "step": 3515 }, { "epoch": 0.67, "grad_norm": 8.6875, "learning_rate": 3.892982903662091e-06, "loss": 0.2687, "step": 3516 }, { "epoch": 0.67, "grad_norm": 13.4375, "learning_rate": 3.892346812144528e-06, "loss": 0.2421, "step": 3517 }, { "epoch": 0.67, "grad_norm": 8.75, "learning_rate": 3.891710589931494e-06, "loss": 0.3283, "step": 3518 }, { "epoch": 0.67, "grad_norm": 11.5, "learning_rate": 3.891074237082707e-06, "loss": 0.3163, "step": 3519 }, { "epoch": 0.67, "grad_norm": 13.1875, "learning_rate": 3.890437753657902e-06, "loss": 0.232, "step": 3520 }, { "epoch": 0.67, "grad_norm": 18.125, "learning_rate": 3.889801139716823e-06, "loss": 0.3069, "step": 3521 }, { "epoch": 0.67, "grad_norm": 11.25, "learning_rate": 3.889164395319227e-06, "loss": 0.3416, "step": 3522 }, { "epoch": 0.67, "grad_norm": 10.1875, "learning_rate": 3.888527520524883e-06, "loss": 0.2453, "step": 3523 }, { "epoch": 0.67, "grad_norm": 9.25, "learning_rate": 3.887890515393573e-06, "loss": 0.2539, "step": 3524 }, { "epoch": 0.67, "grad_norm": 11.5625, "learning_rate": 3.887253379985091e-06, "loss": 0.3489, "step": 3525 }, { "epoch": 0.67, "grad_norm": 10.0, "learning_rate": 3.886616114359242e-06, "loss": 0.2899, "step": 3526 }, { "epoch": 0.67, "grad_norm": 9.6875, "learning_rate": 3.885978718575845e-06, "loss": 0.2704, "step": 3527 }, { "epoch": 0.67, "grad_norm": 10.8125, "learning_rate": 3.885341192694731e-06, "loss": 0.2701, "step": 3528 }, { "epoch": 0.67, "grad_norm": 14.375, "learning_rate": 3.884703536775741e-06, "loss": 0.2972, "step": 3529 }, { "epoch": 0.67, "grad_norm": 11.75, "learning_rate": 3.884065750878733e-06, "loss": 0.3835, "step": 3530 }, { "epoch": 0.67, "grad_norm": 10.25, "learning_rate": 3.88342783506357e-06, "loss": 0.2066, "step": 3531 }, { "epoch": 0.67, "grad_norm": 11.0625, "learning_rate": 3.882789789390134e-06, "loss": 0.3095, "step": 3532 }, { "epoch": 0.67, "grad_norm": 12.5625, "learning_rate": 3.882151613918315e-06, "loss": 0.2673, "step": 3533 }, { "epoch": 0.67, "grad_norm": 9.6875, "learning_rate": 3.881513308708018e-06, "loss": 0.2936, "step": 3534 }, { "epoch": 0.67, "grad_norm": 16.375, "learning_rate": 3.880874873819158e-06, "loss": 0.3231, "step": 3535 }, { "epoch": 0.67, "grad_norm": 19.5, "learning_rate": 3.880236309311663e-06, "loss": 0.205, "step": 3536 }, { "epoch": 0.67, "grad_norm": 9.25, "learning_rate": 3.879597615245474e-06, "loss": 0.2635, "step": 3537 }, { "epoch": 0.67, "grad_norm": 16.375, "learning_rate": 3.878958791680542e-06, "loss": 0.3372, "step": 3538 }, { "epoch": 0.67, "grad_norm": 8.5625, "learning_rate": 3.878319838676832e-06, "loss": 0.2142, "step": 3539 }, { "epoch": 0.67, "grad_norm": 10.3125, "learning_rate": 3.8776807562943205e-06, "loss": 0.2589, "step": 3540 }, { "epoch": 0.67, "grad_norm": 10.9375, "learning_rate": 3.877041544592998e-06, "loss": 0.2493, "step": 3541 }, { "epoch": 0.67, "grad_norm": 10.3125, "learning_rate": 3.876402203632862e-06, "loss": 0.227, "step": 3542 }, { "epoch": 0.67, "grad_norm": 8.3125, "learning_rate": 3.875762733473929e-06, "loss": 0.2242, "step": 3543 }, { "epoch": 0.67, "grad_norm": 9.0625, "learning_rate": 3.875123134176222e-06, "loss": 0.3191, "step": 3544 }, { "epoch": 0.67, "grad_norm": 13.375, "learning_rate": 3.8744834057997784e-06, "loss": 0.2291, "step": 3545 }, { "epoch": 0.67, "grad_norm": 11.25, "learning_rate": 3.873843548404649e-06, "loss": 0.3108, "step": 3546 }, { "epoch": 0.67, "grad_norm": 11.0625, "learning_rate": 3.873203562050894e-06, "loss": 0.2889, "step": 3547 }, { "epoch": 0.67, "grad_norm": 10.3125, "learning_rate": 3.8725634467985875e-06, "loss": 0.3044, "step": 3548 }, { "epoch": 0.67, "grad_norm": 12.0625, "learning_rate": 3.871923202707815e-06, "loss": 0.2888, "step": 3549 }, { "epoch": 0.67, "grad_norm": 7.3125, "learning_rate": 3.871282829838674e-06, "loss": 0.2488, "step": 3550 }, { "epoch": 0.67, "grad_norm": 12.375, "learning_rate": 3.870642328251276e-06, "loss": 0.3486, "step": 3551 }, { "epoch": 0.67, "grad_norm": 9.5, "learning_rate": 3.870001698005742e-06, "loss": 0.2752, "step": 3552 }, { "epoch": 0.67, "grad_norm": 8.375, "learning_rate": 3.869360939162206e-06, "loss": 0.2965, "step": 3553 }, { "epoch": 0.67, "grad_norm": 14.125, "learning_rate": 3.8687200517808135e-06, "loss": 0.2543, "step": 3554 }, { "epoch": 0.67, "grad_norm": 9.25, "learning_rate": 3.868079035921723e-06, "loss": 0.2807, "step": 3555 }, { "epoch": 0.67, "grad_norm": 10.25, "learning_rate": 3.867437891645106e-06, "loss": 0.233, "step": 3556 }, { "epoch": 0.67, "grad_norm": 9.5, "learning_rate": 3.866796619011142e-06, "loss": 0.2747, "step": 3557 }, { "epoch": 0.67, "grad_norm": 7.8125, "learning_rate": 3.86615521808003e-06, "loss": 0.2199, "step": 3558 }, { "epoch": 0.67, "grad_norm": 10.3125, "learning_rate": 3.865513688911972e-06, "loss": 0.3499, "step": 3559 }, { "epoch": 0.67, "grad_norm": 12.375, "learning_rate": 3.864872031567188e-06, "loss": 0.3047, "step": 3560 }, { "epoch": 0.67, "grad_norm": 11.0625, "learning_rate": 3.864230246105909e-06, "loss": 0.2804, "step": 3561 }, { "epoch": 0.67, "grad_norm": 7.84375, "learning_rate": 3.863588332588377e-06, "loss": 0.1541, "step": 3562 }, { "epoch": 0.67, "grad_norm": 11.625, "learning_rate": 3.862946291074847e-06, "loss": 0.3093, "step": 3563 }, { "epoch": 0.67, "grad_norm": 10.75, "learning_rate": 3.862304121625584e-06, "loss": 0.3087, "step": 3564 }, { "epoch": 0.67, "grad_norm": 11.25, "learning_rate": 3.861661824300869e-06, "loss": 0.2949, "step": 3565 }, { "epoch": 0.67, "grad_norm": 9.0625, "learning_rate": 3.86101939916099e-06, "loss": 0.2394, "step": 3566 }, { "epoch": 0.67, "grad_norm": 12.9375, "learning_rate": 3.860376846266252e-06, "loss": 0.3232, "step": 3567 }, { "epoch": 0.67, "grad_norm": 15.4375, "learning_rate": 3.859734165676967e-06, "loss": 0.2922, "step": 3568 }, { "epoch": 0.68, "grad_norm": 10.875, "learning_rate": 3.859091357453464e-06, "loss": 0.2559, "step": 3569 }, { "epoch": 0.68, "grad_norm": 12.875, "learning_rate": 3.858448421656079e-06, "loss": 0.3874, "step": 3570 }, { "epoch": 0.68, "grad_norm": 11.4375, "learning_rate": 3.857805358345165e-06, "loss": 0.3159, "step": 3571 }, { "epoch": 0.68, "grad_norm": 12.875, "learning_rate": 3.857162167581082e-06, "loss": 0.2606, "step": 3572 }, { "epoch": 0.68, "grad_norm": 11.1875, "learning_rate": 3.856518849424207e-06, "loss": 0.2018, "step": 3573 }, { "epoch": 0.68, "grad_norm": 10.125, "learning_rate": 3.855875403934924e-06, "loss": 0.3046, "step": 3574 }, { "epoch": 0.68, "grad_norm": 8.0, "learning_rate": 3.855231831173634e-06, "loss": 0.2375, "step": 3575 }, { "epoch": 0.68, "grad_norm": 7.625, "learning_rate": 3.854588131200744e-06, "loss": 0.2285, "step": 3576 }, { "epoch": 0.68, "grad_norm": 13.25, "learning_rate": 3.853944304076678e-06, "loss": 0.3131, "step": 3577 }, { "epoch": 0.68, "grad_norm": 8.9375, "learning_rate": 3.853300349861871e-06, "loss": 0.2807, "step": 3578 }, { "epoch": 0.68, "grad_norm": 11.4375, "learning_rate": 3.852656268616767e-06, "loss": 0.2264, "step": 3579 }, { "epoch": 0.68, "grad_norm": 22.0, "learning_rate": 3.8520120604018255e-06, "loss": 0.2645, "step": 3580 }, { "epoch": 0.68, "grad_norm": 14.5625, "learning_rate": 3.851367725277516e-06, "loss": 0.341, "step": 3581 }, { "epoch": 0.68, "grad_norm": 8.25, "learning_rate": 3.85072326330432e-06, "loss": 0.2623, "step": 3582 }, { "epoch": 0.68, "grad_norm": 11.25, "learning_rate": 3.8500786745427306e-06, "loss": 0.3163, "step": 3583 }, { "epoch": 0.68, "grad_norm": 10.375, "learning_rate": 3.8494339590532555e-06, "loss": 0.2943, "step": 3584 }, { "epoch": 0.68, "grad_norm": 13.0, "learning_rate": 3.848789116896412e-06, "loss": 0.3029, "step": 3585 }, { "epoch": 0.68, "grad_norm": 15.875, "learning_rate": 3.848144148132728e-06, "loss": 0.2999, "step": 3586 }, { "epoch": 0.68, "grad_norm": 17.25, "learning_rate": 3.847499052822745e-06, "loss": 0.2782, "step": 3587 }, { "epoch": 0.68, "grad_norm": 9.5, "learning_rate": 3.846853831027016e-06, "loss": 0.256, "step": 3588 }, { "epoch": 0.68, "grad_norm": 9.0, "learning_rate": 3.846208482806108e-06, "loss": 0.2607, "step": 3589 }, { "epoch": 0.68, "grad_norm": 9.0625, "learning_rate": 3.845563008220597e-06, "loss": 0.2833, "step": 3590 }, { "epoch": 0.68, "grad_norm": 8.0625, "learning_rate": 3.84491740733107e-06, "loss": 0.2457, "step": 3591 }, { "epoch": 0.68, "grad_norm": 10.375, "learning_rate": 3.84427168019813e-06, "loss": 0.2934, "step": 3592 }, { "epoch": 0.68, "grad_norm": 35.0, "learning_rate": 3.8436258268823875e-06, "loss": 0.2354, "step": 3593 }, { "epoch": 0.68, "grad_norm": 6.9375, "learning_rate": 3.8429798474444676e-06, "loss": 0.2155, "step": 3594 }, { "epoch": 0.68, "grad_norm": 13.1875, "learning_rate": 3.842333741945008e-06, "loss": 0.349, "step": 3595 }, { "epoch": 0.68, "grad_norm": 14.3125, "learning_rate": 3.841687510444655e-06, "loss": 0.2547, "step": 3596 }, { "epoch": 0.68, "grad_norm": 12.0625, "learning_rate": 3.841041153004069e-06, "loss": 0.3205, "step": 3597 }, { "epoch": 0.68, "grad_norm": 11.75, "learning_rate": 3.840394669683922e-06, "loss": 0.2801, "step": 3598 }, { "epoch": 0.68, "grad_norm": 13.9375, "learning_rate": 3.8397480605448956e-06, "loss": 0.2952, "step": 3599 }, { "epoch": 0.68, "grad_norm": 13.3125, "learning_rate": 3.839101325647689e-06, "loss": 0.3256, "step": 3600 }, { "epoch": 0.68, "grad_norm": 11.0625, "learning_rate": 3.838454465053004e-06, "loss": 0.3077, "step": 3601 }, { "epoch": 0.68, "grad_norm": 9.375, "learning_rate": 3.837807478821564e-06, "loss": 0.2829, "step": 3602 }, { "epoch": 0.68, "grad_norm": 10.875, "learning_rate": 3.8371603670140965e-06, "loss": 0.2623, "step": 3603 }, { "epoch": 0.68, "grad_norm": 9.9375, "learning_rate": 3.8365131296913476e-06, "loss": 0.2807, "step": 3604 }, { "epoch": 0.68, "grad_norm": 8.375, "learning_rate": 3.835865766914068e-06, "loss": 0.2979, "step": 3605 }, { "epoch": 0.68, "grad_norm": 11.5625, "learning_rate": 3.835218278743025e-06, "loss": 0.3168, "step": 3606 }, { "epoch": 0.68, "grad_norm": 11.25, "learning_rate": 3.8345706652389974e-06, "loss": 0.2236, "step": 3607 }, { "epoch": 0.68, "grad_norm": 10.8125, "learning_rate": 3.833922926462773e-06, "loss": 0.2834, "step": 3608 }, { "epoch": 0.68, "grad_norm": 9.875, "learning_rate": 3.833275062475155e-06, "loss": 0.2657, "step": 3609 }, { "epoch": 0.68, "grad_norm": 11.875, "learning_rate": 3.832627073336955e-06, "loss": 0.2981, "step": 3610 }, { "epoch": 0.68, "grad_norm": 9.5, "learning_rate": 3.831978959108998e-06, "loss": 0.2802, "step": 3611 }, { "epoch": 0.68, "grad_norm": 11.125, "learning_rate": 3.8313307198521204e-06, "loss": 0.3332, "step": 3612 }, { "epoch": 0.68, "grad_norm": 8.625, "learning_rate": 3.830682355627172e-06, "loss": 0.2247, "step": 3613 }, { "epoch": 0.68, "grad_norm": 17.375, "learning_rate": 3.830033866495011e-06, "loss": 0.318, "step": 3614 }, { "epoch": 0.68, "grad_norm": 8.875, "learning_rate": 3.829385252516511e-06, "loss": 0.2758, "step": 3615 }, { "epoch": 0.68, "grad_norm": 7.6875, "learning_rate": 3.828736513752553e-06, "loss": 0.205, "step": 3616 }, { "epoch": 0.68, "grad_norm": 6.5, "learning_rate": 3.828087650264034e-06, "loss": 0.1957, "step": 3617 }, { "epoch": 0.68, "grad_norm": 11.25, "learning_rate": 3.8274386621118615e-06, "loss": 0.3076, "step": 3618 }, { "epoch": 0.68, "grad_norm": 10.75, "learning_rate": 3.826789549356952e-06, "loss": 0.2431, "step": 3619 }, { "epoch": 0.68, "grad_norm": 10.875, "learning_rate": 3.826140312060238e-06, "loss": 0.2749, "step": 3620 }, { "epoch": 0.68, "grad_norm": 12.5, "learning_rate": 3.82549095028266e-06, "loss": 0.3367, "step": 3621 }, { "epoch": 0.69, "grad_norm": 13.5625, "learning_rate": 3.824841464085171e-06, "loss": 0.3534, "step": 3622 }, { "epoch": 0.69, "grad_norm": 9.5, "learning_rate": 3.824191853528738e-06, "loss": 0.2873, "step": 3623 }, { "epoch": 0.69, "grad_norm": 12.125, "learning_rate": 3.823542118674338e-06, "loss": 0.2136, "step": 3624 }, { "epoch": 0.69, "grad_norm": 11.6875, "learning_rate": 3.822892259582958e-06, "loss": 0.3315, "step": 3625 }, { "epoch": 0.69, "grad_norm": 8.5625, "learning_rate": 3.8222422763156e-06, "loss": 0.312, "step": 3626 }, { "epoch": 0.69, "grad_norm": 11.5, "learning_rate": 3.821592168933276e-06, "loss": 0.3266, "step": 3627 }, { "epoch": 0.69, "grad_norm": 8.875, "learning_rate": 3.820941937497008e-06, "loss": 0.2816, "step": 3628 }, { "epoch": 0.69, "grad_norm": 10.75, "learning_rate": 3.820291582067834e-06, "loss": 0.2039, "step": 3629 }, { "epoch": 0.69, "grad_norm": 16.125, "learning_rate": 3.819641102706797e-06, "loss": 0.3365, "step": 3630 }, { "epoch": 0.69, "grad_norm": 11.0, "learning_rate": 3.818990499474959e-06, "loss": 0.3389, "step": 3631 }, { "epoch": 0.69, "grad_norm": 9.4375, "learning_rate": 3.81833977243339e-06, "loss": 0.2591, "step": 3632 }, { "epoch": 0.69, "grad_norm": 8.75, "learning_rate": 3.81768892164317e-06, "loss": 0.2736, "step": 3633 }, { "epoch": 0.69, "grad_norm": 9.25, "learning_rate": 3.817037947165394e-06, "loss": 0.2573, "step": 3634 }, { "epoch": 0.69, "grad_norm": 12.6875, "learning_rate": 3.8163868490611656e-06, "loss": 0.2895, "step": 3635 }, { "epoch": 0.69, "grad_norm": 12.875, "learning_rate": 3.815735627391602e-06, "loss": 0.3081, "step": 3636 }, { "epoch": 0.69, "grad_norm": 17.25, "learning_rate": 3.815084282217833e-06, "loss": 0.3696, "step": 3637 }, { "epoch": 0.69, "grad_norm": 12.25, "learning_rate": 3.8144328136009967e-06, "loss": 0.3062, "step": 3638 }, { "epoch": 0.69, "grad_norm": 7.84375, "learning_rate": 3.813781221602244e-06, "loss": 0.2709, "step": 3639 }, { "epoch": 0.69, "grad_norm": 12.125, "learning_rate": 3.81312950628274e-06, "loss": 0.3434, "step": 3640 }, { "epoch": 0.69, "grad_norm": 25.25, "learning_rate": 3.8124776677036573e-06, "loss": 0.3374, "step": 3641 }, { "epoch": 0.69, "grad_norm": 12.375, "learning_rate": 3.8118257059261844e-06, "loss": 0.317, "step": 3642 }, { "epoch": 0.69, "grad_norm": 8.3125, "learning_rate": 3.811173621011516e-06, "loss": 0.2259, "step": 3643 }, { "epoch": 0.69, "grad_norm": 12.375, "learning_rate": 3.8105214130208635e-06, "loss": 0.2536, "step": 3644 }, { "epoch": 0.69, "grad_norm": 9.0625, "learning_rate": 3.809869082015447e-06, "loss": 0.269, "step": 3645 }, { "epoch": 0.69, "grad_norm": 12.75, "learning_rate": 3.809216628056499e-06, "loss": 0.2858, "step": 3646 }, { "epoch": 0.69, "grad_norm": 17.125, "learning_rate": 3.8085640512052636e-06, "loss": 0.2291, "step": 3647 }, { "epoch": 0.69, "grad_norm": 17.125, "learning_rate": 3.8079113515229965e-06, "loss": 0.2744, "step": 3648 }, { "epoch": 0.69, "grad_norm": 12.9375, "learning_rate": 3.807258529070964e-06, "loss": 0.254, "step": 3649 }, { "epoch": 0.69, "grad_norm": 11.3125, "learning_rate": 3.8066055839104445e-06, "loss": 0.3082, "step": 3650 }, { "epoch": 0.69, "grad_norm": 11.75, "learning_rate": 3.8059525161027296e-06, "loss": 0.2627, "step": 3651 }, { "epoch": 0.69, "grad_norm": 11.125, "learning_rate": 3.805299325709119e-06, "loss": 0.2569, "step": 3652 }, { "epoch": 0.69, "grad_norm": 11.0625, "learning_rate": 3.804646012790926e-06, "loss": 0.2783, "step": 3653 }, { "epoch": 0.69, "grad_norm": 10.75, "learning_rate": 3.8039925774094756e-06, "loss": 0.2851, "step": 3654 }, { "epoch": 0.69, "grad_norm": 10.0, "learning_rate": 3.8033390196261046e-06, "loss": 0.3178, "step": 3655 }, { "epoch": 0.69, "grad_norm": 9.875, "learning_rate": 3.802685339502159e-06, "loss": 0.2302, "step": 3656 }, { "epoch": 0.69, "grad_norm": 8.9375, "learning_rate": 3.8020315370989992e-06, "loss": 0.2342, "step": 3657 }, { "epoch": 0.69, "grad_norm": 11.9375, "learning_rate": 3.8013776124779945e-06, "loss": 0.1936, "step": 3658 }, { "epoch": 0.69, "grad_norm": 11.0, "learning_rate": 3.800723565700528e-06, "loss": 0.2483, "step": 3659 }, { "epoch": 0.69, "grad_norm": 8.25, "learning_rate": 3.8000693968279923e-06, "loss": 0.2534, "step": 3660 }, { "epoch": 0.69, "grad_norm": 8.8125, "learning_rate": 3.7994151059217925e-06, "loss": 0.2739, "step": 3661 }, { "epoch": 0.69, "grad_norm": 12.4375, "learning_rate": 3.7987606930433448e-06, "loss": 0.3448, "step": 3662 }, { "epoch": 0.69, "grad_norm": 12.125, "learning_rate": 3.7981061582540775e-06, "loss": 0.2416, "step": 3663 }, { "epoch": 0.69, "grad_norm": 16.25, "learning_rate": 3.7974515016154294e-06, "loss": 0.2752, "step": 3664 }, { "epoch": 0.69, "grad_norm": 14.1875, "learning_rate": 3.7967967231888514e-06, "loss": 0.3389, "step": 3665 }, { "epoch": 0.69, "grad_norm": 12.625, "learning_rate": 3.796141823035806e-06, "loss": 0.2878, "step": 3666 }, { "epoch": 0.69, "grad_norm": 10.75, "learning_rate": 3.795486801217766e-06, "loss": 0.3095, "step": 3667 }, { "epoch": 0.69, "grad_norm": 12.875, "learning_rate": 3.7948316577962165e-06, "loss": 0.3065, "step": 3668 }, { "epoch": 0.69, "grad_norm": 13.3125, "learning_rate": 3.7941763928326537e-06, "loss": 0.2961, "step": 3669 }, { "epoch": 0.69, "grad_norm": 9.125, "learning_rate": 3.7935210063885865e-06, "loss": 0.2393, "step": 3670 }, { "epoch": 0.69, "grad_norm": 14.375, "learning_rate": 3.7928654985255333e-06, "loss": 0.2397, "step": 3671 }, { "epoch": 0.69, "grad_norm": 12.0, "learning_rate": 3.7922098693050238e-06, "loss": 0.3015, "step": 3672 }, { "epoch": 0.69, "grad_norm": 11.125, "learning_rate": 3.7915541187886013e-06, "loss": 0.2623, "step": 3673 }, { "epoch": 0.7, "grad_norm": 7.09375, "learning_rate": 3.7908982470378184e-06, "loss": 0.2293, "step": 3674 }, { "epoch": 0.7, "grad_norm": 19.5, "learning_rate": 3.79024225411424e-06, "loss": 0.2895, "step": 3675 }, { "epoch": 0.7, "grad_norm": 9.1875, "learning_rate": 3.789586140079443e-06, "loss": 0.3881, "step": 3676 }, { "epoch": 0.7, "grad_norm": 11.5625, "learning_rate": 3.7889299049950146e-06, "loss": 0.1897, "step": 3677 }, { "epoch": 0.7, "grad_norm": 8.75, "learning_rate": 3.7882735489225527e-06, "loss": 0.2967, "step": 3678 }, { "epoch": 0.7, "grad_norm": 8.5625, "learning_rate": 3.7876170719236684e-06, "loss": 0.2516, "step": 3679 }, { "epoch": 0.7, "grad_norm": 10.0625, "learning_rate": 3.786960474059982e-06, "loss": 0.3216, "step": 3680 }, { "epoch": 0.7, "grad_norm": 8.5, "learning_rate": 3.786303755393128e-06, "loss": 0.2818, "step": 3681 }, { "epoch": 0.7, "grad_norm": 12.3125, "learning_rate": 3.7856469159847498e-06, "loss": 0.3064, "step": 3682 }, { "epoch": 0.7, "grad_norm": 11.1875, "learning_rate": 3.784989955896503e-06, "loss": 0.2174, "step": 3683 }, { "epoch": 0.7, "grad_norm": 13.8125, "learning_rate": 3.7843328751900554e-06, "loss": 0.3319, "step": 3684 }, { "epoch": 0.7, "grad_norm": 11.25, "learning_rate": 3.783675673927083e-06, "loss": 0.2593, "step": 3685 }, { "epoch": 0.7, "grad_norm": 11.8125, "learning_rate": 3.783018352169279e-06, "loss": 0.3172, "step": 3686 }, { "epoch": 0.7, "grad_norm": 9.8125, "learning_rate": 3.78236090997834e-06, "loss": 0.2987, "step": 3687 }, { "epoch": 0.7, "grad_norm": 13.5, "learning_rate": 3.7817033474159814e-06, "loss": 0.2786, "step": 3688 }, { "epoch": 0.7, "grad_norm": 24.625, "learning_rate": 3.7810456645439243e-06, "loss": 0.2521, "step": 3689 }, { "epoch": 0.7, "grad_norm": 7.4375, "learning_rate": 3.7803878614239054e-06, "loss": 0.2051, "step": 3690 }, { "epoch": 0.7, "grad_norm": 8.6875, "learning_rate": 3.7797299381176694e-06, "loss": 0.3082, "step": 3691 }, { "epoch": 0.7, "grad_norm": 36.75, "learning_rate": 3.779071894686975e-06, "loss": 0.3205, "step": 3692 }, { "epoch": 0.7, "grad_norm": 10.25, "learning_rate": 3.7784137311935894e-06, "loss": 0.2437, "step": 3693 }, { "epoch": 0.7, "grad_norm": 12.875, "learning_rate": 3.777755447699293e-06, "loss": 0.2839, "step": 3694 }, { "epoch": 0.7, "grad_norm": 8.125, "learning_rate": 3.7770970442658765e-06, "loss": 0.3104, "step": 3695 }, { "epoch": 0.7, "grad_norm": 10.125, "learning_rate": 3.7764385209551422e-06, "loss": 0.3059, "step": 3696 }, { "epoch": 0.7, "grad_norm": 12.8125, "learning_rate": 3.7757798778289057e-06, "loss": 0.3359, "step": 3697 }, { "epoch": 0.7, "grad_norm": 9.0, "learning_rate": 3.7751211149489897e-06, "loss": 0.2854, "step": 3698 }, { "epoch": 0.7, "grad_norm": 14.9375, "learning_rate": 3.774462232377231e-06, "loss": 0.3068, "step": 3699 }, { "epoch": 0.7, "grad_norm": 23.625, "learning_rate": 3.7738032301754767e-06, "loss": 0.2546, "step": 3700 }, { "epoch": 0.7, "grad_norm": 9.4375, "learning_rate": 3.7731441084055863e-06, "loss": 0.3416, "step": 3701 }, { "epoch": 0.7, "grad_norm": 10.6875, "learning_rate": 3.772484867129428e-06, "loss": 0.2782, "step": 3702 }, { "epoch": 0.7, "grad_norm": 10.3125, "learning_rate": 3.771825506408885e-06, "loss": 0.3276, "step": 3703 }, { "epoch": 0.7, "grad_norm": 18.75, "learning_rate": 3.771166026305847e-06, "loss": 0.2673, "step": 3704 }, { "epoch": 0.7, "grad_norm": 7.625, "learning_rate": 3.77050642688222e-06, "loss": 0.235, "step": 3705 }, { "epoch": 0.7, "grad_norm": 9.9375, "learning_rate": 3.7698467081999158e-06, "loss": 0.2381, "step": 3706 }, { "epoch": 0.7, "grad_norm": 10.875, "learning_rate": 3.7691868703208636e-06, "loss": 0.2417, "step": 3707 }, { "epoch": 0.7, "grad_norm": 13.1875, "learning_rate": 3.7685269133069975e-06, "loss": 0.3368, "step": 3708 }, { "epoch": 0.7, "grad_norm": 14.4375, "learning_rate": 3.767866837220267e-06, "loss": 0.3638, "step": 3709 }, { "epoch": 0.7, "grad_norm": 11.875, "learning_rate": 3.7672066421226316e-06, "loss": 0.2706, "step": 3710 }, { "epoch": 0.7, "grad_norm": 9.3125, "learning_rate": 3.7665463280760616e-06, "loss": 0.2627, "step": 3711 }, { "epoch": 0.7, "grad_norm": 10.5, "learning_rate": 3.765885895142539e-06, "loss": 0.3211, "step": 3712 }, { "epoch": 0.7, "grad_norm": 11.3125, "learning_rate": 3.7652253433840563e-06, "loss": 0.2824, "step": 3713 }, { "epoch": 0.7, "grad_norm": 13.8125, "learning_rate": 3.764564672862617e-06, "loss": 0.2842, "step": 3714 }, { "epoch": 0.7, "grad_norm": 8.875, "learning_rate": 3.7639038836402386e-06, "loss": 0.2475, "step": 3715 }, { "epoch": 0.7, "grad_norm": 9.0625, "learning_rate": 3.7632429757789446e-06, "loss": 0.2729, "step": 3716 }, { "epoch": 0.7, "grad_norm": 10.875, "learning_rate": 3.7625819493407735e-06, "loss": 0.3562, "step": 3717 }, { "epoch": 0.7, "grad_norm": 9.375, "learning_rate": 3.7619208043877742e-06, "loss": 0.2561, "step": 3718 }, { "epoch": 0.7, "grad_norm": 8.625, "learning_rate": 3.761259540982007e-06, "loss": 0.2964, "step": 3719 }, { "epoch": 0.7, "grad_norm": 7.5625, "learning_rate": 3.760598159185541e-06, "loss": 0.2662, "step": 3720 }, { "epoch": 0.7, "grad_norm": 7.375, "learning_rate": 3.75993665906046e-06, "loss": 0.1883, "step": 3721 }, { "epoch": 0.7, "grad_norm": 8.3125, "learning_rate": 3.7592750406688556e-06, "loss": 0.2159, "step": 3722 }, { "epoch": 0.7, "grad_norm": 11.375, "learning_rate": 3.7586133040728324e-06, "loss": 0.3421, "step": 3723 }, { "epoch": 0.7, "grad_norm": 12.8125, "learning_rate": 3.7579514493345067e-06, "loss": 0.3761, "step": 3724 }, { "epoch": 0.7, "grad_norm": 15.125, "learning_rate": 3.7572894765160027e-06, "loss": 0.2354, "step": 3725 }, { "epoch": 0.7, "grad_norm": 18.375, "learning_rate": 3.75662738567946e-06, "loss": 0.2834, "step": 3726 }, { "epoch": 0.71, "grad_norm": 10.9375, "learning_rate": 3.7559651768870265e-06, "loss": 0.2483, "step": 3727 }, { "epoch": 0.71, "grad_norm": 14.0625, "learning_rate": 3.755302850200861e-06, "loss": 0.2934, "step": 3728 }, { "epoch": 0.71, "grad_norm": 11.5625, "learning_rate": 3.754640405683134e-06, "loss": 0.2711, "step": 3729 }, { "epoch": 0.71, "grad_norm": 20.125, "learning_rate": 3.7539778433960287e-06, "loss": 0.2199, "step": 3730 }, { "epoch": 0.71, "grad_norm": 9.375, "learning_rate": 3.753315163401736e-06, "loss": 0.2606, "step": 3731 }, { "epoch": 0.71, "grad_norm": 11.0, "learning_rate": 3.7526523657624626e-06, "loss": 0.3436, "step": 3732 }, { "epoch": 0.71, "grad_norm": 12.4375, "learning_rate": 3.7519894505404196e-06, "loss": 0.2093, "step": 3733 }, { "epoch": 0.71, "grad_norm": 17.75, "learning_rate": 3.751326417797836e-06, "loss": 0.2304, "step": 3734 }, { "epoch": 0.71, "grad_norm": 10.75, "learning_rate": 3.7506632675969467e-06, "loss": 0.3184, "step": 3735 }, { "epoch": 0.71, "grad_norm": 12.4375, "learning_rate": 3.7500000000000005e-06, "loss": 0.3136, "step": 3736 }, { "epoch": 0.71, "grad_norm": 10.1875, "learning_rate": 3.7493366150692566e-06, "loss": 0.293, "step": 3737 }, { "epoch": 0.71, "grad_norm": 23.75, "learning_rate": 3.748673112866984e-06, "loss": 0.2248, "step": 3738 }, { "epoch": 0.71, "grad_norm": 23.125, "learning_rate": 3.7480094934554653e-06, "loss": 0.3516, "step": 3739 }, { "epoch": 0.71, "grad_norm": 11.6875, "learning_rate": 3.74734575689699e-06, "loss": 0.3311, "step": 3740 }, { "epoch": 0.71, "grad_norm": 13.375, "learning_rate": 3.7466819032538636e-06, "loss": 0.2603, "step": 3741 }, { "epoch": 0.71, "grad_norm": 9.25, "learning_rate": 3.7460179325883987e-06, "loss": 0.2464, "step": 3742 }, { "epoch": 0.71, "grad_norm": 14.625, "learning_rate": 3.745353844962922e-06, "loss": 0.3438, "step": 3743 }, { "epoch": 0.71, "grad_norm": 15.1875, "learning_rate": 3.744689640439766e-06, "loss": 0.2858, "step": 3744 }, { "epoch": 0.71, "grad_norm": 8.25, "learning_rate": 3.7440253190812814e-06, "loss": 0.2433, "step": 3745 }, { "epoch": 0.71, "grad_norm": 11.625, "learning_rate": 3.7433608809498234e-06, "loss": 0.2677, "step": 3746 }, { "epoch": 0.71, "grad_norm": 15.625, "learning_rate": 3.742696326107762e-06, "loss": 0.3751, "step": 3747 }, { "epoch": 0.71, "grad_norm": 9.6875, "learning_rate": 3.7420316546174766e-06, "loss": 0.2381, "step": 3748 }, { "epoch": 0.71, "grad_norm": 10.5625, "learning_rate": 3.7413668665413593e-06, "loss": 0.2875, "step": 3749 }, { "epoch": 0.71, "grad_norm": 22.5, "learning_rate": 3.740701961941809e-06, "loss": 0.2351, "step": 3750 }, { "epoch": 0.71, "grad_norm": 11.5, "learning_rate": 3.7400369408812407e-06, "loss": 0.2721, "step": 3751 }, { "epoch": 0.71, "grad_norm": 12.1875, "learning_rate": 3.739371803422077e-06, "loss": 0.3707, "step": 3752 }, { "epoch": 0.71, "grad_norm": 13.125, "learning_rate": 3.7387065496267534e-06, "loss": 0.3766, "step": 3753 }, { "epoch": 0.71, "grad_norm": 13.625, "learning_rate": 3.7380411795577136e-06, "loss": 0.3246, "step": 3754 }, { "epoch": 0.71, "grad_norm": 7.78125, "learning_rate": 3.7373756932774154e-06, "loss": 0.1686, "step": 3755 }, { "epoch": 0.71, "grad_norm": 12.875, "learning_rate": 3.7367100908483252e-06, "loss": 0.2691, "step": 3756 }, { "epoch": 0.71, "grad_norm": 11.3125, "learning_rate": 3.7360443723329216e-06, "loss": 0.3627, "step": 3757 }, { "epoch": 0.71, "grad_norm": 11.375, "learning_rate": 3.735378537793693e-06, "loss": 0.2266, "step": 3758 }, { "epoch": 0.71, "grad_norm": 14.5625, "learning_rate": 3.73471258729314e-06, "loss": 0.3378, "step": 3759 }, { "epoch": 0.71, "grad_norm": 13.4375, "learning_rate": 3.7340465208937726e-06, "loss": 0.3742, "step": 3760 }, { "epoch": 0.71, "grad_norm": 11.4375, "learning_rate": 3.7333803386581134e-06, "loss": 0.3161, "step": 3761 }, { "epoch": 0.71, "grad_norm": 9.875, "learning_rate": 3.732714040648695e-06, "loss": 0.2277, "step": 3762 }, { "epoch": 0.71, "grad_norm": 11.5625, "learning_rate": 3.73204762692806e-06, "loss": 0.3626, "step": 3763 }, { "epoch": 0.71, "grad_norm": 20.75, "learning_rate": 3.731381097558763e-06, "loss": 0.2742, "step": 3764 }, { "epoch": 0.71, "grad_norm": 10.4375, "learning_rate": 3.730714452603369e-06, "loss": 0.2835, "step": 3765 }, { "epoch": 0.71, "grad_norm": 12.625, "learning_rate": 3.730047692124454e-06, "loss": 0.289, "step": 3766 }, { "epoch": 0.71, "grad_norm": 9.875, "learning_rate": 3.7293808161846063e-06, "loss": 0.3058, "step": 3767 }, { "epoch": 0.71, "grad_norm": 9.0, "learning_rate": 3.728713824846421e-06, "loss": 0.2923, "step": 3768 }, { "epoch": 0.71, "grad_norm": 10.875, "learning_rate": 3.728046718172509e-06, "loss": 0.2814, "step": 3769 }, { "epoch": 0.71, "grad_norm": 11.8125, "learning_rate": 3.727379496225487e-06, "loss": 0.2684, "step": 3770 }, { "epoch": 0.71, "grad_norm": 7.96875, "learning_rate": 3.726712159067988e-06, "loss": 0.2294, "step": 3771 }, { "epoch": 0.71, "grad_norm": 11.25, "learning_rate": 3.726044706762651e-06, "loss": 0.3279, "step": 3772 }, { "epoch": 0.71, "grad_norm": 9.5625, "learning_rate": 3.7253771393721284e-06, "loss": 0.3248, "step": 3773 }, { "epoch": 0.71, "grad_norm": 11.75, "learning_rate": 3.7247094569590832e-06, "loss": 0.2828, "step": 3774 }, { "epoch": 0.71, "grad_norm": 10.375, "learning_rate": 3.7240416595861873e-06, "loss": 0.2933, "step": 3775 }, { "epoch": 0.71, "grad_norm": 10.5625, "learning_rate": 3.723373747316127e-06, "loss": 0.316, "step": 3776 }, { "epoch": 0.71, "grad_norm": 13.5625, "learning_rate": 3.7227057202115956e-06, "loss": 0.3038, "step": 3777 }, { "epoch": 0.71, "grad_norm": 12.0625, "learning_rate": 3.7220375783352995e-06, "loss": 0.2836, "step": 3778 }, { "epoch": 0.71, "grad_norm": 9.125, "learning_rate": 3.7213693217499554e-06, "loss": 0.2041, "step": 3779 }, { "epoch": 0.72, "grad_norm": 13.9375, "learning_rate": 3.7207009505182902e-06, "loss": 0.2651, "step": 3780 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.720032464703041e-06, "loss": 0.3236, "step": 3781 }, { "epoch": 0.72, "grad_norm": 9.5, "learning_rate": 3.719363864366958e-06, "loss": 0.2695, "step": 3782 }, { "epoch": 0.72, "grad_norm": 10.4375, "learning_rate": 3.718695149572801e-06, "loss": 0.2673, "step": 3783 }, { "epoch": 0.72, "grad_norm": 10.375, "learning_rate": 3.7180263203833396e-06, "loss": 0.2464, "step": 3784 }, { "epoch": 0.72, "grad_norm": 10.0625, "learning_rate": 3.7173573768613543e-06, "loss": 0.2105, "step": 3785 }, { "epoch": 0.72, "grad_norm": 12.375, "learning_rate": 3.7166883190696367e-06, "loss": 0.3482, "step": 3786 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.7160191470709904e-06, "loss": 0.2132, "step": 3787 }, { "epoch": 0.72, "grad_norm": 7.3125, "learning_rate": 3.715349860928228e-06, "loss": 0.2673, "step": 3788 }, { "epoch": 0.72, "grad_norm": 9.8125, "learning_rate": 3.714680460704174e-06, "loss": 0.2338, "step": 3789 }, { "epoch": 0.72, "grad_norm": 8.5625, "learning_rate": 3.714010946461662e-06, "loss": 0.2142, "step": 3790 }, { "epoch": 0.72, "grad_norm": 9.25, "learning_rate": 3.7133413182635385e-06, "loss": 0.292, "step": 3791 }, { "epoch": 0.72, "grad_norm": 8.8125, "learning_rate": 3.7126715761726578e-06, "loss": 0.2348, "step": 3792 }, { "epoch": 0.72, "grad_norm": 26.25, "learning_rate": 3.7120017202518886e-06, "loss": 0.3163, "step": 3793 }, { "epoch": 0.72, "grad_norm": 12.375, "learning_rate": 3.7113317505641067e-06, "loss": 0.2465, "step": 3794 }, { "epoch": 0.72, "grad_norm": 12.875, "learning_rate": 3.7106616671722017e-06, "loss": 0.3618, "step": 3795 }, { "epoch": 0.72, "grad_norm": 9.875, "learning_rate": 3.7099914701390703e-06, "loss": 0.2881, "step": 3796 }, { "epoch": 0.72, "grad_norm": 8.3125, "learning_rate": 3.709321159527624e-06, "loss": 0.2301, "step": 3797 }, { "epoch": 0.72, "grad_norm": 10.1875, "learning_rate": 3.7086507354007816e-06, "loss": 0.3109, "step": 3798 }, { "epoch": 0.72, "grad_norm": 17.0, "learning_rate": 3.707980197821474e-06, "loss": 0.2754, "step": 3799 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.707309546852643e-06, "loss": 0.2403, "step": 3800 }, { "epoch": 0.72, "grad_norm": 11.6875, "learning_rate": 3.70663878255724e-06, "loss": 0.3097, "step": 3801 }, { "epoch": 0.72, "grad_norm": 13.75, "learning_rate": 3.705967904998229e-06, "loss": 0.2932, "step": 3802 }, { "epoch": 0.72, "grad_norm": 14.875, "learning_rate": 3.7052969142385816e-06, "loss": 0.2746, "step": 3803 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.7046258103412826e-06, "loss": 0.3643, "step": 3804 }, { "epoch": 0.72, "grad_norm": 9.6875, "learning_rate": 3.7039545933693262e-06, "loss": 0.2404, "step": 3805 }, { "epoch": 0.72, "grad_norm": 10.25, "learning_rate": 3.7032832633857178e-06, "loss": 0.2421, "step": 3806 }, { "epoch": 0.72, "grad_norm": 8.625, "learning_rate": 3.7026118204534727e-06, "loss": 0.2044, "step": 3807 }, { "epoch": 0.72, "grad_norm": 11.5625, "learning_rate": 3.701940264635618e-06, "loss": 0.2376, "step": 3808 }, { "epoch": 0.72, "grad_norm": 14.625, "learning_rate": 3.7012685959951906e-06, "loss": 0.3345, "step": 3809 }, { "epoch": 0.72, "grad_norm": 15.6875, "learning_rate": 3.7005968145952374e-06, "loss": 0.3639, "step": 3810 }, { "epoch": 0.72, "grad_norm": 9.875, "learning_rate": 3.699924920498817e-06, "loss": 0.2995, "step": 3811 }, { "epoch": 0.72, "grad_norm": 12.0625, "learning_rate": 3.699252913768998e-06, "loss": 0.2929, "step": 3812 }, { "epoch": 0.72, "grad_norm": 11.8125, "learning_rate": 3.698580794468859e-06, "loss": 0.3305, "step": 3813 }, { "epoch": 0.72, "grad_norm": 11.625, "learning_rate": 3.6979085626614914e-06, "loss": 0.2795, "step": 3814 }, { "epoch": 0.72, "grad_norm": 15.8125, "learning_rate": 3.6972362184099943e-06, "loss": 0.2376, "step": 3815 }, { "epoch": 0.72, "grad_norm": 13.375, "learning_rate": 3.6965637617774793e-06, "loss": 0.2719, "step": 3816 }, { "epoch": 0.72, "grad_norm": 12.375, "learning_rate": 3.6958911928270674e-06, "loss": 0.2592, "step": 3817 }, { "epoch": 0.72, "grad_norm": 9.5625, "learning_rate": 3.695218511621891e-06, "loss": 0.3071, "step": 3818 }, { "epoch": 0.72, "grad_norm": 9.1875, "learning_rate": 3.6945457182250926e-06, "loss": 0.2443, "step": 3819 }, { "epoch": 0.72, "grad_norm": 13.125, "learning_rate": 3.6938728126998258e-06, "loss": 0.3617, "step": 3820 }, { "epoch": 0.72, "grad_norm": 9.125, "learning_rate": 3.6931997951092535e-06, "loss": 0.2734, "step": 3821 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.6925266655165498e-06, "loss": 0.2536, "step": 3822 }, { "epoch": 0.72, "grad_norm": 11.0, "learning_rate": 3.6918534239849e-06, "loss": 0.3918, "step": 3823 }, { "epoch": 0.72, "grad_norm": 11.25, "learning_rate": 3.6911800705774993e-06, "loss": 0.3078, "step": 3824 }, { "epoch": 0.72, "grad_norm": 10.5625, "learning_rate": 3.690506605357553e-06, "loss": 0.2434, "step": 3825 }, { "epoch": 0.72, "grad_norm": 19.875, "learning_rate": 3.6898330283882775e-06, "loss": 0.3167, "step": 3826 }, { "epoch": 0.72, "grad_norm": 12.75, "learning_rate": 3.6891593397328997e-06, "loss": 0.2308, "step": 3827 }, { "epoch": 0.72, "grad_norm": 9.5625, "learning_rate": 3.688485539454656e-06, "loss": 0.2936, "step": 3828 }, { "epoch": 0.72, "grad_norm": 9.8125, "learning_rate": 3.6878116276167946e-06, "loss": 0.2572, "step": 3829 }, { "epoch": 0.72, "grad_norm": 10.25, "learning_rate": 3.687137604282573e-06, "loss": 0.2548, "step": 3830 }, { "epoch": 0.72, "grad_norm": 10.625, "learning_rate": 3.6864634695152613e-06, "loss": 0.2357, "step": 3831 }, { "epoch": 0.72, "grad_norm": 7.59375, "learning_rate": 3.6857892233781366e-06, "loss": 0.2707, "step": 3832 }, { "epoch": 0.73, "grad_norm": 9.25, "learning_rate": 3.68511486593449e-06, "loss": 0.2754, "step": 3833 }, { "epoch": 0.73, "grad_norm": 13.9375, "learning_rate": 3.68444039724762e-06, "loss": 0.1856, "step": 3834 }, { "epoch": 0.73, "grad_norm": 9.4375, "learning_rate": 3.683765817380839e-06, "loss": 0.2691, "step": 3835 }, { "epoch": 0.73, "grad_norm": 12.5625, "learning_rate": 3.6830911263974644e-06, "loss": 0.4208, "step": 3836 }, { "epoch": 0.73, "grad_norm": 9.75, "learning_rate": 3.682416324360832e-06, "loss": 0.3365, "step": 3837 }, { "epoch": 0.73, "grad_norm": 8.375, "learning_rate": 3.6817414113342792e-06, "loss": 0.2503, "step": 3838 }, { "epoch": 0.73, "grad_norm": 9.375, "learning_rate": 3.6810663873811614e-06, "loss": 0.2676, "step": 3839 }, { "epoch": 0.73, "grad_norm": 11.875, "learning_rate": 3.6803912525648375e-06, "loss": 0.296, "step": 3840 }, { "epoch": 0.73, "grad_norm": 8.5, "learning_rate": 3.6797160069486845e-06, "loss": 0.2806, "step": 3841 }, { "epoch": 0.73, "grad_norm": 12.125, "learning_rate": 3.679040650596083e-06, "loss": 0.2133, "step": 3842 }, { "epoch": 0.73, "grad_norm": 13.0, "learning_rate": 3.678365183570428e-06, "loss": 0.2497, "step": 3843 }, { "epoch": 0.73, "grad_norm": 13.4375, "learning_rate": 3.677689605935122e-06, "loss": 0.3488, "step": 3844 }, { "epoch": 0.73, "grad_norm": 17.25, "learning_rate": 3.677013917753581e-06, "loss": 0.3108, "step": 3845 }, { "epoch": 0.73, "grad_norm": 12.0625, "learning_rate": 3.67633811908923e-06, "loss": 0.3082, "step": 3846 }, { "epoch": 0.73, "grad_norm": 8.125, "learning_rate": 3.675662210005503e-06, "loss": 0.242, "step": 3847 }, { "epoch": 0.73, "grad_norm": 21.125, "learning_rate": 3.674986190565847e-06, "loss": 0.3038, "step": 3848 }, { "epoch": 0.73, "grad_norm": 15.0, "learning_rate": 3.6743100608337157e-06, "loss": 0.2968, "step": 3849 }, { "epoch": 0.73, "grad_norm": 13.875, "learning_rate": 3.6736338208725775e-06, "loss": 0.2654, "step": 3850 }, { "epoch": 0.73, "grad_norm": 9.75, "learning_rate": 3.6729574707459077e-06, "loss": 0.2496, "step": 3851 }, { "epoch": 0.73, "grad_norm": 11.5, "learning_rate": 3.672281010517195e-06, "loss": 0.2314, "step": 3852 }, { "epoch": 0.73, "grad_norm": 8.25, "learning_rate": 3.6716044402499346e-06, "loss": 0.314, "step": 3853 }, { "epoch": 0.73, "grad_norm": 13.1875, "learning_rate": 3.6709277600076354e-06, "loss": 0.2291, "step": 3854 }, { "epoch": 0.73, "grad_norm": 15.3125, "learning_rate": 3.6702509698538157e-06, "loss": 0.2716, "step": 3855 }, { "epoch": 0.73, "grad_norm": 9.1875, "learning_rate": 3.6695740698520027e-06, "loss": 0.2163, "step": 3856 }, { "epoch": 0.73, "grad_norm": 12.9375, "learning_rate": 3.6688970600657358e-06, "loss": 0.2485, "step": 3857 }, { "epoch": 0.73, "grad_norm": 9.125, "learning_rate": 3.668219940558564e-06, "loss": 0.2478, "step": 3858 }, { "epoch": 0.73, "grad_norm": 11.0, "learning_rate": 3.6675427113940453e-06, "loss": 0.2391, "step": 3859 }, { "epoch": 0.73, "grad_norm": 10.25, "learning_rate": 3.6668653726357505e-06, "loss": 0.2642, "step": 3860 }, { "epoch": 0.73, "grad_norm": 14.8125, "learning_rate": 3.6661879243472593e-06, "loss": 0.3009, "step": 3861 }, { "epoch": 0.73, "grad_norm": 10.375, "learning_rate": 3.665510366592161e-06, "loss": 0.2705, "step": 3862 }, { "epoch": 0.73, "grad_norm": 10.6875, "learning_rate": 3.6648326994340567e-06, "loss": 0.2405, "step": 3863 }, { "epoch": 0.73, "grad_norm": 10.5625, "learning_rate": 3.664154922936556e-06, "loss": 0.3177, "step": 3864 }, { "epoch": 0.73, "grad_norm": 9.5, "learning_rate": 3.663477037163281e-06, "loss": 0.3056, "step": 3865 }, { "epoch": 0.73, "grad_norm": 12.0, "learning_rate": 3.662799042177862e-06, "loss": 0.2733, "step": 3866 }, { "epoch": 0.73, "grad_norm": 24.25, "learning_rate": 3.6621209380439402e-06, "loss": 0.2737, "step": 3867 }, { "epoch": 0.73, "grad_norm": 9.25, "learning_rate": 3.6614427248251683e-06, "loss": 0.2747, "step": 3868 }, { "epoch": 0.73, "grad_norm": 12.75, "learning_rate": 3.6607644025852074e-06, "loss": 0.2191, "step": 3869 }, { "epoch": 0.73, "grad_norm": 8.9375, "learning_rate": 3.660085971387729e-06, "loss": 0.2311, "step": 3870 }, { "epoch": 0.73, "grad_norm": 17.0, "learning_rate": 3.6594074312964167e-06, "loss": 0.2781, "step": 3871 }, { "epoch": 0.73, "grad_norm": 21.875, "learning_rate": 3.658728782374963e-06, "loss": 0.3469, "step": 3872 }, { "epoch": 0.73, "grad_norm": 9.25, "learning_rate": 3.6580500246870697e-06, "loss": 0.2422, "step": 3873 }, { "epoch": 0.73, "grad_norm": 8.3125, "learning_rate": 3.6573711582964498e-06, "loss": 0.245, "step": 3874 }, { "epoch": 0.73, "grad_norm": 10.9375, "learning_rate": 3.656692183266828e-06, "loss": 0.2217, "step": 3875 }, { "epoch": 0.73, "grad_norm": 9.5, "learning_rate": 3.6560130996619346e-06, "loss": 0.3077, "step": 3876 }, { "epoch": 0.73, "grad_norm": 16.0, "learning_rate": 3.6553339075455174e-06, "loss": 0.2478, "step": 3877 }, { "epoch": 0.73, "grad_norm": 15.5625, "learning_rate": 3.6546546069813264e-06, "loss": 0.3127, "step": 3878 }, { "epoch": 0.73, "grad_norm": 11.6875, "learning_rate": 3.6539751980331284e-06, "loss": 0.2877, "step": 3879 }, { "epoch": 0.73, "grad_norm": 10.0, "learning_rate": 3.653295680764695e-06, "loss": 0.2666, "step": 3880 }, { "epoch": 0.73, "grad_norm": 12.9375, "learning_rate": 3.6526160552398123e-06, "loss": 0.252, "step": 3881 }, { "epoch": 0.73, "grad_norm": 10.0, "learning_rate": 3.6519363215222725e-06, "loss": 0.4018, "step": 3882 }, { "epoch": 0.73, "grad_norm": 16.375, "learning_rate": 3.651256479675884e-06, "loss": 0.2819, "step": 3883 }, { "epoch": 0.73, "grad_norm": 16.75, "learning_rate": 3.6505765297644575e-06, "loss": 0.3461, "step": 3884 }, { "epoch": 0.73, "grad_norm": 11.0625, "learning_rate": 3.649896471851821e-06, "loss": 0.3834, "step": 3885 }, { "epoch": 0.74, "grad_norm": 13.6875, "learning_rate": 3.649216306001807e-06, "loss": 0.2996, "step": 3886 }, { "epoch": 0.74, "grad_norm": 10.125, "learning_rate": 3.6485360322782627e-06, "loss": 0.2351, "step": 3887 }, { "epoch": 0.74, "grad_norm": 11.8125, "learning_rate": 3.647855650745042e-06, "loss": 0.2683, "step": 3888 }, { "epoch": 0.74, "grad_norm": 8.6875, "learning_rate": 3.6471751614660113e-06, "loss": 0.2921, "step": 3889 }, { "epoch": 0.74, "grad_norm": 9.5625, "learning_rate": 3.6464945645050454e-06, "loss": 0.3253, "step": 3890 }, { "epoch": 0.74, "grad_norm": 9.75, "learning_rate": 3.6458138599260297e-06, "loss": 0.2764, "step": 3891 }, { "epoch": 0.74, "grad_norm": 22.375, "learning_rate": 3.6451330477928616e-06, "loss": 0.3156, "step": 3892 }, { "epoch": 0.74, "grad_norm": 10.375, "learning_rate": 3.6444521281694447e-06, "loss": 0.3133, "step": 3893 }, { "epoch": 0.74, "grad_norm": 11.5625, "learning_rate": 3.643771101119696e-06, "loss": 0.347, "step": 3894 }, { "epoch": 0.74, "grad_norm": 12.6875, "learning_rate": 3.6430899667075416e-06, "loss": 0.2797, "step": 3895 }, { "epoch": 0.74, "grad_norm": 8.9375, "learning_rate": 3.6424087249969175e-06, "loss": 0.262, "step": 3896 }, { "epoch": 0.74, "grad_norm": 8.0625, "learning_rate": 3.641727376051769e-06, "loss": 0.2758, "step": 3897 }, { "epoch": 0.74, "grad_norm": 9.75, "learning_rate": 3.6410459199360533e-06, "loss": 0.2855, "step": 3898 }, { "epoch": 0.74, "grad_norm": 9.8125, "learning_rate": 3.6403643567137364e-06, "loss": 0.2173, "step": 3899 }, { "epoch": 0.74, "grad_norm": 9.8125, "learning_rate": 3.639682686448795e-06, "loss": 0.3411, "step": 3900 }, { "epoch": 0.74, "grad_norm": 8.9375, "learning_rate": 3.639000909205215e-06, "loss": 0.2955, "step": 3901 }, { "epoch": 0.74, "grad_norm": 13.8125, "learning_rate": 3.6383190250469924e-06, "loss": 0.2237, "step": 3902 }, { "epoch": 0.74, "grad_norm": 9.5, "learning_rate": 3.637637034038135e-06, "loss": 0.2976, "step": 3903 }, { "epoch": 0.74, "grad_norm": 12.0, "learning_rate": 3.6369549362426577e-06, "loss": 0.2849, "step": 3904 }, { "epoch": 0.74, "grad_norm": 10.875, "learning_rate": 3.636272731724588e-06, "loss": 0.3319, "step": 3905 }, { "epoch": 0.74, "grad_norm": 11.5625, "learning_rate": 3.635590420547962e-06, "loss": 0.2767, "step": 3906 }, { "epoch": 0.74, "grad_norm": 9.875, "learning_rate": 3.6349080027768263e-06, "loss": 0.2573, "step": 3907 }, { "epoch": 0.74, "grad_norm": 16.125, "learning_rate": 3.6342254784752377e-06, "loss": 0.2529, "step": 3908 }, { "epoch": 0.74, "grad_norm": 10.9375, "learning_rate": 3.6335428477072625e-06, "loss": 0.3484, "step": 3909 }, { "epoch": 0.74, "grad_norm": 11.4375, "learning_rate": 3.632860110536977e-06, "loss": 0.3113, "step": 3910 }, { "epoch": 0.74, "grad_norm": 10.125, "learning_rate": 3.6321772670284683e-06, "loss": 0.2296, "step": 3911 }, { "epoch": 0.74, "grad_norm": 11.25, "learning_rate": 3.631494317245832e-06, "loss": 0.2531, "step": 3912 }, { "epoch": 0.74, "grad_norm": 9.875, "learning_rate": 3.630811261253176e-06, "loss": 0.278, "step": 3913 }, { "epoch": 0.74, "grad_norm": 11.1875, "learning_rate": 3.6301280991146156e-06, "loss": 0.2509, "step": 3914 }, { "epoch": 0.74, "grad_norm": 19.75, "learning_rate": 3.629444830894277e-06, "loss": 0.2199, "step": 3915 }, { "epoch": 0.74, "grad_norm": 9.25, "learning_rate": 3.6287614566562975e-06, "loss": 0.2762, "step": 3916 }, { "epoch": 0.74, "grad_norm": 11.875, "learning_rate": 3.6280779764648227e-06, "loss": 0.2687, "step": 3917 }, { "epoch": 0.74, "grad_norm": 10.4375, "learning_rate": 3.627394390384009e-06, "loss": 0.23, "step": 3918 }, { "epoch": 0.74, "grad_norm": 8.6875, "learning_rate": 3.626710698478023e-06, "loss": 0.3286, "step": 3919 }, { "epoch": 0.74, "grad_norm": 9.5, "learning_rate": 3.6260269008110407e-06, "loss": 0.2577, "step": 3920 }, { "epoch": 0.74, "grad_norm": 13.125, "learning_rate": 3.625342997447248e-06, "loss": 0.326, "step": 3921 }, { "epoch": 0.74, "grad_norm": 13.0625, "learning_rate": 3.6246589884508406e-06, "loss": 0.4032, "step": 3922 }, { "epoch": 0.74, "grad_norm": 19.375, "learning_rate": 3.6239748738860254e-06, "loss": 0.3264, "step": 3923 }, { "epoch": 0.74, "grad_norm": 12.25, "learning_rate": 3.6232906538170166e-06, "loss": 0.2698, "step": 3924 }, { "epoch": 0.74, "grad_norm": 11.0625, "learning_rate": 3.6226063283080417e-06, "loss": 0.2761, "step": 3925 }, { "epoch": 0.74, "grad_norm": 10.0, "learning_rate": 3.621921897423335e-06, "loss": 0.281, "step": 3926 }, { "epoch": 0.74, "grad_norm": 8.3125, "learning_rate": 3.621237361227143e-06, "loss": 0.2486, "step": 3927 }, { "epoch": 0.74, "grad_norm": 11.25, "learning_rate": 3.620552719783719e-06, "loss": 0.2628, "step": 3928 }, { "epoch": 0.74, "grad_norm": 11.3125, "learning_rate": 3.6198679731573317e-06, "loss": 0.3051, "step": 3929 }, { "epoch": 0.74, "grad_norm": 10.6875, "learning_rate": 3.619183121412253e-06, "loss": 0.2722, "step": 3930 }, { "epoch": 0.74, "grad_norm": 8.1875, "learning_rate": 3.6184981646127693e-06, "loss": 0.2733, "step": 3931 }, { "epoch": 0.74, "grad_norm": 10.6875, "learning_rate": 3.6178131028231765e-06, "loss": 0.2707, "step": 3932 }, { "epoch": 0.74, "grad_norm": 15.9375, "learning_rate": 3.617127936107777e-06, "loss": 0.3309, "step": 3933 }, { "epoch": 0.74, "grad_norm": 14.3125, "learning_rate": 3.6164426645308876e-06, "loss": 0.2833, "step": 3934 }, { "epoch": 0.74, "grad_norm": 10.1875, "learning_rate": 3.6157572881568305e-06, "loss": 0.2027, "step": 3935 }, { "epoch": 0.74, "grad_norm": 10.5, "learning_rate": 3.6150718070499426e-06, "loss": 0.2986, "step": 3936 }, { "epoch": 0.74, "grad_norm": 13.3125, "learning_rate": 3.6143862212745653e-06, "loss": 0.303, "step": 3937 }, { "epoch": 0.74, "grad_norm": 9.9375, "learning_rate": 3.6137005308950545e-06, "loss": 0.2532, "step": 3938 }, { "epoch": 0.75, "grad_norm": 9.125, "learning_rate": 3.6130147359757724e-06, "loss": 0.2221, "step": 3939 }, { "epoch": 0.75, "grad_norm": 10.3125, "learning_rate": 3.6123288365810944e-06, "loss": 0.3234, "step": 3940 }, { "epoch": 0.75, "grad_norm": 12.0625, "learning_rate": 3.6116428327754017e-06, "loss": 0.4318, "step": 3941 }, { "epoch": 0.75, "grad_norm": 16.375, "learning_rate": 3.610956724623089e-06, "loss": 0.2076, "step": 3942 }, { "epoch": 0.75, "grad_norm": 9.625, "learning_rate": 3.6102705121885586e-06, "loss": 0.3005, "step": 3943 }, { "epoch": 0.75, "grad_norm": 8.5625, "learning_rate": 3.609584195536223e-06, "loss": 0.2836, "step": 3944 }, { "epoch": 0.75, "grad_norm": 8.5625, "learning_rate": 3.608897774730506e-06, "loss": 0.2332, "step": 3945 }, { "epoch": 0.75, "grad_norm": 9.625, "learning_rate": 3.6082112498358386e-06, "loss": 0.2667, "step": 3946 }, { "epoch": 0.75, "grad_norm": 12.125, "learning_rate": 3.6075246209166626e-06, "loss": 0.3823, "step": 3947 }, { "epoch": 0.75, "grad_norm": 29.25, "learning_rate": 3.606837888037431e-06, "loss": 0.2973, "step": 3948 }, { "epoch": 0.75, "grad_norm": 9.3125, "learning_rate": 3.606151051262605e-06, "loss": 0.2564, "step": 3949 }, { "epoch": 0.75, "grad_norm": 11.4375, "learning_rate": 3.6054641106566556e-06, "loss": 0.2808, "step": 3950 }, { "epoch": 0.75, "grad_norm": 10.6875, "learning_rate": 3.604777066284064e-06, "loss": 0.2453, "step": 3951 }, { "epoch": 0.75, "grad_norm": 8.3125, "learning_rate": 3.6040899182093213e-06, "loss": 0.2564, "step": 3952 }, { "epoch": 0.75, "grad_norm": 17.75, "learning_rate": 3.6034026664969274e-06, "loss": 0.3089, "step": 3953 }, { "epoch": 0.75, "grad_norm": 10.625, "learning_rate": 3.602715311211393e-06, "loss": 0.3421, "step": 3954 }, { "epoch": 0.75, "grad_norm": 13.375, "learning_rate": 3.6020278524172385e-06, "loss": 0.2793, "step": 3955 }, { "epoch": 0.75, "grad_norm": 11.125, "learning_rate": 3.601340290178993e-06, "loss": 0.2197, "step": 3956 }, { "epoch": 0.75, "grad_norm": 8.625, "learning_rate": 3.600652624561196e-06, "loss": 0.2522, "step": 3957 }, { "epoch": 0.75, "grad_norm": 13.9375, "learning_rate": 3.599964855628397e-06, "loss": 0.3009, "step": 3958 }, { "epoch": 0.75, "grad_norm": 8.6875, "learning_rate": 3.599276983445155e-06, "loss": 0.3202, "step": 3959 }, { "epoch": 0.75, "grad_norm": 8.875, "learning_rate": 3.598589008076038e-06, "loss": 0.3197, "step": 3960 }, { "epoch": 0.75, "grad_norm": 8.8125, "learning_rate": 3.597900929585624e-06, "loss": 0.2593, "step": 3961 }, { "epoch": 0.75, "grad_norm": 8.6875, "learning_rate": 3.597212748038501e-06, "loss": 0.2719, "step": 3962 }, { "epoch": 0.75, "grad_norm": 16.875, "learning_rate": 3.5965244634992676e-06, "loss": 0.2779, "step": 3963 }, { "epoch": 0.75, "grad_norm": 8.6875, "learning_rate": 3.5958360760325294e-06, "loss": 0.2711, "step": 3964 }, { "epoch": 0.75, "grad_norm": 8.4375, "learning_rate": 3.595147585702905e-06, "loss": 0.2078, "step": 3965 }, { "epoch": 0.75, "grad_norm": 9.8125, "learning_rate": 3.594458992575019e-06, "loss": 0.3258, "step": 3966 }, { "epoch": 0.75, "grad_norm": 13.25, "learning_rate": 3.5937702967135097e-06, "loss": 0.2507, "step": 3967 }, { "epoch": 0.75, "grad_norm": 10.75, "learning_rate": 3.593081498183021e-06, "loss": 0.2745, "step": 3968 }, { "epoch": 0.75, "grad_norm": 10.0625, "learning_rate": 3.5923925970482104e-06, "loss": 0.2357, "step": 3969 }, { "epoch": 0.75, "grad_norm": 12.625, "learning_rate": 3.591703593373741e-06, "loss": 0.3429, "step": 3970 }, { "epoch": 0.75, "grad_norm": 10.4375, "learning_rate": 3.591014487224289e-06, "loss": 0.3133, "step": 3971 }, { "epoch": 0.75, "grad_norm": 9.75, "learning_rate": 3.590325278664538e-06, "loss": 0.2984, "step": 3972 }, { "epoch": 0.75, "grad_norm": 7.34375, "learning_rate": 3.5896359677591825e-06, "loss": 0.21, "step": 3973 }, { "epoch": 0.75, "grad_norm": 11.875, "learning_rate": 3.5889465545729256e-06, "loss": 0.1972, "step": 3974 }, { "epoch": 0.75, "grad_norm": 10.375, "learning_rate": 3.5882570391704808e-06, "loss": 0.3471, "step": 3975 }, { "epoch": 0.75, "grad_norm": 10.8125, "learning_rate": 3.58756742161657e-06, "loss": 0.2602, "step": 3976 }, { "epoch": 0.75, "grad_norm": 8.9375, "learning_rate": 3.586877701975926e-06, "loss": 0.2269, "step": 3977 }, { "epoch": 0.75, "grad_norm": 9.75, "learning_rate": 3.5861878803132922e-06, "loss": 0.2645, "step": 3978 }, { "epoch": 0.75, "grad_norm": 12.0625, "learning_rate": 3.585497956693418e-06, "loss": 0.3861, "step": 3979 }, { "epoch": 0.75, "grad_norm": 9.875, "learning_rate": 3.5848079311810656e-06, "loss": 0.2936, "step": 3980 }, { "epoch": 0.75, "grad_norm": 13.8125, "learning_rate": 3.584117803841005e-06, "loss": 0.3612, "step": 3981 }, { "epoch": 0.75, "grad_norm": 11.8125, "learning_rate": 3.5834275747380183e-06, "loss": 0.2801, "step": 3982 }, { "epoch": 0.75, "grad_norm": 8.8125, "learning_rate": 3.5827372439368923e-06, "loss": 0.3113, "step": 3983 }, { "epoch": 0.75, "grad_norm": 17.375, "learning_rate": 3.582046811502429e-06, "loss": 0.2852, "step": 3984 }, { "epoch": 0.75, "grad_norm": 8.5, "learning_rate": 3.581356277499435e-06, "loss": 0.3151, "step": 3985 }, { "epoch": 0.75, "grad_norm": 12.9375, "learning_rate": 3.5806656419927303e-06, "loss": 0.2916, "step": 3986 }, { "epoch": 0.75, "grad_norm": 10.375, "learning_rate": 3.5799749050471416e-06, "loss": 0.2557, "step": 3987 }, { "epoch": 0.75, "grad_norm": 8.375, "learning_rate": 3.5792840667275075e-06, "loss": 0.2626, "step": 3988 }, { "epoch": 0.75, "grad_norm": 9.1875, "learning_rate": 3.5785931270986748e-06, "loss": 0.2966, "step": 3989 }, { "epoch": 0.75, "grad_norm": 8.6875, "learning_rate": 3.5779020862254995e-06, "loss": 0.2528, "step": 3990 }, { "epoch": 0.75, "grad_norm": 10.375, "learning_rate": 3.577210944172847e-06, "loss": 0.2617, "step": 3991 }, { "epoch": 0.76, "grad_norm": 10.375, "learning_rate": 3.576519701005594e-06, "loss": 0.2239, "step": 3992 }, { "epoch": 0.76, "grad_norm": 14.75, "learning_rate": 3.575828356788624e-06, "loss": 0.3893, "step": 3993 }, { "epoch": 0.76, "grad_norm": 11.6875, "learning_rate": 3.5751369115868334e-06, "loss": 0.2714, "step": 3994 }, { "epoch": 0.76, "grad_norm": 11.8125, "learning_rate": 3.5744453654651246e-06, "loss": 0.3139, "step": 3995 }, { "epoch": 0.76, "grad_norm": 15.1875, "learning_rate": 3.5737537184884115e-06, "loss": 0.2892, "step": 3996 }, { "epoch": 0.76, "grad_norm": 8.5625, "learning_rate": 3.5730619707216162e-06, "loss": 0.2633, "step": 3997 }, { "epoch": 0.76, "grad_norm": 10.1875, "learning_rate": 3.5723701222296727e-06, "loss": 0.2306, "step": 3998 }, { "epoch": 0.76, "grad_norm": 8.0, "learning_rate": 3.571678173077521e-06, "loss": 0.2406, "step": 3999 }, { "epoch": 0.76, "grad_norm": 10.8125, "learning_rate": 3.5709861233301137e-06, "loss": 0.2498, "step": 4000 }, { "epoch": 0.76, "grad_norm": 10.6875, "learning_rate": 3.57029397305241e-06, "loss": 0.2634, "step": 4001 }, { "epoch": 0.76, "grad_norm": 8.6875, "learning_rate": 3.569601722309381e-06, "loss": 0.2746, "step": 4002 }, { "epoch": 0.76, "grad_norm": 10.625, "learning_rate": 3.5689093711660062e-06, "loss": 0.3193, "step": 4003 }, { "epoch": 0.76, "grad_norm": 11.9375, "learning_rate": 3.568216919687274e-06, "loss": 0.2485, "step": 4004 }, { "epoch": 0.76, "grad_norm": 9.5, "learning_rate": 3.567524367938184e-06, "loss": 0.2944, "step": 4005 }, { "epoch": 0.76, "grad_norm": 10.1875, "learning_rate": 3.566831715983742e-06, "loss": 0.2495, "step": 4006 }, { "epoch": 0.76, "grad_norm": 20.625, "learning_rate": 3.5661389638889675e-06, "loss": 0.2656, "step": 4007 }, { "epoch": 0.76, "grad_norm": 12.6875, "learning_rate": 3.5654461117188856e-06, "loss": 0.2696, "step": 4008 }, { "epoch": 0.76, "grad_norm": 15.25, "learning_rate": 3.564753159538532e-06, "loss": 0.3793, "step": 4009 }, { "epoch": 0.76, "grad_norm": 7.65625, "learning_rate": 3.564060107412952e-06, "loss": 0.2558, "step": 4010 }, { "epoch": 0.76, "grad_norm": 17.75, "learning_rate": 3.563366955407203e-06, "loss": 0.286, "step": 4011 }, { "epoch": 0.76, "grad_norm": 11.875, "learning_rate": 3.5626737035863456e-06, "loss": 0.3639, "step": 4012 }, { "epoch": 0.76, "grad_norm": 9.4375, "learning_rate": 3.561980352015456e-06, "loss": 0.2324, "step": 4013 }, { "epoch": 0.76, "grad_norm": 11.3125, "learning_rate": 3.5612869007596145e-06, "loss": 0.2608, "step": 4014 }, { "epoch": 0.76, "grad_norm": 14.0, "learning_rate": 3.5605933498839163e-06, "loss": 0.3274, "step": 4015 }, { "epoch": 0.76, "grad_norm": 12.4375, "learning_rate": 3.5598996994534602e-06, "loss": 0.2956, "step": 4016 }, { "epoch": 0.76, "grad_norm": 17.125, "learning_rate": 3.5592059495333585e-06, "loss": 0.3569, "step": 4017 }, { "epoch": 0.76, "grad_norm": 9.5, "learning_rate": 3.5585121001887314e-06, "loss": 0.2894, "step": 4018 }, { "epoch": 0.76, "grad_norm": 8.625, "learning_rate": 3.557818151484708e-06, "loss": 0.2261, "step": 4019 }, { "epoch": 0.76, "grad_norm": 11.5, "learning_rate": 3.557124103486428e-06, "loss": 0.2901, "step": 4020 }, { "epoch": 0.76, "grad_norm": 13.125, "learning_rate": 3.5564299562590385e-06, "loss": 0.3223, "step": 4021 }, { "epoch": 0.76, "grad_norm": 6.40625, "learning_rate": 3.555735709867699e-06, "loss": 0.1572, "step": 4022 }, { "epoch": 0.76, "grad_norm": 11.0625, "learning_rate": 3.5550413643775734e-06, "loss": 0.3008, "step": 4023 }, { "epoch": 0.76, "grad_norm": 14.25, "learning_rate": 3.5543469198538406e-06, "loss": 0.2554, "step": 4024 }, { "epoch": 0.76, "grad_norm": 11.0, "learning_rate": 3.5536523763616843e-06, "loss": 0.3436, "step": 4025 }, { "epoch": 0.76, "grad_norm": 10.375, "learning_rate": 3.5529577339663007e-06, "loss": 0.341, "step": 4026 }, { "epoch": 0.76, "grad_norm": 13.375, "learning_rate": 3.552262992732892e-06, "loss": 0.3735, "step": 4027 }, { "epoch": 0.76, "grad_norm": 13.75, "learning_rate": 3.551568152726674e-06, "loss": 0.294, "step": 4028 }, { "epoch": 0.76, "grad_norm": 12.5, "learning_rate": 3.550873214012866e-06, "loss": 0.3047, "step": 4029 }, { "epoch": 0.76, "grad_norm": 10.75, "learning_rate": 3.5501781766567033e-06, "loss": 0.3262, "step": 4030 }, { "epoch": 0.76, "grad_norm": 8.1875, "learning_rate": 3.5494830407234237e-06, "loss": 0.1884, "step": 4031 }, { "epoch": 0.76, "grad_norm": 14.125, "learning_rate": 3.5487878062782806e-06, "loss": 0.2773, "step": 4032 }, { "epoch": 0.76, "grad_norm": 9.4375, "learning_rate": 3.5480924733865313e-06, "loss": 0.2899, "step": 4033 }, { "epoch": 0.76, "grad_norm": 12.3125, "learning_rate": 3.547397042113446e-06, "loss": 0.2908, "step": 4034 }, { "epoch": 0.76, "grad_norm": 10.9375, "learning_rate": 3.5467015125243022e-06, "loss": 0.2708, "step": 4035 }, { "epoch": 0.76, "grad_norm": 10.0, "learning_rate": 3.546005884684387e-06, "loss": 0.2338, "step": 4036 }, { "epoch": 0.76, "grad_norm": 8.9375, "learning_rate": 3.545310158658998e-06, "loss": 0.2829, "step": 4037 }, { "epoch": 0.76, "grad_norm": 16.375, "learning_rate": 3.544614334513439e-06, "loss": 0.244, "step": 4038 }, { "epoch": 0.76, "grad_norm": 11.0625, "learning_rate": 3.5439184123130273e-06, "loss": 0.3275, "step": 4039 }, { "epoch": 0.76, "grad_norm": 10.25, "learning_rate": 3.543222392123085e-06, "loss": 0.3306, "step": 4040 }, { "epoch": 0.76, "grad_norm": 9.75, "learning_rate": 3.5425262740089466e-06, "loss": 0.2268, "step": 4041 }, { "epoch": 0.76, "grad_norm": 9.5, "learning_rate": 3.5418300580359545e-06, "loss": 0.2853, "step": 4042 }, { "epoch": 0.76, "grad_norm": 11.375, "learning_rate": 3.5411337442694605e-06, "loss": 0.3144, "step": 4043 }, { "epoch": 0.77, "grad_norm": 12.75, "learning_rate": 3.540437332774825e-06, "loss": 0.284, "step": 4044 }, { "epoch": 0.77, "grad_norm": 9.8125, "learning_rate": 3.5397408236174185e-06, "loss": 0.2446, "step": 4045 }, { "epoch": 0.77, "grad_norm": 11.8125, "learning_rate": 3.53904421686262e-06, "loss": 0.2794, "step": 4046 }, { "epoch": 0.77, "grad_norm": 18.75, "learning_rate": 3.538347512575818e-06, "loss": 0.3267, "step": 4047 }, { "epoch": 0.77, "grad_norm": 12.625, "learning_rate": 3.5376507108224107e-06, "loss": 0.2941, "step": 4048 }, { "epoch": 0.77, "grad_norm": 12.9375, "learning_rate": 3.5369538116678038e-06, "loss": 0.2522, "step": 4049 }, { "epoch": 0.77, "grad_norm": 10.0625, "learning_rate": 3.5362568151774137e-06, "loss": 0.2962, "step": 4050 }, { "epoch": 0.77, "grad_norm": 11.5, "learning_rate": 3.5355597214166652e-06, "loss": 0.2606, "step": 4051 }, { "epoch": 0.77, "grad_norm": 10.6875, "learning_rate": 3.5348625304509926e-06, "loss": 0.2884, "step": 4052 }, { "epoch": 0.77, "grad_norm": 9.625, "learning_rate": 3.5341652423458396e-06, "loss": 0.283, "step": 4053 }, { "epoch": 0.77, "grad_norm": 13.8125, "learning_rate": 3.533467857166657e-06, "loss": 0.2893, "step": 4054 }, { "epoch": 0.77, "grad_norm": 10.0, "learning_rate": 3.532770374978908e-06, "loss": 0.2403, "step": 4055 }, { "epoch": 0.77, "grad_norm": 11.75, "learning_rate": 3.5320727958480622e-06, "loss": 0.2734, "step": 4056 }, { "epoch": 0.77, "grad_norm": 11.4375, "learning_rate": 3.5313751198396e-06, "loss": 0.3271, "step": 4057 }, { "epoch": 0.77, "grad_norm": 10.75, "learning_rate": 3.53067734701901e-06, "loss": 0.325, "step": 4058 }, { "epoch": 0.77, "grad_norm": 11.125, "learning_rate": 3.5299794774517892e-06, "loss": 0.2325, "step": 4059 }, { "epoch": 0.77, "grad_norm": 11.3125, "learning_rate": 3.5292815112034457e-06, "loss": 0.2355, "step": 4060 }, { "epoch": 0.77, "grad_norm": 12.4375, "learning_rate": 3.5285834483394947e-06, "loss": 0.397, "step": 4061 }, { "epoch": 0.77, "grad_norm": 10.5625, "learning_rate": 3.5278852889254624e-06, "loss": 0.285, "step": 4062 }, { "epoch": 0.77, "grad_norm": 11.5, "learning_rate": 3.5271870330268808e-06, "loss": 0.2803, "step": 4063 }, { "epoch": 0.77, "grad_norm": 9.8125, "learning_rate": 3.5264886807092957e-06, "loss": 0.2802, "step": 4064 }, { "epoch": 0.77, "grad_norm": 9.8125, "learning_rate": 3.5257902320382574e-06, "loss": 0.2966, "step": 4065 }, { "epoch": 0.77, "grad_norm": 8.5625, "learning_rate": 3.5250916870793284e-06, "loss": 0.2088, "step": 4066 }, { "epoch": 0.77, "grad_norm": 8.8125, "learning_rate": 3.524393045898078e-06, "loss": 0.3011, "step": 4067 }, { "epoch": 0.77, "grad_norm": 12.625, "learning_rate": 3.5236943085600873e-06, "loss": 0.361, "step": 4068 }, { "epoch": 0.77, "grad_norm": 14.5, "learning_rate": 3.5229954751309426e-06, "loss": 0.351, "step": 4069 }, { "epoch": 0.77, "grad_norm": 15.0, "learning_rate": 3.522296545676243e-06, "loss": 0.2621, "step": 4070 }, { "epoch": 0.77, "grad_norm": 12.0625, "learning_rate": 3.5215975202615936e-06, "loss": 0.3234, "step": 4071 }, { "epoch": 0.77, "grad_norm": 13.5625, "learning_rate": 3.5208983989526113e-06, "loss": 0.31, "step": 4072 }, { "epoch": 0.77, "grad_norm": 10.25, "learning_rate": 3.520199181814919e-06, "loss": 0.3105, "step": 4073 }, { "epoch": 0.77, "grad_norm": 9.0625, "learning_rate": 3.5194998689141517e-06, "loss": 0.2582, "step": 4074 }, { "epoch": 0.77, "grad_norm": 7.71875, "learning_rate": 3.5188004603159506e-06, "loss": 0.2514, "step": 4075 }, { "epoch": 0.77, "grad_norm": 19.375, "learning_rate": 3.5181009560859674e-06, "loss": 0.2773, "step": 4076 }, { "epoch": 0.77, "grad_norm": 7.9375, "learning_rate": 3.5174013562898625e-06, "loss": 0.2049, "step": 4077 }, { "epoch": 0.77, "grad_norm": 9.25, "learning_rate": 3.5167016609933056e-06, "loss": 0.236, "step": 4078 }, { "epoch": 0.77, "grad_norm": 9.0625, "learning_rate": 3.5160018702619747e-06, "loss": 0.272, "step": 4079 }, { "epoch": 0.77, "grad_norm": 10.125, "learning_rate": 3.5153019841615576e-06, "loss": 0.2551, "step": 4080 }, { "epoch": 0.77, "grad_norm": 8.25, "learning_rate": 3.51460200275775e-06, "loss": 0.2461, "step": 4081 }, { "epoch": 0.77, "grad_norm": 7.6875, "learning_rate": 3.5139019261162573e-06, "loss": 0.2652, "step": 4082 }, { "epoch": 0.77, "grad_norm": 17.875, "learning_rate": 3.513201754302793e-06, "loss": 0.3061, "step": 4083 }, { "epoch": 0.77, "grad_norm": 12.375, "learning_rate": 3.512501487383081e-06, "loss": 0.329, "step": 4084 }, { "epoch": 0.77, "grad_norm": 10.0625, "learning_rate": 3.511801125422853e-06, "loss": 0.2502, "step": 4085 }, { "epoch": 0.77, "grad_norm": 13.5625, "learning_rate": 3.5111006684878503e-06, "loss": 0.3755, "step": 4086 }, { "epoch": 0.77, "grad_norm": 9.6875, "learning_rate": 3.5104001166438224e-06, "loss": 0.295, "step": 4087 }, { "epoch": 0.77, "grad_norm": 8.625, "learning_rate": 3.509699469956527e-06, "loss": 0.2559, "step": 4088 }, { "epoch": 0.77, "grad_norm": 14.9375, "learning_rate": 3.5089987284917325e-06, "loss": 0.3514, "step": 4089 }, { "epoch": 0.77, "grad_norm": 13.1875, "learning_rate": 3.508297892315217e-06, "loss": 0.3933, "step": 4090 }, { "epoch": 0.77, "grad_norm": 17.375, "learning_rate": 3.507596961492763e-06, "loss": 0.2881, "step": 4091 }, { "epoch": 0.77, "grad_norm": 8.8125, "learning_rate": 3.5068959360901666e-06, "loss": 0.2104, "step": 4092 }, { "epoch": 0.77, "grad_norm": 10.4375, "learning_rate": 3.506194816173231e-06, "loss": 0.3021, "step": 4093 }, { "epoch": 0.77, "grad_norm": 7.90625, "learning_rate": 3.5054936018077684e-06, "loss": 0.2478, "step": 4094 }, { "epoch": 0.77, "grad_norm": 11.0625, "learning_rate": 3.504792293059599e-06, "loss": 0.3079, "step": 4095 }, { "epoch": 0.77, "grad_norm": 13.0, "learning_rate": 3.5040908899945526e-06, "loss": 0.1978, "step": 4096 }, { "epoch": 0.78, "grad_norm": 12.875, "learning_rate": 3.5033893926784677e-06, "loss": 0.2262, "step": 4097 }, { "epoch": 0.78, "grad_norm": 19.75, "learning_rate": 3.5026878011771926e-06, "loss": 0.2259, "step": 4098 }, { "epoch": 0.78, "grad_norm": 11.0625, "learning_rate": 3.5019861155565826e-06, "loss": 0.2178, "step": 4099 }, { "epoch": 0.78, "grad_norm": 10.75, "learning_rate": 3.5012843358825044e-06, "loss": 0.2856, "step": 4100 }, { "epoch": 0.78, "grad_norm": 9.5625, "learning_rate": 3.50058246222083e-06, "loss": 0.2495, "step": 4101 }, { "epoch": 0.78, "grad_norm": 13.8125, "learning_rate": 3.499880494637443e-06, "loss": 0.26, "step": 4102 }, { "epoch": 0.78, "grad_norm": 13.875, "learning_rate": 3.499178433198236e-06, "loss": 0.3525, "step": 4103 }, { "epoch": 0.78, "grad_norm": 22.0, "learning_rate": 3.4984762779691085e-06, "loss": 0.3246, "step": 4104 }, { "epoch": 0.78, "grad_norm": 16.125, "learning_rate": 3.4977740290159694e-06, "loss": 0.2643, "step": 4105 }, { "epoch": 0.78, "grad_norm": 12.375, "learning_rate": 3.4970716864047377e-06, "loss": 0.3002, "step": 4106 }, { "epoch": 0.78, "grad_norm": 9.6875, "learning_rate": 3.4963692502013395e-06, "loss": 0.2947, "step": 4107 }, { "epoch": 0.78, "grad_norm": 11.375, "learning_rate": 3.4956667204717103e-06, "loss": 0.3246, "step": 4108 }, { "epoch": 0.78, "grad_norm": 9.5, "learning_rate": 3.4949640972817957e-06, "loss": 0.2607, "step": 4109 }, { "epoch": 0.78, "grad_norm": 8.0625, "learning_rate": 3.4942613806975467e-06, "loss": 0.2549, "step": 4110 }, { "epoch": 0.78, "grad_norm": 14.125, "learning_rate": 3.493558570784927e-06, "loss": 0.3112, "step": 4111 }, { "epoch": 0.78, "grad_norm": 17.25, "learning_rate": 3.4928556676099068e-06, "loss": 0.2704, "step": 4112 }, { "epoch": 0.78, "grad_norm": 11.375, "learning_rate": 3.4921526712384653e-06, "loss": 0.3101, "step": 4113 }, { "epoch": 0.78, "grad_norm": 15.8125, "learning_rate": 3.491449581736591e-06, "loss": 0.3463, "step": 4114 }, { "epoch": 0.78, "grad_norm": 9.5, "learning_rate": 3.4907463991702805e-06, "loss": 0.2749, "step": 4115 }, { "epoch": 0.78, "grad_norm": 22.125, "learning_rate": 3.490043123605541e-06, "loss": 0.2832, "step": 4116 }, { "epoch": 0.78, "grad_norm": 8.8125, "learning_rate": 3.4893397551083835e-06, "loss": 0.2634, "step": 4117 }, { "epoch": 0.78, "grad_norm": 9.0, "learning_rate": 3.488636293744835e-06, "loss": 0.2805, "step": 4118 }, { "epoch": 0.78, "grad_norm": 15.4375, "learning_rate": 3.4879327395809244e-06, "loss": 0.2834, "step": 4119 }, { "epoch": 0.78, "grad_norm": 56.0, "learning_rate": 3.487229092682694e-06, "loss": 0.3797, "step": 4120 }, { "epoch": 0.78, "grad_norm": 9.1875, "learning_rate": 3.4865253531161913e-06, "loss": 0.2206, "step": 4121 }, { "epoch": 0.78, "grad_norm": 13.0625, "learning_rate": 3.4858215209474765e-06, "loss": 0.2726, "step": 4122 }, { "epoch": 0.78, "grad_norm": 21.25, "learning_rate": 3.485117596242615e-06, "loss": 0.298, "step": 4123 }, { "epoch": 0.78, "grad_norm": 10.9375, "learning_rate": 3.484413579067682e-06, "loss": 0.2594, "step": 4124 }, { "epoch": 0.78, "grad_norm": 8.4375, "learning_rate": 3.483709469488762e-06, "loss": 0.2929, "step": 4125 }, { "epoch": 0.78, "grad_norm": 20.75, "learning_rate": 3.483005267571947e-06, "loss": 0.361, "step": 4126 }, { "epoch": 0.78, "grad_norm": 8.75, "learning_rate": 3.482300973383339e-06, "loss": 0.3574, "step": 4127 }, { "epoch": 0.78, "grad_norm": 7.46875, "learning_rate": 3.4815965869890477e-06, "loss": 0.2712, "step": 4128 }, { "epoch": 0.78, "grad_norm": 11.375, "learning_rate": 3.4808921084551923e-06, "loss": 0.204, "step": 4129 }, { "epoch": 0.78, "grad_norm": 11.1875, "learning_rate": 3.4801875378478988e-06, "loss": 0.3006, "step": 4130 }, { "epoch": 0.78, "grad_norm": 10.375, "learning_rate": 3.4794828752333048e-06, "loss": 0.2626, "step": 4131 }, { "epoch": 0.78, "grad_norm": 14.6875, "learning_rate": 3.4787781206775534e-06, "loss": 0.3226, "step": 4132 }, { "epoch": 0.78, "grad_norm": 7.8125, "learning_rate": 3.478073274246799e-06, "loss": 0.1969, "step": 4133 }, { "epoch": 0.78, "grad_norm": 12.25, "learning_rate": 3.477368336007203e-06, "loss": 0.275, "step": 4134 }, { "epoch": 0.78, "grad_norm": 17.0, "learning_rate": 3.4766633060249356e-06, "loss": 0.3145, "step": 4135 }, { "epoch": 0.78, "grad_norm": 13.9375, "learning_rate": 3.4759581843661768e-06, "loss": 0.3065, "step": 4136 }, { "epoch": 0.78, "grad_norm": 14.0, "learning_rate": 3.475252971097113e-06, "loss": 0.3785, "step": 4137 }, { "epoch": 0.78, "grad_norm": 9.1875, "learning_rate": 3.474547666283941e-06, "loss": 0.3196, "step": 4138 }, { "epoch": 0.78, "grad_norm": 11.6875, "learning_rate": 3.4738422699928658e-06, "loss": 0.2504, "step": 4139 }, { "epoch": 0.78, "grad_norm": 11.8125, "learning_rate": 3.473136782290101e-06, "loss": 0.3072, "step": 4140 }, { "epoch": 0.78, "grad_norm": 13.3125, "learning_rate": 3.4724312032418683e-06, "loss": 0.375, "step": 4141 }, { "epoch": 0.78, "grad_norm": 11.4375, "learning_rate": 3.4717255329143988e-06, "loss": 0.3545, "step": 4142 }, { "epoch": 0.78, "grad_norm": 12.5, "learning_rate": 3.4710197713739313e-06, "loss": 0.3107, "step": 4143 }, { "epoch": 0.78, "grad_norm": 9.875, "learning_rate": 3.4703139186867137e-06, "loss": 0.2939, "step": 4144 }, { "epoch": 0.78, "grad_norm": 18.625, "learning_rate": 3.4696079749190016e-06, "loss": 0.3309, "step": 4145 }, { "epoch": 0.78, "grad_norm": 15.625, "learning_rate": 3.468901940137061e-06, "loss": 0.2249, "step": 4146 }, { "epoch": 0.78, "grad_norm": 14.5, "learning_rate": 3.4681958144071645e-06, "loss": 0.315, "step": 4147 }, { "epoch": 0.78, "grad_norm": 9.125, "learning_rate": 3.467489597795594e-06, "loss": 0.2764, "step": 4148 }, { "epoch": 0.78, "grad_norm": 10.875, "learning_rate": 3.466783290368641e-06, "loss": 0.2682, "step": 4149 }, { "epoch": 0.79, "grad_norm": 9.5625, "learning_rate": 3.4660768921926036e-06, "loss": 0.245, "step": 4150 }, { "epoch": 0.79, "grad_norm": 12.75, "learning_rate": 3.4653704033337888e-06, "loss": 0.3751, "step": 4151 }, { "epoch": 0.79, "grad_norm": 15.0, "learning_rate": 3.464663823858514e-06, "loss": 0.3063, "step": 4152 }, { "epoch": 0.79, "grad_norm": 10.1875, "learning_rate": 3.4639571538331023e-06, "loss": 0.1903, "step": 4153 }, { "epoch": 0.79, "grad_norm": 10.3125, "learning_rate": 3.4632503933238875e-06, "loss": 0.2419, "step": 4154 }, { "epoch": 0.79, "grad_norm": 11.125, "learning_rate": 3.462543542397211e-06, "loss": 0.2885, "step": 4155 }, { "epoch": 0.79, "grad_norm": 10.1875, "learning_rate": 3.4618366011194237e-06, "loss": 0.2572, "step": 4156 }, { "epoch": 0.79, "grad_norm": 11.6875, "learning_rate": 3.461129569556882e-06, "loss": 0.3164, "step": 4157 }, { "epoch": 0.79, "grad_norm": 9.875, "learning_rate": 3.460422447775955e-06, "loss": 0.3095, "step": 4158 }, { "epoch": 0.79, "grad_norm": 12.625, "learning_rate": 3.4597152358430163e-06, "loss": 0.2628, "step": 4159 }, { "epoch": 0.79, "grad_norm": 9.1875, "learning_rate": 3.4590079338244516e-06, "loss": 0.2868, "step": 4160 }, { "epoch": 0.79, "grad_norm": 8.25, "learning_rate": 3.458300541786651e-06, "loss": 0.2644, "step": 4161 }, { "epoch": 0.79, "grad_norm": 9.4375, "learning_rate": 3.4575930597960183e-06, "loss": 0.2462, "step": 4162 }, { "epoch": 0.79, "grad_norm": 10.875, "learning_rate": 3.4568854879189595e-06, "loss": 0.2721, "step": 4163 }, { "epoch": 0.79, "grad_norm": 11.9375, "learning_rate": 3.4561778262218954e-06, "loss": 0.3457, "step": 4164 }, { "epoch": 0.79, "grad_norm": 12.75, "learning_rate": 3.4554700747712496e-06, "loss": 0.3038, "step": 4165 }, { "epoch": 0.79, "grad_norm": 9.0625, "learning_rate": 3.454762233633458e-06, "loss": 0.2576, "step": 4166 }, { "epoch": 0.79, "grad_norm": 7.9375, "learning_rate": 3.4540543028749623e-06, "loss": 0.2034, "step": 4167 }, { "epoch": 0.79, "grad_norm": 10.25, "learning_rate": 3.4533462825622147e-06, "loss": 0.2996, "step": 4168 }, { "epoch": 0.79, "grad_norm": 10.375, "learning_rate": 3.452638172761676e-06, "loss": 0.3163, "step": 4169 }, { "epoch": 0.79, "grad_norm": 12.1875, "learning_rate": 3.4519299735398125e-06, "loss": 0.2458, "step": 4170 }, { "epoch": 0.79, "grad_norm": 11.3125, "learning_rate": 3.4512216849631027e-06, "loss": 0.2513, "step": 4171 }, { "epoch": 0.79, "grad_norm": 12.0625, "learning_rate": 3.450513307098029e-06, "loss": 0.2855, "step": 4172 }, { "epoch": 0.79, "grad_norm": 16.375, "learning_rate": 3.4498048400110874e-06, "loss": 0.3247, "step": 4173 }, { "epoch": 0.79, "grad_norm": 13.8125, "learning_rate": 3.4490962837687776e-06, "loss": 0.24, "step": 4174 }, { "epoch": 0.79, "grad_norm": 11.6875, "learning_rate": 3.448387638437611e-06, "loss": 0.2394, "step": 4175 }, { "epoch": 0.79, "grad_norm": 9.3125, "learning_rate": 3.4476789040841046e-06, "loss": 0.3023, "step": 4176 }, { "epoch": 0.79, "grad_norm": 8.4375, "learning_rate": 3.446970080774787e-06, "loss": 0.3122, "step": 4177 }, { "epoch": 0.79, "grad_norm": 14.4375, "learning_rate": 3.4462611685761917e-06, "loss": 0.27, "step": 4178 }, { "epoch": 0.79, "grad_norm": 9.6875, "learning_rate": 3.4455521675548632e-06, "loss": 0.3093, "step": 4179 }, { "epoch": 0.79, "grad_norm": 12.375, "learning_rate": 3.4448430777773534e-06, "loss": 0.2769, "step": 4180 }, { "epoch": 0.79, "grad_norm": 12.8125, "learning_rate": 3.444133899310222e-06, "loss": 0.2775, "step": 4181 }, { "epoch": 0.79, "grad_norm": 14.0, "learning_rate": 3.443424632220037e-06, "loss": 0.3083, "step": 4182 }, { "epoch": 0.79, "grad_norm": 9.375, "learning_rate": 3.442715276573376e-06, "loss": 0.2838, "step": 4183 }, { "epoch": 0.79, "grad_norm": 12.0, "learning_rate": 3.4420058324368244e-06, "loss": 0.313, "step": 4184 }, { "epoch": 0.79, "grad_norm": 10.25, "learning_rate": 3.441296299876975e-06, "loss": 0.2476, "step": 4185 }, { "epoch": 0.79, "grad_norm": 10.875, "learning_rate": 3.44058667896043e-06, "loss": 0.2949, "step": 4186 }, { "epoch": 0.79, "grad_norm": 10.1875, "learning_rate": 3.439876969753798e-06, "loss": 0.2638, "step": 4187 }, { "epoch": 0.79, "grad_norm": 11.6875, "learning_rate": 3.4391671723236995e-06, "loss": 0.3244, "step": 4188 }, { "epoch": 0.79, "grad_norm": 18.25, "learning_rate": 3.4384572867367595e-06, "loss": 0.3436, "step": 4189 }, { "epoch": 0.79, "grad_norm": 14.6875, "learning_rate": 3.437747313059613e-06, "loss": 0.2864, "step": 4190 }, { "epoch": 0.79, "grad_norm": 11.0625, "learning_rate": 3.4370372513589035e-06, "loss": 0.2851, "step": 4191 }, { "epoch": 0.79, "grad_norm": 10.3125, "learning_rate": 3.4363271017012827e-06, "loss": 0.3109, "step": 4192 }, { "epoch": 0.79, "grad_norm": 11.1875, "learning_rate": 3.4356168641534095e-06, "loss": 0.3518, "step": 4193 }, { "epoch": 0.79, "grad_norm": 12.8125, "learning_rate": 3.4349065387819526e-06, "loss": 0.2604, "step": 4194 }, { "epoch": 0.79, "grad_norm": 11.5, "learning_rate": 3.434196125653587e-06, "loss": 0.2487, "step": 4195 }, { "epoch": 0.79, "grad_norm": 10.75, "learning_rate": 3.4334856248349985e-06, "loss": 0.2816, "step": 4196 }, { "epoch": 0.79, "grad_norm": 12.9375, "learning_rate": 3.4327750363928785e-06, "loss": 0.2558, "step": 4197 }, { "epoch": 0.79, "grad_norm": 10.25, "learning_rate": 3.4320643603939284e-06, "loss": 0.3188, "step": 4198 }, { "epoch": 0.79, "grad_norm": 9.375, "learning_rate": 3.4313535969048575e-06, "loss": 0.3105, "step": 4199 }, { "epoch": 0.79, "grad_norm": 13.375, "learning_rate": 3.430642745992382e-06, "loss": 0.2647, "step": 4200 }, { "epoch": 0.79, "grad_norm": 9.8125, "learning_rate": 3.429931807723228e-06, "loss": 0.2473, "step": 4201 }, { "epoch": 0.79, "grad_norm": 11.25, "learning_rate": 3.4292207821641303e-06, "loss": 0.2784, "step": 4202 }, { "epoch": 0.8, "grad_norm": 12.375, "learning_rate": 3.4285096693818286e-06, "loss": 0.2827, "step": 4203 }, { "epoch": 0.8, "grad_norm": 11.9375, "learning_rate": 3.4277984694430756e-06, "loss": 0.3307, "step": 4204 }, { "epoch": 0.8, "grad_norm": 11.5625, "learning_rate": 3.427087182414627e-06, "loss": 0.2622, "step": 4205 }, { "epoch": 0.8, "grad_norm": 9.0, "learning_rate": 3.4263758083632503e-06, "loss": 0.2897, "step": 4206 }, { "epoch": 0.8, "grad_norm": 9.5, "learning_rate": 3.42566434735572e-06, "loss": 0.2428, "step": 4207 }, { "epoch": 0.8, "grad_norm": 12.0625, "learning_rate": 3.4249527994588193e-06, "loss": 0.3045, "step": 4208 }, { "epoch": 0.8, "grad_norm": 10.0625, "learning_rate": 3.4242411647393374e-06, "loss": 0.2297, "step": 4209 }, { "epoch": 0.8, "grad_norm": 15.0625, "learning_rate": 3.4235294432640766e-06, "loss": 0.3659, "step": 4210 }, { "epoch": 0.8, "grad_norm": 9.5, "learning_rate": 3.42281763509984e-06, "loss": 0.3208, "step": 4211 }, { "epoch": 0.8, "grad_norm": 13.375, "learning_rate": 3.422105740313446e-06, "loss": 0.3488, "step": 4212 }, { "epoch": 0.8, "grad_norm": 10.25, "learning_rate": 3.421393758971717e-06, "loss": 0.2774, "step": 4213 }, { "epoch": 0.8, "grad_norm": 19.625, "learning_rate": 3.420681691141484e-06, "loss": 0.2359, "step": 4214 }, { "epoch": 0.8, "grad_norm": 9.8125, "learning_rate": 3.4199695368895886e-06, "loss": 0.2614, "step": 4215 }, { "epoch": 0.8, "grad_norm": 12.25, "learning_rate": 3.419257296282876e-06, "loss": 0.2118, "step": 4216 }, { "epoch": 0.8, "grad_norm": 8.875, "learning_rate": 3.4185449693882044e-06, "loss": 0.2489, "step": 4217 }, { "epoch": 0.8, "grad_norm": 12.0625, "learning_rate": 3.4178325562724355e-06, "loss": 0.328, "step": 4218 }, { "epoch": 0.8, "grad_norm": 8.375, "learning_rate": 3.417120057002445e-06, "loss": 0.2218, "step": 4219 }, { "epoch": 0.8, "grad_norm": 15.3125, "learning_rate": 3.416407471645109e-06, "loss": 0.3148, "step": 4220 }, { "epoch": 0.8, "grad_norm": 10.875, "learning_rate": 3.4156948002673185e-06, "loss": 0.2453, "step": 4221 }, { "epoch": 0.8, "grad_norm": 10.8125, "learning_rate": 3.4149820429359682e-06, "loss": 0.2662, "step": 4222 }, { "epoch": 0.8, "grad_norm": 9.1875, "learning_rate": 3.414269199717964e-06, "loss": 0.2344, "step": 4223 }, { "epoch": 0.8, "grad_norm": 12.625, "learning_rate": 3.4135562706802172e-06, "loss": 0.2819, "step": 4224 }, { "epoch": 0.8, "grad_norm": 20.375, "learning_rate": 3.4128432558896498e-06, "loss": 0.3413, "step": 4225 }, { "epoch": 0.8, "grad_norm": 11.75, "learning_rate": 3.4121301554131887e-06, "loss": 0.2521, "step": 4226 }, { "epoch": 0.8, "grad_norm": 7.9375, "learning_rate": 3.4114169693177713e-06, "loss": 0.2412, "step": 4227 }, { "epoch": 0.8, "grad_norm": 12.5625, "learning_rate": 3.4107036976703423e-06, "loss": 0.3446, "step": 4228 }, { "epoch": 0.8, "grad_norm": 10.6875, "learning_rate": 3.409990340537854e-06, "loss": 0.3085, "step": 4229 }, { "epoch": 0.8, "grad_norm": 11.1875, "learning_rate": 3.409276897987268e-06, "loss": 0.2395, "step": 4230 }, { "epoch": 0.8, "grad_norm": 9.1875, "learning_rate": 3.4085633700855518e-06, "loss": 0.2827, "step": 4231 }, { "epoch": 0.8, "grad_norm": 9.4375, "learning_rate": 3.407849756899683e-06, "loss": 0.2527, "step": 4232 }, { "epoch": 0.8, "grad_norm": 8.8125, "learning_rate": 3.407136058496647e-06, "loss": 0.2845, "step": 4233 }, { "epoch": 0.8, "grad_norm": 12.5, "learning_rate": 3.406422274943435e-06, "loss": 0.2279, "step": 4234 }, { "epoch": 0.8, "grad_norm": 11.9375, "learning_rate": 3.405708406307049e-06, "loss": 0.3002, "step": 4235 }, { "epoch": 0.8, "grad_norm": 11.3125, "learning_rate": 3.4049944526544966e-06, "loss": 0.2915, "step": 4236 }, { "epoch": 0.8, "grad_norm": 7.9375, "learning_rate": 3.404280414052795e-06, "loss": 0.2438, "step": 4237 }, { "epoch": 0.8, "grad_norm": 11.0625, "learning_rate": 3.4035662905689696e-06, "loss": 0.26, "step": 4238 }, { "epoch": 0.8, "grad_norm": 9.25, "learning_rate": 3.4028520822700524e-06, "loss": 0.2868, "step": 4239 }, { "epoch": 0.8, "grad_norm": 9.0, "learning_rate": 3.402137789223084e-06, "loss": 0.2646, "step": 4240 }, { "epoch": 0.8, "grad_norm": 10.75, "learning_rate": 3.401423411495113e-06, "loss": 0.2358, "step": 4241 }, { "epoch": 0.8, "grad_norm": 9.6875, "learning_rate": 3.4007089491531957e-06, "loss": 0.2579, "step": 4242 }, { "epoch": 0.8, "grad_norm": 8.625, "learning_rate": 3.399994402264397e-06, "loss": 0.255, "step": 4243 }, { "epoch": 0.8, "grad_norm": 12.875, "learning_rate": 3.3992797708957897e-06, "loss": 0.2689, "step": 4244 }, { "epoch": 0.8, "grad_norm": 9.625, "learning_rate": 3.3985650551144527e-06, "loss": 0.2258, "step": 4245 }, { "epoch": 0.8, "grad_norm": 12.1875, "learning_rate": 3.3978502549874764e-06, "loss": 0.2781, "step": 4246 }, { "epoch": 0.8, "grad_norm": 15.375, "learning_rate": 3.3971353705819542e-06, "loss": 0.2748, "step": 4247 }, { "epoch": 0.8, "grad_norm": 11.125, "learning_rate": 3.3964204019649933e-06, "loss": 0.2693, "step": 4248 }, { "epoch": 0.8, "grad_norm": 11.3125, "learning_rate": 3.3957053492037027e-06, "loss": 0.2925, "step": 4249 }, { "epoch": 0.8, "grad_norm": 11.375, "learning_rate": 3.3949902123652044e-06, "loss": 0.2946, "step": 4250 }, { "epoch": 0.8, "grad_norm": 10.5, "learning_rate": 3.394274991516625e-06, "loss": 0.267, "step": 4251 }, { "epoch": 0.8, "grad_norm": 11.1875, "learning_rate": 3.3935596867251014e-06, "loss": 0.2851, "step": 4252 }, { "epoch": 0.8, "grad_norm": 12.625, "learning_rate": 3.392844298057776e-06, "loss": 0.2964, "step": 4253 }, { "epoch": 0.8, "grad_norm": 13.125, "learning_rate": 3.3921288255818e-06, "loss": 0.2479, "step": 4254 }, { "epoch": 0.8, "grad_norm": 10.125, "learning_rate": 3.3914132693643343e-06, "loss": 0.2699, "step": 4255 }, { "epoch": 0.81, "grad_norm": 11.9375, "learning_rate": 3.3906976294725433e-06, "loss": 0.2478, "step": 4256 }, { "epoch": 0.81, "grad_norm": 7.09375, "learning_rate": 3.3899819059736056e-06, "loss": 0.1503, "step": 4257 }, { "epoch": 0.81, "grad_norm": 11.9375, "learning_rate": 3.3892660989347004e-06, "loss": 0.333, "step": 4258 }, { "epoch": 0.81, "grad_norm": 16.375, "learning_rate": 3.3885502084230217e-06, "loss": 0.264, "step": 4259 }, { "epoch": 0.81, "grad_norm": 7.3125, "learning_rate": 3.3878342345057654e-06, "loss": 0.2425, "step": 4260 }, { "epoch": 0.81, "grad_norm": 16.75, "learning_rate": 3.3871181772501393e-06, "loss": 0.3108, "step": 4261 }, { "epoch": 0.81, "grad_norm": 35.0, "learning_rate": 3.386402036723357e-06, "loss": 0.3467, "step": 4262 }, { "epoch": 0.81, "grad_norm": 13.375, "learning_rate": 3.385685812992641e-06, "loss": 0.2412, "step": 4263 }, { "epoch": 0.81, "grad_norm": 7.28125, "learning_rate": 3.38496950612522e-06, "loss": 0.2225, "step": 4264 }, { "epoch": 0.81, "grad_norm": 9.9375, "learning_rate": 3.3842531161883337e-06, "loss": 0.2723, "step": 4265 }, { "epoch": 0.81, "grad_norm": 15.625, "learning_rate": 3.3835366432492245e-06, "loss": 0.2499, "step": 4266 }, { "epoch": 0.81, "grad_norm": 12.0625, "learning_rate": 3.382820087375149e-06, "loss": 0.219, "step": 4267 }, { "epoch": 0.81, "grad_norm": 7.5, "learning_rate": 3.3821034486333646e-06, "loss": 0.2688, "step": 4268 }, { "epoch": 0.81, "grad_norm": 13.625, "learning_rate": 3.3813867270911426e-06, "loss": 0.4226, "step": 4269 }, { "epoch": 0.81, "grad_norm": 11.875, "learning_rate": 3.3806699228157593e-06, "loss": 0.2508, "step": 4270 }, { "epoch": 0.81, "grad_norm": 10.375, "learning_rate": 3.3799530358744982e-06, "loss": 0.2747, "step": 4271 }, { "epoch": 0.81, "grad_norm": 10.6875, "learning_rate": 3.3792360663346512e-06, "loss": 0.2385, "step": 4272 }, { "epoch": 0.81, "grad_norm": 15.875, "learning_rate": 3.3785190142635192e-06, "loss": 0.3631, "step": 4273 }, { "epoch": 0.81, "grad_norm": 10.625, "learning_rate": 3.377801879728409e-06, "loss": 0.2418, "step": 4274 }, { "epoch": 0.81, "grad_norm": 9.5625, "learning_rate": 3.3770846627966365e-06, "loss": 0.2572, "step": 4275 }, { "epoch": 0.81, "grad_norm": 12.25, "learning_rate": 3.3763673635355233e-06, "loss": 0.2798, "step": 4276 }, { "epoch": 0.81, "grad_norm": 11.75, "learning_rate": 3.3756499820124016e-06, "loss": 0.292, "step": 4277 }, { "epoch": 0.81, "grad_norm": 11.125, "learning_rate": 3.3749325182946096e-06, "loss": 0.2632, "step": 4278 }, { "epoch": 0.81, "grad_norm": 9.6875, "learning_rate": 3.374214972449493e-06, "loss": 0.3416, "step": 4279 }, { "epoch": 0.81, "grad_norm": 12.5, "learning_rate": 3.373497344544406e-06, "loss": 0.2731, "step": 4280 }, { "epoch": 0.81, "grad_norm": 10.75, "learning_rate": 3.3727796346467102e-06, "loss": 0.2638, "step": 4281 }, { "epoch": 0.81, "grad_norm": 9.1875, "learning_rate": 3.372061842823775e-06, "loss": 0.2597, "step": 4282 }, { "epoch": 0.81, "grad_norm": 14.0625, "learning_rate": 3.3713439691429774e-06, "loss": 0.2591, "step": 4283 }, { "epoch": 0.81, "grad_norm": 8.875, "learning_rate": 3.3706260136717016e-06, "loss": 0.1739, "step": 4284 }, { "epoch": 0.81, "grad_norm": 9.375, "learning_rate": 3.36990797647734e-06, "loss": 0.3016, "step": 4285 }, { "epoch": 0.81, "grad_norm": 9.5625, "learning_rate": 3.369189857627294e-06, "loss": 0.2956, "step": 4286 }, { "epoch": 0.81, "grad_norm": 10.375, "learning_rate": 3.3684716571889697e-06, "loss": 0.2467, "step": 4287 }, { "epoch": 0.81, "grad_norm": 9.3125, "learning_rate": 3.3677533752297832e-06, "loss": 0.2421, "step": 4288 }, { "epoch": 0.81, "grad_norm": 10.0, "learning_rate": 3.367035011817157e-06, "loss": 0.2249, "step": 4289 }, { "epoch": 0.81, "grad_norm": 11.8125, "learning_rate": 3.3663165670185216e-06, "loss": 0.2922, "step": 4290 }, { "epoch": 0.81, "grad_norm": 12.0, "learning_rate": 3.3655980409013156e-06, "loss": 0.2939, "step": 4291 }, { "epoch": 0.81, "grad_norm": 16.125, "learning_rate": 3.364879433532986e-06, "loss": 0.2963, "step": 4292 }, { "epoch": 0.81, "grad_norm": 9.5, "learning_rate": 3.3641607449809842e-06, "loss": 0.2575, "step": 4293 }, { "epoch": 0.81, "grad_norm": 9.25, "learning_rate": 3.3634419753127743e-06, "loss": 0.2615, "step": 4294 }, { "epoch": 0.81, "grad_norm": 13.5, "learning_rate": 3.362723124595822e-06, "loss": 0.3096, "step": 4295 }, { "epoch": 0.81, "grad_norm": 7.96875, "learning_rate": 3.3620041928976056e-06, "loss": 0.2724, "step": 4296 }, { "epoch": 0.81, "grad_norm": 10.125, "learning_rate": 3.3612851802856084e-06, "loss": 0.2591, "step": 4297 }, { "epoch": 0.81, "grad_norm": 12.5625, "learning_rate": 3.3605660868273217e-06, "loss": 0.3153, "step": 4298 }, { "epoch": 0.81, "grad_norm": 12.9375, "learning_rate": 3.3598469125902455e-06, "loss": 0.3004, "step": 4299 }, { "epoch": 0.81, "grad_norm": 11.375, "learning_rate": 3.359127657641886e-06, "loss": 0.3079, "step": 4300 }, { "epoch": 0.81, "grad_norm": 14.5, "learning_rate": 3.3584083220497575e-06, "loss": 0.2714, "step": 4301 }, { "epoch": 0.81, "grad_norm": 11.3125, "learning_rate": 3.3576889058813818e-06, "loss": 0.2873, "step": 4302 }, { "epoch": 0.81, "grad_norm": 10.125, "learning_rate": 3.356969409204289e-06, "loss": 0.3257, "step": 4303 }, { "epoch": 0.81, "grad_norm": 13.5, "learning_rate": 3.3562498320860154e-06, "loss": 0.308, "step": 4304 }, { "epoch": 0.81, "grad_norm": 10.9375, "learning_rate": 3.3555301745941067e-06, "loss": 0.2816, "step": 4305 }, { "epoch": 0.81, "grad_norm": 11.3125, "learning_rate": 3.3548104367961134e-06, "loss": 0.3196, "step": 4306 }, { "epoch": 0.81, "grad_norm": 9.5625, "learning_rate": 3.354090618759596e-06, "loss": 0.2556, "step": 4307 }, { "epoch": 0.81, "grad_norm": 11.75, "learning_rate": 3.3533707205521215e-06, "loss": 0.3054, "step": 4308 }, { "epoch": 0.82, "grad_norm": 11.875, "learning_rate": 3.352650742241266e-06, "loss": 0.3312, "step": 4309 }, { "epoch": 0.82, "grad_norm": 8.3125, "learning_rate": 3.351930683894609e-06, "loss": 0.2573, "step": 4310 }, { "epoch": 0.82, "grad_norm": 19.0, "learning_rate": 3.3512105455797435e-06, "loss": 0.3129, "step": 4311 }, { "epoch": 0.82, "grad_norm": 13.875, "learning_rate": 3.3504903273642632e-06, "loss": 0.3821, "step": 4312 }, { "epoch": 0.82, "grad_norm": 11.0625, "learning_rate": 3.3497700293157753e-06, "loss": 0.2507, "step": 4313 }, { "epoch": 0.82, "grad_norm": 14.3125, "learning_rate": 3.349049651501892e-06, "loss": 0.267, "step": 4314 }, { "epoch": 0.82, "grad_norm": 9.6875, "learning_rate": 3.348329193990232e-06, "loss": 0.2099, "step": 4315 }, { "epoch": 0.82, "grad_norm": 14.5625, "learning_rate": 3.3476086568484225e-06, "loss": 0.2936, "step": 4316 }, { "epoch": 0.82, "grad_norm": 13.0625, "learning_rate": 3.346888040144099e-06, "loss": 0.3046, "step": 4317 }, { "epoch": 0.82, "grad_norm": 11.125, "learning_rate": 3.3461673439449034e-06, "loss": 0.2559, "step": 4318 }, { "epoch": 0.82, "grad_norm": 8.6875, "learning_rate": 3.345446568318485e-06, "loss": 0.232, "step": 4319 }, { "epoch": 0.82, "grad_norm": 7.875, "learning_rate": 3.344725713332502e-06, "loss": 0.2486, "step": 4320 }, { "epoch": 0.82, "grad_norm": 10.25, "learning_rate": 3.3440047790546164e-06, "loss": 0.2008, "step": 4321 }, { "epoch": 0.82, "grad_norm": 12.3125, "learning_rate": 3.343283765552503e-06, "loss": 0.29, "step": 4322 }, { "epoch": 0.82, "grad_norm": 14.6875, "learning_rate": 3.3425626728938397e-06, "loss": 0.2483, "step": 4323 }, { "epoch": 0.82, "grad_norm": 10.1875, "learning_rate": 3.3418415011463135e-06, "loss": 0.2332, "step": 4324 }, { "epoch": 0.82, "grad_norm": 11.375, "learning_rate": 3.3411202503776184e-06, "loss": 0.224, "step": 4325 }, { "epoch": 0.82, "grad_norm": 12.0625, "learning_rate": 3.340398920655457e-06, "loss": 0.2719, "step": 4326 }, { "epoch": 0.82, "grad_norm": 11.1875, "learning_rate": 3.3396775120475377e-06, "loss": 0.2625, "step": 4327 }, { "epoch": 0.82, "grad_norm": 21.0, "learning_rate": 3.3389560246215768e-06, "loss": 0.3295, "step": 4328 }, { "epoch": 0.82, "grad_norm": 15.5, "learning_rate": 3.3382344584452987e-06, "loss": 0.272, "step": 4329 }, { "epoch": 0.82, "grad_norm": 16.25, "learning_rate": 3.337512813586435e-06, "loss": 0.3217, "step": 4330 }, { "epoch": 0.82, "grad_norm": 10.1875, "learning_rate": 3.336791090112723e-06, "loss": 0.2983, "step": 4331 }, { "epoch": 0.82, "grad_norm": 21.625, "learning_rate": 3.33606928809191e-06, "loss": 0.2694, "step": 4332 }, { "epoch": 0.82, "grad_norm": 8.4375, "learning_rate": 3.3353474075917493e-06, "loss": 0.259, "step": 4333 }, { "epoch": 0.82, "grad_norm": 11.375, "learning_rate": 3.334625448680001e-06, "loss": 0.3133, "step": 4334 }, { "epoch": 0.82, "grad_norm": 8.4375, "learning_rate": 3.3339034114244327e-06, "loss": 0.2623, "step": 4335 }, { "epoch": 0.82, "grad_norm": 11.5625, "learning_rate": 3.3331812958928224e-06, "loss": 0.2926, "step": 4336 }, { "epoch": 0.82, "grad_norm": 8.9375, "learning_rate": 3.33245910215295e-06, "loss": 0.3126, "step": 4337 }, { "epoch": 0.82, "grad_norm": 10.5, "learning_rate": 3.3317368302726084e-06, "loss": 0.2295, "step": 4338 }, { "epoch": 0.82, "grad_norm": 13.0, "learning_rate": 3.331014480319593e-06, "loss": 0.3241, "step": 4339 }, { "epoch": 0.82, "grad_norm": 9.0, "learning_rate": 3.33029205236171e-06, "loss": 0.2181, "step": 4340 }, { "epoch": 0.82, "grad_norm": 10.0625, "learning_rate": 3.329569546466771e-06, "loss": 0.2851, "step": 4341 }, { "epoch": 0.82, "grad_norm": 11.6875, "learning_rate": 3.328846962702595e-06, "loss": 0.2343, "step": 4342 }, { "epoch": 0.82, "grad_norm": 11.875, "learning_rate": 3.3281243011370096e-06, "loss": 0.2633, "step": 4343 }, { "epoch": 0.82, "grad_norm": 10.125, "learning_rate": 3.3274015618378487e-06, "loss": 0.2661, "step": 4344 }, { "epoch": 0.82, "grad_norm": 10.0625, "learning_rate": 3.3266787448729532e-06, "loss": 0.2613, "step": 4345 }, { "epoch": 0.82, "grad_norm": 13.375, "learning_rate": 3.3259558503101718e-06, "loss": 0.2865, "step": 4346 }, { "epoch": 0.82, "grad_norm": 9.9375, "learning_rate": 3.325232878217362e-06, "loss": 0.3065, "step": 4347 }, { "epoch": 0.82, "grad_norm": 8.1875, "learning_rate": 3.3245098286623846e-06, "loss": 0.2687, "step": 4348 }, { "epoch": 0.82, "grad_norm": 19.375, "learning_rate": 3.323786701713113e-06, "loss": 0.2391, "step": 4349 }, { "epoch": 0.82, "grad_norm": 9.6875, "learning_rate": 3.3230634974374223e-06, "loss": 0.2714, "step": 4350 }, { "epoch": 0.82, "grad_norm": 9.375, "learning_rate": 3.322340215903199e-06, "loss": 0.2449, "step": 4351 }, { "epoch": 0.82, "grad_norm": 9.375, "learning_rate": 3.321616857178335e-06, "loss": 0.2503, "step": 4352 }, { "epoch": 0.82, "grad_norm": 11.0, "learning_rate": 3.32089342133073e-06, "loss": 0.2444, "step": 4353 }, { "epoch": 0.82, "grad_norm": 12.875, "learning_rate": 3.32016990842829e-06, "loss": 0.3056, "step": 4354 }, { "epoch": 0.82, "grad_norm": 10.3125, "learning_rate": 3.3194463185389307e-06, "loss": 0.3172, "step": 4355 }, { "epoch": 0.82, "grad_norm": 9.125, "learning_rate": 3.3187226517305716e-06, "loss": 0.298, "step": 4356 }, { "epoch": 0.82, "grad_norm": 11.75, "learning_rate": 3.3179989080711424e-06, "loss": 0.2725, "step": 4357 }, { "epoch": 0.82, "grad_norm": 9.4375, "learning_rate": 3.3172750876285776e-06, "loss": 0.2605, "step": 4358 }, { "epoch": 0.82, "grad_norm": 12.75, "learning_rate": 3.3165511904708215e-06, "loss": 0.3091, "step": 4359 }, { "epoch": 0.82, "grad_norm": 17.375, "learning_rate": 3.3158272166658227e-06, "loss": 0.2415, "step": 4360 }, { "epoch": 0.82, "grad_norm": 8.75, "learning_rate": 3.31510316628154e-06, "loss": 0.2404, "step": 4361 }, { "epoch": 0.83, "grad_norm": 13.8125, "learning_rate": 3.3143790393859364e-06, "loss": 0.2443, "step": 4362 }, { "epoch": 0.83, "grad_norm": 10.5, "learning_rate": 3.3136548360469843e-06, "loss": 0.2217, "step": 4363 }, { "epoch": 0.83, "grad_norm": 10.9375, "learning_rate": 3.312930556332663e-06, "loss": 0.3, "step": 4364 }, { "epoch": 0.83, "grad_norm": 13.0625, "learning_rate": 3.312206200310958e-06, "loss": 0.3229, "step": 4365 }, { "epoch": 0.83, "grad_norm": 19.5, "learning_rate": 3.3114817680498613e-06, "loss": 0.2813, "step": 4366 }, { "epoch": 0.83, "grad_norm": 10.5625, "learning_rate": 3.3107572596173754e-06, "loss": 0.3169, "step": 4367 }, { "epoch": 0.83, "grad_norm": 13.4375, "learning_rate": 3.310032675081506e-06, "loss": 0.2781, "step": 4368 }, { "epoch": 0.83, "grad_norm": 10.625, "learning_rate": 3.3093080145102686e-06, "loss": 0.274, "step": 4369 }, { "epoch": 0.83, "grad_norm": 12.6875, "learning_rate": 3.308583277971684e-06, "loss": 0.2548, "step": 4370 }, { "epoch": 0.83, "grad_norm": 10.3125, "learning_rate": 3.307858465533783e-06, "loss": 0.2179, "step": 4371 }, { "epoch": 0.83, "grad_norm": 7.71875, "learning_rate": 3.3071335772645998e-06, "loss": 0.1483, "step": 4372 }, { "epoch": 0.83, "grad_norm": 10.6875, "learning_rate": 3.306408613232178e-06, "loss": 0.3807, "step": 4373 }, { "epoch": 0.83, "grad_norm": 9.6875, "learning_rate": 3.3056835735045678e-06, "loss": 0.2582, "step": 4374 }, { "epoch": 0.83, "grad_norm": 13.125, "learning_rate": 3.3049584581498274e-06, "loss": 0.2562, "step": 4375 }, { "epoch": 0.83, "grad_norm": 16.625, "learning_rate": 3.3042332672360195e-06, "loss": 0.2852, "step": 4376 }, { "epoch": 0.83, "grad_norm": 13.1875, "learning_rate": 3.303508000831217e-06, "loss": 0.3792, "step": 4377 }, { "epoch": 0.83, "grad_norm": 13.25, "learning_rate": 3.3027826590034985e-06, "loss": 0.2677, "step": 4378 }, { "epoch": 0.83, "grad_norm": 8.875, "learning_rate": 3.3020572418209485e-06, "loss": 0.2464, "step": 4379 }, { "epoch": 0.83, "grad_norm": 11.9375, "learning_rate": 3.3013317493516613e-06, "loss": 0.24, "step": 4380 }, { "epoch": 0.83, "grad_norm": 13.0625, "learning_rate": 3.3006061816637347e-06, "loss": 0.3099, "step": 4381 }, { "epoch": 0.83, "grad_norm": 12.625, "learning_rate": 3.2998805388252785e-06, "loss": 0.3516, "step": 4382 }, { "epoch": 0.83, "grad_norm": 11.625, "learning_rate": 3.2991548209044043e-06, "loss": 0.2833, "step": 4383 }, { "epoch": 0.83, "grad_norm": 15.1875, "learning_rate": 3.2984290279692344e-06, "loss": 0.2719, "step": 4384 }, { "epoch": 0.83, "grad_norm": 14.3125, "learning_rate": 3.2977031600878956e-06, "loss": 0.2659, "step": 4385 }, { "epoch": 0.83, "grad_norm": 8.1875, "learning_rate": 3.2969772173285236e-06, "loss": 0.263, "step": 4386 }, { "epoch": 0.83, "grad_norm": 14.4375, "learning_rate": 3.2962511997592606e-06, "loss": 0.2545, "step": 4387 }, { "epoch": 0.83, "grad_norm": 17.625, "learning_rate": 3.295525107448256e-06, "loss": 0.3367, "step": 4388 }, { "epoch": 0.83, "grad_norm": 10.375, "learning_rate": 3.2947989404636647e-06, "loss": 0.2583, "step": 4389 }, { "epoch": 0.83, "grad_norm": 10.0, "learning_rate": 3.2940726988736516e-06, "loss": 0.2844, "step": 4390 }, { "epoch": 0.83, "grad_norm": 18.125, "learning_rate": 3.2933463827463857e-06, "loss": 0.2559, "step": 4391 }, { "epoch": 0.83, "grad_norm": 10.3125, "learning_rate": 3.292619992150044e-06, "loss": 0.266, "step": 4392 }, { "epoch": 0.83, "grad_norm": 16.75, "learning_rate": 3.291893527152812e-06, "loss": 0.3452, "step": 4393 }, { "epoch": 0.83, "grad_norm": 8.8125, "learning_rate": 3.2911669878228787e-06, "loss": 0.2505, "step": 4394 }, { "epoch": 0.83, "grad_norm": 11.5, "learning_rate": 3.290440374228444e-06, "loss": 0.3252, "step": 4395 }, { "epoch": 0.83, "grad_norm": 9.75, "learning_rate": 3.2897136864377122e-06, "loss": 0.2089, "step": 4396 }, { "epoch": 0.83, "grad_norm": 13.125, "learning_rate": 3.288986924518896e-06, "loss": 0.257, "step": 4397 }, { "epoch": 0.83, "grad_norm": 14.3125, "learning_rate": 3.2882600885402126e-06, "loss": 0.2724, "step": 4398 }, { "epoch": 0.83, "grad_norm": 10.6875, "learning_rate": 3.2875331785698907e-06, "loss": 0.3229, "step": 4399 }, { "epoch": 0.83, "grad_norm": 10.125, "learning_rate": 3.28680619467616e-06, "loss": 0.2719, "step": 4400 }, { "epoch": 0.83, "grad_norm": 17.5, "learning_rate": 3.2860791369272634e-06, "loss": 0.3075, "step": 4401 }, { "epoch": 0.83, "grad_norm": 11.6875, "learning_rate": 3.285352005391445e-06, "loss": 0.2842, "step": 4402 }, { "epoch": 0.83, "grad_norm": 10.8125, "learning_rate": 3.2846248001369606e-06, "loss": 0.2483, "step": 4403 }, { "epoch": 0.83, "grad_norm": 7.21875, "learning_rate": 3.2838975212320695e-06, "loss": 0.2305, "step": 4404 }, { "epoch": 0.83, "grad_norm": 16.25, "learning_rate": 3.28317016874504e-06, "loss": 0.3139, "step": 4405 }, { "epoch": 0.83, "grad_norm": 13.5, "learning_rate": 3.2824427427441463e-06, "loss": 0.3728, "step": 4406 }, { "epoch": 0.83, "grad_norm": 10.9375, "learning_rate": 3.2817152432976696e-06, "loss": 0.301, "step": 4407 }, { "epoch": 0.83, "grad_norm": 11.5, "learning_rate": 3.2809876704738975e-06, "loss": 0.2395, "step": 4408 }, { "epoch": 0.83, "grad_norm": 10.5625, "learning_rate": 3.2802600243411264e-06, "loss": 0.2732, "step": 4409 }, { "epoch": 0.83, "grad_norm": 8.125, "learning_rate": 3.2795323049676568e-06, "loss": 0.2637, "step": 4410 }, { "epoch": 0.83, "grad_norm": 13.3125, "learning_rate": 3.2788045124217992e-06, "loss": 0.2334, "step": 4411 }, { "epoch": 0.83, "grad_norm": 12.5, "learning_rate": 3.2780766467718683e-06, "loss": 0.1953, "step": 4412 }, { "epoch": 0.83, "grad_norm": 13.8125, "learning_rate": 3.2773487080861865e-06, "loss": 0.3336, "step": 4413 }, { "epoch": 0.84, "grad_norm": 12.6875, "learning_rate": 3.276620696433084e-06, "loss": 0.2487, "step": 4414 }, { "epoch": 0.84, "grad_norm": 16.375, "learning_rate": 3.2758926118808963e-06, "loss": 0.2723, "step": 4415 }, { "epoch": 0.84, "grad_norm": 18.625, "learning_rate": 3.2751644544979677e-06, "loss": 0.3174, "step": 4416 }, { "epoch": 0.84, "grad_norm": 12.5, "learning_rate": 3.2744362243526474e-06, "loss": 0.3425, "step": 4417 }, { "epoch": 0.84, "grad_norm": 8.3125, "learning_rate": 3.2737079215132917e-06, "loss": 0.2562, "step": 4418 }, { "epoch": 0.84, "grad_norm": 10.8125, "learning_rate": 3.2729795460482647e-06, "loss": 0.238, "step": 4419 }, { "epoch": 0.84, "grad_norm": 12.375, "learning_rate": 3.272251098025938e-06, "loss": 0.2524, "step": 4420 }, { "epoch": 0.84, "grad_norm": 7.78125, "learning_rate": 3.271522577514687e-06, "loss": 0.2504, "step": 4421 }, { "epoch": 0.84, "grad_norm": 8.9375, "learning_rate": 3.2707939845828966e-06, "loss": 0.324, "step": 4422 }, { "epoch": 0.84, "grad_norm": 11.0625, "learning_rate": 3.270065319298958e-06, "loss": 0.3244, "step": 4423 }, { "epoch": 0.84, "grad_norm": 10.6875, "learning_rate": 3.2693365817312683e-06, "loss": 0.2476, "step": 4424 }, { "epoch": 0.84, "grad_norm": 12.5, "learning_rate": 3.2686077719482316e-06, "loss": 0.2223, "step": 4425 }, { "epoch": 0.84, "grad_norm": 11.875, "learning_rate": 3.2678788900182603e-06, "loss": 0.3416, "step": 4426 }, { "epoch": 0.84, "grad_norm": 8.875, "learning_rate": 3.2671499360097713e-06, "loss": 0.2745, "step": 4427 }, { "epoch": 0.84, "grad_norm": 9.625, "learning_rate": 3.2664209099911904e-06, "loss": 0.3009, "step": 4428 }, { "epoch": 0.84, "grad_norm": 11.25, "learning_rate": 3.265691812030948e-06, "loss": 0.2985, "step": 4429 }, { "epoch": 0.84, "grad_norm": 9.3125, "learning_rate": 3.2649626421974826e-06, "loss": 0.3604, "step": 4430 }, { "epoch": 0.84, "grad_norm": 11.875, "learning_rate": 3.26423340055924e-06, "loss": 0.28, "step": 4431 }, { "epoch": 0.84, "grad_norm": 18.75, "learning_rate": 3.2635040871846713e-06, "loss": 0.3639, "step": 4432 }, { "epoch": 0.84, "grad_norm": 11.0625, "learning_rate": 3.2627747021422353e-06, "loss": 0.2621, "step": 4433 }, { "epoch": 0.84, "grad_norm": 14.5625, "learning_rate": 3.2620452455003965e-06, "loss": 0.3998, "step": 4434 }, { "epoch": 0.84, "grad_norm": 12.875, "learning_rate": 3.261315717327628e-06, "loss": 0.2942, "step": 4435 }, { "epoch": 0.84, "grad_norm": 9.8125, "learning_rate": 3.2605861176924063e-06, "loss": 0.2345, "step": 4436 }, { "epoch": 0.84, "grad_norm": 11.0625, "learning_rate": 3.2598564466632197e-06, "loss": 0.2967, "step": 4437 }, { "epoch": 0.84, "grad_norm": 9.75, "learning_rate": 3.259126704308558e-06, "loss": 0.2584, "step": 4438 }, { "epoch": 0.84, "grad_norm": 11.6875, "learning_rate": 3.258396890696922e-06, "loss": 0.2903, "step": 4439 }, { "epoch": 0.84, "grad_norm": 11.5625, "learning_rate": 3.2576670058968146e-06, "loss": 0.3055, "step": 4440 }, { "epoch": 0.84, "grad_norm": 9.5, "learning_rate": 3.2569370499767507e-06, "loss": 0.2942, "step": 4441 }, { "epoch": 0.84, "grad_norm": 13.125, "learning_rate": 3.256207023005246e-06, "loss": 0.2595, "step": 4442 }, { "epoch": 0.84, "grad_norm": 10.875, "learning_rate": 3.255476925050829e-06, "loss": 0.2445, "step": 4443 }, { "epoch": 0.84, "grad_norm": 8.5, "learning_rate": 3.254746756182029e-06, "loss": 0.2669, "step": 4444 }, { "epoch": 0.84, "grad_norm": 9.0, "learning_rate": 3.2540165164673876e-06, "loss": 0.2566, "step": 4445 }, { "epoch": 0.84, "grad_norm": 13.9375, "learning_rate": 3.253286205975448e-06, "loss": 0.2795, "step": 4446 }, { "epoch": 0.84, "grad_norm": 12.0625, "learning_rate": 3.2525558247747635e-06, "loss": 0.248, "step": 4447 }, { "epoch": 0.84, "grad_norm": 10.4375, "learning_rate": 3.2518253729338924e-06, "loss": 0.2622, "step": 4448 }, { "epoch": 0.84, "grad_norm": 11.9375, "learning_rate": 3.2510948505214003e-06, "loss": 0.2686, "step": 4449 }, { "epoch": 0.84, "grad_norm": 9.5, "learning_rate": 3.250364257605859e-06, "loss": 0.2212, "step": 4450 }, { "epoch": 0.84, "grad_norm": 12.0625, "learning_rate": 3.249633594255847e-06, "loss": 0.2814, "step": 4451 }, { "epoch": 0.84, "grad_norm": 9.4375, "learning_rate": 3.2489028605399498e-06, "loss": 0.2381, "step": 4452 }, { "epoch": 0.84, "grad_norm": 12.625, "learning_rate": 3.2481720565267583e-06, "loss": 0.2379, "step": 4453 }, { "epoch": 0.84, "grad_norm": 13.0625, "learning_rate": 3.247441182284873e-06, "loss": 0.2934, "step": 4454 }, { "epoch": 0.84, "grad_norm": 12.1875, "learning_rate": 3.2467102378828963e-06, "loss": 0.287, "step": 4455 }, { "epoch": 0.84, "grad_norm": 9.375, "learning_rate": 3.2459792233894417e-06, "loss": 0.2924, "step": 4456 }, { "epoch": 0.84, "grad_norm": 9.875, "learning_rate": 3.2452481388731256e-06, "loss": 0.3, "step": 4457 }, { "epoch": 0.84, "grad_norm": 10.875, "learning_rate": 3.2445169844025747e-06, "loss": 0.2778, "step": 4458 }, { "epoch": 0.84, "grad_norm": 14.75, "learning_rate": 3.243785760046419e-06, "loss": 0.2539, "step": 4459 }, { "epoch": 0.84, "grad_norm": 12.375, "learning_rate": 3.243054465873297e-06, "loss": 0.2962, "step": 4460 }, { "epoch": 0.84, "grad_norm": 9.4375, "learning_rate": 3.242323101951853e-06, "loss": 0.235, "step": 4461 }, { "epoch": 0.84, "grad_norm": 14.0625, "learning_rate": 3.2415916683507376e-06, "loss": 0.3534, "step": 4462 }, { "epoch": 0.84, "grad_norm": 11.125, "learning_rate": 3.240860165138609e-06, "loss": 0.3275, "step": 4463 }, { "epoch": 0.84, "grad_norm": 10.1875, "learning_rate": 3.2401285923841308e-06, "loss": 0.3286, "step": 4464 }, { "epoch": 0.84, "grad_norm": 15.0625, "learning_rate": 3.239396950155973e-06, "loss": 0.3345, "step": 4465 }, { "epoch": 0.84, "grad_norm": 10.75, "learning_rate": 3.2386652385228136e-06, "loss": 0.2131, "step": 4466 }, { "epoch": 0.85, "grad_norm": 9.25, "learning_rate": 3.2379334575533362e-06, "loss": 0.2686, "step": 4467 }, { "epoch": 0.85, "grad_norm": 18.5, "learning_rate": 3.2372016073162297e-06, "loss": 0.3387, "step": 4468 }, { "epoch": 0.85, "grad_norm": 12.6875, "learning_rate": 3.2364696878801927e-06, "loss": 0.2388, "step": 4469 }, { "epoch": 0.85, "grad_norm": 11.1875, "learning_rate": 3.2357376993139268e-06, "loss": 0.2578, "step": 4470 }, { "epoch": 0.85, "grad_norm": 9.625, "learning_rate": 3.2350056416861416e-06, "loss": 0.2734, "step": 4471 }, { "epoch": 0.85, "grad_norm": 10.625, "learning_rate": 3.2342735150655534e-06, "loss": 0.3099, "step": 4472 }, { "epoch": 0.85, "grad_norm": 16.625, "learning_rate": 3.233541319520886e-06, "loss": 0.2828, "step": 4473 }, { "epoch": 0.85, "grad_norm": 13.0625, "learning_rate": 3.2328090551208666e-06, "loss": 0.3052, "step": 4474 }, { "epoch": 0.85, "grad_norm": 10.1875, "learning_rate": 3.2320767219342313e-06, "loss": 0.2647, "step": 4475 }, { "epoch": 0.85, "grad_norm": 7.40625, "learning_rate": 3.231344320029723e-06, "loss": 0.2429, "step": 4476 }, { "epoch": 0.85, "grad_norm": 8.4375, "learning_rate": 3.2306118494760886e-06, "loss": 0.2268, "step": 4477 }, { "epoch": 0.85, "grad_norm": 14.5, "learning_rate": 3.229879310342084e-06, "loss": 0.2674, "step": 4478 }, { "epoch": 0.85, "grad_norm": 9.375, "learning_rate": 3.2291467026964702e-06, "loss": 0.2502, "step": 4479 }, { "epoch": 0.85, "grad_norm": 10.25, "learning_rate": 3.2284140266080144e-06, "loss": 0.212, "step": 4480 }, { "epoch": 0.85, "grad_norm": 9.25, "learning_rate": 3.2276812821454916e-06, "loss": 0.2473, "step": 4481 }, { "epoch": 0.85, "grad_norm": 9.875, "learning_rate": 3.226948469377681e-06, "loss": 0.2515, "step": 4482 }, { "epoch": 0.85, "grad_norm": 10.3125, "learning_rate": 3.226215588373372e-06, "loss": 0.2532, "step": 4483 }, { "epoch": 0.85, "grad_norm": 13.75, "learning_rate": 3.225482639201355e-06, "loss": 0.2898, "step": 4484 }, { "epoch": 0.85, "grad_norm": 11.0625, "learning_rate": 3.2247496219304324e-06, "loss": 0.2995, "step": 4485 }, { "epoch": 0.85, "grad_norm": 10.25, "learning_rate": 3.2240165366294084e-06, "loss": 0.2189, "step": 4486 }, { "epoch": 0.85, "grad_norm": 9.9375, "learning_rate": 3.2232833833670974e-06, "loss": 0.2756, "step": 4487 }, { "epoch": 0.85, "grad_norm": 10.75, "learning_rate": 3.2225501622123157e-06, "loss": 0.2326, "step": 4488 }, { "epoch": 0.85, "grad_norm": 11.6875, "learning_rate": 3.2218168732338915e-06, "loss": 0.3118, "step": 4489 }, { "epoch": 0.85, "grad_norm": 11.8125, "learning_rate": 3.2210835165006544e-06, "loss": 0.3763, "step": 4490 }, { "epoch": 0.85, "grad_norm": 10.375, "learning_rate": 3.2203500920814434e-06, "loss": 0.3314, "step": 4491 }, { "epoch": 0.85, "grad_norm": 12.25, "learning_rate": 3.2196166000451025e-06, "loss": 0.2658, "step": 4492 }, { "epoch": 0.85, "grad_norm": 19.0, "learning_rate": 3.218883040460482e-06, "loss": 0.2788, "step": 4493 }, { "epoch": 0.85, "grad_norm": 9.1875, "learning_rate": 3.218149413396441e-06, "loss": 0.2002, "step": 4494 }, { "epoch": 0.85, "grad_norm": 14.5625, "learning_rate": 3.2174157189218403e-06, "loss": 0.2493, "step": 4495 }, { "epoch": 0.85, "grad_norm": 12.1875, "learning_rate": 3.2166819571055515e-06, "loss": 0.2713, "step": 4496 }, { "epoch": 0.85, "grad_norm": 9.375, "learning_rate": 3.215948128016448e-06, "loss": 0.2504, "step": 4497 }, { "epoch": 0.85, "grad_norm": 20.0, "learning_rate": 3.2152142317234167e-06, "loss": 0.2573, "step": 4498 }, { "epoch": 0.85, "grad_norm": 11.5, "learning_rate": 3.214480268295342e-06, "loss": 0.2935, "step": 4499 }, { "epoch": 0.85, "grad_norm": 16.5, "learning_rate": 3.213746237801121e-06, "loss": 0.2631, "step": 4500 }, { "epoch": 0.85, "grad_norm": 15.0625, "learning_rate": 3.2130121403096536e-06, "loss": 0.3079, "step": 4501 }, { "epoch": 0.85, "grad_norm": 11.3125, "learning_rate": 3.2122779758898496e-06, "loss": 0.2726, "step": 4502 }, { "epoch": 0.85, "grad_norm": 11.375, "learning_rate": 3.2115437446106205e-06, "loss": 0.2594, "step": 4503 }, { "epoch": 0.85, "grad_norm": 12.125, "learning_rate": 3.2108094465408877e-06, "loss": 0.2756, "step": 4504 }, { "epoch": 0.85, "grad_norm": 14.5, "learning_rate": 3.2100750817495773e-06, "loss": 0.2854, "step": 4505 }, { "epoch": 0.85, "grad_norm": 11.25, "learning_rate": 3.209340650305622e-06, "loss": 0.2818, "step": 4506 }, { "epoch": 0.85, "grad_norm": 15.4375, "learning_rate": 3.2086061522779606e-06, "loss": 0.2562, "step": 4507 }, { "epoch": 0.85, "grad_norm": 9.375, "learning_rate": 3.207871587735538e-06, "loss": 0.23, "step": 4508 }, { "epoch": 0.85, "grad_norm": 8.125, "learning_rate": 3.207136956747306e-06, "loss": 0.2321, "step": 4509 }, { "epoch": 0.85, "grad_norm": 9.3125, "learning_rate": 3.206402259382222e-06, "loss": 0.2628, "step": 4510 }, { "epoch": 0.85, "grad_norm": 15.3125, "learning_rate": 3.2056674957092503e-06, "loss": 0.2685, "step": 4511 }, { "epoch": 0.85, "grad_norm": 10.4375, "learning_rate": 3.2049326657973606e-06, "loss": 0.2806, "step": 4512 }, { "epoch": 0.85, "grad_norm": 14.125, "learning_rate": 3.204197769715529e-06, "loss": 0.3606, "step": 4513 }, { "epoch": 0.85, "grad_norm": 10.0625, "learning_rate": 3.2034628075327386e-06, "loss": 0.2434, "step": 4514 }, { "epoch": 0.85, "grad_norm": 11.875, "learning_rate": 3.2027277793179778e-06, "loss": 0.2373, "step": 4515 }, { "epoch": 0.85, "grad_norm": 17.5, "learning_rate": 3.201992685140241e-06, "loss": 0.2948, "step": 4516 }, { "epoch": 0.85, "grad_norm": 11.4375, "learning_rate": 3.2012575250685306e-06, "loss": 0.2885, "step": 4517 }, { "epoch": 0.85, "grad_norm": 7.625, "learning_rate": 3.2005222991718526e-06, "loss": 0.2199, "step": 4518 }, { "epoch": 0.85, "grad_norm": 10.3125, "learning_rate": 3.1997870075192213e-06, "loss": 0.3378, "step": 4519 }, { "epoch": 0.86, "grad_norm": 11.125, "learning_rate": 3.199051650179656e-06, "loss": 0.2681, "step": 4520 }, { "epoch": 0.86, "grad_norm": 9.75, "learning_rate": 3.1983162272221823e-06, "loss": 0.2292, "step": 4521 }, { "epoch": 0.86, "grad_norm": 12.375, "learning_rate": 3.197580738715833e-06, "loss": 0.2686, "step": 4522 }, { "epoch": 0.86, "grad_norm": 11.25, "learning_rate": 3.1968451847296457e-06, "loss": 0.296, "step": 4523 }, { "epoch": 0.86, "grad_norm": 8.25, "learning_rate": 3.196109565332664e-06, "loss": 0.1945, "step": 4524 }, { "epoch": 0.86, "grad_norm": 13.0, "learning_rate": 3.1953738805939396e-06, "loss": 0.2577, "step": 4525 }, { "epoch": 0.86, "grad_norm": 8.8125, "learning_rate": 3.1946381305825273e-06, "loss": 0.2324, "step": 4526 }, { "epoch": 0.86, "grad_norm": 9.3125, "learning_rate": 3.1939023153674925e-06, "loss": 0.2497, "step": 4527 }, { "epoch": 0.86, "grad_norm": 10.875, "learning_rate": 3.1931664350179013e-06, "loss": 0.1919, "step": 4528 }, { "epoch": 0.86, "grad_norm": 11.125, "learning_rate": 3.192430489602831e-06, "loss": 0.3864, "step": 4529 }, { "epoch": 0.86, "grad_norm": 9.9375, "learning_rate": 3.1916944791913594e-06, "loss": 0.2733, "step": 4530 }, { "epoch": 0.86, "grad_norm": 24.875, "learning_rate": 3.1909584038525776e-06, "loss": 0.3375, "step": 4531 }, { "epoch": 0.86, "grad_norm": 12.1875, "learning_rate": 3.1902222636555746e-06, "loss": 0.2368, "step": 4532 }, { "epoch": 0.86, "grad_norm": 10.0625, "learning_rate": 3.189486058669454e-06, "loss": 0.2812, "step": 4533 }, { "epoch": 0.86, "grad_norm": 17.75, "learning_rate": 3.1887497889633173e-06, "loss": 0.3184, "step": 4534 }, { "epoch": 0.86, "grad_norm": 12.1875, "learning_rate": 3.188013454606279e-06, "loss": 0.3633, "step": 4535 }, { "epoch": 0.86, "grad_norm": 12.4375, "learning_rate": 3.187277055667454e-06, "loss": 0.2817, "step": 4536 }, { "epoch": 0.86, "grad_norm": 13.25, "learning_rate": 3.186540592215968e-06, "loss": 0.3161, "step": 4537 }, { "epoch": 0.86, "grad_norm": 14.0, "learning_rate": 3.1858040643209505e-06, "loss": 0.2441, "step": 4538 }, { "epoch": 0.86, "grad_norm": 13.5, "learning_rate": 3.185067472051535e-06, "loss": 0.2502, "step": 4539 }, { "epoch": 0.86, "grad_norm": 13.375, "learning_rate": 3.1843308154768667e-06, "loss": 0.3021, "step": 4540 }, { "epoch": 0.86, "grad_norm": 12.0625, "learning_rate": 3.183594094666089e-06, "loss": 0.3289, "step": 4541 }, { "epoch": 0.86, "grad_norm": 10.9375, "learning_rate": 3.182857309688361e-06, "loss": 0.2932, "step": 4542 }, { "epoch": 0.86, "grad_norm": 16.25, "learning_rate": 3.1821204606128373e-06, "loss": 0.2603, "step": 4543 }, { "epoch": 0.86, "grad_norm": 15.1875, "learning_rate": 3.181383547508687e-06, "loss": 0.2376, "step": 4544 }, { "epoch": 0.86, "grad_norm": 9.6875, "learning_rate": 3.1806465704450805e-06, "loss": 0.1886, "step": 4545 }, { "epoch": 0.86, "grad_norm": 10.3125, "learning_rate": 3.1799095294911974e-06, "loss": 0.2228, "step": 4546 }, { "epoch": 0.86, "grad_norm": 10.5, "learning_rate": 3.179172424716219e-06, "loss": 0.2749, "step": 4547 }, { "epoch": 0.86, "grad_norm": 9.9375, "learning_rate": 3.1784352561893376e-06, "loss": 0.2559, "step": 4548 }, { "epoch": 0.86, "grad_norm": 9.9375, "learning_rate": 3.177698023979748e-06, "loss": 0.2997, "step": 4549 }, { "epoch": 0.86, "grad_norm": 11.375, "learning_rate": 3.176960728156651e-06, "loss": 0.2306, "step": 4550 }, { "epoch": 0.86, "grad_norm": 10.0, "learning_rate": 3.1762233687892562e-06, "loss": 0.2746, "step": 4551 }, { "epoch": 0.86, "grad_norm": 14.4375, "learning_rate": 3.175485945946776e-06, "loss": 0.2139, "step": 4552 }, { "epoch": 0.86, "grad_norm": 15.0625, "learning_rate": 3.1747484596984307e-06, "loss": 0.2898, "step": 4553 }, { "epoch": 0.86, "grad_norm": 13.3125, "learning_rate": 3.1740109101134457e-06, "loss": 0.2903, "step": 4554 }, { "epoch": 0.86, "grad_norm": 12.1875, "learning_rate": 3.1732732972610524e-06, "loss": 0.3383, "step": 4555 }, { "epoch": 0.86, "grad_norm": 8.9375, "learning_rate": 3.172535621210489e-06, "loss": 0.1968, "step": 4556 }, { "epoch": 0.86, "grad_norm": 13.0, "learning_rate": 3.1717978820309993e-06, "loss": 0.2321, "step": 4557 }, { "epoch": 0.86, "grad_norm": 15.3125, "learning_rate": 3.1710600797918307e-06, "loss": 0.2809, "step": 4558 }, { "epoch": 0.86, "grad_norm": 12.875, "learning_rate": 3.1703222145622404e-06, "loss": 0.2869, "step": 4559 }, { "epoch": 0.86, "grad_norm": 14.125, "learning_rate": 3.169584286411489e-06, "loss": 0.2713, "step": 4560 }, { "epoch": 0.86, "grad_norm": 9.9375, "learning_rate": 3.168846295408844e-06, "loss": 0.3025, "step": 4561 }, { "epoch": 0.86, "grad_norm": 12.375, "learning_rate": 3.1681082416235775e-06, "loss": 0.2501, "step": 4562 }, { "epoch": 0.86, "grad_norm": 13.0625, "learning_rate": 3.167370125124969e-06, "loss": 0.1825, "step": 4563 }, { "epoch": 0.86, "grad_norm": 15.9375, "learning_rate": 3.1666319459823037e-06, "loss": 0.3524, "step": 4564 }, { "epoch": 0.86, "grad_norm": 15.3125, "learning_rate": 3.165893704264872e-06, "loss": 0.3064, "step": 4565 }, { "epoch": 0.86, "grad_norm": 17.125, "learning_rate": 3.16515540004197e-06, "loss": 0.3055, "step": 4566 }, { "epoch": 0.86, "grad_norm": 14.1875, "learning_rate": 3.1644170333829004e-06, "loss": 0.3418, "step": 4567 }, { "epoch": 0.86, "grad_norm": 9.1875, "learning_rate": 3.163678604356972e-06, "loss": 0.2392, "step": 4568 }, { "epoch": 0.86, "grad_norm": 8.25, "learning_rate": 3.1629401130334984e-06, "loss": 0.2344, "step": 4569 }, { "epoch": 0.86, "grad_norm": 8.1875, "learning_rate": 3.162201559481799e-06, "loss": 0.2059, "step": 4570 }, { "epoch": 0.86, "grad_norm": 13.4375, "learning_rate": 3.161462943771202e-06, "loss": 0.3281, "step": 4571 }, { "epoch": 0.86, "grad_norm": 9.4375, "learning_rate": 3.160724265971036e-06, "loss": 0.2656, "step": 4572 }, { "epoch": 0.87, "grad_norm": 12.25, "learning_rate": 3.1599855261506406e-06, "loss": 0.2928, "step": 4573 }, { "epoch": 0.87, "grad_norm": 9.75, "learning_rate": 3.159246724379358e-06, "loss": 0.3402, "step": 4574 }, { "epoch": 0.87, "grad_norm": 10.25, "learning_rate": 3.1585078607265383e-06, "loss": 0.2602, "step": 4575 }, { "epoch": 0.87, "grad_norm": 11.75, "learning_rate": 3.157768935261536e-06, "loss": 0.2483, "step": 4576 }, { "epoch": 0.87, "grad_norm": 11.0625, "learning_rate": 3.1570299480537114e-06, "loss": 0.2595, "step": 4577 }, { "epoch": 0.87, "grad_norm": 12.875, "learning_rate": 3.1562908991724312e-06, "loss": 0.2915, "step": 4578 }, { "epoch": 0.87, "grad_norm": 9.4375, "learning_rate": 3.155551788687069e-06, "loss": 0.2511, "step": 4579 }, { "epoch": 0.87, "grad_norm": 15.5, "learning_rate": 3.154812616667001e-06, "loss": 0.296, "step": 4580 }, { "epoch": 0.87, "grad_norm": 8.5, "learning_rate": 3.1540733831816125e-06, "loss": 0.2511, "step": 4581 }, { "epoch": 0.87, "grad_norm": 11.6875, "learning_rate": 3.1533340883002917e-06, "loss": 0.3003, "step": 4582 }, { "epoch": 0.87, "grad_norm": 9.5, "learning_rate": 3.152594732092435e-06, "loss": 0.2233, "step": 4583 }, { "epoch": 0.87, "grad_norm": 9.5, "learning_rate": 3.1518553146274445e-06, "loss": 0.2972, "step": 4584 }, { "epoch": 0.87, "grad_norm": 11.125, "learning_rate": 3.151115835974725e-06, "loss": 0.2892, "step": 4585 }, { "epoch": 0.87, "grad_norm": 10.375, "learning_rate": 3.1503762962036914e-06, "loss": 0.2901, "step": 4586 }, { "epoch": 0.87, "grad_norm": 10.1875, "learning_rate": 3.14963669538376e-06, "loss": 0.2534, "step": 4587 }, { "epoch": 0.87, "grad_norm": 21.375, "learning_rate": 3.148897033584357e-06, "loss": 0.2745, "step": 4588 }, { "epoch": 0.87, "grad_norm": 11.875, "learning_rate": 3.1481573108749098e-06, "loss": 0.3061, "step": 4589 }, { "epoch": 0.87, "grad_norm": 8.625, "learning_rate": 3.147417527324857e-06, "loss": 0.1971, "step": 4590 }, { "epoch": 0.87, "grad_norm": 23.0, "learning_rate": 3.146677683003637e-06, "loss": 0.3609, "step": 4591 }, { "epoch": 0.87, "grad_norm": 10.0, "learning_rate": 3.145937777980699e-06, "loss": 0.2685, "step": 4592 }, { "epoch": 0.87, "grad_norm": 13.9375, "learning_rate": 3.145197812325494e-06, "loss": 0.211, "step": 4593 }, { "epoch": 0.87, "grad_norm": 9.75, "learning_rate": 3.1444577861074814e-06, "loss": 0.2824, "step": 4594 }, { "epoch": 0.87, "grad_norm": 8.375, "learning_rate": 3.143717699396125e-06, "loss": 0.2434, "step": 4595 }, { "epoch": 0.87, "grad_norm": 10.125, "learning_rate": 3.142977552260895e-06, "loss": 0.1946, "step": 4596 }, { "epoch": 0.87, "grad_norm": 9.1875, "learning_rate": 3.1422373447712664e-06, "loss": 0.2545, "step": 4597 }, { "epoch": 0.87, "grad_norm": 9.875, "learning_rate": 3.1414970769967197e-06, "loss": 0.2901, "step": 4598 }, { "epoch": 0.87, "grad_norm": 13.75, "learning_rate": 3.1407567490067426e-06, "loss": 0.3172, "step": 4599 }, { "epoch": 0.87, "grad_norm": 11.375, "learning_rate": 3.1400163608708266e-06, "loss": 0.2646, "step": 4600 }, { "epoch": 0.87, "grad_norm": 14.0625, "learning_rate": 3.139275912658471e-06, "loss": 0.2774, "step": 4601 }, { "epoch": 0.87, "grad_norm": 7.5625, "learning_rate": 3.138535404439178e-06, "loss": 0.2384, "step": 4602 }, { "epoch": 0.87, "grad_norm": 12.75, "learning_rate": 3.1377948362824585e-06, "loss": 0.3236, "step": 4603 }, { "epoch": 0.87, "grad_norm": 12.6875, "learning_rate": 3.1370542082578257e-06, "loss": 0.2949, "step": 4604 }, { "epoch": 0.87, "grad_norm": 12.875, "learning_rate": 3.1363135204348018e-06, "loss": 0.2385, "step": 4605 }, { "epoch": 0.87, "grad_norm": 27.625, "learning_rate": 3.135572772882912e-06, "loss": 0.2708, "step": 4606 }, { "epoch": 0.87, "grad_norm": 12.75, "learning_rate": 3.1348319656716885e-06, "loss": 0.2644, "step": 4607 }, { "epoch": 0.87, "grad_norm": 13.375, "learning_rate": 3.134091098870668e-06, "loss": 0.308, "step": 4608 }, { "epoch": 0.87, "grad_norm": 12.75, "learning_rate": 3.133350172549394e-06, "loss": 0.2745, "step": 4609 }, { "epoch": 0.87, "grad_norm": 8.1875, "learning_rate": 3.1326091867774157e-06, "loss": 0.2016, "step": 4610 }, { "epoch": 0.87, "grad_norm": 16.625, "learning_rate": 3.1318681416242857e-06, "loss": 0.2308, "step": 4611 }, { "epoch": 0.87, "grad_norm": 11.5625, "learning_rate": 3.1311270371595646e-06, "loss": 0.2818, "step": 4612 }, { "epoch": 0.87, "grad_norm": 8.625, "learning_rate": 3.1303858734528188e-06, "loss": 0.2397, "step": 4613 }, { "epoch": 0.87, "grad_norm": 12.125, "learning_rate": 3.1296446505736167e-06, "loss": 0.3199, "step": 4614 }, { "epoch": 0.87, "grad_norm": 10.0625, "learning_rate": 3.1289033685915366e-06, "loss": 0.2421, "step": 4615 }, { "epoch": 0.87, "grad_norm": 10.8125, "learning_rate": 3.1281620275761594e-06, "loss": 0.2124, "step": 4616 }, { "epoch": 0.87, "grad_norm": 8.875, "learning_rate": 3.127420627597074e-06, "loss": 0.1995, "step": 4617 }, { "epoch": 0.87, "grad_norm": 9.3125, "learning_rate": 3.126679168723871e-06, "loss": 0.2621, "step": 4618 }, { "epoch": 0.87, "grad_norm": 13.125, "learning_rate": 3.125937651026152e-06, "loss": 0.3076, "step": 4619 }, { "epoch": 0.87, "grad_norm": 11.25, "learning_rate": 3.1251960745735176e-06, "loss": 0.2067, "step": 4620 }, { "epoch": 0.87, "grad_norm": 9.3125, "learning_rate": 3.1244544394355807e-06, "loss": 0.2879, "step": 4621 }, { "epoch": 0.87, "grad_norm": 11.6875, "learning_rate": 3.1237127456819545e-06, "loss": 0.2604, "step": 4622 }, { "epoch": 0.87, "grad_norm": 19.625, "learning_rate": 3.1229709933822597e-06, "loss": 0.2304, "step": 4623 }, { "epoch": 0.87, "grad_norm": 9.8125, "learning_rate": 3.1222291826061234e-06, "loss": 0.258, "step": 4624 }, { "epoch": 0.87, "grad_norm": 7.90625, "learning_rate": 3.121487313423176e-06, "loss": 0.1963, "step": 4625 }, { "epoch": 0.88, "grad_norm": 9.5, "learning_rate": 3.1207453859030555e-06, "loss": 0.2479, "step": 4626 }, { "epoch": 0.88, "grad_norm": 9.125, "learning_rate": 3.1200034001154034e-06, "loss": 0.2059, "step": 4627 }, { "epoch": 0.88, "grad_norm": 9.5625, "learning_rate": 3.119261356129869e-06, "loss": 0.2729, "step": 4628 }, { "epoch": 0.88, "grad_norm": 31.5, "learning_rate": 3.1185192540161043e-06, "loss": 0.2823, "step": 4629 }, { "epoch": 0.88, "grad_norm": 10.125, "learning_rate": 3.1177770938437696e-06, "loss": 0.2361, "step": 4630 }, { "epoch": 0.88, "grad_norm": 10.6875, "learning_rate": 3.117034875682528e-06, "loss": 0.3086, "step": 4631 }, { "epoch": 0.88, "grad_norm": 17.875, "learning_rate": 3.116292599602051e-06, "loss": 0.2991, "step": 4632 }, { "epoch": 0.88, "grad_norm": 14.0625, "learning_rate": 3.115550265672012e-06, "loss": 0.2753, "step": 4633 }, { "epoch": 0.88, "grad_norm": 12.625, "learning_rate": 3.1148078739620934e-06, "loss": 0.2829, "step": 4634 }, { "epoch": 0.88, "grad_norm": 14.6875, "learning_rate": 3.11406542454198e-06, "loss": 0.3025, "step": 4635 }, { "epoch": 0.88, "grad_norm": 11.4375, "learning_rate": 3.113322917481364e-06, "loss": 0.2705, "step": 4636 }, { "epoch": 0.88, "grad_norm": 11.875, "learning_rate": 3.1125803528499416e-06, "loss": 0.3098, "step": 4637 }, { "epoch": 0.88, "grad_norm": 12.4375, "learning_rate": 3.1118377307174156e-06, "loss": 0.3228, "step": 4638 }, { "epoch": 0.88, "grad_norm": 18.375, "learning_rate": 3.1110950511534943e-06, "loss": 0.2908, "step": 4639 }, { "epoch": 0.88, "grad_norm": 9.8125, "learning_rate": 3.1103523142278908e-06, "loss": 0.233, "step": 4640 }, { "epoch": 0.88, "grad_norm": 11.75, "learning_rate": 3.109609520010323e-06, "loss": 0.2861, "step": 4641 }, { "epoch": 0.88, "grad_norm": 9.375, "learning_rate": 3.108866668570514e-06, "loss": 0.2611, "step": 4642 }, { "epoch": 0.88, "grad_norm": 9.3125, "learning_rate": 3.108123759978195e-06, "loss": 0.2956, "step": 4643 }, { "epoch": 0.88, "grad_norm": 11.875, "learning_rate": 3.1073807943030994e-06, "loss": 0.2604, "step": 4644 }, { "epoch": 0.88, "grad_norm": 9.125, "learning_rate": 3.106637771614967e-06, "loss": 0.2093, "step": 4645 }, { "epoch": 0.88, "grad_norm": 8.625, "learning_rate": 3.1058946919835443e-06, "loss": 0.2934, "step": 4646 }, { "epoch": 0.88, "grad_norm": 8.1875, "learning_rate": 3.1051515554785805e-06, "loss": 0.237, "step": 4647 }, { "epoch": 0.88, "grad_norm": 9.3125, "learning_rate": 3.104408362169833e-06, "loss": 0.2343, "step": 4648 }, { "epoch": 0.88, "grad_norm": 16.25, "learning_rate": 3.1036651121270622e-06, "loss": 0.2453, "step": 4649 }, { "epoch": 0.88, "grad_norm": 9.75, "learning_rate": 3.1029218054200354e-06, "loss": 0.315, "step": 4650 }, { "epoch": 0.88, "grad_norm": 11.5, "learning_rate": 3.102178442118524e-06, "loss": 0.296, "step": 4651 }, { "epoch": 0.88, "grad_norm": 13.875, "learning_rate": 3.1014350222923063e-06, "loss": 0.2852, "step": 4652 }, { "epoch": 0.88, "grad_norm": 9.75, "learning_rate": 3.100691546011163e-06, "loss": 0.2622, "step": 4653 }, { "epoch": 0.88, "grad_norm": 11.625, "learning_rate": 3.0999480133448845e-06, "loss": 0.2538, "step": 4654 }, { "epoch": 0.88, "grad_norm": 9.875, "learning_rate": 3.0992044243632622e-06, "loss": 0.2744, "step": 4655 }, { "epoch": 0.88, "grad_norm": 17.875, "learning_rate": 3.0984607791360955e-06, "loss": 0.2181, "step": 4656 }, { "epoch": 0.88, "grad_norm": 11.5, "learning_rate": 3.097717077733187e-06, "loss": 0.2381, "step": 4657 }, { "epoch": 0.88, "grad_norm": 12.25, "learning_rate": 3.096973320224347e-06, "loss": 0.3512, "step": 4658 }, { "epoch": 0.88, "grad_norm": 10.125, "learning_rate": 3.09622950667939e-06, "loss": 0.2243, "step": 4659 }, { "epoch": 0.88, "grad_norm": 10.0, "learning_rate": 3.095485637168134e-06, "loss": 0.2646, "step": 4660 }, { "epoch": 0.88, "grad_norm": 10.5, "learning_rate": 3.094741711760406e-06, "loss": 0.275, "step": 4661 }, { "epoch": 0.88, "grad_norm": 10.5625, "learning_rate": 3.0939977305260337e-06, "loss": 0.3096, "step": 4662 }, { "epoch": 0.88, "grad_norm": 14.75, "learning_rate": 3.093253693534855e-06, "loss": 0.2702, "step": 4663 }, { "epoch": 0.88, "grad_norm": 16.25, "learning_rate": 3.0925096008567083e-06, "loss": 0.2361, "step": 4664 }, { "epoch": 0.88, "grad_norm": 8.5625, "learning_rate": 3.0917654525614404e-06, "loss": 0.2477, "step": 4665 }, { "epoch": 0.88, "grad_norm": 15.875, "learning_rate": 3.0910212487189016e-06, "loss": 0.3058, "step": 4666 }, { "epoch": 0.88, "grad_norm": 14.1875, "learning_rate": 3.090276989398949e-06, "loss": 0.2346, "step": 4667 }, { "epoch": 0.88, "grad_norm": 9.875, "learning_rate": 3.0895326746714426e-06, "loss": 0.2145, "step": 4668 }, { "epoch": 0.88, "grad_norm": 9.5625, "learning_rate": 3.0887883046062506e-06, "loss": 0.3124, "step": 4669 }, { "epoch": 0.88, "grad_norm": 11.75, "learning_rate": 3.0880438792732443e-06, "loss": 0.304, "step": 4670 }, { "epoch": 0.88, "grad_norm": 12.0, "learning_rate": 3.0872993987423e-06, "loss": 0.3051, "step": 4671 }, { "epoch": 0.88, "grad_norm": 12.125, "learning_rate": 3.0865548630833015e-06, "loss": 0.237, "step": 4672 }, { "epoch": 0.88, "grad_norm": 8.1875, "learning_rate": 3.0858102723661335e-06, "loss": 0.2239, "step": 4673 }, { "epoch": 0.88, "grad_norm": 12.0, "learning_rate": 3.085065626660692e-06, "loss": 0.3173, "step": 4674 }, { "epoch": 0.88, "grad_norm": 10.5625, "learning_rate": 3.084320926036872e-06, "loss": 0.3264, "step": 4675 }, { "epoch": 0.88, "grad_norm": 15.25, "learning_rate": 3.0835761705645773e-06, "loss": 0.3188, "step": 4676 }, { "epoch": 0.88, "grad_norm": 12.875, "learning_rate": 3.082831360313715e-06, "loss": 0.2868, "step": 4677 }, { "epoch": 0.88, "grad_norm": 11.1875, "learning_rate": 3.0820864953542007e-06, "loss": 0.3108, "step": 4678 }, { "epoch": 0.89, "grad_norm": 13.3125, "learning_rate": 3.081341575755949e-06, "loss": 0.3298, "step": 4679 }, { "epoch": 0.89, "grad_norm": 13.0, "learning_rate": 3.0805966015888866e-06, "loss": 0.3274, "step": 4680 }, { "epoch": 0.89, "grad_norm": 11.0, "learning_rate": 3.07985157292294e-06, "loss": 0.2178, "step": 4681 }, { "epoch": 0.89, "grad_norm": 10.875, "learning_rate": 3.079106489828044e-06, "loss": 0.2751, "step": 4682 }, { "epoch": 0.89, "grad_norm": 10.9375, "learning_rate": 3.0783613523741366e-06, "loss": 0.274, "step": 4683 }, { "epoch": 0.89, "grad_norm": 11.9375, "learning_rate": 3.077616160631162e-06, "loss": 0.2437, "step": 4684 }, { "epoch": 0.89, "grad_norm": 12.125, "learning_rate": 3.076870914669069e-06, "loss": 0.2501, "step": 4685 }, { "epoch": 0.89, "grad_norm": 11.5, "learning_rate": 3.0761256145578123e-06, "loss": 0.2762, "step": 4686 }, { "epoch": 0.89, "grad_norm": 26.875, "learning_rate": 3.0753802603673495e-06, "loss": 0.2899, "step": 4687 }, { "epoch": 0.89, "grad_norm": 10.4375, "learning_rate": 3.074634852167646e-06, "loss": 0.2371, "step": 4688 }, { "epoch": 0.89, "grad_norm": 16.25, "learning_rate": 3.073889390028671e-06, "loss": 0.1946, "step": 4689 }, { "epoch": 0.89, "grad_norm": 11.4375, "learning_rate": 3.0731438740203982e-06, "loss": 0.3208, "step": 4690 }, { "epoch": 0.89, "grad_norm": 10.6875, "learning_rate": 3.0723983042128075e-06, "loss": 0.31, "step": 4691 }, { "epoch": 0.89, "grad_norm": 11.75, "learning_rate": 3.071652680675883e-06, "loss": 0.2407, "step": 4692 }, { "epoch": 0.89, "grad_norm": 9.375, "learning_rate": 3.0709070034796152e-06, "loss": 0.2858, "step": 4693 }, { "epoch": 0.89, "grad_norm": 11.25, "learning_rate": 3.0701612726939965e-06, "loss": 0.2466, "step": 4694 }, { "epoch": 0.89, "grad_norm": 8.0625, "learning_rate": 3.0694154883890287e-06, "loss": 0.2672, "step": 4695 }, { "epoch": 0.89, "grad_norm": 17.625, "learning_rate": 3.0686696506347146e-06, "loss": 0.3281, "step": 4696 }, { "epoch": 0.89, "grad_norm": 13.4375, "learning_rate": 3.067923759501065e-06, "loss": 0.3838, "step": 4697 }, { "epoch": 0.89, "grad_norm": 10.9375, "learning_rate": 3.0671778150580943e-06, "loss": 0.2096, "step": 4698 }, { "epoch": 0.89, "grad_norm": 12.375, "learning_rate": 3.066431817375821e-06, "loss": 0.3318, "step": 4699 }, { "epoch": 0.89, "grad_norm": 16.5, "learning_rate": 3.065685766524271e-06, "loss": 0.2431, "step": 4700 }, { "epoch": 0.89, "grad_norm": 13.875, "learning_rate": 3.064939662573474e-06, "loss": 0.3198, "step": 4701 }, { "epoch": 0.89, "grad_norm": 9.8125, "learning_rate": 3.0641935055934633e-06, "loss": 0.2394, "step": 4702 }, { "epoch": 0.89, "grad_norm": 10.4375, "learning_rate": 3.0634472956542795e-06, "loss": 0.3114, "step": 4703 }, { "epoch": 0.89, "grad_norm": 12.25, "learning_rate": 3.0627010328259667e-06, "loss": 0.3415, "step": 4704 }, { "epoch": 0.89, "grad_norm": 12.5625, "learning_rate": 3.061954717178574e-06, "loss": 0.3242, "step": 4705 }, { "epoch": 0.89, "grad_norm": 9.6875, "learning_rate": 3.0612083487821564e-06, "loss": 0.2985, "step": 4706 }, { "epoch": 0.89, "grad_norm": 12.875, "learning_rate": 3.060461927706773e-06, "loss": 0.258, "step": 4707 }, { "epoch": 0.89, "grad_norm": 7.28125, "learning_rate": 3.0597154540224888e-06, "loss": 0.2388, "step": 4708 }, { "epoch": 0.89, "grad_norm": 9.5, "learning_rate": 3.058968927799373e-06, "loss": 0.2517, "step": 4709 }, { "epoch": 0.89, "grad_norm": 13.125, "learning_rate": 3.058222349107498e-06, "loss": 0.2819, "step": 4710 }, { "epoch": 0.89, "grad_norm": 11.9375, "learning_rate": 3.057475718016945e-06, "loss": 0.263, "step": 4711 }, { "epoch": 0.89, "grad_norm": 10.75, "learning_rate": 3.0567290345977977e-06, "loss": 0.318, "step": 4712 }, { "epoch": 0.89, "grad_norm": 12.25, "learning_rate": 3.055982298920144e-06, "loss": 0.3138, "step": 4713 }, { "epoch": 0.89, "grad_norm": 11.0625, "learning_rate": 3.0552355110540794e-06, "loss": 0.2728, "step": 4714 }, { "epoch": 0.89, "grad_norm": 12.75, "learning_rate": 3.0544886710697013e-06, "loss": 0.2906, "step": 4715 }, { "epoch": 0.89, "grad_norm": 13.125, "learning_rate": 3.0537417790371137e-06, "loss": 0.3199, "step": 4716 }, { "epoch": 0.89, "grad_norm": 11.0, "learning_rate": 3.0529948350264247e-06, "loss": 0.2559, "step": 4717 }, { "epoch": 0.89, "grad_norm": 7.8125, "learning_rate": 3.052247839107749e-06, "loss": 0.2256, "step": 4718 }, { "epoch": 0.89, "grad_norm": 15.9375, "learning_rate": 3.051500791351204e-06, "loss": 0.2713, "step": 4719 }, { "epoch": 0.89, "grad_norm": 9.5, "learning_rate": 3.0507536918269133e-06, "loss": 0.3341, "step": 4720 }, { "epoch": 0.89, "grad_norm": 14.3125, "learning_rate": 3.050006540605004e-06, "loss": 0.2697, "step": 4721 }, { "epoch": 0.89, "grad_norm": 9.875, "learning_rate": 3.0492593377556105e-06, "loss": 0.3104, "step": 4722 }, { "epoch": 0.89, "grad_norm": 13.0625, "learning_rate": 3.0485120833488683e-06, "loss": 0.2314, "step": 4723 }, { "epoch": 0.89, "grad_norm": 14.8125, "learning_rate": 3.047764777454923e-06, "loss": 0.2321, "step": 4724 }, { "epoch": 0.89, "grad_norm": 16.25, "learning_rate": 3.047017420143919e-06, "loss": 0.2341, "step": 4725 }, { "epoch": 0.89, "grad_norm": 11.0, "learning_rate": 3.0462700114860107e-06, "loss": 0.3405, "step": 4726 }, { "epoch": 0.89, "grad_norm": 15.8125, "learning_rate": 3.0455225515513533e-06, "loss": 0.2794, "step": 4727 }, { "epoch": 0.89, "grad_norm": 9.875, "learning_rate": 3.04477504041011e-06, "loss": 0.2758, "step": 4728 }, { "epoch": 0.89, "grad_norm": 11.0, "learning_rate": 3.044027478132447e-06, "loss": 0.2837, "step": 4729 }, { "epoch": 0.89, "grad_norm": 11.875, "learning_rate": 3.043279864788536e-06, "loss": 0.2689, "step": 4730 }, { "epoch": 0.89, "grad_norm": 9.25, "learning_rate": 3.042532200448553e-06, "loss": 0.1546, "step": 4731 }, { "epoch": 0.9, "grad_norm": 17.25, "learning_rate": 3.041784485182679e-06, "loss": 0.2706, "step": 4732 }, { "epoch": 0.9, "grad_norm": 10.5625, "learning_rate": 3.0410367190611e-06, "loss": 0.2902, "step": 4733 }, { "epoch": 0.9, "grad_norm": 10.0, "learning_rate": 3.040288902154006e-06, "loss": 0.2867, "step": 4734 }, { "epoch": 0.9, "grad_norm": 12.5625, "learning_rate": 3.039541034531594e-06, "loss": 0.3053, "step": 4735 }, { "epoch": 0.9, "grad_norm": 12.875, "learning_rate": 3.038793116264062e-06, "loss": 0.2639, "step": 4736 }, { "epoch": 0.9, "grad_norm": 9.625, "learning_rate": 3.0380451474216165e-06, "loss": 0.2614, "step": 4737 }, { "epoch": 0.9, "grad_norm": 10.875, "learning_rate": 3.0372971280744655e-06, "loss": 0.2487, "step": 4738 }, { "epoch": 0.9, "grad_norm": 14.3125, "learning_rate": 3.036549058292825e-06, "loss": 0.353, "step": 4739 }, { "epoch": 0.9, "grad_norm": 11.875, "learning_rate": 3.0358009381469132e-06, "loss": 0.2976, "step": 4740 }, { "epoch": 0.9, "grad_norm": 10.9375, "learning_rate": 3.0350527677069547e-06, "loss": 0.2305, "step": 4741 }, { "epoch": 0.9, "grad_norm": 10.3125, "learning_rate": 3.034304547043177e-06, "loss": 0.2906, "step": 4742 }, { "epoch": 0.9, "grad_norm": 9.9375, "learning_rate": 3.033556276225814e-06, "loss": 0.3226, "step": 4743 }, { "epoch": 0.9, "grad_norm": 9.6875, "learning_rate": 3.0328079553251033e-06, "loss": 0.2672, "step": 4744 }, { "epoch": 0.9, "grad_norm": 9.8125, "learning_rate": 3.0320595844112884e-06, "loss": 0.2579, "step": 4745 }, { "epoch": 0.9, "grad_norm": 15.75, "learning_rate": 3.0313111635546156e-06, "loss": 0.2073, "step": 4746 }, { "epoch": 0.9, "grad_norm": 14.25, "learning_rate": 3.0305626928253375e-06, "loss": 0.2587, "step": 4747 }, { "epoch": 0.9, "grad_norm": 10.1875, "learning_rate": 3.0298141722937103e-06, "loss": 0.2496, "step": 4748 }, { "epoch": 0.9, "grad_norm": 11.375, "learning_rate": 3.0290656020299963e-06, "loss": 0.2927, "step": 4749 }, { "epoch": 0.9, "grad_norm": 12.125, "learning_rate": 3.028316982104461e-06, "loss": 0.3005, "step": 4750 }, { "epoch": 0.9, "grad_norm": 12.375, "learning_rate": 3.027568312587375e-06, "loss": 0.3224, "step": 4751 }, { "epoch": 0.9, "grad_norm": 12.0, "learning_rate": 3.026819593549014e-06, "loss": 0.2955, "step": 4752 }, { "epoch": 0.9, "grad_norm": 12.0625, "learning_rate": 3.0260708250596583e-06, "loss": 0.2026, "step": 4753 }, { "epoch": 0.9, "grad_norm": 8.875, "learning_rate": 3.025322007189592e-06, "loss": 0.255, "step": 4754 }, { "epoch": 0.9, "grad_norm": 7.875, "learning_rate": 3.024573140009105e-06, "loss": 0.2616, "step": 4755 }, { "epoch": 0.9, "grad_norm": 10.0, "learning_rate": 3.0238242235884906e-06, "loss": 0.2621, "step": 4756 }, { "epoch": 0.9, "grad_norm": 13.75, "learning_rate": 3.0230752579980477e-06, "loss": 0.329, "step": 4757 }, { "epoch": 0.9, "grad_norm": 19.375, "learning_rate": 3.0223262433080793e-06, "loss": 0.2393, "step": 4758 }, { "epoch": 0.9, "grad_norm": 10.9375, "learning_rate": 3.0215771795888936e-06, "loss": 0.3024, "step": 4759 }, { "epoch": 0.9, "grad_norm": 11.5625, "learning_rate": 3.0208280669108026e-06, "loss": 0.2844, "step": 4760 }, { "epoch": 0.9, "grad_norm": 11.6875, "learning_rate": 3.0200789053441225e-06, "loss": 0.1718, "step": 4761 }, { "epoch": 0.9, "grad_norm": 18.5, "learning_rate": 3.0193296949591775e-06, "loss": 0.3128, "step": 4762 }, { "epoch": 0.9, "grad_norm": 9.375, "learning_rate": 3.018580435826291e-06, "loss": 0.2278, "step": 4763 }, { "epoch": 0.9, "grad_norm": 9.5, "learning_rate": 3.017831128015796e-06, "loss": 0.2663, "step": 4764 }, { "epoch": 0.9, "grad_norm": 13.4375, "learning_rate": 3.017081771598025e-06, "loss": 0.2782, "step": 4765 }, { "epoch": 0.9, "grad_norm": 14.5, "learning_rate": 3.016332366643321e-06, "loss": 0.2539, "step": 4766 }, { "epoch": 0.9, "grad_norm": 7.375, "learning_rate": 3.015582913222025e-06, "loss": 0.2272, "step": 4767 }, { "epoch": 0.9, "grad_norm": 19.125, "learning_rate": 3.0148334114044897e-06, "loss": 0.2504, "step": 4768 }, { "epoch": 0.9, "grad_norm": 9.625, "learning_rate": 3.0140838612610656e-06, "loss": 0.2861, "step": 4769 }, { "epoch": 0.9, "grad_norm": 11.4375, "learning_rate": 3.013334262862113e-06, "loss": 0.3094, "step": 4770 }, { "epoch": 0.9, "grad_norm": 12.9375, "learning_rate": 3.012584616277992e-06, "loss": 0.3354, "step": 4771 }, { "epoch": 0.9, "grad_norm": 9.0, "learning_rate": 3.011834921579072e-06, "loss": 0.2615, "step": 4772 }, { "epoch": 0.9, "grad_norm": 9.375, "learning_rate": 3.011085178835724e-06, "loss": 0.2288, "step": 4773 }, { "epoch": 0.9, "grad_norm": 12.5625, "learning_rate": 3.0103353881183236e-06, "loss": 0.2685, "step": 4774 }, { "epoch": 0.9, "grad_norm": 11.125, "learning_rate": 3.0095855494972515e-06, "loss": 0.2447, "step": 4775 }, { "epoch": 0.9, "grad_norm": 11.8125, "learning_rate": 3.0088356630428935e-06, "loss": 0.3175, "step": 4776 }, { "epoch": 0.9, "grad_norm": 13.25, "learning_rate": 3.008085728825639e-06, "loss": 0.282, "step": 4777 }, { "epoch": 0.9, "grad_norm": 11.625, "learning_rate": 3.007335746915881e-06, "loss": 0.2136, "step": 4778 }, { "epoch": 0.9, "grad_norm": 17.25, "learning_rate": 3.0065857173840197e-06, "loss": 0.2142, "step": 4779 }, { "epoch": 0.9, "grad_norm": 11.0, "learning_rate": 3.0058356403004563e-06, "loss": 0.2538, "step": 4780 }, { "epoch": 0.9, "grad_norm": 13.875, "learning_rate": 3.0050855157356006e-06, "loss": 0.3116, "step": 4781 }, { "epoch": 0.9, "grad_norm": 11.1875, "learning_rate": 3.004335343759863e-06, "loss": 0.29, "step": 4782 }, { "epoch": 0.9, "grad_norm": 12.875, "learning_rate": 3.0035851244436604e-06, "loss": 0.2429, "step": 4783 }, { "epoch": 0.91, "grad_norm": 8.5625, "learning_rate": 3.002834857857413e-06, "loss": 0.1989, "step": 4784 }, { "epoch": 0.91, "grad_norm": 9.875, "learning_rate": 3.0020845440715466e-06, "loss": 0.2863, "step": 4785 }, { "epoch": 0.91, "grad_norm": 18.875, "learning_rate": 3.001334183156492e-06, "loss": 0.2968, "step": 4786 }, { "epoch": 0.91, "grad_norm": 12.0, "learning_rate": 3.0005837751826815e-06, "loss": 0.2691, "step": 4787 }, { "epoch": 0.91, "grad_norm": 10.0625, "learning_rate": 2.999833320220555e-06, "loss": 0.2521, "step": 4788 }, { "epoch": 0.91, "grad_norm": 15.5, "learning_rate": 2.999082818340555e-06, "loss": 0.2703, "step": 4789 }, { "epoch": 0.91, "grad_norm": 13.6875, "learning_rate": 2.998332269613129e-06, "loss": 0.3263, "step": 4790 }, { "epoch": 0.91, "grad_norm": 12.125, "learning_rate": 2.9975816741087282e-06, "loss": 0.2931, "step": 4791 }, { "epoch": 0.91, "grad_norm": 11.9375, "learning_rate": 2.9968310318978095e-06, "loss": 0.2943, "step": 4792 }, { "epoch": 0.91, "grad_norm": 14.4375, "learning_rate": 2.9960803430508334e-06, "loss": 0.2669, "step": 4793 }, { "epoch": 0.91, "grad_norm": 8.9375, "learning_rate": 2.995329607638265e-06, "loss": 0.2688, "step": 4794 }, { "epoch": 0.91, "grad_norm": 17.625, "learning_rate": 2.9945788257305725e-06, "loss": 0.2882, "step": 4795 }, { "epoch": 0.91, "grad_norm": 12.125, "learning_rate": 2.993827997398231e-06, "loss": 0.281, "step": 4796 }, { "epoch": 0.91, "grad_norm": 12.375, "learning_rate": 2.9930771227117173e-06, "loss": 0.316, "step": 4797 }, { "epoch": 0.91, "grad_norm": 11.5, "learning_rate": 2.992326201741515e-06, "loss": 0.3075, "step": 4798 }, { "epoch": 0.91, "grad_norm": 8.25, "learning_rate": 2.99157523455811e-06, "loss": 0.2148, "step": 4799 }, { "epoch": 0.91, "grad_norm": 8.0, "learning_rate": 2.9908242212319933e-06, "loss": 0.2607, "step": 4800 }, { "epoch": 0.91, "grad_norm": 13.125, "learning_rate": 2.9900731618336608e-06, "loss": 0.3436, "step": 4801 }, { "epoch": 0.91, "grad_norm": 14.125, "learning_rate": 2.9893220564336117e-06, "loss": 0.3203, "step": 4802 }, { "epoch": 0.91, "grad_norm": 15.8125, "learning_rate": 2.988570905102351e-06, "loss": 0.2824, "step": 4803 }, { "epoch": 0.91, "grad_norm": 14.9375, "learning_rate": 2.9878197079103855e-06, "loss": 0.2912, "step": 4804 }, { "epoch": 0.91, "grad_norm": 11.4375, "learning_rate": 2.9870684649282296e-06, "loss": 0.2892, "step": 4805 }, { "epoch": 0.91, "grad_norm": 12.6875, "learning_rate": 2.9863171762263985e-06, "loss": 0.3081, "step": 4806 }, { "epoch": 0.91, "grad_norm": 7.25, "learning_rate": 2.985565841875415e-06, "loss": 0.2538, "step": 4807 }, { "epoch": 0.91, "grad_norm": 10.0625, "learning_rate": 2.984814461945804e-06, "loss": 0.3388, "step": 4808 }, { "epoch": 0.91, "grad_norm": 12.1875, "learning_rate": 2.984063036508094e-06, "loss": 0.3746, "step": 4809 }, { "epoch": 0.91, "grad_norm": 13.0625, "learning_rate": 2.983311565632822e-06, "loss": 0.3191, "step": 4810 }, { "epoch": 0.91, "grad_norm": 11.75, "learning_rate": 2.982560049390524e-06, "loss": 0.237, "step": 4811 }, { "epoch": 0.91, "grad_norm": 11.625, "learning_rate": 2.981808487851744e-06, "loss": 0.2994, "step": 4812 }, { "epoch": 0.91, "grad_norm": 14.4375, "learning_rate": 2.9810568810870266e-06, "loss": 0.2783, "step": 4813 }, { "epoch": 0.91, "grad_norm": 11.875, "learning_rate": 2.980305229166926e-06, "loss": 0.2886, "step": 4814 }, { "epoch": 0.91, "grad_norm": 15.8125, "learning_rate": 2.979553532161995e-06, "loss": 0.3271, "step": 4815 }, { "epoch": 0.91, "grad_norm": 8.5, "learning_rate": 2.9788017901427946e-06, "loss": 0.2073, "step": 4816 }, { "epoch": 0.91, "grad_norm": 10.8125, "learning_rate": 2.9780500031798877e-06, "loss": 0.2911, "step": 4817 }, { "epoch": 0.91, "grad_norm": 20.375, "learning_rate": 2.977298171343843e-06, "loss": 0.2312, "step": 4818 }, { "epoch": 0.91, "grad_norm": 9.875, "learning_rate": 2.9765462947052335e-06, "loss": 0.2569, "step": 4819 }, { "epoch": 0.91, "grad_norm": 11.0, "learning_rate": 2.9757943733346333e-06, "loss": 0.2801, "step": 4820 }, { "epoch": 0.91, "grad_norm": 11.25, "learning_rate": 2.975042407302626e-06, "loss": 0.2907, "step": 4821 }, { "epoch": 0.91, "grad_norm": 8.5625, "learning_rate": 2.9742903966797926e-06, "loss": 0.235, "step": 4822 }, { "epoch": 0.91, "grad_norm": 18.625, "learning_rate": 2.973538341536726e-06, "loss": 0.2345, "step": 4823 }, { "epoch": 0.91, "grad_norm": 15.25, "learning_rate": 2.9727862419440166e-06, "loss": 0.2964, "step": 4824 }, { "epoch": 0.91, "grad_norm": 7.21875, "learning_rate": 2.972034097972264e-06, "loss": 0.2621, "step": 4825 }, { "epoch": 0.91, "grad_norm": 13.4375, "learning_rate": 2.9712819096920676e-06, "loss": 0.2358, "step": 4826 }, { "epoch": 0.91, "grad_norm": 10.5, "learning_rate": 2.970529677174035e-06, "loss": 0.2757, "step": 4827 }, { "epoch": 0.91, "grad_norm": 12.125, "learning_rate": 2.9697774004887737e-06, "loss": 0.2362, "step": 4828 }, { "epoch": 0.91, "grad_norm": 10.3125, "learning_rate": 2.9690250797068997e-06, "loss": 0.3141, "step": 4829 }, { "epoch": 0.91, "grad_norm": 10.8125, "learning_rate": 2.968272714899031e-06, "loss": 0.2438, "step": 4830 }, { "epoch": 0.91, "grad_norm": 11.1875, "learning_rate": 2.9675203061357893e-06, "loss": 0.2836, "step": 4831 }, { "epoch": 0.91, "grad_norm": 11.5, "learning_rate": 2.9667678534878006e-06, "loss": 0.2263, "step": 4832 }, { "epoch": 0.91, "grad_norm": 13.875, "learning_rate": 2.966015357025696e-06, "loss": 0.2806, "step": 4833 }, { "epoch": 0.91, "grad_norm": 8.3125, "learning_rate": 2.9652628168201104e-06, "loss": 0.2222, "step": 4834 }, { "epoch": 0.91, "grad_norm": 9.25, "learning_rate": 2.964510232941682e-06, "loss": 0.3314, "step": 4835 }, { "epoch": 0.91, "grad_norm": 8.6875, "learning_rate": 2.9637576054610533e-06, "loss": 0.2368, "step": 4836 }, { "epoch": 0.92, "grad_norm": 11.875, "learning_rate": 2.963004934448872e-06, "loss": 0.3069, "step": 4837 }, { "epoch": 0.92, "grad_norm": 7.53125, "learning_rate": 2.9622522199757887e-06, "loss": 0.2119, "step": 4838 }, { "epoch": 0.92, "grad_norm": 7.1875, "learning_rate": 2.9614994621124583e-06, "loss": 0.2135, "step": 4839 }, { "epoch": 0.92, "grad_norm": 11.8125, "learning_rate": 2.9607466609295405e-06, "loss": 0.2476, "step": 4840 }, { "epoch": 0.92, "grad_norm": 28.125, "learning_rate": 2.9599938164976976e-06, "loss": 0.2981, "step": 4841 }, { "epoch": 0.92, "grad_norm": 7.96875, "learning_rate": 2.959240928887598e-06, "loss": 0.2409, "step": 4842 }, { "epoch": 0.92, "grad_norm": 13.375, "learning_rate": 2.9584879981699123e-06, "loss": 0.2951, "step": 4843 }, { "epoch": 0.92, "grad_norm": 12.25, "learning_rate": 2.957735024415316e-06, "loss": 0.2764, "step": 4844 }, { "epoch": 0.92, "grad_norm": 11.625, "learning_rate": 2.956982007694488e-06, "loss": 0.2449, "step": 4845 }, { "epoch": 0.92, "grad_norm": 11.125, "learning_rate": 2.956228948078113e-06, "loss": 0.2659, "step": 4846 }, { "epoch": 0.92, "grad_norm": 12.875, "learning_rate": 2.955475845636878e-06, "loss": 0.2477, "step": 4847 }, { "epoch": 0.92, "grad_norm": 9.875, "learning_rate": 2.954722700441474e-06, "loss": 0.2825, "step": 4848 }, { "epoch": 0.92, "grad_norm": 8.9375, "learning_rate": 2.9539695125625967e-06, "loss": 0.2528, "step": 4849 }, { "epoch": 0.92, "grad_norm": 13.25, "learning_rate": 2.9532162820709457e-06, "loss": 0.3044, "step": 4850 }, { "epoch": 0.92, "grad_norm": 10.0625, "learning_rate": 2.952463009037224e-06, "loss": 0.2617, "step": 4851 }, { "epoch": 0.92, "grad_norm": 20.25, "learning_rate": 2.9517096935321408e-06, "loss": 0.3117, "step": 4852 }, { "epoch": 0.92, "grad_norm": 9.0625, "learning_rate": 2.9509563356264057e-06, "loss": 0.3148, "step": 4853 }, { "epoch": 0.92, "grad_norm": 11.75, "learning_rate": 2.9502029353907358e-06, "loss": 0.3366, "step": 4854 }, { "epoch": 0.92, "grad_norm": 9.8125, "learning_rate": 2.9494494928958484e-06, "loss": 0.2347, "step": 4855 }, { "epoch": 0.92, "grad_norm": 15.1875, "learning_rate": 2.9486960082124695e-06, "loss": 0.3204, "step": 4856 }, { "epoch": 0.92, "grad_norm": 10.8125, "learning_rate": 2.947942481411324e-06, "loss": 0.3213, "step": 4857 }, { "epoch": 0.92, "grad_norm": 10.5625, "learning_rate": 2.947188912563145e-06, "loss": 0.2892, "step": 4858 }, { "epoch": 0.92, "grad_norm": 16.125, "learning_rate": 2.9464353017386664e-06, "loss": 0.2962, "step": 4859 }, { "epoch": 0.92, "grad_norm": 10.625, "learning_rate": 2.9456816490086294e-06, "loss": 0.2493, "step": 4860 }, { "epoch": 0.92, "grad_norm": 10.125, "learning_rate": 2.9449279544437753e-06, "loss": 0.2714, "step": 4861 }, { "epoch": 0.92, "grad_norm": 10.125, "learning_rate": 2.9441742181148513e-06, "loss": 0.3074, "step": 4862 }, { "epoch": 0.92, "grad_norm": 12.9375, "learning_rate": 2.9434204400926104e-06, "loss": 0.3193, "step": 4863 }, { "epoch": 0.92, "grad_norm": 12.625, "learning_rate": 2.9426666204478037e-06, "loss": 0.2952, "step": 4864 }, { "epoch": 0.92, "grad_norm": 8.3125, "learning_rate": 2.9419127592511948e-06, "loss": 0.2477, "step": 4865 }, { "epoch": 0.92, "grad_norm": 12.125, "learning_rate": 2.9411588565735422e-06, "loss": 0.2656, "step": 4866 }, { "epoch": 0.92, "grad_norm": 12.5625, "learning_rate": 2.940404912485615e-06, "loss": 0.3115, "step": 4867 }, { "epoch": 0.92, "grad_norm": 8.75, "learning_rate": 2.9396509270581827e-06, "loss": 0.2214, "step": 4868 }, { "epoch": 0.92, "grad_norm": 12.1875, "learning_rate": 2.9388969003620203e-06, "loss": 0.3037, "step": 4869 }, { "epoch": 0.92, "grad_norm": 12.0625, "learning_rate": 2.938142832467905e-06, "loss": 0.2781, "step": 4870 }, { "epoch": 0.92, "grad_norm": 11.3125, "learning_rate": 2.9373887234466205e-06, "loss": 0.2947, "step": 4871 }, { "epoch": 0.92, "grad_norm": 8.3125, "learning_rate": 2.9366345733689505e-06, "loss": 0.2001, "step": 4872 }, { "epoch": 0.92, "grad_norm": 8.0625, "learning_rate": 2.9358803823056873e-06, "loss": 0.2751, "step": 4873 }, { "epoch": 0.92, "grad_norm": 8.6875, "learning_rate": 2.9351261503276233e-06, "loss": 0.2516, "step": 4874 }, { "epoch": 0.92, "grad_norm": 12.9375, "learning_rate": 2.9343718775055564e-06, "loss": 0.3335, "step": 4875 }, { "epoch": 0.92, "grad_norm": 9.9375, "learning_rate": 2.9336175639102877e-06, "loss": 0.2647, "step": 4876 }, { "epoch": 0.92, "grad_norm": 9.6875, "learning_rate": 2.9328632096126227e-06, "loss": 0.2559, "step": 4877 }, { "epoch": 0.92, "grad_norm": 10.0625, "learning_rate": 2.93210881468337e-06, "loss": 0.2055, "step": 4878 }, { "epoch": 0.92, "grad_norm": 15.625, "learning_rate": 2.931354379193342e-06, "loss": 0.2857, "step": 4879 }, { "epoch": 0.92, "grad_norm": 10.625, "learning_rate": 2.9305999032133563e-06, "loss": 0.2657, "step": 4880 }, { "epoch": 0.92, "grad_norm": 8.375, "learning_rate": 2.929845386814233e-06, "loss": 0.2062, "step": 4881 }, { "epoch": 0.92, "grad_norm": 8.875, "learning_rate": 2.929090830066796e-06, "loss": 0.1681, "step": 4882 }, { "epoch": 0.92, "grad_norm": 10.3125, "learning_rate": 2.928336233041874e-06, "loss": 0.2149, "step": 4883 }, { "epoch": 0.92, "grad_norm": 12.0, "learning_rate": 2.9275815958102984e-06, "loss": 0.2466, "step": 4884 }, { "epoch": 0.92, "grad_norm": 16.25, "learning_rate": 2.9268269184429043e-06, "loss": 0.2866, "step": 4885 }, { "epoch": 0.92, "grad_norm": 15.9375, "learning_rate": 2.926072201010532e-06, "loss": 0.306, "step": 4886 }, { "epoch": 0.92, "grad_norm": 10.9375, "learning_rate": 2.9253174435840237e-06, "loss": 0.3075, "step": 4887 }, { "epoch": 0.92, "grad_norm": 11.375, "learning_rate": 2.9245626462342264e-06, "loss": 0.2388, "step": 4888 }, { "epoch": 0.92, "grad_norm": 11.6875, "learning_rate": 2.923807809031991e-06, "loss": 0.3082, "step": 4889 }, { "epoch": 0.93, "grad_norm": 11.0, "learning_rate": 2.9230529320481716e-06, "loss": 0.2374, "step": 4890 }, { "epoch": 0.93, "grad_norm": 9.9375, "learning_rate": 2.9222980153536266e-06, "loss": 0.2732, "step": 4891 }, { "epoch": 0.93, "grad_norm": 12.1875, "learning_rate": 2.9215430590192174e-06, "loss": 0.2656, "step": 4892 }, { "epoch": 0.93, "grad_norm": 10.0, "learning_rate": 2.92078806311581e-06, "loss": 0.2631, "step": 4893 }, { "epoch": 0.93, "grad_norm": 11.5625, "learning_rate": 2.920033027714273e-06, "loss": 0.2604, "step": 4894 }, { "epoch": 0.93, "grad_norm": 13.625, "learning_rate": 2.9192779528854796e-06, "loss": 0.2749, "step": 4895 }, { "epoch": 0.93, "grad_norm": 13.125, "learning_rate": 2.9185228387003067e-06, "loss": 0.275, "step": 4896 }, { "epoch": 0.93, "grad_norm": 11.625, "learning_rate": 2.917767685229634e-06, "loss": 0.274, "step": 4897 }, { "epoch": 0.93, "grad_norm": 11.0625, "learning_rate": 2.9170124925443468e-06, "loss": 0.2297, "step": 4898 }, { "epoch": 0.93, "grad_norm": 14.625, "learning_rate": 2.9162572607153307e-06, "loss": 0.3972, "step": 4899 }, { "epoch": 0.93, "grad_norm": 11.8125, "learning_rate": 2.9155019898134796e-06, "loss": 0.2415, "step": 4900 }, { "epoch": 0.93, "grad_norm": 12.1875, "learning_rate": 2.9147466799096857e-06, "loss": 0.277, "step": 4901 }, { "epoch": 0.93, "grad_norm": 11.0625, "learning_rate": 2.9139913310748514e-06, "loss": 0.2786, "step": 4902 }, { "epoch": 0.93, "grad_norm": 13.0, "learning_rate": 2.913235943379875e-06, "loss": 0.3081, "step": 4903 }, { "epoch": 0.93, "grad_norm": 8.9375, "learning_rate": 2.9124805168956655e-06, "loss": 0.276, "step": 4904 }, { "epoch": 0.93, "grad_norm": 13.3125, "learning_rate": 2.9117250516931312e-06, "loss": 0.2231, "step": 4905 }, { "epoch": 0.93, "grad_norm": 18.875, "learning_rate": 2.9109695478431852e-06, "loss": 0.3371, "step": 4906 }, { "epoch": 0.93, "grad_norm": 10.0, "learning_rate": 2.9102140054167453e-06, "loss": 0.2574, "step": 4907 }, { "epoch": 0.93, "grad_norm": 10.875, "learning_rate": 2.9094584244847313e-06, "loss": 0.252, "step": 4908 }, { "epoch": 0.93, "grad_norm": 11.25, "learning_rate": 2.908702805118068e-06, "loss": 0.2593, "step": 4909 }, { "epoch": 0.93, "grad_norm": 9.875, "learning_rate": 2.9079471473876825e-06, "loss": 0.2442, "step": 4910 }, { "epoch": 0.93, "grad_norm": 8.4375, "learning_rate": 2.9071914513645074e-06, "loss": 0.2758, "step": 4911 }, { "epoch": 0.93, "grad_norm": 13.0, "learning_rate": 2.906435717119475e-06, "loss": 0.2793, "step": 4912 }, { "epoch": 0.93, "grad_norm": 10.375, "learning_rate": 2.9056799447235267e-06, "loss": 0.3321, "step": 4913 }, { "epoch": 0.93, "grad_norm": 13.6875, "learning_rate": 2.904924134247603e-06, "loss": 0.3061, "step": 4914 }, { "epoch": 0.93, "grad_norm": 10.0, "learning_rate": 2.904168285762651e-06, "loss": 0.2635, "step": 4915 }, { "epoch": 0.93, "grad_norm": 11.8125, "learning_rate": 2.903412399339617e-06, "loss": 0.2673, "step": 4916 }, { "epoch": 0.93, "grad_norm": 8.875, "learning_rate": 2.9026564750494578e-06, "loss": 0.2554, "step": 4917 }, { "epoch": 0.93, "grad_norm": 13.5, "learning_rate": 2.9019005129631263e-06, "loss": 0.2026, "step": 4918 }, { "epoch": 0.93, "grad_norm": 10.1875, "learning_rate": 2.9011445131515847e-06, "loss": 0.2687, "step": 4919 }, { "epoch": 0.93, "grad_norm": 10.4375, "learning_rate": 2.900388475685795e-06, "loss": 0.2518, "step": 4920 }, { "epoch": 0.93, "grad_norm": 12.25, "learning_rate": 2.8996324006367256e-06, "loss": 0.2611, "step": 4921 }, { "epoch": 0.93, "grad_norm": 14.5625, "learning_rate": 2.8988762880753464e-06, "loss": 0.3399, "step": 4922 }, { "epoch": 0.93, "grad_norm": 10.375, "learning_rate": 2.8981201380726308e-06, "loss": 0.2918, "step": 4923 }, { "epoch": 0.93, "grad_norm": 11.5, "learning_rate": 2.897363950699557e-06, "loss": 0.3045, "step": 4924 }, { "epoch": 0.93, "grad_norm": 8.6875, "learning_rate": 2.8966077260271065e-06, "loss": 0.2628, "step": 4925 }, { "epoch": 0.93, "grad_norm": 10.625, "learning_rate": 2.8958514641262623e-06, "loss": 0.2318, "step": 4926 }, { "epoch": 0.93, "grad_norm": 10.0625, "learning_rate": 2.8950951650680144e-06, "loss": 0.2349, "step": 4927 }, { "epoch": 0.93, "grad_norm": 21.25, "learning_rate": 2.8943388289233534e-06, "loss": 0.2538, "step": 4928 }, { "epoch": 0.93, "grad_norm": 10.375, "learning_rate": 2.8935824557632742e-06, "loss": 0.2518, "step": 4929 }, { "epoch": 0.93, "grad_norm": 9.375, "learning_rate": 2.8928260456587752e-06, "loss": 0.2625, "step": 4930 }, { "epoch": 0.93, "grad_norm": 17.75, "learning_rate": 2.892069598680859e-06, "loss": 0.2709, "step": 4931 }, { "epoch": 0.93, "grad_norm": 10.25, "learning_rate": 2.891313114900531e-06, "loss": 0.2305, "step": 4932 }, { "epoch": 0.93, "grad_norm": 9.1875, "learning_rate": 2.8905565943887993e-06, "loss": 0.2029, "step": 4933 }, { "epoch": 0.93, "grad_norm": 11.5625, "learning_rate": 2.889800037216677e-06, "loss": 0.2872, "step": 4934 }, { "epoch": 0.93, "grad_norm": 12.125, "learning_rate": 2.8890434434551793e-06, "loss": 0.3184, "step": 4935 }, { "epoch": 0.93, "grad_norm": 10.5625, "learning_rate": 2.888286813175326e-06, "loss": 0.2993, "step": 4936 }, { "epoch": 0.93, "grad_norm": 11.125, "learning_rate": 2.88753014644814e-06, "loss": 0.3178, "step": 4937 }, { "epoch": 0.93, "grad_norm": 8.1875, "learning_rate": 2.886773443344646e-06, "loss": 0.2451, "step": 4938 }, { "epoch": 0.93, "grad_norm": 11.125, "learning_rate": 2.8860167039358745e-06, "loss": 0.3061, "step": 4939 }, { "epoch": 0.93, "grad_norm": 9.875, "learning_rate": 2.8852599282928583e-06, "loss": 0.2648, "step": 4940 }, { "epoch": 0.93, "grad_norm": 10.75, "learning_rate": 2.884503116486633e-06, "loss": 0.2639, "step": 4941 }, { "epoch": 0.93, "grad_norm": 10.625, "learning_rate": 2.88374626858824e-06, "loss": 0.3333, "step": 4942 }, { "epoch": 0.94, "grad_norm": 11.6875, "learning_rate": 2.882989384668721e-06, "loss": 0.3079, "step": 4943 }, { "epoch": 0.94, "grad_norm": 9.875, "learning_rate": 2.882232464799123e-06, "loss": 0.2609, "step": 4944 }, { "epoch": 0.94, "grad_norm": 12.8125, "learning_rate": 2.881475509050495e-06, "loss": 0.3488, "step": 4945 }, { "epoch": 0.94, "grad_norm": 12.875, "learning_rate": 2.8807185174938905e-06, "loss": 0.364, "step": 4946 }, { "epoch": 0.94, "grad_norm": 18.0, "learning_rate": 2.879961490200367e-06, "loss": 0.2953, "step": 4947 }, { "epoch": 0.94, "grad_norm": 12.5, "learning_rate": 2.879204427240983e-06, "loss": 0.3213, "step": 4948 }, { "epoch": 0.94, "grad_norm": 11.625, "learning_rate": 2.8784473286868027e-06, "loss": 0.3205, "step": 4949 }, { "epoch": 0.94, "grad_norm": 10.25, "learning_rate": 2.8776901946088924e-06, "loss": 0.258, "step": 4950 }, { "epoch": 0.94, "grad_norm": 8.25, "learning_rate": 2.8769330250783224e-06, "loss": 0.2031, "step": 4951 }, { "epoch": 0.94, "grad_norm": 9.1875, "learning_rate": 2.876175820166165e-06, "loss": 0.2317, "step": 4952 }, { "epoch": 0.94, "grad_norm": 21.5, "learning_rate": 2.875418579943498e-06, "loss": 0.2762, "step": 4953 }, { "epoch": 0.94, "grad_norm": 8.75, "learning_rate": 2.8746613044814e-06, "loss": 0.213, "step": 4954 }, { "epoch": 0.94, "grad_norm": 8.125, "learning_rate": 2.8739039938509565e-06, "loss": 0.2499, "step": 4955 }, { "epoch": 0.94, "grad_norm": 9.4375, "learning_rate": 2.873146648123251e-06, "loss": 0.2272, "step": 4956 }, { "epoch": 0.94, "grad_norm": 12.375, "learning_rate": 2.8723892673693753e-06, "loss": 0.3865, "step": 4957 }, { "epoch": 0.94, "grad_norm": 12.5625, "learning_rate": 2.871631851660422e-06, "loss": 0.313, "step": 4958 }, { "epoch": 0.94, "grad_norm": 15.375, "learning_rate": 2.870874401067488e-06, "loss": 0.264, "step": 4959 }, { "epoch": 0.94, "grad_norm": 13.5625, "learning_rate": 2.8701169156616703e-06, "loss": 0.293, "step": 4960 }, { "epoch": 0.94, "grad_norm": 12.625, "learning_rate": 2.869359395514076e-06, "loss": 0.2805, "step": 4961 }, { "epoch": 0.94, "grad_norm": 8.8125, "learning_rate": 2.8686018406958078e-06, "loss": 0.1674, "step": 4962 }, { "epoch": 0.94, "grad_norm": 14.5625, "learning_rate": 2.867844251277977e-06, "loss": 0.2638, "step": 4963 }, { "epoch": 0.94, "grad_norm": 11.1875, "learning_rate": 2.8670866273316955e-06, "loss": 0.3042, "step": 4964 }, { "epoch": 0.94, "grad_norm": 11.5, "learning_rate": 2.866328968928079e-06, "loss": 0.2936, "step": 4965 }, { "epoch": 0.94, "grad_norm": 13.6875, "learning_rate": 2.865571276138248e-06, "loss": 0.3409, "step": 4966 }, { "epoch": 0.94, "grad_norm": 14.5625, "learning_rate": 2.864813549033324e-06, "loss": 0.2955, "step": 4967 }, { "epoch": 0.94, "grad_norm": 10.4375, "learning_rate": 2.8640557876844323e-06, "loss": 0.2347, "step": 4968 }, { "epoch": 0.94, "grad_norm": 10.0625, "learning_rate": 2.863297992162702e-06, "loss": 0.2931, "step": 4969 }, { "epoch": 0.94, "grad_norm": 13.125, "learning_rate": 2.862540162539265e-06, "loss": 0.3401, "step": 4970 }, { "epoch": 0.94, "grad_norm": 30.375, "learning_rate": 2.861782298885257e-06, "loss": 0.3032, "step": 4971 }, { "epoch": 0.94, "grad_norm": 9.625, "learning_rate": 2.861024401271815e-06, "loss": 0.2827, "step": 4972 }, { "epoch": 0.94, "grad_norm": 11.875, "learning_rate": 2.8602664697700827e-06, "loss": 0.3213, "step": 4973 }, { "epoch": 0.94, "grad_norm": 9.9375, "learning_rate": 2.8595085044512037e-06, "loss": 0.2722, "step": 4974 }, { "epoch": 0.94, "grad_norm": 10.5625, "learning_rate": 2.8587505053863258e-06, "loss": 0.2595, "step": 4975 }, { "epoch": 0.94, "grad_norm": 8.75, "learning_rate": 2.8579924726466006e-06, "loss": 0.2636, "step": 4976 }, { "epoch": 0.94, "grad_norm": 10.3125, "learning_rate": 2.857234406303182e-06, "loss": 0.2188, "step": 4977 }, { "epoch": 0.94, "grad_norm": 13.4375, "learning_rate": 2.8564763064272277e-06, "loss": 0.3715, "step": 4978 }, { "epoch": 0.94, "grad_norm": 18.0, "learning_rate": 2.8557181730898988e-06, "loss": 0.2842, "step": 4979 }, { "epoch": 0.94, "grad_norm": 12.375, "learning_rate": 2.854960006362358e-06, "loss": 0.3533, "step": 4980 }, { "epoch": 0.94, "grad_norm": 11.25, "learning_rate": 2.8542018063157723e-06, "loss": 0.2644, "step": 4981 }, { "epoch": 0.94, "grad_norm": 9.0625, "learning_rate": 2.8534435730213127e-06, "loss": 0.234, "step": 4982 }, { "epoch": 0.94, "grad_norm": 12.4375, "learning_rate": 2.8526853065501514e-06, "loss": 0.2466, "step": 4983 }, { "epoch": 0.94, "grad_norm": 13.0625, "learning_rate": 2.851927006973465e-06, "loss": 0.2871, "step": 4984 }, { "epoch": 0.94, "grad_norm": 14.5, "learning_rate": 2.8511686743624323e-06, "loss": 0.3419, "step": 4985 }, { "epoch": 0.94, "grad_norm": 10.25, "learning_rate": 2.850410308788236e-06, "loss": 0.285, "step": 4986 }, { "epoch": 0.94, "grad_norm": 12.5, "learning_rate": 2.8496519103220625e-06, "loss": 0.2942, "step": 4987 }, { "epoch": 0.94, "grad_norm": 11.5, "learning_rate": 2.8488934790351e-06, "loss": 0.2735, "step": 4988 }, { "epoch": 0.94, "grad_norm": 9.125, "learning_rate": 2.848135014998539e-06, "loss": 0.2344, "step": 4989 }, { "epoch": 0.94, "grad_norm": 12.6875, "learning_rate": 2.847376518283575e-06, "loss": 0.2852, "step": 4990 }, { "epoch": 0.94, "grad_norm": 13.4375, "learning_rate": 2.8466179889614066e-06, "loss": 0.2983, "step": 4991 }, { "epoch": 0.94, "grad_norm": 11.125, "learning_rate": 2.8458594271032343e-06, "loss": 0.3549, "step": 4992 }, { "epoch": 0.94, "grad_norm": 11.5, "learning_rate": 2.845100832780261e-06, "loss": 0.2773, "step": 4993 }, { "epoch": 0.94, "grad_norm": 13.0625, "learning_rate": 2.8443422060636956e-06, "loss": 0.2211, "step": 4994 }, { "epoch": 0.94, "grad_norm": 12.125, "learning_rate": 2.8435835470247464e-06, "loss": 0.2211, "step": 4995 }, { "epoch": 0.95, "grad_norm": 7.90625, "learning_rate": 2.842824855734627e-06, "loss": 0.2729, "step": 4996 }, { "epoch": 0.95, "grad_norm": 9.625, "learning_rate": 2.8420661322645545e-06, "loss": 0.2846, "step": 4997 }, { "epoch": 0.95, "grad_norm": 8.75, "learning_rate": 2.8413073766857463e-06, "loss": 0.188, "step": 4998 }, { "epoch": 0.95, "grad_norm": 11.9375, "learning_rate": 2.8405485890694263e-06, "loss": 0.2216, "step": 4999 }, { "epoch": 0.95, "grad_norm": 10.1875, "learning_rate": 2.839789769486818e-06, "loss": 0.257, "step": 5000 }, { "epoch": 0.95, "grad_norm": 14.5, "learning_rate": 2.8390309180091513e-06, "loss": 0.3053, "step": 5001 }, { "epoch": 0.95, "grad_norm": 13.4375, "learning_rate": 2.8382720347076554e-06, "loss": 0.3618, "step": 5002 }, { "epoch": 0.95, "grad_norm": 10.3125, "learning_rate": 2.8375131196535666e-06, "loss": 0.2903, "step": 5003 }, { "epoch": 0.95, "grad_norm": 11.0625, "learning_rate": 2.8367541729181196e-06, "loss": 0.3039, "step": 5004 }, { "epoch": 0.95, "grad_norm": 9.75, "learning_rate": 2.8359951945725568e-06, "loss": 0.2767, "step": 5005 }, { "epoch": 0.95, "grad_norm": 8.75, "learning_rate": 2.835236184688119e-06, "loss": 0.2636, "step": 5006 }, { "epoch": 0.95, "grad_norm": 8.25, "learning_rate": 2.8344771433360547e-06, "loss": 0.2732, "step": 5007 }, { "epoch": 0.95, "grad_norm": 8.5625, "learning_rate": 2.83371807058761e-06, "loss": 0.1903, "step": 5008 }, { "epoch": 0.95, "grad_norm": 12.5, "learning_rate": 2.8329589665140396e-06, "loss": 0.21, "step": 5009 }, { "epoch": 0.95, "grad_norm": 7.96875, "learning_rate": 2.832199831186596e-06, "loss": 0.1642, "step": 5010 }, { "epoch": 0.95, "grad_norm": 10.8125, "learning_rate": 2.831440664676539e-06, "loss": 0.2287, "step": 5011 }, { "epoch": 0.95, "grad_norm": 8.25, "learning_rate": 2.830681467055128e-06, "loss": 0.2194, "step": 5012 }, { "epoch": 0.95, "grad_norm": 9.1875, "learning_rate": 2.8299222383936274e-06, "loss": 0.2301, "step": 5013 }, { "epoch": 0.95, "grad_norm": 12.1875, "learning_rate": 2.829162978763303e-06, "loss": 0.2885, "step": 5014 }, { "epoch": 0.95, "grad_norm": 21.0, "learning_rate": 2.828403688235425e-06, "loss": 0.3618, "step": 5015 }, { "epoch": 0.95, "grad_norm": 8.9375, "learning_rate": 2.8276443668812654e-06, "loss": 0.2431, "step": 5016 }, { "epoch": 0.95, "grad_norm": 8.25, "learning_rate": 2.826885014772099e-06, "loss": 0.2066, "step": 5017 }, { "epoch": 0.95, "grad_norm": 18.25, "learning_rate": 2.8261256319792047e-06, "loss": 0.3583, "step": 5018 }, { "epoch": 0.95, "grad_norm": 11.375, "learning_rate": 2.8253662185738624e-06, "loss": 0.2591, "step": 5019 }, { "epoch": 0.95, "grad_norm": 8.75, "learning_rate": 2.8246067746273577e-06, "loss": 0.2012, "step": 5020 }, { "epoch": 0.95, "grad_norm": 9.625, "learning_rate": 2.8238473002109756e-06, "loss": 0.2283, "step": 5021 }, { "epoch": 0.95, "grad_norm": 8.5625, "learning_rate": 2.8230877953960074e-06, "loss": 0.2584, "step": 5022 }, { "epoch": 0.95, "grad_norm": 9.4375, "learning_rate": 2.822328260253744e-06, "loss": 0.1667, "step": 5023 }, { "epoch": 0.95, "grad_norm": 10.5, "learning_rate": 2.8215686948554816e-06, "loss": 0.2978, "step": 5024 }, { "epoch": 0.95, "grad_norm": 8.6875, "learning_rate": 2.8208090992725183e-06, "loss": 0.2351, "step": 5025 }, { "epoch": 0.95, "grad_norm": 10.375, "learning_rate": 2.820049473576155e-06, "loss": 0.3038, "step": 5026 }, { "epoch": 0.95, "grad_norm": 10.875, "learning_rate": 2.8192898178376953e-06, "loss": 0.2914, "step": 5027 }, { "epoch": 0.95, "grad_norm": 14.25, "learning_rate": 2.8185301321284464e-06, "loss": 0.3326, "step": 5028 }, { "epoch": 0.95, "grad_norm": 8.0625, "learning_rate": 2.8177704165197167e-06, "loss": 0.2734, "step": 5029 }, { "epoch": 0.95, "grad_norm": 12.4375, "learning_rate": 2.8170106710828192e-06, "loss": 0.2162, "step": 5030 }, { "epoch": 0.95, "grad_norm": 10.9375, "learning_rate": 2.8162508958890694e-06, "loss": 0.3033, "step": 5031 }, { "epoch": 0.95, "grad_norm": 9.875, "learning_rate": 2.8154910910097845e-06, "loss": 0.2661, "step": 5032 }, { "epoch": 0.95, "grad_norm": 12.875, "learning_rate": 2.814731256516285e-06, "loss": 0.2461, "step": 5033 }, { "epoch": 0.95, "grad_norm": 12.3125, "learning_rate": 2.8139713924798938e-06, "loss": 0.2597, "step": 5034 }, { "epoch": 0.95, "grad_norm": 8.875, "learning_rate": 2.813211498971939e-06, "loss": 0.2738, "step": 5035 }, { "epoch": 0.95, "grad_norm": 9.8125, "learning_rate": 2.8124515760637476e-06, "loss": 0.3152, "step": 5036 }, { "epoch": 0.95, "grad_norm": 10.625, "learning_rate": 2.811691623826653e-06, "loss": 0.2958, "step": 5037 }, { "epoch": 0.95, "grad_norm": 12.375, "learning_rate": 2.810931642331988e-06, "loss": 0.2912, "step": 5038 }, { "epoch": 0.95, "grad_norm": 9.875, "learning_rate": 2.810171631651091e-06, "loss": 0.2208, "step": 5039 }, { "epoch": 0.95, "grad_norm": 8.875, "learning_rate": 2.8094115918553012e-06, "loss": 0.2342, "step": 5040 }, { "epoch": 0.95, "grad_norm": 10.9375, "learning_rate": 2.808651523015962e-06, "loss": 0.2519, "step": 5041 }, { "epoch": 0.95, "grad_norm": 12.875, "learning_rate": 2.807891425204418e-06, "loss": 0.2797, "step": 5042 }, { "epoch": 0.95, "grad_norm": 13.6875, "learning_rate": 2.8071312984920177e-06, "loss": 0.3292, "step": 5043 }, { "epoch": 0.95, "grad_norm": 12.25, "learning_rate": 2.8063711429501123e-06, "loss": 0.2957, "step": 5044 }, { "epoch": 0.95, "grad_norm": 9.125, "learning_rate": 2.805610958650055e-06, "loss": 0.273, "step": 5045 }, { "epoch": 0.95, "grad_norm": 12.25, "learning_rate": 2.8048507456632018e-06, "loss": 0.3421, "step": 5046 }, { "epoch": 0.95, "grad_norm": 11.25, "learning_rate": 2.804090504060913e-06, "loss": 0.263, "step": 5047 }, { "epoch": 0.95, "grad_norm": 7.53125, "learning_rate": 2.803330233914548e-06, "loss": 0.2599, "step": 5048 }, { "epoch": 0.96, "grad_norm": 9.875, "learning_rate": 2.8025699352954726e-06, "loss": 0.2579, "step": 5049 }, { "epoch": 0.96, "grad_norm": 12.5625, "learning_rate": 2.8018096082750533e-06, "loss": 0.2466, "step": 5050 }, { "epoch": 0.96, "grad_norm": 22.625, "learning_rate": 2.801049252924661e-06, "loss": 0.3414, "step": 5051 }, { "epoch": 0.96, "grad_norm": 7.5, "learning_rate": 2.8002888693156653e-06, "loss": 0.2184, "step": 5052 }, { "epoch": 0.96, "grad_norm": 10.625, "learning_rate": 2.7995284575194436e-06, "loss": 0.3341, "step": 5053 }, { "epoch": 0.96, "grad_norm": 7.8125, "learning_rate": 2.7987680176073727e-06, "loss": 0.2615, "step": 5054 }, { "epoch": 0.96, "grad_norm": 9.3125, "learning_rate": 2.7980075496508325e-06, "loss": 0.2416, "step": 5055 }, { "epoch": 0.96, "grad_norm": 13.25, "learning_rate": 2.7972470537212066e-06, "loss": 0.3323, "step": 5056 }, { "epoch": 0.96, "grad_norm": 12.125, "learning_rate": 2.7964865298898797e-06, "loss": 0.3188, "step": 5057 }, { "epoch": 0.96, "grad_norm": 17.0, "learning_rate": 2.795725978228241e-06, "loss": 0.2833, "step": 5058 }, { "epoch": 0.96, "grad_norm": 14.8125, "learning_rate": 2.7949653988076798e-06, "loss": 0.2516, "step": 5059 }, { "epoch": 0.96, "grad_norm": 12.0625, "learning_rate": 2.7942047916995908e-06, "loss": 0.26, "step": 5060 }, { "epoch": 0.96, "grad_norm": 15.875, "learning_rate": 2.7934441569753685e-06, "loss": 0.2712, "step": 5061 }, { "epoch": 0.96, "grad_norm": 11.6875, "learning_rate": 2.7926834947064128e-06, "loss": 0.2637, "step": 5062 }, { "epoch": 0.96, "grad_norm": 15.0625, "learning_rate": 2.791922804964124e-06, "loss": 0.2097, "step": 5063 }, { "epoch": 0.96, "grad_norm": 10.5625, "learning_rate": 2.7911620878199057e-06, "loss": 0.2308, "step": 5064 }, { "epoch": 0.96, "grad_norm": 13.0, "learning_rate": 2.7904013433451654e-06, "loss": 0.2878, "step": 5065 }, { "epoch": 0.96, "grad_norm": 14.375, "learning_rate": 2.7896405716113107e-06, "loss": 0.2623, "step": 5066 }, { "epoch": 0.96, "grad_norm": 11.0625, "learning_rate": 2.7888797726897533e-06, "loss": 0.2637, "step": 5067 }, { "epoch": 0.96, "grad_norm": 12.3125, "learning_rate": 2.788118946651907e-06, "loss": 0.2278, "step": 5068 }, { "epoch": 0.96, "grad_norm": 13.375, "learning_rate": 2.7873580935691885e-06, "loss": 0.2759, "step": 5069 }, { "epoch": 0.96, "grad_norm": 9.75, "learning_rate": 2.786597213513017e-06, "loss": 0.2515, "step": 5070 }, { "epoch": 0.96, "grad_norm": 14.1875, "learning_rate": 2.785836306554813e-06, "loss": 0.3256, "step": 5071 }, { "epoch": 0.96, "grad_norm": 22.125, "learning_rate": 2.785075372766003e-06, "loss": 0.2942, "step": 5072 }, { "epoch": 0.96, "grad_norm": 8.9375, "learning_rate": 2.784314412218011e-06, "loss": 0.2096, "step": 5073 }, { "epoch": 0.96, "grad_norm": 11.25, "learning_rate": 2.783553424982267e-06, "loss": 0.2335, "step": 5074 }, { "epoch": 0.96, "grad_norm": 8.0625, "learning_rate": 2.7827924111302033e-06, "loss": 0.2248, "step": 5075 }, { "epoch": 0.96, "grad_norm": 8.5, "learning_rate": 2.782031370733253e-06, "loss": 0.2491, "step": 5076 }, { "epoch": 0.96, "grad_norm": 10.3125, "learning_rate": 2.7812703038628537e-06, "loss": 0.3085, "step": 5077 }, { "epoch": 0.96, "grad_norm": 12.6875, "learning_rate": 2.7805092105904434e-06, "loss": 0.2847, "step": 5078 }, { "epoch": 0.96, "grad_norm": 14.875, "learning_rate": 2.7797480909874647e-06, "loss": 0.2514, "step": 5079 }, { "epoch": 0.96, "grad_norm": 15.5625, "learning_rate": 2.7789869451253617e-06, "loss": 0.3012, "step": 5080 }, { "epoch": 0.96, "grad_norm": 10.75, "learning_rate": 2.778225773075579e-06, "loss": 0.2539, "step": 5081 }, { "epoch": 0.96, "grad_norm": 11.375, "learning_rate": 2.777464574909568e-06, "loss": 0.2493, "step": 5082 }, { "epoch": 0.96, "grad_norm": 11.875, "learning_rate": 2.776703350698779e-06, "loss": 0.2394, "step": 5083 }, { "epoch": 0.96, "grad_norm": 15.875, "learning_rate": 2.775942100514666e-06, "loss": 0.2872, "step": 5084 }, { "epoch": 0.96, "grad_norm": 10.375, "learning_rate": 2.7751808244286847e-06, "loss": 0.253, "step": 5085 }, { "epoch": 0.96, "grad_norm": 13.3125, "learning_rate": 2.774419522512294e-06, "loss": 0.2714, "step": 5086 }, { "epoch": 0.96, "grad_norm": 11.5, "learning_rate": 2.7736581948369566e-06, "loss": 0.2705, "step": 5087 }, { "epoch": 0.96, "grad_norm": 8.9375, "learning_rate": 2.7728968414741344e-06, "loss": 0.222, "step": 5088 }, { "epoch": 0.96, "grad_norm": 12.875, "learning_rate": 2.772135462495294e-06, "loss": 0.2428, "step": 5089 }, { "epoch": 0.96, "grad_norm": 13.625, "learning_rate": 2.7713740579719035e-06, "loss": 0.3273, "step": 5090 }, { "epoch": 0.96, "grad_norm": 15.5, "learning_rate": 2.770612627975434e-06, "loss": 0.3409, "step": 5091 }, { "epoch": 0.96, "grad_norm": 11.0, "learning_rate": 2.769851172577358e-06, "loss": 0.2976, "step": 5092 }, { "epoch": 0.96, "grad_norm": 11.5, "learning_rate": 2.769089691849153e-06, "loss": 0.2784, "step": 5093 }, { "epoch": 0.96, "grad_norm": 8.625, "learning_rate": 2.768328185862294e-06, "loss": 0.2031, "step": 5094 }, { "epoch": 0.96, "grad_norm": 10.1875, "learning_rate": 2.767566654688264e-06, "loss": 0.2724, "step": 5095 }, { "epoch": 0.96, "grad_norm": 10.25, "learning_rate": 2.766805098398544e-06, "loss": 0.3502, "step": 5096 }, { "epoch": 0.96, "grad_norm": 10.875, "learning_rate": 2.7660435170646198e-06, "loss": 0.3143, "step": 5097 }, { "epoch": 0.96, "grad_norm": 12.875, "learning_rate": 2.7652819107579783e-06, "loss": 0.2734, "step": 5098 }, { "epoch": 0.96, "grad_norm": 9.375, "learning_rate": 2.7645202795501097e-06, "loss": 0.2705, "step": 5099 }, { "epoch": 0.96, "grad_norm": 18.375, "learning_rate": 2.763758623512506e-06, "loss": 0.1683, "step": 5100 }, { "epoch": 0.96, "grad_norm": 10.3125, "learning_rate": 2.762996942716661e-06, "loss": 0.2341, "step": 5101 }, { "epoch": 0.97, "grad_norm": 10.4375, "learning_rate": 2.7622352372340727e-06, "loss": 0.3248, "step": 5102 }, { "epoch": 0.97, "grad_norm": 10.125, "learning_rate": 2.761473507136238e-06, "loss": 0.2865, "step": 5103 }, { "epoch": 0.97, "grad_norm": 8.4375, "learning_rate": 2.760711752494661e-06, "loss": 0.1965, "step": 5104 }, { "epoch": 0.97, "grad_norm": 8.8125, "learning_rate": 2.759949973380843e-06, "loss": 0.2727, "step": 5105 }, { "epoch": 0.97, "grad_norm": 10.25, "learning_rate": 2.759188169866291e-06, "loss": 0.2795, "step": 5106 }, { "epoch": 0.97, "grad_norm": 12.75, "learning_rate": 2.7584263420225126e-06, "loss": 0.255, "step": 5107 }, { "epoch": 0.97, "grad_norm": 11.125, "learning_rate": 2.7576644899210197e-06, "loss": 0.4128, "step": 5108 }, { "epoch": 0.97, "grad_norm": 11.9375, "learning_rate": 2.756902613633324e-06, "loss": 0.2283, "step": 5109 }, { "epoch": 0.97, "grad_norm": 13.0625, "learning_rate": 2.7561407132309405e-06, "loss": 0.257, "step": 5110 }, { "epoch": 0.97, "grad_norm": 13.0, "learning_rate": 2.7553787887853868e-06, "loss": 0.3433, "step": 5111 }, { "epoch": 0.97, "grad_norm": 12.375, "learning_rate": 2.7546168403681827e-06, "loss": 0.2924, "step": 5112 }, { "epoch": 0.97, "grad_norm": 12.9375, "learning_rate": 2.7538548680508497e-06, "loss": 0.3436, "step": 5113 }, { "epoch": 0.97, "grad_norm": 14.1875, "learning_rate": 2.753092871904913e-06, "loss": 0.3397, "step": 5114 }, { "epoch": 0.97, "grad_norm": 10.875, "learning_rate": 2.7523308520018965e-06, "loss": 0.3238, "step": 5115 }, { "epoch": 0.97, "grad_norm": 12.3125, "learning_rate": 2.7515688084133313e-06, "loss": 0.2734, "step": 5116 }, { "epoch": 0.97, "grad_norm": 8.4375, "learning_rate": 2.7508067412107462e-06, "loss": 0.2744, "step": 5117 }, { "epoch": 0.97, "grad_norm": 9.4375, "learning_rate": 2.7500446504656757e-06, "loss": 0.2007, "step": 5118 }, { "epoch": 0.97, "grad_norm": 12.5, "learning_rate": 2.7492825362496543e-06, "loss": 0.2515, "step": 5119 }, { "epoch": 0.97, "grad_norm": 16.5, "learning_rate": 2.7485203986342194e-06, "loss": 0.2678, "step": 5120 }, { "epoch": 0.97, "grad_norm": 13.25, "learning_rate": 2.747758237690911e-06, "loss": 0.2402, "step": 5121 }, { "epoch": 0.97, "grad_norm": 13.25, "learning_rate": 2.7469960534912703e-06, "loss": 0.2232, "step": 5122 }, { "epoch": 0.97, "grad_norm": 13.1875, "learning_rate": 2.7462338461068416e-06, "loss": 0.2517, "step": 5123 }, { "epoch": 0.97, "grad_norm": 10.0625, "learning_rate": 2.7454716156091716e-06, "loss": 0.2116, "step": 5124 }, { "epoch": 0.97, "grad_norm": 12.8125, "learning_rate": 2.7447093620698074e-06, "loss": 0.2846, "step": 5125 }, { "epoch": 0.97, "grad_norm": 8.5, "learning_rate": 2.7439470855603e-06, "loss": 0.2441, "step": 5126 }, { "epoch": 0.97, "grad_norm": 10.5, "learning_rate": 2.743184786152202e-06, "loss": 0.2874, "step": 5127 }, { "epoch": 0.97, "grad_norm": 11.0, "learning_rate": 2.7424224639170687e-06, "loss": 0.2516, "step": 5128 }, { "epoch": 0.97, "grad_norm": 10.6875, "learning_rate": 2.741660118926457e-06, "loss": 0.2934, "step": 5129 }, { "epoch": 0.97, "grad_norm": 10.8125, "learning_rate": 2.740897751251925e-06, "loss": 0.256, "step": 5130 }, { "epoch": 0.97, "grad_norm": 22.875, "learning_rate": 2.740135360965035e-06, "loss": 0.3547, "step": 5131 }, { "epoch": 0.97, "grad_norm": 16.75, "learning_rate": 2.7393729481373486e-06, "loss": 0.2264, "step": 5132 }, { "epoch": 0.97, "grad_norm": 10.125, "learning_rate": 2.738610512840434e-06, "loss": 0.2007, "step": 5133 }, { "epoch": 0.97, "grad_norm": 8.6875, "learning_rate": 2.7378480551458563e-06, "loss": 0.2945, "step": 5134 }, { "epoch": 0.97, "grad_norm": 17.125, "learning_rate": 2.7370855751251867e-06, "loss": 0.2987, "step": 5135 }, { "epoch": 0.97, "grad_norm": 14.125, "learning_rate": 2.7363230728499957e-06, "loss": 0.2827, "step": 5136 }, { "epoch": 0.97, "grad_norm": 9.625, "learning_rate": 2.735560548391859e-06, "loss": 0.2597, "step": 5137 }, { "epoch": 0.97, "grad_norm": 10.3125, "learning_rate": 2.7347980018223495e-06, "loss": 0.2706, "step": 5138 }, { "epoch": 0.97, "grad_norm": 16.625, "learning_rate": 2.734035433213049e-06, "loss": 0.2804, "step": 5139 }, { "epoch": 0.97, "grad_norm": 10.3125, "learning_rate": 2.733272842635533e-06, "loss": 0.3196, "step": 5140 }, { "epoch": 0.97, "grad_norm": 15.4375, "learning_rate": 2.7325102301613883e-06, "loss": 0.2515, "step": 5141 }, { "epoch": 0.97, "grad_norm": 10.8125, "learning_rate": 2.7317475958621953e-06, "loss": 0.2848, "step": 5142 }, { "epoch": 0.97, "grad_norm": 9.1875, "learning_rate": 2.7309849398095423e-06, "loss": 0.1947, "step": 5143 }, { "epoch": 0.97, "grad_norm": 13.125, "learning_rate": 2.730222262075018e-06, "loss": 0.2614, "step": 5144 }, { "epoch": 0.97, "grad_norm": 13.1875, "learning_rate": 2.7294595627302105e-06, "loss": 0.2924, "step": 5145 }, { "epoch": 0.97, "grad_norm": 13.5, "learning_rate": 2.7286968418467146e-06, "loss": 0.2934, "step": 5146 }, { "epoch": 0.97, "grad_norm": 10.625, "learning_rate": 2.7279340994961223e-06, "loss": 0.2475, "step": 5147 }, { "epoch": 0.97, "grad_norm": 10.5, "learning_rate": 2.7271713357500322e-06, "loss": 0.2503, "step": 5148 }, { "epoch": 0.97, "grad_norm": 12.1875, "learning_rate": 2.726408550680041e-06, "loss": 0.2965, "step": 5149 }, { "epoch": 0.97, "grad_norm": 13.1875, "learning_rate": 2.72564574435775e-06, "loss": 0.3523, "step": 5150 }, { "epoch": 0.97, "grad_norm": 14.125, "learning_rate": 2.7248829168547602e-06, "loss": 0.2932, "step": 5151 }, { "epoch": 0.97, "grad_norm": 9.625, "learning_rate": 2.724120068242678e-06, "loss": 0.2314, "step": 5152 }, { "epoch": 0.97, "grad_norm": 11.125, "learning_rate": 2.7233571985931086e-06, "loss": 0.2808, "step": 5153 }, { "epoch": 0.97, "grad_norm": 15.3125, "learning_rate": 2.72259430797766e-06, "loss": 0.3012, "step": 5154 }, { "epoch": 0.98, "grad_norm": 9.8125, "learning_rate": 2.7218313964679432e-06, "loss": 0.2711, "step": 5155 }, { "epoch": 0.98, "grad_norm": 15.8125, "learning_rate": 2.7210684641355706e-06, "loss": 0.2667, "step": 5156 }, { "epoch": 0.98, "grad_norm": 11.1875, "learning_rate": 2.720305511052156e-06, "loss": 0.3199, "step": 5157 }, { "epoch": 0.98, "grad_norm": 7.84375, "learning_rate": 2.7195425372893147e-06, "loss": 0.2492, "step": 5158 }, { "epoch": 0.98, "grad_norm": 11.6875, "learning_rate": 2.7187795429186663e-06, "loss": 0.2805, "step": 5159 }, { "epoch": 0.98, "grad_norm": 14.1875, "learning_rate": 2.71801652801183e-06, "loss": 0.2192, "step": 5160 }, { "epoch": 0.98, "grad_norm": 9.8125, "learning_rate": 2.7172534926404277e-06, "loss": 0.2645, "step": 5161 }, { "epoch": 0.98, "grad_norm": 11.625, "learning_rate": 2.7164904368760834e-06, "loss": 0.3133, "step": 5162 }, { "epoch": 0.98, "grad_norm": 12.125, "learning_rate": 2.7157273607904234e-06, "loss": 0.3153, "step": 5163 }, { "epoch": 0.98, "grad_norm": 15.5625, "learning_rate": 2.714964264455075e-06, "loss": 0.2956, "step": 5164 }, { "epoch": 0.98, "grad_norm": 15.625, "learning_rate": 2.7142011479416675e-06, "loss": 0.2834, "step": 5165 }, { "epoch": 0.98, "grad_norm": 16.5, "learning_rate": 2.7134380113218324e-06, "loss": 0.3286, "step": 5166 }, { "epoch": 0.98, "grad_norm": 8.0, "learning_rate": 2.712674854667203e-06, "loss": 0.2661, "step": 5167 }, { "epoch": 0.98, "grad_norm": 12.9375, "learning_rate": 2.7119116780494155e-06, "loss": 0.2989, "step": 5168 }, { "epoch": 0.98, "grad_norm": 10.5, "learning_rate": 2.711148481540106e-06, "loss": 0.3188, "step": 5169 }, { "epoch": 0.98, "grad_norm": 7.6875, "learning_rate": 2.7103852652109143e-06, "loss": 0.1915, "step": 5170 }, { "epoch": 0.98, "grad_norm": 9.9375, "learning_rate": 2.70962202913348e-06, "loss": 0.1829, "step": 5171 }, { "epoch": 0.98, "grad_norm": 11.25, "learning_rate": 2.7088587733794473e-06, "loss": 0.2542, "step": 5172 }, { "epoch": 0.98, "grad_norm": 10.875, "learning_rate": 2.70809549802046e-06, "loss": 0.2538, "step": 5173 }, { "epoch": 0.98, "grad_norm": 12.875, "learning_rate": 2.7073322031281645e-06, "loss": 0.2715, "step": 5174 }, { "epoch": 0.98, "grad_norm": 10.25, "learning_rate": 2.706568888774209e-06, "loss": 0.2649, "step": 5175 }, { "epoch": 0.98, "grad_norm": 14.5625, "learning_rate": 2.7058055550302437e-06, "loss": 0.2874, "step": 5176 }, { "epoch": 0.98, "grad_norm": 12.0625, "learning_rate": 2.705042201967921e-06, "loss": 0.3844, "step": 5177 }, { "epoch": 0.98, "grad_norm": 11.125, "learning_rate": 2.7042788296588932e-06, "loss": 0.286, "step": 5178 }, { "epoch": 0.98, "grad_norm": 16.0, "learning_rate": 2.7035154381748174e-06, "loss": 0.4324, "step": 5179 }, { "epoch": 0.98, "grad_norm": 14.8125, "learning_rate": 2.70275202758735e-06, "loss": 0.2906, "step": 5180 }, { "epoch": 0.98, "grad_norm": 8.375, "learning_rate": 2.7019885979681504e-06, "loss": 0.2327, "step": 5181 }, { "epoch": 0.98, "grad_norm": 10.3125, "learning_rate": 2.701225149388878e-06, "loss": 0.2337, "step": 5182 }, { "epoch": 0.98, "grad_norm": 10.5625, "learning_rate": 2.7004616819211988e-06, "loss": 0.2765, "step": 5183 }, { "epoch": 0.98, "grad_norm": 13.8125, "learning_rate": 2.6996981956367736e-06, "loss": 0.3262, "step": 5184 }, { "epoch": 0.98, "grad_norm": 11.6875, "learning_rate": 2.698934690607271e-06, "loss": 0.2647, "step": 5185 }, { "epoch": 0.98, "grad_norm": 12.75, "learning_rate": 2.698171166904358e-06, "loss": 0.3079, "step": 5186 }, { "epoch": 0.98, "grad_norm": 9.25, "learning_rate": 2.697407624599704e-06, "loss": 0.2681, "step": 5187 }, { "epoch": 0.98, "grad_norm": 9.75, "learning_rate": 2.6966440637649817e-06, "loss": 0.2335, "step": 5188 }, { "epoch": 0.98, "grad_norm": 15.0625, "learning_rate": 2.695880484471863e-06, "loss": 0.4249, "step": 5189 }, { "epoch": 0.98, "grad_norm": 13.25, "learning_rate": 2.6951168867920235e-06, "loss": 0.3121, "step": 5190 }, { "epoch": 0.98, "grad_norm": 12.5, "learning_rate": 2.6943532707971394e-06, "loss": 0.2682, "step": 5191 }, { "epoch": 0.98, "grad_norm": 12.6875, "learning_rate": 2.6935896365588897e-06, "loss": 0.2339, "step": 5192 }, { "epoch": 0.98, "grad_norm": 9.75, "learning_rate": 2.6928259841489534e-06, "loss": 0.3058, "step": 5193 }, { "epoch": 0.98, "grad_norm": 11.1875, "learning_rate": 2.6920623136390144e-06, "loss": 0.2954, "step": 5194 }, { "epoch": 0.98, "grad_norm": 9.1875, "learning_rate": 2.6912986251007533e-06, "loss": 0.1992, "step": 5195 }, { "epoch": 0.98, "grad_norm": 10.75, "learning_rate": 2.6905349186058583e-06, "loss": 0.2504, "step": 5196 }, { "epoch": 0.98, "grad_norm": 9.5625, "learning_rate": 2.6897711942260133e-06, "loss": 0.3046, "step": 5197 }, { "epoch": 0.98, "grad_norm": 8.875, "learning_rate": 2.68900745203291e-06, "loss": 0.2417, "step": 5198 }, { "epoch": 0.98, "grad_norm": 10.5625, "learning_rate": 2.688243692098235e-06, "loss": 0.3117, "step": 5199 }, { "epoch": 0.98, "grad_norm": 12.125, "learning_rate": 2.6874799144936837e-06, "loss": 0.2544, "step": 5200 }, { "epoch": 0.98, "grad_norm": 9.6875, "learning_rate": 2.6867161192909475e-06, "loss": 0.272, "step": 5201 }, { "epoch": 0.98, "grad_norm": 10.625, "learning_rate": 2.685952306561722e-06, "loss": 0.2574, "step": 5202 }, { "epoch": 0.98, "grad_norm": 8.0625, "learning_rate": 2.685188476377704e-06, "loss": 0.2512, "step": 5203 }, { "epoch": 0.98, "grad_norm": 12.8125, "learning_rate": 2.684424628810593e-06, "loss": 0.2868, "step": 5204 }, { "epoch": 0.98, "grad_norm": 7.875, "learning_rate": 2.6836607639320877e-06, "loss": 0.2145, "step": 5205 }, { "epoch": 0.98, "grad_norm": 10.4375, "learning_rate": 2.6828968818138904e-06, "loss": 0.2471, "step": 5206 }, { "epoch": 0.99, "grad_norm": 9.5, "learning_rate": 2.6821329825277046e-06, "loss": 0.2427, "step": 5207 }, { "epoch": 0.99, "grad_norm": 10.6875, "learning_rate": 2.6813690661452353e-06, "loss": 0.2788, "step": 5208 }, { "epoch": 0.99, "grad_norm": 12.8125, "learning_rate": 2.680605132738189e-06, "loss": 0.2989, "step": 5209 }, { "epoch": 0.99, "grad_norm": 13.25, "learning_rate": 2.6798411823782733e-06, "loss": 0.2963, "step": 5210 }, { "epoch": 0.99, "grad_norm": 15.0625, "learning_rate": 2.6790772151371985e-06, "loss": 0.3145, "step": 5211 }, { "epoch": 0.99, "grad_norm": 13.375, "learning_rate": 2.678313231086676e-06, "loss": 0.2563, "step": 5212 }, { "epoch": 0.99, "grad_norm": 9.9375, "learning_rate": 2.6775492302984184e-06, "loss": 0.2214, "step": 5213 }, { "epoch": 0.99, "grad_norm": 16.75, "learning_rate": 2.676785212844141e-06, "loss": 0.2779, "step": 5214 }, { "epoch": 0.99, "grad_norm": 7.53125, "learning_rate": 2.6760211787955586e-06, "loss": 0.2279, "step": 5215 }, { "epoch": 0.99, "grad_norm": 12.375, "learning_rate": 2.6752571282243896e-06, "loss": 0.3067, "step": 5216 }, { "epoch": 0.99, "grad_norm": 12.8125, "learning_rate": 2.674493061202353e-06, "loss": 0.245, "step": 5217 }, { "epoch": 0.99, "grad_norm": 13.875, "learning_rate": 2.6737289778011694e-06, "loss": 0.3373, "step": 5218 }, { "epoch": 0.99, "grad_norm": 10.1875, "learning_rate": 2.6729648780925616e-06, "loss": 0.2779, "step": 5219 }, { "epoch": 0.99, "grad_norm": 10.875, "learning_rate": 2.672200762148253e-06, "loss": 0.2697, "step": 5220 }, { "epoch": 0.99, "grad_norm": 17.5, "learning_rate": 2.671436630039968e-06, "loss": 0.3028, "step": 5221 }, { "epoch": 0.99, "grad_norm": 13.875, "learning_rate": 2.6706724818394343e-06, "loss": 0.2653, "step": 5222 }, { "epoch": 0.99, "grad_norm": 9.875, "learning_rate": 2.669908317618381e-06, "loss": 0.2698, "step": 5223 }, { "epoch": 0.99, "grad_norm": 13.625, "learning_rate": 2.669144137448536e-06, "loss": 0.3368, "step": 5224 }, { "epoch": 0.99, "grad_norm": 15.125, "learning_rate": 2.668379941401633e-06, "loss": 0.2517, "step": 5225 }, { "epoch": 0.99, "grad_norm": 12.625, "learning_rate": 2.667615729549403e-06, "loss": 0.2508, "step": 5226 }, { "epoch": 0.99, "grad_norm": 12.5, "learning_rate": 2.666851501963581e-06, "loss": 0.3326, "step": 5227 }, { "epoch": 0.99, "grad_norm": 10.1875, "learning_rate": 2.6660872587159033e-06, "loss": 0.2984, "step": 5228 }, { "epoch": 0.99, "grad_norm": 10.375, "learning_rate": 2.6653229998781053e-06, "loss": 0.2846, "step": 5229 }, { "epoch": 0.99, "grad_norm": 8.875, "learning_rate": 2.6645587255219273e-06, "loss": 0.2255, "step": 5230 }, { "epoch": 0.99, "grad_norm": 11.3125, "learning_rate": 2.6637944357191094e-06, "loss": 0.2665, "step": 5231 }, { "epoch": 0.99, "grad_norm": 34.0, "learning_rate": 2.663030130541393e-06, "loss": 0.3855, "step": 5232 }, { "epoch": 0.99, "grad_norm": 11.75, "learning_rate": 2.66226581006052e-06, "loss": 0.2402, "step": 5233 }, { "epoch": 0.99, "grad_norm": 15.5, "learning_rate": 2.6615014743482376e-06, "loss": 0.1852, "step": 5234 }, { "epoch": 0.99, "grad_norm": 8.8125, "learning_rate": 2.6607371234762886e-06, "loss": 0.2677, "step": 5235 }, { "epoch": 0.99, "grad_norm": 10.5, "learning_rate": 2.659972757516423e-06, "loss": 0.2504, "step": 5236 }, { "epoch": 0.99, "grad_norm": 11.8125, "learning_rate": 2.6592083765403877e-06, "loss": 0.2966, "step": 5237 }, { "epoch": 0.99, "grad_norm": 10.9375, "learning_rate": 2.6584439806199342e-06, "loss": 0.3186, "step": 5238 }, { "epoch": 0.99, "grad_norm": 8.4375, "learning_rate": 2.6576795698268125e-06, "loss": 0.2107, "step": 5239 }, { "epoch": 0.99, "grad_norm": 14.8125, "learning_rate": 2.656915144232778e-06, "loss": 0.2841, "step": 5240 }, { "epoch": 0.99, "grad_norm": 10.4375, "learning_rate": 2.6561507039095818e-06, "loss": 0.2505, "step": 5241 }, { "epoch": 0.99, "grad_norm": 10.5625, "learning_rate": 2.6553862489289835e-06, "loss": 0.2972, "step": 5242 }, { "epoch": 0.99, "grad_norm": 10.3125, "learning_rate": 2.6546217793627362e-06, "loss": 0.2607, "step": 5243 }, { "epoch": 0.99, "grad_norm": 11.375, "learning_rate": 2.6538572952826014e-06, "loss": 0.2477, "step": 5244 }, { "epoch": 0.99, "grad_norm": 11.625, "learning_rate": 2.653092796760338e-06, "loss": 0.2949, "step": 5245 }, { "epoch": 0.99, "grad_norm": 10.8125, "learning_rate": 2.652328283867707e-06, "loss": 0.3065, "step": 5246 }, { "epoch": 0.99, "grad_norm": 11.5625, "learning_rate": 2.6515637566764714e-06, "loss": 0.2424, "step": 5247 }, { "epoch": 0.99, "grad_norm": 9.25, "learning_rate": 2.6507992152583945e-06, "loss": 0.2757, "step": 5248 }, { "epoch": 0.99, "grad_norm": 8.6875, "learning_rate": 2.650034659685242e-06, "loss": 0.2618, "step": 5249 }, { "epoch": 0.99, "grad_norm": 10.25, "learning_rate": 2.6492700900287804e-06, "loss": 0.2463, "step": 5250 }, { "epoch": 0.99, "grad_norm": 11.4375, "learning_rate": 2.648505506360777e-06, "loss": 0.2938, "step": 5251 }, { "epoch": 0.99, "grad_norm": 9.0, "learning_rate": 2.647740908753003e-06, "loss": 0.229, "step": 5252 }, { "epoch": 0.99, "grad_norm": 9.75, "learning_rate": 2.646976297277226e-06, "loss": 0.2595, "step": 5253 }, { "epoch": 0.99, "grad_norm": 8.5625, "learning_rate": 2.6462116720052196e-06, "loss": 0.2553, "step": 5254 }, { "epoch": 0.99, "grad_norm": 11.8125, "learning_rate": 2.6454470330087566e-06, "loss": 0.2793, "step": 5255 }, { "epoch": 0.99, "grad_norm": 11.0, "learning_rate": 2.6446823803596115e-06, "loss": 0.2492, "step": 5256 }, { "epoch": 0.99, "grad_norm": 12.375, "learning_rate": 2.6439177141295595e-06, "loss": 0.2253, "step": 5257 }, { "epoch": 0.99, "grad_norm": 9.75, "learning_rate": 2.6431530343903785e-06, "loss": 0.2614, "step": 5258 }, { "epoch": 0.99, "grad_norm": 10.9375, "learning_rate": 2.6423883412138456e-06, "loss": 0.2725, "step": 5259 }, { "epoch": 1.0, "grad_norm": 12.125, "learning_rate": 2.641623634671741e-06, "loss": 0.2157, "step": 5260 }, { "epoch": 1.0, "grad_norm": 15.0625, "learning_rate": 2.6408589148358456e-06, "loss": 0.2183, "step": 5261 }, { "epoch": 1.0, "grad_norm": 7.625, "learning_rate": 2.6400941817779404e-06, "loss": 0.2165, "step": 5262 }, { "epoch": 1.0, "grad_norm": 10.625, "learning_rate": 2.6393294355698097e-06, "loss": 0.2993, "step": 5263 }, { "epoch": 1.0, "grad_norm": 14.8125, "learning_rate": 2.6385646762832374e-06, "loss": 0.3825, "step": 5264 }, { "epoch": 1.0, "grad_norm": 13.25, "learning_rate": 2.6377999039900093e-06, "loss": 0.2452, "step": 5265 }, { "epoch": 1.0, "grad_norm": 9.125, "learning_rate": 2.6370351187619113e-06, "loss": 0.2014, "step": 5266 }, { "epoch": 1.0, "grad_norm": 13.375, "learning_rate": 2.636270320670734e-06, "loss": 0.2738, "step": 5267 }, { "epoch": 1.0, "grad_norm": 12.625, "learning_rate": 2.635505509788264e-06, "loss": 0.2481, "step": 5268 }, { "epoch": 1.0, "grad_norm": 13.5, "learning_rate": 2.634740686186295e-06, "loss": 0.2406, "step": 5269 }, { "epoch": 1.0, "grad_norm": 12.0, "learning_rate": 2.6339758499366152e-06, "loss": 0.2915, "step": 5270 }, { "epoch": 1.0, "grad_norm": 10.1875, "learning_rate": 2.6332110011110202e-06, "loss": 0.2736, "step": 5271 }, { "epoch": 1.0, "grad_norm": 9.625, "learning_rate": 2.632446139781302e-06, "loss": 0.2591, "step": 5272 }, { "epoch": 1.0, "grad_norm": 14.6875, "learning_rate": 2.6316812660192575e-06, "loss": 0.2647, "step": 5273 }, { "epoch": 1.0, "grad_norm": 10.0, "learning_rate": 2.6309163798966825e-06, "loss": 0.2081, "step": 5274 }, { "epoch": 1.0, "grad_norm": 17.375, "learning_rate": 2.6301514814853748e-06, "loss": 0.3388, "step": 5275 }, { "epoch": 1.0, "grad_norm": 11.0625, "learning_rate": 2.6293865708571325e-06, "loss": 0.268, "step": 5276 }, { "epoch": 1.0, "grad_norm": 12.4375, "learning_rate": 2.6286216480837558e-06, "loss": 0.2645, "step": 5277 }, { "epoch": 1.0, "grad_norm": 21.375, "learning_rate": 2.627856713237047e-06, "loss": 0.2865, "step": 5278 }, { "epoch": 1.0, "grad_norm": 13.5, "learning_rate": 2.6270917663888057e-06, "loss": 0.2526, "step": 5279 }, { "epoch": 1.0, "grad_norm": 10.1875, "learning_rate": 2.6263268076108383e-06, "loss": 0.2599, "step": 5280 }, { "epoch": 1.0, "grad_norm": 11.6875, "learning_rate": 2.6255618369749463e-06, "loss": 0.2433, "step": 5281 }, { "epoch": 1.0, "grad_norm": 10.4375, "learning_rate": 2.6247968545529377e-06, "loss": 0.2632, "step": 5282 }, { "epoch": 1.0, "grad_norm": 11.3125, "learning_rate": 2.6240318604166167e-06, "loss": 0.2255, "step": 5283 }, { "epoch": 1.0, "grad_norm": 10.0625, "learning_rate": 2.623266854637793e-06, "loss": 0.323, "step": 5284 }, { "epoch": 1.0, "grad_norm": 11.25, "learning_rate": 2.622501837288274e-06, "loss": 0.2102, "step": 5285 }, { "epoch": 1.0, "grad_norm": 19.125, "learning_rate": 2.621736808439872e-06, "loss": 0.3662, "step": 5286 }, { "epoch": 1.0, "grad_norm": 11.125, "learning_rate": 2.6209717681643948e-06, "loss": 0.2524, "step": 5287 }, { "epoch": 1.0, "grad_norm": 7.21875, "learning_rate": 2.6202067165336563e-06, "loss": 0.1969, "step": 5288 }, { "epoch": 1.0, "grad_norm": 10.0625, "learning_rate": 2.6194416536194696e-06, "loss": 0.2757, "step": 5289 }, { "epoch": 1.0, "grad_norm": 8.0, "learning_rate": 2.6186765794936486e-06, "loss": 0.2446, "step": 5290 }, { "epoch": 1.0, "grad_norm": 5.8125, "learning_rate": 2.617911494228008e-06, "loss": 0.1812, "step": 5291 }, { "epoch": 1.0, "grad_norm": 11.125, "learning_rate": 2.6171463978943655e-06, "loss": 0.2355, "step": 5292 }, { "epoch": 1.0, "grad_norm": 7.34375, "learning_rate": 2.616381290564537e-06, "loss": 0.1952, "step": 5293 }, { "epoch": 1.0, "grad_norm": 8.875, "learning_rate": 2.6156161723103417e-06, "loss": 0.2858, "step": 5294 }, { "epoch": 1.0, "grad_norm": 10.375, "learning_rate": 2.6148510432035985e-06, "loss": 0.2136, "step": 5295 }, { "epoch": 1.0, "grad_norm": 9.125, "learning_rate": 2.6140859033161283e-06, "loss": 0.2529, "step": 5296 }, { "epoch": 1.0, "grad_norm": 11.375, "learning_rate": 2.6133207527197518e-06, "loss": 0.2431, "step": 5297 }, { "epoch": 1.0, "grad_norm": 8.5, "learning_rate": 2.612555591486292e-06, "loss": 0.2769, "step": 5298 }, { "epoch": 1.0, "grad_norm": 10.0625, "learning_rate": 2.6117904196875733e-06, "loss": 0.1702, "step": 5299 }, { "epoch": 1.0, "grad_norm": 9.0625, "learning_rate": 2.611025237395417e-06, "loss": 0.2126, "step": 5300 }, { "epoch": 1.0, "grad_norm": 10.9375, "learning_rate": 2.610260044681651e-06, "loss": 0.2351, "step": 5301 }, { "epoch": 1.0, "grad_norm": 14.8125, "learning_rate": 2.609494841618102e-06, "loss": 0.1868, "step": 5302 }, { "epoch": 1.0, "grad_norm": 11.5, "learning_rate": 2.608729628276596e-06, "loss": 0.2093, "step": 5303 }, { "epoch": 1.0, "grad_norm": 11.75, "learning_rate": 2.607964404728962e-06, "loss": 0.251, "step": 5304 }, { "epoch": 1.0, "grad_norm": 10.25, "learning_rate": 2.607199171047029e-06, "loss": 0.2596, "step": 5305 }, { "epoch": 1.0, "grad_norm": 8.0625, "learning_rate": 2.6064339273026267e-06, "loss": 0.2123, "step": 5306 }, { "epoch": 1.0, "grad_norm": 12.8125, "learning_rate": 2.605668673567588e-06, "loss": 0.2311, "step": 5307 }, { "epoch": 1.0, "grad_norm": 13.125, "learning_rate": 2.6049034099137425e-06, "loss": 0.248, "step": 5308 }, { "epoch": 1.0, "grad_norm": 12.125, "learning_rate": 2.604138136412926e-06, "loss": 0.2266, "step": 5309 }, { "epoch": 1.0, "grad_norm": 8.875, "learning_rate": 2.60337285313697e-06, "loss": 0.1874, "step": 5310 }, { "epoch": 1.0, "grad_norm": 8.5, "learning_rate": 2.602607560157711e-06, "loss": 0.195, "step": 5311 }, { "epoch": 1.0, "grad_norm": 9.0625, "learning_rate": 2.601842257546984e-06, "loss": 0.2359, "step": 5312 }, { "epoch": 1.01, "grad_norm": 11.4375, "learning_rate": 2.601076945376626e-06, "loss": 0.2032, "step": 5313 }, { "epoch": 1.01, "grad_norm": 11.1875, "learning_rate": 2.600311623718474e-06, "loss": 0.2559, "step": 5314 }, { "epoch": 1.01, "grad_norm": 8.875, "learning_rate": 2.5995462926443677e-06, "loss": 0.1834, "step": 5315 }, { "epoch": 1.01, "grad_norm": 19.0, "learning_rate": 2.5987809522261455e-06, "loss": 0.2376, "step": 5316 }, { "epoch": 1.01, "grad_norm": 10.9375, "learning_rate": 2.598015602535647e-06, "loss": 0.1925, "step": 5317 }, { "epoch": 1.01, "grad_norm": 8.875, "learning_rate": 2.597250243644715e-06, "loss": 0.2281, "step": 5318 }, { "epoch": 1.01, "grad_norm": 9.375, "learning_rate": 2.596484875625191e-06, "loss": 0.1911, "step": 5319 }, { "epoch": 1.01, "grad_norm": 12.125, "learning_rate": 2.595719498548917e-06, "loss": 0.2234, "step": 5320 }, { "epoch": 1.01, "grad_norm": 10.9375, "learning_rate": 2.594954112487737e-06, "loss": 0.2583, "step": 5321 }, { "epoch": 1.01, "grad_norm": 10.0625, "learning_rate": 2.5941887175134956e-06, "loss": 0.197, "step": 5322 }, { "epoch": 1.01, "grad_norm": 10.8125, "learning_rate": 2.593423313698038e-06, "loss": 0.2689, "step": 5323 }, { "epoch": 1.01, "grad_norm": 8.6875, "learning_rate": 2.5926579011132117e-06, "loss": 0.2224, "step": 5324 }, { "epoch": 1.01, "grad_norm": 9.125, "learning_rate": 2.5918924798308614e-06, "loss": 0.2144, "step": 5325 }, { "epoch": 1.01, "grad_norm": 11.0, "learning_rate": 2.5911270499228376e-06, "loss": 0.2629, "step": 5326 }, { "epoch": 1.01, "grad_norm": 14.0, "learning_rate": 2.590361611460986e-06, "loss": 0.2224, "step": 5327 }, { "epoch": 1.01, "grad_norm": 9.0, "learning_rate": 2.5895961645171587e-06, "loss": 0.1809, "step": 5328 }, { "epoch": 1.01, "grad_norm": 10.625, "learning_rate": 2.5888307091632038e-06, "loss": 0.2529, "step": 5329 }, { "epoch": 1.01, "grad_norm": 9.3125, "learning_rate": 2.5880652454709743e-06, "loss": 0.2471, "step": 5330 }, { "epoch": 1.01, "grad_norm": 9.25, "learning_rate": 2.58729977351232e-06, "loss": 0.2151, "step": 5331 }, { "epoch": 1.01, "grad_norm": 10.75, "learning_rate": 2.5865342933590957e-06, "loss": 0.2383, "step": 5332 }, { "epoch": 1.01, "grad_norm": 11.125, "learning_rate": 2.5857688050831525e-06, "loss": 0.1925, "step": 5333 }, { "epoch": 1.01, "grad_norm": 11.5625, "learning_rate": 2.5850033087563457e-06, "loss": 0.2392, "step": 5334 }, { "epoch": 1.01, "grad_norm": 9.875, "learning_rate": 2.5842378044505306e-06, "loss": 0.1583, "step": 5335 }, { "epoch": 1.01, "grad_norm": 10.75, "learning_rate": 2.583472292237562e-06, "loss": 0.2073, "step": 5336 }, { "epoch": 1.01, "grad_norm": 8.75, "learning_rate": 2.582706772189297e-06, "loss": 0.2548, "step": 5337 }, { "epoch": 1.01, "grad_norm": 11.125, "learning_rate": 2.5819412443775915e-06, "loss": 0.2549, "step": 5338 }, { "epoch": 1.01, "grad_norm": 10.6875, "learning_rate": 2.581175708874304e-06, "loss": 0.2184, "step": 5339 }, { "epoch": 1.01, "grad_norm": 11.1875, "learning_rate": 2.5804101657512935e-06, "loss": 0.2659, "step": 5340 }, { "epoch": 1.01, "grad_norm": 7.9375, "learning_rate": 2.5796446150804194e-06, "loss": 0.1689, "step": 5341 }, { "epoch": 1.01, "grad_norm": 10.4375, "learning_rate": 2.5788790569335402e-06, "loss": 0.2412, "step": 5342 }, { "epoch": 1.01, "grad_norm": 8.375, "learning_rate": 2.5781134913825182e-06, "loss": 0.23, "step": 5343 }, { "epoch": 1.01, "grad_norm": 10.1875, "learning_rate": 2.5773479184992135e-06, "loss": 0.2541, "step": 5344 }, { "epoch": 1.01, "grad_norm": 15.5, "learning_rate": 2.5765823383554893e-06, "loss": 0.1977, "step": 5345 }, { "epoch": 1.01, "grad_norm": 10.25, "learning_rate": 2.5758167510232073e-06, "loss": 0.2609, "step": 5346 }, { "epoch": 1.01, "grad_norm": 8.6875, "learning_rate": 2.575051156574232e-06, "loss": 0.203, "step": 5347 }, { "epoch": 1.01, "grad_norm": 11.5, "learning_rate": 2.5742855550804265e-06, "loss": 0.195, "step": 5348 }, { "epoch": 1.01, "grad_norm": 10.8125, "learning_rate": 2.573519946613656e-06, "loss": 0.2038, "step": 5349 }, { "epoch": 1.01, "grad_norm": 7.34375, "learning_rate": 2.572754331245786e-06, "loss": 0.1826, "step": 5350 }, { "epoch": 1.01, "grad_norm": 9.3125, "learning_rate": 2.571988709048682e-06, "loss": 0.1898, "step": 5351 }, { "epoch": 1.01, "grad_norm": 16.125, "learning_rate": 2.5712230800942116e-06, "loss": 0.2, "step": 5352 }, { "epoch": 1.01, "grad_norm": 11.5625, "learning_rate": 2.5704574444542413e-06, "loss": 0.2283, "step": 5353 }, { "epoch": 1.01, "grad_norm": 8.6875, "learning_rate": 2.5696918022006395e-06, "loss": 0.1874, "step": 5354 }, { "epoch": 1.01, "grad_norm": 8.625, "learning_rate": 2.5689261534052743e-06, "loss": 0.2379, "step": 5355 }, { "epoch": 1.01, "grad_norm": 18.125, "learning_rate": 2.568160498140016e-06, "loss": 0.2247, "step": 5356 }, { "epoch": 1.01, "grad_norm": 10.5625, "learning_rate": 2.567394836476733e-06, "loss": 0.1752, "step": 5357 }, { "epoch": 1.01, "grad_norm": 13.3125, "learning_rate": 2.5666291684872964e-06, "loss": 0.2633, "step": 5358 }, { "epoch": 1.01, "grad_norm": 11.3125, "learning_rate": 2.5658634942435772e-06, "loss": 0.1938, "step": 5359 }, { "epoch": 1.01, "grad_norm": 7.75, "learning_rate": 2.5650978138174463e-06, "loss": 0.2153, "step": 5360 }, { "epoch": 1.01, "grad_norm": 7.71875, "learning_rate": 2.564332127280777e-06, "loss": 0.2001, "step": 5361 }, { "epoch": 1.01, "grad_norm": 10.875, "learning_rate": 2.563566434705441e-06, "loss": 0.1917, "step": 5362 }, { "epoch": 1.01, "grad_norm": 12.4375, "learning_rate": 2.562800736163313e-06, "loss": 0.2344, "step": 5363 }, { "epoch": 1.01, "grad_norm": 8.8125, "learning_rate": 2.562035031726265e-06, "loss": 0.2168, "step": 5364 }, { "epoch": 1.01, "grad_norm": 11.375, "learning_rate": 2.561269321466172e-06, "loss": 0.185, "step": 5365 }, { "epoch": 1.02, "grad_norm": 10.125, "learning_rate": 2.5605036054549098e-06, "loss": 0.2159, "step": 5366 }, { "epoch": 1.02, "grad_norm": 16.125, "learning_rate": 2.559737883764353e-06, "loss": 0.2604, "step": 5367 }, { "epoch": 1.02, "grad_norm": 7.3125, "learning_rate": 2.558972156466379e-06, "loss": 0.1699, "step": 5368 }, { "epoch": 1.02, "grad_norm": 9.4375, "learning_rate": 2.558206423632862e-06, "loss": 0.251, "step": 5369 }, { "epoch": 1.02, "grad_norm": 8.75, "learning_rate": 2.5574406853356817e-06, "loss": 0.1788, "step": 5370 }, { "epoch": 1.02, "grad_norm": 17.875, "learning_rate": 2.5566749416467123e-06, "loss": 0.2098, "step": 5371 }, { "epoch": 1.02, "grad_norm": 10.9375, "learning_rate": 2.555909192637836e-06, "loss": 0.1486, "step": 5372 }, { "epoch": 1.02, "grad_norm": 7.65625, "learning_rate": 2.5551434383809282e-06, "loss": 0.1798, "step": 5373 }, { "epoch": 1.02, "grad_norm": 11.6875, "learning_rate": 2.5543776789478703e-06, "loss": 0.2247, "step": 5374 }, { "epoch": 1.02, "grad_norm": 10.75, "learning_rate": 2.5536119144105393e-06, "loss": 0.1982, "step": 5375 }, { "epoch": 1.02, "grad_norm": 9.3125, "learning_rate": 2.5528461448408183e-06, "loss": 0.2035, "step": 5376 }, { "epoch": 1.02, "grad_norm": 9.5, "learning_rate": 2.5520803703105844e-06, "loss": 0.1896, "step": 5377 }, { "epoch": 1.02, "grad_norm": 8.5, "learning_rate": 2.5513145908917214e-06, "loss": 0.1785, "step": 5378 }, { "epoch": 1.02, "grad_norm": 10.375, "learning_rate": 2.55054880665611e-06, "loss": 0.2631, "step": 5379 }, { "epoch": 1.02, "grad_norm": 9.25, "learning_rate": 2.549783017675631e-06, "loss": 0.196, "step": 5380 }, { "epoch": 1.02, "grad_norm": 9.125, "learning_rate": 2.5490172240221676e-06, "loss": 0.2025, "step": 5381 }, { "epoch": 1.02, "grad_norm": 9.125, "learning_rate": 2.5482514257676034e-06, "loss": 0.2323, "step": 5382 }, { "epoch": 1.02, "grad_norm": 7.875, "learning_rate": 2.5474856229838206e-06, "loss": 0.1825, "step": 5383 }, { "epoch": 1.02, "grad_norm": 9.125, "learning_rate": 2.5467198157427027e-06, "loss": 0.1967, "step": 5384 }, { "epoch": 1.02, "grad_norm": 7.59375, "learning_rate": 2.5459540041161344e-06, "loss": 0.1669, "step": 5385 }, { "epoch": 1.02, "grad_norm": 8.1875, "learning_rate": 2.5451881881759988e-06, "loss": 0.1505, "step": 5386 }, { "epoch": 1.02, "grad_norm": 12.4375, "learning_rate": 2.544422367994183e-06, "loss": 0.2072, "step": 5387 }, { "epoch": 1.02, "grad_norm": 9.75, "learning_rate": 2.54365654364257e-06, "loss": 0.2269, "step": 5388 }, { "epoch": 1.02, "grad_norm": 7.875, "learning_rate": 2.542890715193047e-06, "loss": 0.1564, "step": 5389 }, { "epoch": 1.02, "grad_norm": 9.5, "learning_rate": 2.5421248827175e-06, "loss": 0.1714, "step": 5390 }, { "epoch": 1.02, "grad_norm": 11.125, "learning_rate": 2.5413590462878144e-06, "loss": 0.214, "step": 5391 }, { "epoch": 1.02, "grad_norm": 8.1875, "learning_rate": 2.540593205975878e-06, "loss": 0.2042, "step": 5392 }, { "epoch": 1.02, "grad_norm": 8.5625, "learning_rate": 2.539827361853578e-06, "loss": 0.2085, "step": 5393 }, { "epoch": 1.02, "grad_norm": 9.1875, "learning_rate": 2.539061513992801e-06, "loss": 0.1704, "step": 5394 }, { "epoch": 1.02, "grad_norm": 11.125, "learning_rate": 2.5382956624654353e-06, "loss": 0.2306, "step": 5395 }, { "epoch": 1.02, "grad_norm": 10.25, "learning_rate": 2.537529807343369e-06, "loss": 0.2188, "step": 5396 }, { "epoch": 1.02, "grad_norm": 10.625, "learning_rate": 2.5367639486984914e-06, "loss": 0.2132, "step": 5397 }, { "epoch": 1.02, "grad_norm": 9.8125, "learning_rate": 2.53599808660269e-06, "loss": 0.2332, "step": 5398 }, { "epoch": 1.02, "grad_norm": 9.8125, "learning_rate": 2.5352322211278556e-06, "loss": 0.1925, "step": 5399 }, { "epoch": 1.02, "grad_norm": 11.625, "learning_rate": 2.534466352345877e-06, "loss": 0.1946, "step": 5400 }, { "epoch": 1.02, "grad_norm": 11.5625, "learning_rate": 2.5337004803286437e-06, "loss": 0.2526, "step": 5401 }, { "epoch": 1.02, "grad_norm": 15.9375, "learning_rate": 2.5329346051480457e-06, "loss": 0.2038, "step": 5402 }, { "epoch": 1.02, "grad_norm": 9.875, "learning_rate": 2.5321687268759743e-06, "loss": 0.2342, "step": 5403 }, { "epoch": 1.02, "grad_norm": 14.0, "learning_rate": 2.53140284558432e-06, "loss": 0.2241, "step": 5404 }, { "epoch": 1.02, "grad_norm": 12.5625, "learning_rate": 2.530636961344973e-06, "loss": 0.2425, "step": 5405 }, { "epoch": 1.02, "grad_norm": 8.75, "learning_rate": 2.5298710742298255e-06, "loss": 0.2086, "step": 5406 }, { "epoch": 1.02, "grad_norm": 7.65625, "learning_rate": 2.529105184310768e-06, "loss": 0.1443, "step": 5407 }, { "epoch": 1.02, "grad_norm": 10.625, "learning_rate": 2.5283392916596946e-06, "loss": 0.1593, "step": 5408 }, { "epoch": 1.02, "grad_norm": 8.9375, "learning_rate": 2.5275733963484946e-06, "loss": 0.1946, "step": 5409 }, { "epoch": 1.02, "grad_norm": 9.875, "learning_rate": 2.526807498449062e-06, "loss": 0.1518, "step": 5410 }, { "epoch": 1.02, "grad_norm": 10.0625, "learning_rate": 2.5260415980332888e-06, "loss": 0.1651, "step": 5411 }, { "epoch": 1.02, "grad_norm": 10.4375, "learning_rate": 2.5252756951730677e-06, "loss": 0.2196, "step": 5412 }, { "epoch": 1.02, "grad_norm": 14.375, "learning_rate": 2.5245097899402917e-06, "loss": 0.187, "step": 5413 }, { "epoch": 1.02, "grad_norm": 10.5, "learning_rate": 2.5237438824068556e-06, "loss": 0.2003, "step": 5414 }, { "epoch": 1.02, "grad_norm": 11.25, "learning_rate": 2.52297797264465e-06, "loss": 0.2005, "step": 5415 }, { "epoch": 1.02, "grad_norm": 9.875, "learning_rate": 2.5222120607255713e-06, "loss": 0.2019, "step": 5416 }, { "epoch": 1.02, "grad_norm": 30.375, "learning_rate": 2.521446146721511e-06, "loss": 0.2533, "step": 5417 }, { "epoch": 1.02, "grad_norm": 10.5625, "learning_rate": 2.5206802307043667e-06, "loss": 0.1952, "step": 5418 }, { "epoch": 1.03, "grad_norm": 10.75, "learning_rate": 2.519914312746028e-06, "loss": 0.2348, "step": 5419 }, { "epoch": 1.03, "grad_norm": 8.375, "learning_rate": 2.519148392918393e-06, "loss": 0.2111, "step": 5420 }, { "epoch": 1.03, "grad_norm": 10.875, "learning_rate": 2.5183824712933546e-06, "loss": 0.177, "step": 5421 }, { "epoch": 1.03, "grad_norm": 10.4375, "learning_rate": 2.517616547942809e-06, "loss": 0.2119, "step": 5422 }, { "epoch": 1.03, "grad_norm": 8.125, "learning_rate": 2.5168506229386487e-06, "loss": 0.2338, "step": 5423 }, { "epoch": 1.03, "grad_norm": 14.4375, "learning_rate": 2.516084696352771e-06, "loss": 0.1926, "step": 5424 }, { "epoch": 1.03, "grad_norm": 12.75, "learning_rate": 2.5153187682570713e-06, "loss": 0.2617, "step": 5425 }, { "epoch": 1.03, "grad_norm": 12.5, "learning_rate": 2.5145528387234426e-06, "loss": 0.2651, "step": 5426 }, { "epoch": 1.03, "grad_norm": 12.1875, "learning_rate": 2.5137869078237836e-06, "loss": 0.269, "step": 5427 }, { "epoch": 1.03, "grad_norm": 11.375, "learning_rate": 2.5130209756299874e-06, "loss": 0.2436, "step": 5428 }, { "epoch": 1.03, "grad_norm": 11.8125, "learning_rate": 2.5122550422139515e-06, "loss": 0.223, "step": 5429 }, { "epoch": 1.03, "grad_norm": 12.375, "learning_rate": 2.5114891076475702e-06, "loss": 0.2014, "step": 5430 }, { "epoch": 1.03, "grad_norm": 9.375, "learning_rate": 2.5107231720027407e-06, "loss": 0.2031, "step": 5431 }, { "epoch": 1.03, "grad_norm": 9.625, "learning_rate": 2.5099572353513587e-06, "loss": 0.1771, "step": 5432 }, { "epoch": 1.03, "grad_norm": 11.25, "learning_rate": 2.5091912977653215e-06, "loss": 0.2207, "step": 5433 }, { "epoch": 1.03, "grad_norm": 14.8125, "learning_rate": 2.5084253593165233e-06, "loss": 0.262, "step": 5434 }, { "epoch": 1.03, "grad_norm": 11.4375, "learning_rate": 2.507659420076862e-06, "loss": 0.1651, "step": 5435 }, { "epoch": 1.03, "grad_norm": 8.125, "learning_rate": 2.5068934801182332e-06, "loss": 0.1882, "step": 5436 }, { "epoch": 1.03, "grad_norm": 9.9375, "learning_rate": 2.5061275395125338e-06, "loss": 0.2368, "step": 5437 }, { "epoch": 1.03, "grad_norm": 8.8125, "learning_rate": 2.505361598331661e-06, "loss": 0.135, "step": 5438 }, { "epoch": 1.03, "grad_norm": 8.3125, "learning_rate": 2.5045956566475106e-06, "loss": 0.1998, "step": 5439 }, { "epoch": 1.03, "grad_norm": 10.75, "learning_rate": 2.5038297145319795e-06, "loss": 0.2636, "step": 5440 }, { "epoch": 1.03, "grad_norm": 11.75, "learning_rate": 2.5030637720569643e-06, "loss": 0.1946, "step": 5441 }, { "epoch": 1.03, "grad_norm": 9.875, "learning_rate": 2.5022978292943624e-06, "loss": 0.2542, "step": 5442 }, { "epoch": 1.03, "grad_norm": 8.3125, "learning_rate": 2.50153188631607e-06, "loss": 0.2408, "step": 5443 }, { "epoch": 1.03, "grad_norm": 7.9375, "learning_rate": 2.5007659431939833e-06, "loss": 0.1817, "step": 5444 }, { "epoch": 1.03, "grad_norm": 18.75, "learning_rate": 2.5e-06, "loss": 0.2307, "step": 5445 }, { "epoch": 1.03, "grad_norm": 8.125, "learning_rate": 2.499234056806017e-06, "loss": 0.2058, "step": 5446 }, { "epoch": 1.03, "grad_norm": 11.125, "learning_rate": 2.49846811368393e-06, "loss": 0.2376, "step": 5447 }, { "epoch": 1.03, "grad_norm": 8.125, "learning_rate": 2.4977021707056384e-06, "loss": 0.1757, "step": 5448 }, { "epoch": 1.03, "grad_norm": 14.4375, "learning_rate": 2.496936227943036e-06, "loss": 0.2015, "step": 5449 }, { "epoch": 1.03, "grad_norm": 8.375, "learning_rate": 2.4961702854680213e-06, "loss": 0.2031, "step": 5450 }, { "epoch": 1.03, "grad_norm": 14.6875, "learning_rate": 2.4954043433524894e-06, "loss": 0.1953, "step": 5451 }, { "epoch": 1.03, "grad_norm": 10.0, "learning_rate": 2.4946384016683396e-06, "loss": 0.2192, "step": 5452 }, { "epoch": 1.03, "grad_norm": 9.9375, "learning_rate": 2.4938724604874666e-06, "loss": 0.19, "step": 5453 }, { "epoch": 1.03, "grad_norm": 12.8125, "learning_rate": 2.4931065198817684e-06, "loss": 0.2917, "step": 5454 }, { "epoch": 1.03, "grad_norm": 13.8125, "learning_rate": 2.4923405799231386e-06, "loss": 0.2448, "step": 5455 }, { "epoch": 1.03, "grad_norm": 8.3125, "learning_rate": 2.491574640683478e-06, "loss": 0.159, "step": 5456 }, { "epoch": 1.03, "grad_norm": 11.8125, "learning_rate": 2.49080870223468e-06, "loss": 0.2312, "step": 5457 }, { "epoch": 1.03, "grad_norm": 10.6875, "learning_rate": 2.4900427646486413e-06, "loss": 0.2403, "step": 5458 }, { "epoch": 1.03, "grad_norm": 12.875, "learning_rate": 2.4892768279972597e-06, "loss": 0.2277, "step": 5459 }, { "epoch": 1.03, "grad_norm": 8.125, "learning_rate": 2.488510892352431e-06, "loss": 0.1614, "step": 5460 }, { "epoch": 1.03, "grad_norm": 8.875, "learning_rate": 2.48774495778605e-06, "loss": 0.2074, "step": 5461 }, { "epoch": 1.03, "grad_norm": 10.125, "learning_rate": 2.486979024370013e-06, "loss": 0.2472, "step": 5462 }, { "epoch": 1.03, "grad_norm": 10.875, "learning_rate": 2.4862130921762172e-06, "loss": 0.2302, "step": 5463 }, { "epoch": 1.03, "grad_norm": 14.0, "learning_rate": 2.4854471612765578e-06, "loss": 0.2289, "step": 5464 }, { "epoch": 1.03, "grad_norm": 7.1875, "learning_rate": 2.4846812317429304e-06, "loss": 0.1676, "step": 5465 }, { "epoch": 1.03, "grad_norm": 9.5, "learning_rate": 2.4839153036472295e-06, "loss": 0.2129, "step": 5466 }, { "epoch": 1.03, "grad_norm": 8.25, "learning_rate": 2.483149377061352e-06, "loss": 0.2357, "step": 5467 }, { "epoch": 1.03, "grad_norm": 8.4375, "learning_rate": 2.4823834520571928e-06, "loss": 0.1705, "step": 5468 }, { "epoch": 1.03, "grad_norm": 10.3125, "learning_rate": 2.481617528706646e-06, "loss": 0.2221, "step": 5469 }, { "epoch": 1.03, "grad_norm": 7.96875, "learning_rate": 2.4808516070816075e-06, "loss": 0.208, "step": 5470 }, { "epoch": 1.03, "grad_norm": 9.4375, "learning_rate": 2.4800856872539727e-06, "loss": 0.216, "step": 5471 }, { "epoch": 1.04, "grad_norm": 9.8125, "learning_rate": 2.479319769295635e-06, "loss": 0.2313, "step": 5472 }, { "epoch": 1.04, "grad_norm": 11.4375, "learning_rate": 2.478553853278489e-06, "loss": 0.2688, "step": 5473 }, { "epoch": 1.04, "grad_norm": 11.125, "learning_rate": 2.477787939274429e-06, "loss": 0.268, "step": 5474 }, { "epoch": 1.04, "grad_norm": 11.9375, "learning_rate": 2.4770220273553508e-06, "loss": 0.176, "step": 5475 }, { "epoch": 1.04, "grad_norm": 14.0625, "learning_rate": 2.476256117593146e-06, "loss": 0.2032, "step": 5476 }, { "epoch": 1.04, "grad_norm": 11.25, "learning_rate": 2.4754902100597083e-06, "loss": 0.2449, "step": 5477 }, { "epoch": 1.04, "grad_norm": 8.125, "learning_rate": 2.4747243048269327e-06, "loss": 0.2218, "step": 5478 }, { "epoch": 1.04, "grad_norm": 8.4375, "learning_rate": 2.4739584019667125e-06, "loss": 0.1598, "step": 5479 }, { "epoch": 1.04, "grad_norm": 13.0625, "learning_rate": 2.473192501550939e-06, "loss": 0.2912, "step": 5480 }, { "epoch": 1.04, "grad_norm": 12.5, "learning_rate": 2.4724266036515063e-06, "loss": 0.1882, "step": 5481 }, { "epoch": 1.04, "grad_norm": 10.0625, "learning_rate": 2.4716607083403067e-06, "loss": 0.2384, "step": 5482 }, { "epoch": 1.04, "grad_norm": 7.78125, "learning_rate": 2.470894815689233e-06, "loss": 0.131, "step": 5483 }, { "epoch": 1.04, "grad_norm": 10.0, "learning_rate": 2.4701289257701754e-06, "loss": 0.2336, "step": 5484 }, { "epoch": 1.04, "grad_norm": 12.75, "learning_rate": 2.469363038655028e-06, "loss": 0.2329, "step": 5485 }, { "epoch": 1.04, "grad_norm": 13.875, "learning_rate": 2.468597154415681e-06, "loss": 0.2191, "step": 5486 }, { "epoch": 1.04, "grad_norm": 9.8125, "learning_rate": 2.467831273124027e-06, "loss": 0.2124, "step": 5487 }, { "epoch": 1.04, "grad_norm": 12.6875, "learning_rate": 2.467065394851955e-06, "loss": 0.2645, "step": 5488 }, { "epoch": 1.04, "grad_norm": 11.375, "learning_rate": 2.4662995196713575e-06, "loss": 0.2063, "step": 5489 }, { "epoch": 1.04, "grad_norm": 12.3125, "learning_rate": 2.4655336476541243e-06, "loss": 0.238, "step": 5490 }, { "epoch": 1.04, "grad_norm": 9.9375, "learning_rate": 2.464767778872145e-06, "loss": 0.2062, "step": 5491 }, { "epoch": 1.04, "grad_norm": 10.5, "learning_rate": 2.4640019133973102e-06, "loss": 0.1812, "step": 5492 }, { "epoch": 1.04, "grad_norm": 8.9375, "learning_rate": 2.46323605130151e-06, "loss": 0.2118, "step": 5493 }, { "epoch": 1.04, "grad_norm": 12.5625, "learning_rate": 2.462470192656632e-06, "loss": 0.2522, "step": 5494 }, { "epoch": 1.04, "grad_norm": 20.375, "learning_rate": 2.461704337534565e-06, "loss": 0.2643, "step": 5495 }, { "epoch": 1.04, "grad_norm": 13.25, "learning_rate": 2.4609384860072e-06, "loss": 0.2599, "step": 5496 }, { "epoch": 1.04, "grad_norm": 13.375, "learning_rate": 2.460172638146423e-06, "loss": 0.2113, "step": 5497 }, { "epoch": 1.04, "grad_norm": 7.75, "learning_rate": 2.459406794024123e-06, "loss": 0.1618, "step": 5498 }, { "epoch": 1.04, "grad_norm": 9.3125, "learning_rate": 2.4586409537121856e-06, "loss": 0.1966, "step": 5499 }, { "epoch": 1.04, "grad_norm": 12.5, "learning_rate": 2.457875117282501e-06, "loss": 0.2115, "step": 5500 }, { "epoch": 1.04, "grad_norm": 9.375, "learning_rate": 2.4571092848069534e-06, "loss": 0.2338, "step": 5501 }, { "epoch": 1.04, "grad_norm": 9.4375, "learning_rate": 2.4563434563574303e-06, "loss": 0.2049, "step": 5502 }, { "epoch": 1.04, "grad_norm": 12.375, "learning_rate": 2.4555776320058174e-06, "loss": 0.2374, "step": 5503 }, { "epoch": 1.04, "grad_norm": 14.0, "learning_rate": 2.454811811824002e-06, "loss": 0.2771, "step": 5504 }, { "epoch": 1.04, "grad_norm": 11.9375, "learning_rate": 2.454045995883867e-06, "loss": 0.2113, "step": 5505 }, { "epoch": 1.04, "grad_norm": 14.0, "learning_rate": 2.4532801842572977e-06, "loss": 0.1984, "step": 5506 }, { "epoch": 1.04, "grad_norm": 8.625, "learning_rate": 2.45251437701618e-06, "loss": 0.2397, "step": 5507 }, { "epoch": 1.04, "grad_norm": 7.4375, "learning_rate": 2.4517485742323974e-06, "loss": 0.1516, "step": 5508 }, { "epoch": 1.04, "grad_norm": 9.4375, "learning_rate": 2.450982775977833e-06, "loss": 0.2201, "step": 5509 }, { "epoch": 1.04, "grad_norm": 8.0625, "learning_rate": 2.450216982324369e-06, "loss": 0.2091, "step": 5510 }, { "epoch": 1.04, "grad_norm": 12.3125, "learning_rate": 2.449451193343891e-06, "loss": 0.2182, "step": 5511 }, { "epoch": 1.04, "grad_norm": 18.625, "learning_rate": 2.4486854091082795e-06, "loss": 0.2179, "step": 5512 }, { "epoch": 1.04, "grad_norm": 9.25, "learning_rate": 2.4479196296894156e-06, "loss": 0.2128, "step": 5513 }, { "epoch": 1.04, "grad_norm": 10.875, "learning_rate": 2.4471538551591825e-06, "loss": 0.2214, "step": 5514 }, { "epoch": 1.04, "grad_norm": 13.25, "learning_rate": 2.4463880855894615e-06, "loss": 0.2108, "step": 5515 }, { "epoch": 1.04, "grad_norm": 11.75, "learning_rate": 2.445622321052131e-06, "loss": 0.1828, "step": 5516 }, { "epoch": 1.04, "grad_norm": 11.125, "learning_rate": 2.4448565616190718e-06, "loss": 0.2294, "step": 5517 }, { "epoch": 1.04, "grad_norm": 20.25, "learning_rate": 2.444090807362165e-06, "loss": 0.2298, "step": 5518 }, { "epoch": 1.04, "grad_norm": 13.5, "learning_rate": 2.443325058353288e-06, "loss": 0.197, "step": 5519 }, { "epoch": 1.04, "grad_norm": 11.0625, "learning_rate": 2.44255931466432e-06, "loss": 0.2392, "step": 5520 }, { "epoch": 1.04, "grad_norm": 8.75, "learning_rate": 2.4417935763671384e-06, "loss": 0.2042, "step": 5521 }, { "epoch": 1.04, "grad_norm": 9.625, "learning_rate": 2.441027843533622e-06, "loss": 0.2335, "step": 5522 }, { "epoch": 1.04, "grad_norm": 16.375, "learning_rate": 2.4402621162356475e-06, "loss": 0.2489, "step": 5523 }, { "epoch": 1.04, "grad_norm": 11.25, "learning_rate": 2.439496394545091e-06, "loss": 0.2405, "step": 5524 }, { "epoch": 1.05, "grad_norm": 11.25, "learning_rate": 2.4387306785338287e-06, "loss": 0.2037, "step": 5525 }, { "epoch": 1.05, "grad_norm": 18.875, "learning_rate": 2.437964968273736e-06, "loss": 0.2173, "step": 5526 }, { "epoch": 1.05, "grad_norm": 21.5, "learning_rate": 2.4371992638366887e-06, "loss": 0.1988, "step": 5527 }, { "epoch": 1.05, "grad_norm": 8.875, "learning_rate": 2.4364335652945593e-06, "loss": 0.195, "step": 5528 }, { "epoch": 1.05, "grad_norm": 13.875, "learning_rate": 2.435667872719224e-06, "loss": 0.2345, "step": 5529 }, { "epoch": 1.05, "grad_norm": 10.1875, "learning_rate": 2.434902186182554e-06, "loss": 0.2406, "step": 5530 }, { "epoch": 1.05, "grad_norm": 10.0, "learning_rate": 2.4341365057564244e-06, "loss": 0.1929, "step": 5531 }, { "epoch": 1.05, "grad_norm": 10.125, "learning_rate": 2.4333708315127044e-06, "loss": 0.2607, "step": 5532 }, { "epoch": 1.05, "grad_norm": 8.625, "learning_rate": 2.4326051635232677e-06, "loss": 0.1732, "step": 5533 }, { "epoch": 1.05, "grad_norm": 10.9375, "learning_rate": 2.4318395018599847e-06, "loss": 0.2657, "step": 5534 }, { "epoch": 1.05, "grad_norm": 10.4375, "learning_rate": 2.4310738465947252e-06, "loss": 0.2129, "step": 5535 }, { "epoch": 1.05, "grad_norm": 9.8125, "learning_rate": 2.4303081977993613e-06, "loss": 0.1896, "step": 5536 }, { "epoch": 1.05, "grad_norm": 9.125, "learning_rate": 2.429542555545759e-06, "loss": 0.2189, "step": 5537 }, { "epoch": 1.05, "grad_norm": 11.625, "learning_rate": 2.4287769199057897e-06, "loss": 0.194, "step": 5538 }, { "epoch": 1.05, "grad_norm": 35.75, "learning_rate": 2.428011290951318e-06, "loss": 0.2362, "step": 5539 }, { "epoch": 1.05, "grad_norm": 12.5625, "learning_rate": 2.427245668754215e-06, "loss": 0.1816, "step": 5540 }, { "epoch": 1.05, "grad_norm": 10.3125, "learning_rate": 2.426480053386345e-06, "loss": 0.2344, "step": 5541 }, { "epoch": 1.05, "grad_norm": 11.0625, "learning_rate": 2.4257144449195748e-06, "loss": 0.2262, "step": 5542 }, { "epoch": 1.05, "grad_norm": 11.125, "learning_rate": 2.424948843425768e-06, "loss": 0.2566, "step": 5543 }, { "epoch": 1.05, "grad_norm": 10.625, "learning_rate": 2.424183248976793e-06, "loss": 0.2064, "step": 5544 }, { "epoch": 1.05, "grad_norm": 7.9375, "learning_rate": 2.4234176616445116e-06, "loss": 0.1764, "step": 5545 }, { "epoch": 1.05, "grad_norm": 11.875, "learning_rate": 2.422652081500787e-06, "loss": 0.248, "step": 5546 }, { "epoch": 1.05, "grad_norm": 11.1875, "learning_rate": 2.421886508617482e-06, "loss": 0.2295, "step": 5547 }, { "epoch": 1.05, "grad_norm": 11.25, "learning_rate": 2.4211209430664606e-06, "loss": 0.1886, "step": 5548 }, { "epoch": 1.05, "grad_norm": 10.3125, "learning_rate": 2.420355384919582e-06, "loss": 0.1895, "step": 5549 }, { "epoch": 1.05, "grad_norm": 10.625, "learning_rate": 2.4195898342487065e-06, "loss": 0.2057, "step": 5550 }, { "epoch": 1.05, "grad_norm": 12.75, "learning_rate": 2.4188242911256965e-06, "loss": 0.2346, "step": 5551 }, { "epoch": 1.05, "grad_norm": 7.71875, "learning_rate": 2.4180587556224093e-06, "loss": 0.1796, "step": 5552 }, { "epoch": 1.05, "grad_norm": 11.5625, "learning_rate": 2.4172932278107044e-06, "loss": 0.1698, "step": 5553 }, { "epoch": 1.05, "grad_norm": 13.125, "learning_rate": 2.4165277077624384e-06, "loss": 0.2338, "step": 5554 }, { "epoch": 1.05, "grad_norm": 9.5625, "learning_rate": 2.4157621955494703e-06, "loss": 0.2275, "step": 5555 }, { "epoch": 1.05, "grad_norm": 19.125, "learning_rate": 2.4149966912436547e-06, "loss": 0.1896, "step": 5556 }, { "epoch": 1.05, "grad_norm": 11.9375, "learning_rate": 2.4142311949168483e-06, "loss": 0.2195, "step": 5557 }, { "epoch": 1.05, "grad_norm": 15.4375, "learning_rate": 2.413465706640905e-06, "loss": 0.2125, "step": 5558 }, { "epoch": 1.05, "grad_norm": 10.25, "learning_rate": 2.4127002264876805e-06, "loss": 0.2118, "step": 5559 }, { "epoch": 1.05, "grad_norm": 11.6875, "learning_rate": 2.4119347545290265e-06, "loss": 0.2005, "step": 5560 }, { "epoch": 1.05, "grad_norm": 11.375, "learning_rate": 2.4111692908367962e-06, "loss": 0.2971, "step": 5561 }, { "epoch": 1.05, "grad_norm": 7.96875, "learning_rate": 2.4104038354828417e-06, "loss": 0.194, "step": 5562 }, { "epoch": 1.05, "grad_norm": 11.1875, "learning_rate": 2.409638388539015e-06, "loss": 0.2588, "step": 5563 }, { "epoch": 1.05, "grad_norm": 10.6875, "learning_rate": 2.4088729500771637e-06, "loss": 0.2617, "step": 5564 }, { "epoch": 1.05, "grad_norm": 11.125, "learning_rate": 2.4081075201691385e-06, "loss": 0.1881, "step": 5565 }, { "epoch": 1.05, "grad_norm": 9.8125, "learning_rate": 2.407342098886789e-06, "loss": 0.1701, "step": 5566 }, { "epoch": 1.05, "grad_norm": 15.25, "learning_rate": 2.4065766863019625e-06, "loss": 0.2355, "step": 5567 }, { "epoch": 1.05, "grad_norm": 11.0625, "learning_rate": 2.4058112824865056e-06, "loss": 0.2411, "step": 5568 }, { "epoch": 1.05, "grad_norm": 12.9375, "learning_rate": 2.405045887512264e-06, "loss": 0.293, "step": 5569 }, { "epoch": 1.05, "grad_norm": 9.75, "learning_rate": 2.404280501451084e-06, "loss": 0.2137, "step": 5570 }, { "epoch": 1.05, "grad_norm": 8.25, "learning_rate": 2.4035151243748106e-06, "loss": 0.2061, "step": 5571 }, { "epoch": 1.05, "grad_norm": 11.3125, "learning_rate": 2.4027497563552854e-06, "loss": 0.2421, "step": 5572 }, { "epoch": 1.05, "grad_norm": 12.25, "learning_rate": 2.4019843974643533e-06, "loss": 0.2193, "step": 5573 }, { "epoch": 1.05, "grad_norm": 8.5625, "learning_rate": 2.401219047773856e-06, "loss": 0.1854, "step": 5574 }, { "epoch": 1.05, "grad_norm": 9.9375, "learning_rate": 2.400453707355634e-06, "loss": 0.1672, "step": 5575 }, { "epoch": 1.05, "grad_norm": 9.4375, "learning_rate": 2.3996883762815265e-06, "loss": 0.1998, "step": 5576 }, { "epoch": 1.06, "grad_norm": 8.9375, "learning_rate": 2.3989230546233747e-06, "loss": 0.1955, "step": 5577 }, { "epoch": 1.06, "grad_norm": 9.5, "learning_rate": 2.398157742453017e-06, "loss": 0.185, "step": 5578 }, { "epoch": 1.06, "grad_norm": 8.875, "learning_rate": 2.3973924398422903e-06, "loss": 0.1748, "step": 5579 }, { "epoch": 1.06, "grad_norm": 12.9375, "learning_rate": 2.3966271468630305e-06, "loss": 0.215, "step": 5580 }, { "epoch": 1.06, "grad_norm": 10.1875, "learning_rate": 2.395861863587075e-06, "loss": 0.1853, "step": 5581 }, { "epoch": 1.06, "grad_norm": 10.25, "learning_rate": 2.3950965900862583e-06, "loss": 0.1843, "step": 5582 }, { "epoch": 1.06, "grad_norm": 9.5625, "learning_rate": 2.3943313264324125e-06, "loss": 0.2406, "step": 5583 }, { "epoch": 1.06, "grad_norm": 8.6875, "learning_rate": 2.3935660726973737e-06, "loss": 0.1721, "step": 5584 }, { "epoch": 1.06, "grad_norm": 10.1875, "learning_rate": 2.392800828952972e-06, "loss": 0.2581, "step": 5585 }, { "epoch": 1.06, "grad_norm": 12.9375, "learning_rate": 2.3920355952710393e-06, "loss": 0.2714, "step": 5586 }, { "epoch": 1.06, "grad_norm": 10.8125, "learning_rate": 2.3912703717234043e-06, "loss": 0.2053, "step": 5587 }, { "epoch": 1.06, "grad_norm": 13.625, "learning_rate": 2.390505158381899e-06, "loss": 0.2233, "step": 5588 }, { "epoch": 1.06, "grad_norm": 10.875, "learning_rate": 2.3897399553183492e-06, "loss": 0.1878, "step": 5589 }, { "epoch": 1.06, "grad_norm": 7.25, "learning_rate": 2.3889747626045842e-06, "loss": 0.1334, "step": 5590 }, { "epoch": 1.06, "grad_norm": 9.6875, "learning_rate": 2.388209580312428e-06, "loss": 0.2053, "step": 5591 }, { "epoch": 1.06, "grad_norm": 11.5625, "learning_rate": 2.3874444085137085e-06, "loss": 0.2457, "step": 5592 }, { "epoch": 1.06, "grad_norm": 10.375, "learning_rate": 2.386679247280249e-06, "loss": 0.222, "step": 5593 }, { "epoch": 1.06, "grad_norm": 10.9375, "learning_rate": 2.385914096683872e-06, "loss": 0.1367, "step": 5594 }, { "epoch": 1.06, "grad_norm": 11.0, "learning_rate": 2.3851489567964024e-06, "loss": 0.2352, "step": 5595 }, { "epoch": 1.06, "grad_norm": 10.1875, "learning_rate": 2.384383827689659e-06, "loss": 0.1991, "step": 5596 }, { "epoch": 1.06, "grad_norm": 11.375, "learning_rate": 2.383618709435464e-06, "loss": 0.2435, "step": 5597 }, { "epoch": 1.06, "grad_norm": 9.1875, "learning_rate": 2.382853602105635e-06, "loss": 0.191, "step": 5598 }, { "epoch": 1.06, "grad_norm": 7.0625, "learning_rate": 2.3820885057719923e-06, "loss": 0.1914, "step": 5599 }, { "epoch": 1.06, "grad_norm": 14.25, "learning_rate": 2.3813234205063523e-06, "loss": 0.2248, "step": 5600 }, { "epoch": 1.06, "grad_norm": 12.0, "learning_rate": 2.3805583463805313e-06, "loss": 0.2265, "step": 5601 }, { "epoch": 1.06, "grad_norm": 11.0, "learning_rate": 2.3797932834663437e-06, "loss": 0.2284, "step": 5602 }, { "epoch": 1.06, "grad_norm": 15.1875, "learning_rate": 2.379028231835606e-06, "loss": 0.1964, "step": 5603 }, { "epoch": 1.06, "grad_norm": 13.5, "learning_rate": 2.378263191560129e-06, "loss": 0.2463, "step": 5604 }, { "epoch": 1.06, "grad_norm": 9.875, "learning_rate": 2.377498162711726e-06, "loss": 0.2103, "step": 5605 }, { "epoch": 1.06, "grad_norm": 9.9375, "learning_rate": 2.3767331453622073e-06, "loss": 0.2369, "step": 5606 }, { "epoch": 1.06, "grad_norm": 9.6875, "learning_rate": 2.3759681395833837e-06, "loss": 0.2555, "step": 5607 }, { "epoch": 1.06, "grad_norm": 8.4375, "learning_rate": 2.3752031454470636e-06, "loss": 0.1625, "step": 5608 }, { "epoch": 1.06, "grad_norm": 10.625, "learning_rate": 2.3744381630250537e-06, "loss": 0.2193, "step": 5609 }, { "epoch": 1.06, "grad_norm": 9.4375, "learning_rate": 2.3736731923891625e-06, "loss": 0.1764, "step": 5610 }, { "epoch": 1.06, "grad_norm": 17.875, "learning_rate": 2.3729082336111947e-06, "loss": 0.2683, "step": 5611 }, { "epoch": 1.06, "grad_norm": 14.4375, "learning_rate": 2.372143286762954e-06, "loss": 0.1831, "step": 5612 }, { "epoch": 1.06, "grad_norm": 9.0625, "learning_rate": 2.3713783519162442e-06, "loss": 0.2851, "step": 5613 }, { "epoch": 1.06, "grad_norm": 8.0, "learning_rate": 2.3706134291428683e-06, "loss": 0.2202, "step": 5614 }, { "epoch": 1.06, "grad_norm": 12.1875, "learning_rate": 2.3698485185146265e-06, "loss": 0.2404, "step": 5615 }, { "epoch": 1.06, "grad_norm": 10.0, "learning_rate": 2.3690836201033183e-06, "loss": 0.2553, "step": 5616 }, { "epoch": 1.06, "grad_norm": 9.9375, "learning_rate": 2.368318733980743e-06, "loss": 0.2015, "step": 5617 }, { "epoch": 1.06, "grad_norm": 17.125, "learning_rate": 2.3675538602186983e-06, "loss": 0.1778, "step": 5618 }, { "epoch": 1.06, "grad_norm": 10.0625, "learning_rate": 2.3667889988889815e-06, "loss": 0.1833, "step": 5619 }, { "epoch": 1.06, "grad_norm": 9.625, "learning_rate": 2.366024150063385e-06, "loss": 0.2107, "step": 5620 }, { "epoch": 1.06, "grad_norm": 10.0625, "learning_rate": 2.3652593138137057e-06, "loss": 0.1952, "step": 5621 }, { "epoch": 1.06, "grad_norm": 11.8125, "learning_rate": 2.3644944902117363e-06, "loss": 0.1503, "step": 5622 }, { "epoch": 1.06, "grad_norm": 10.75, "learning_rate": 2.3637296793292672e-06, "loss": 0.2223, "step": 5623 }, { "epoch": 1.06, "grad_norm": 10.9375, "learning_rate": 2.3629648812380887e-06, "loss": 0.209, "step": 5624 }, { "epoch": 1.06, "grad_norm": 9.5, "learning_rate": 2.3622000960099915e-06, "loss": 0.2111, "step": 5625 }, { "epoch": 1.06, "grad_norm": 11.0625, "learning_rate": 2.3614353237167642e-06, "loss": 0.223, "step": 5626 }, { "epoch": 1.06, "grad_norm": 17.5, "learning_rate": 2.3606705644301907e-06, "loss": 0.306, "step": 5627 }, { "epoch": 1.06, "grad_norm": 21.125, "learning_rate": 2.35990581822206e-06, "loss": 0.1949, "step": 5628 }, { "epoch": 1.06, "grad_norm": 10.125, "learning_rate": 2.3591410851641553e-06, "loss": 0.2235, "step": 5629 }, { "epoch": 1.07, "grad_norm": 11.3125, "learning_rate": 2.3583763653282603e-06, "loss": 0.2236, "step": 5630 }, { "epoch": 1.07, "grad_norm": 8.75, "learning_rate": 2.3576116587861544e-06, "loss": 0.2156, "step": 5631 }, { "epoch": 1.07, "grad_norm": 7.90625, "learning_rate": 2.356846965609622e-06, "loss": 0.1845, "step": 5632 }, { "epoch": 1.07, "grad_norm": 7.8125, "learning_rate": 2.356082285870441e-06, "loss": 0.1772, "step": 5633 }, { "epoch": 1.07, "grad_norm": 13.9375, "learning_rate": 2.3553176196403897e-06, "loss": 0.2751, "step": 5634 }, { "epoch": 1.07, "grad_norm": 15.875, "learning_rate": 2.354552966991244e-06, "loss": 0.2206, "step": 5635 }, { "epoch": 1.07, "grad_norm": 25.5, "learning_rate": 2.353788327994781e-06, "loss": 0.2609, "step": 5636 }, { "epoch": 1.07, "grad_norm": 9.0, "learning_rate": 2.353023702722775e-06, "loss": 0.1894, "step": 5637 }, { "epoch": 1.07, "grad_norm": 13.6875, "learning_rate": 2.3522590912469976e-06, "loss": 0.2164, "step": 5638 }, { "epoch": 1.07, "grad_norm": 14.875, "learning_rate": 2.3514944936392232e-06, "loss": 0.176, "step": 5639 }, { "epoch": 1.07, "grad_norm": 7.75, "learning_rate": 2.3507299099712204e-06, "loss": 0.1995, "step": 5640 }, { "epoch": 1.07, "grad_norm": 12.375, "learning_rate": 2.3499653403147587e-06, "loss": 0.2515, "step": 5641 }, { "epoch": 1.07, "grad_norm": 7.78125, "learning_rate": 2.3492007847416055e-06, "loss": 0.1954, "step": 5642 }, { "epoch": 1.07, "grad_norm": 11.5625, "learning_rate": 2.3484362433235294e-06, "loss": 0.1637, "step": 5643 }, { "epoch": 1.07, "grad_norm": 12.125, "learning_rate": 2.3476717161322937e-06, "loss": 0.1911, "step": 5644 }, { "epoch": 1.07, "grad_norm": 10.6875, "learning_rate": 2.346907203239663e-06, "loss": 0.2855, "step": 5645 }, { "epoch": 1.07, "grad_norm": 12.6875, "learning_rate": 2.3461427047173986e-06, "loss": 0.2747, "step": 5646 }, { "epoch": 1.07, "grad_norm": 11.875, "learning_rate": 2.345378220637264e-06, "loss": 0.2505, "step": 5647 }, { "epoch": 1.07, "grad_norm": 13.9375, "learning_rate": 2.344613751071018e-06, "loss": 0.2452, "step": 5648 }, { "epoch": 1.07, "grad_norm": 8.625, "learning_rate": 2.3438492960904182e-06, "loss": 0.1947, "step": 5649 }, { "epoch": 1.07, "grad_norm": 11.625, "learning_rate": 2.3430848557672232e-06, "loss": 0.2571, "step": 5650 }, { "epoch": 1.07, "grad_norm": 9.75, "learning_rate": 2.3423204301731884e-06, "loss": 0.2125, "step": 5651 }, { "epoch": 1.07, "grad_norm": 11.875, "learning_rate": 2.3415560193800675e-06, "loss": 0.219, "step": 5652 }, { "epoch": 1.07, "grad_norm": 18.0, "learning_rate": 2.3407916234596127e-06, "loss": 0.236, "step": 5653 }, { "epoch": 1.07, "grad_norm": 11.1875, "learning_rate": 2.340027242483578e-06, "loss": 0.2095, "step": 5654 }, { "epoch": 1.07, "grad_norm": 11.625, "learning_rate": 2.3392628765237122e-06, "loss": 0.2068, "step": 5655 }, { "epoch": 1.07, "grad_norm": 18.75, "learning_rate": 2.338498525651764e-06, "loss": 0.2211, "step": 5656 }, { "epoch": 1.07, "grad_norm": 10.9375, "learning_rate": 2.33773418993948e-06, "loss": 0.2296, "step": 5657 }, { "epoch": 1.07, "grad_norm": 12.1875, "learning_rate": 2.336969869458608e-06, "loss": 0.2113, "step": 5658 }, { "epoch": 1.07, "grad_norm": 8.3125, "learning_rate": 2.336205564280892e-06, "loss": 0.1445, "step": 5659 }, { "epoch": 1.07, "grad_norm": 10.875, "learning_rate": 2.335441274478073e-06, "loss": 0.2151, "step": 5660 }, { "epoch": 1.07, "grad_norm": 9.1875, "learning_rate": 2.3346770001218955e-06, "loss": 0.2176, "step": 5661 }, { "epoch": 1.07, "grad_norm": 10.25, "learning_rate": 2.333912741284098e-06, "loss": 0.2415, "step": 5662 }, { "epoch": 1.07, "grad_norm": 12.75, "learning_rate": 2.33314849803642e-06, "loss": 0.2769, "step": 5663 }, { "epoch": 1.07, "grad_norm": 11.875, "learning_rate": 2.332384270450597e-06, "loss": 0.2052, "step": 5664 }, { "epoch": 1.07, "grad_norm": 6.59375, "learning_rate": 2.3316200585983678e-06, "loss": 0.1509, "step": 5665 }, { "epoch": 1.07, "grad_norm": 9.0625, "learning_rate": 2.330855862551465e-06, "loss": 0.2575, "step": 5666 }, { "epoch": 1.07, "grad_norm": 11.625, "learning_rate": 2.33009168238162e-06, "loss": 0.2879, "step": 5667 }, { "epoch": 1.07, "grad_norm": 14.375, "learning_rate": 2.329327518160566e-06, "loss": 0.1757, "step": 5668 }, { "epoch": 1.07, "grad_norm": 13.5625, "learning_rate": 2.328563369960033e-06, "loss": 0.2407, "step": 5669 }, { "epoch": 1.07, "grad_norm": 9.6875, "learning_rate": 2.327799237851749e-06, "loss": 0.2117, "step": 5670 }, { "epoch": 1.07, "grad_norm": 10.875, "learning_rate": 2.327035121907439e-06, "loss": 0.2782, "step": 5671 }, { "epoch": 1.07, "grad_norm": 16.75, "learning_rate": 2.3262710221988314e-06, "loss": 0.2475, "step": 5672 }, { "epoch": 1.07, "grad_norm": 11.6875, "learning_rate": 2.325506938797648e-06, "loss": 0.2275, "step": 5673 }, { "epoch": 1.07, "grad_norm": 9.3125, "learning_rate": 2.3247428717756117e-06, "loss": 0.229, "step": 5674 }, { "epoch": 1.07, "grad_norm": 11.5, "learning_rate": 2.323978821204442e-06, "loss": 0.2652, "step": 5675 }, { "epoch": 1.07, "grad_norm": 11.5, "learning_rate": 2.32321478715586e-06, "loss": 0.2822, "step": 5676 }, { "epoch": 1.07, "grad_norm": 10.6875, "learning_rate": 2.3224507697015824e-06, "loss": 0.2035, "step": 5677 }, { "epoch": 1.07, "grad_norm": 11.625, "learning_rate": 2.321686768913325e-06, "loss": 0.2261, "step": 5678 }, { "epoch": 1.07, "grad_norm": 13.375, "learning_rate": 2.3209227848628023e-06, "loss": 0.2153, "step": 5679 }, { "epoch": 1.07, "grad_norm": 10.25, "learning_rate": 2.3201588176217275e-06, "loss": 0.2185, "step": 5680 }, { "epoch": 1.07, "grad_norm": 11.25, "learning_rate": 2.319394867261812e-06, "loss": 0.1957, "step": 5681 }, { "epoch": 1.07, "grad_norm": 9.1875, "learning_rate": 2.318630933854765e-06, "loss": 0.243, "step": 5682 }, { "epoch": 1.08, "grad_norm": 11.9375, "learning_rate": 2.317867017472296e-06, "loss": 0.2709, "step": 5683 }, { "epoch": 1.08, "grad_norm": 9.1875, "learning_rate": 2.31710311818611e-06, "loss": 0.245, "step": 5684 }, { "epoch": 1.08, "grad_norm": 10.125, "learning_rate": 2.3163392360679128e-06, "loss": 0.2185, "step": 5685 }, { "epoch": 1.08, "grad_norm": 9.125, "learning_rate": 2.315575371189407e-06, "loss": 0.2125, "step": 5686 }, { "epoch": 1.08, "grad_norm": 9.75, "learning_rate": 2.3148115236222963e-06, "loss": 0.2092, "step": 5687 }, { "epoch": 1.08, "grad_norm": 16.0, "learning_rate": 2.3140476934382785e-06, "loss": 0.2818, "step": 5688 }, { "epoch": 1.08, "grad_norm": 8.25, "learning_rate": 2.3132838807090533e-06, "loss": 0.1946, "step": 5689 }, { "epoch": 1.08, "grad_norm": 10.4375, "learning_rate": 2.3125200855063167e-06, "loss": 0.1989, "step": 5690 }, { "epoch": 1.08, "grad_norm": 9.75, "learning_rate": 2.3117563079017653e-06, "loss": 0.1859, "step": 5691 }, { "epoch": 1.08, "grad_norm": 9.9375, "learning_rate": 2.3109925479670916e-06, "loss": 0.2479, "step": 5692 }, { "epoch": 1.08, "grad_norm": 32.0, "learning_rate": 2.3102288057739866e-06, "loss": 0.2336, "step": 5693 }, { "epoch": 1.08, "grad_norm": 8.9375, "learning_rate": 2.309465081394142e-06, "loss": 0.1967, "step": 5694 }, { "epoch": 1.08, "grad_norm": 10.6875, "learning_rate": 2.308701374899247e-06, "loss": 0.2122, "step": 5695 }, { "epoch": 1.08, "grad_norm": 9.1875, "learning_rate": 2.307937686360987e-06, "loss": 0.207, "step": 5696 }, { "epoch": 1.08, "grad_norm": 9.8125, "learning_rate": 2.3071740158510466e-06, "loss": 0.1481, "step": 5697 }, { "epoch": 1.08, "grad_norm": 14.9375, "learning_rate": 2.3064103634411107e-06, "loss": 0.2281, "step": 5698 }, { "epoch": 1.08, "grad_norm": 10.75, "learning_rate": 2.305646729202862e-06, "loss": 0.217, "step": 5699 }, { "epoch": 1.08, "grad_norm": 21.875, "learning_rate": 2.3048831132079777e-06, "loss": 0.2454, "step": 5700 }, { "epoch": 1.08, "grad_norm": 11.125, "learning_rate": 2.3041195155281376e-06, "loss": 0.1665, "step": 5701 }, { "epoch": 1.08, "grad_norm": 11.5625, "learning_rate": 2.303355936235019e-06, "loss": 0.2231, "step": 5702 }, { "epoch": 1.08, "grad_norm": 20.5, "learning_rate": 2.302592375400297e-06, "loss": 0.2166, "step": 5703 }, { "epoch": 1.08, "grad_norm": 11.4375, "learning_rate": 2.3018288330956427e-06, "loss": 0.2329, "step": 5704 }, { "epoch": 1.08, "grad_norm": 10.1875, "learning_rate": 2.3010653093927295e-06, "loss": 0.2152, "step": 5705 }, { "epoch": 1.08, "grad_norm": 8.8125, "learning_rate": 2.3003018043632272e-06, "loss": 0.1485, "step": 5706 }, { "epoch": 1.08, "grad_norm": 8.625, "learning_rate": 2.299538318078803e-06, "loss": 0.1986, "step": 5707 }, { "epoch": 1.08, "grad_norm": 12.1875, "learning_rate": 2.298774850611122e-06, "loss": 0.2264, "step": 5708 }, { "epoch": 1.08, "grad_norm": 8.25, "learning_rate": 2.2980114020318505e-06, "loss": 0.2065, "step": 5709 }, { "epoch": 1.08, "grad_norm": 10.6875, "learning_rate": 2.2972479724126513e-06, "loss": 0.1844, "step": 5710 }, { "epoch": 1.08, "grad_norm": 12.875, "learning_rate": 2.2964845618251835e-06, "loss": 0.1984, "step": 5711 }, { "epoch": 1.08, "grad_norm": 11.75, "learning_rate": 2.295721170341107e-06, "loss": 0.232, "step": 5712 }, { "epoch": 1.08, "grad_norm": 13.3125, "learning_rate": 2.2949577980320793e-06, "loss": 0.2094, "step": 5713 }, { "epoch": 1.08, "grad_norm": 13.0625, "learning_rate": 2.294194444969757e-06, "loss": 0.2337, "step": 5714 }, { "epoch": 1.08, "grad_norm": 9.6875, "learning_rate": 2.293431111225791e-06, "loss": 0.2419, "step": 5715 }, { "epoch": 1.08, "grad_norm": 11.4375, "learning_rate": 2.292667796871836e-06, "loss": 0.1696, "step": 5716 }, { "epoch": 1.08, "grad_norm": 12.375, "learning_rate": 2.2919045019795404e-06, "loss": 0.2138, "step": 5717 }, { "epoch": 1.08, "grad_norm": 10.1875, "learning_rate": 2.291141226620554e-06, "loss": 0.2281, "step": 5718 }, { "epoch": 1.08, "grad_norm": 11.25, "learning_rate": 2.2903779708665203e-06, "loss": 0.2079, "step": 5719 }, { "epoch": 1.08, "grad_norm": 10.125, "learning_rate": 2.2896147347890865e-06, "loss": 0.2482, "step": 5720 }, { "epoch": 1.08, "grad_norm": 9.25, "learning_rate": 2.2888515184598948e-06, "loss": 0.2236, "step": 5721 }, { "epoch": 1.08, "grad_norm": 9.375, "learning_rate": 2.2880883219505857e-06, "loss": 0.1716, "step": 5722 }, { "epoch": 1.08, "grad_norm": 8.25, "learning_rate": 2.2873251453327973e-06, "loss": 0.2011, "step": 5723 }, { "epoch": 1.08, "grad_norm": 13.75, "learning_rate": 2.2865619886781685e-06, "loss": 0.182, "step": 5724 }, { "epoch": 1.08, "grad_norm": 10.875, "learning_rate": 2.2857988520583333e-06, "loss": 0.239, "step": 5725 }, { "epoch": 1.08, "grad_norm": 12.5, "learning_rate": 2.2850357355449254e-06, "loss": 0.1865, "step": 5726 }, { "epoch": 1.08, "grad_norm": 9.4375, "learning_rate": 2.284272639209577e-06, "loss": 0.219, "step": 5727 }, { "epoch": 1.08, "grad_norm": 8.8125, "learning_rate": 2.283509563123917e-06, "loss": 0.1939, "step": 5728 }, { "epoch": 1.08, "grad_norm": 11.0, "learning_rate": 2.2827465073595727e-06, "loss": 0.1497, "step": 5729 }, { "epoch": 1.08, "grad_norm": 9.4375, "learning_rate": 2.28198347198817e-06, "loss": 0.1822, "step": 5730 }, { "epoch": 1.08, "grad_norm": 14.375, "learning_rate": 2.2812204570813345e-06, "loss": 0.2097, "step": 5731 }, { "epoch": 1.08, "grad_norm": 12.9375, "learning_rate": 2.280457462710686e-06, "loss": 0.2534, "step": 5732 }, { "epoch": 1.08, "grad_norm": 8.4375, "learning_rate": 2.279694488947845e-06, "loss": 0.1633, "step": 5733 }, { "epoch": 1.08, "grad_norm": 14.0, "learning_rate": 2.2789315358644294e-06, "loss": 0.2046, "step": 5734 }, { "epoch": 1.08, "grad_norm": 9.25, "learning_rate": 2.278168603532057e-06, "loss": 0.2397, "step": 5735 }, { "epoch": 1.09, "grad_norm": 9.3125, "learning_rate": 2.2774056920223404e-06, "loss": 0.199, "step": 5736 }, { "epoch": 1.09, "grad_norm": 10.1875, "learning_rate": 2.276642801406892e-06, "loss": 0.2484, "step": 5737 }, { "epoch": 1.09, "grad_norm": 11.0625, "learning_rate": 2.2758799317573222e-06, "loss": 0.1907, "step": 5738 }, { "epoch": 1.09, "grad_norm": 8.1875, "learning_rate": 2.2751170831452406e-06, "loss": 0.1858, "step": 5739 }, { "epoch": 1.09, "grad_norm": 13.875, "learning_rate": 2.2743542556422517e-06, "loss": 0.2198, "step": 5740 }, { "epoch": 1.09, "grad_norm": 11.6875, "learning_rate": 2.2735914493199594e-06, "loss": 0.1885, "step": 5741 }, { "epoch": 1.09, "grad_norm": 20.0, "learning_rate": 2.2728286642499686e-06, "loss": 0.2404, "step": 5742 }, { "epoch": 1.09, "grad_norm": 10.0, "learning_rate": 2.2720659005038785e-06, "loss": 0.1852, "step": 5743 }, { "epoch": 1.09, "grad_norm": 10.9375, "learning_rate": 2.2713031581532867e-06, "loss": 0.2128, "step": 5744 }, { "epoch": 1.09, "grad_norm": 13.8125, "learning_rate": 2.2705404372697895e-06, "loss": 0.2147, "step": 5745 }, { "epoch": 1.09, "grad_norm": 9.75, "learning_rate": 2.2697777379249827e-06, "loss": 0.1892, "step": 5746 }, { "epoch": 1.09, "grad_norm": 9.8125, "learning_rate": 2.269015060190458e-06, "loss": 0.1867, "step": 5747 }, { "epoch": 1.09, "grad_norm": 10.875, "learning_rate": 2.2682524041378047e-06, "loss": 0.159, "step": 5748 }, { "epoch": 1.09, "grad_norm": 10.8125, "learning_rate": 2.2674897698386125e-06, "loss": 0.1795, "step": 5749 }, { "epoch": 1.09, "grad_norm": 13.375, "learning_rate": 2.2667271573644677e-06, "loss": 0.202, "step": 5750 }, { "epoch": 1.09, "grad_norm": 11.4375, "learning_rate": 2.265964566786953e-06, "loss": 0.264, "step": 5751 }, { "epoch": 1.09, "grad_norm": 10.8125, "learning_rate": 2.2652019981776505e-06, "loss": 0.1823, "step": 5752 }, { "epoch": 1.09, "grad_norm": 11.8125, "learning_rate": 2.264439451608142e-06, "loss": 0.2382, "step": 5753 }, { "epoch": 1.09, "grad_norm": 10.0625, "learning_rate": 2.2636769271500047e-06, "loss": 0.218, "step": 5754 }, { "epoch": 1.09, "grad_norm": 7.46875, "learning_rate": 2.262914424874814e-06, "loss": 0.1751, "step": 5755 }, { "epoch": 1.09, "grad_norm": 10.0, "learning_rate": 2.262151944854144e-06, "loss": 0.1853, "step": 5756 }, { "epoch": 1.09, "grad_norm": 14.0, "learning_rate": 2.2613894871595663e-06, "loss": 0.1937, "step": 5757 }, { "epoch": 1.09, "grad_norm": 9.0625, "learning_rate": 2.260627051862652e-06, "loss": 0.1567, "step": 5758 }, { "epoch": 1.09, "grad_norm": 13.0625, "learning_rate": 2.2598646390349667e-06, "loss": 0.2259, "step": 5759 }, { "epoch": 1.09, "grad_norm": 15.875, "learning_rate": 2.2591022487480756e-06, "loss": 0.2613, "step": 5760 }, { "epoch": 1.09, "grad_norm": 12.625, "learning_rate": 2.2583398810735442e-06, "loss": 0.1809, "step": 5761 }, { "epoch": 1.09, "grad_norm": 8.1875, "learning_rate": 2.2575775360829326e-06, "loss": 0.1932, "step": 5762 }, { "epoch": 1.09, "grad_norm": 12.8125, "learning_rate": 2.2568152138477984e-06, "loss": 0.2422, "step": 5763 }, { "epoch": 1.09, "grad_norm": 16.25, "learning_rate": 2.2560529144397007e-06, "loss": 0.2521, "step": 5764 }, { "epoch": 1.09, "grad_norm": 15.875, "learning_rate": 2.255290637930194e-06, "loss": 0.2229, "step": 5765 }, { "epoch": 1.09, "grad_norm": 10.25, "learning_rate": 2.25452838439083e-06, "loss": 0.1783, "step": 5766 }, { "epoch": 1.09, "grad_norm": 12.625, "learning_rate": 2.253766153893159e-06, "loss": 0.214, "step": 5767 }, { "epoch": 1.09, "grad_norm": 12.875, "learning_rate": 2.25300394650873e-06, "loss": 0.2589, "step": 5768 }, { "epoch": 1.09, "grad_norm": 8.4375, "learning_rate": 2.2522417623090894e-06, "loss": 0.1774, "step": 5769 }, { "epoch": 1.09, "grad_norm": 12.5, "learning_rate": 2.2514796013657814e-06, "loss": 0.2225, "step": 5770 }, { "epoch": 1.09, "grad_norm": 9.1875, "learning_rate": 2.250717463750346e-06, "loss": 0.2084, "step": 5771 }, { "epoch": 1.09, "grad_norm": 12.9375, "learning_rate": 2.2499553495343247e-06, "loss": 0.2705, "step": 5772 }, { "epoch": 1.09, "grad_norm": 14.875, "learning_rate": 2.249193258789254e-06, "loss": 0.2291, "step": 5773 }, { "epoch": 1.09, "grad_norm": 14.125, "learning_rate": 2.248431191586669e-06, "loss": 0.2617, "step": 5774 }, { "epoch": 1.09, "grad_norm": 15.75, "learning_rate": 2.247669147998104e-06, "loss": 0.2591, "step": 5775 }, { "epoch": 1.09, "grad_norm": 10.3125, "learning_rate": 2.2469071280950884e-06, "loss": 0.1904, "step": 5776 }, { "epoch": 1.09, "grad_norm": 9.0625, "learning_rate": 2.246145131949151e-06, "loss": 0.2023, "step": 5777 }, { "epoch": 1.09, "grad_norm": 8.9375, "learning_rate": 2.2453831596318173e-06, "loss": 0.2194, "step": 5778 }, { "epoch": 1.09, "grad_norm": 11.3125, "learning_rate": 2.2446212112146136e-06, "loss": 0.2, "step": 5779 }, { "epoch": 1.09, "grad_norm": 9.8125, "learning_rate": 2.2438592867690603e-06, "loss": 0.2019, "step": 5780 }, { "epoch": 1.09, "grad_norm": 10.875, "learning_rate": 2.243097386366678e-06, "loss": 0.2452, "step": 5781 }, { "epoch": 1.09, "grad_norm": 10.25, "learning_rate": 2.2423355100789807e-06, "loss": 0.2138, "step": 5782 }, { "epoch": 1.09, "grad_norm": 9.4375, "learning_rate": 2.241573657977488e-06, "loss": 0.198, "step": 5783 }, { "epoch": 1.09, "grad_norm": 8.625, "learning_rate": 2.2408118301337102e-06, "loss": 0.236, "step": 5784 }, { "epoch": 1.09, "grad_norm": 11.625, "learning_rate": 2.2400500266191575e-06, "loss": 0.1766, "step": 5785 }, { "epoch": 1.09, "grad_norm": 17.875, "learning_rate": 2.23928824750534e-06, "loss": 0.214, "step": 5786 }, { "epoch": 1.09, "grad_norm": 9.8125, "learning_rate": 2.2385264928637625e-06, "loss": 0.1761, "step": 5787 }, { "epoch": 1.09, "grad_norm": 10.625, "learning_rate": 2.237764762765929e-06, "loss": 0.2157, "step": 5788 }, { "epoch": 1.1, "grad_norm": 15.0625, "learning_rate": 2.237003057283339e-06, "loss": 0.2819, "step": 5789 }, { "epoch": 1.1, "grad_norm": 11.4375, "learning_rate": 2.236241376487495e-06, "loss": 0.218, "step": 5790 }, { "epoch": 1.1, "grad_norm": 8.6875, "learning_rate": 2.235479720449891e-06, "loss": 0.2112, "step": 5791 }, { "epoch": 1.1, "grad_norm": 10.375, "learning_rate": 2.2347180892420226e-06, "loss": 0.1719, "step": 5792 }, { "epoch": 1.1, "grad_norm": 9.3125, "learning_rate": 2.2339564829353806e-06, "loss": 0.1433, "step": 5793 }, { "epoch": 1.1, "grad_norm": 13.5, "learning_rate": 2.233194901601457e-06, "loss": 0.2232, "step": 5794 }, { "epoch": 1.1, "grad_norm": 13.4375, "learning_rate": 2.232433345311737e-06, "loss": 0.2091, "step": 5795 }, { "epoch": 1.1, "grad_norm": 10.75, "learning_rate": 2.231671814137706e-06, "loss": 0.1675, "step": 5796 }, { "epoch": 1.1, "grad_norm": 12.875, "learning_rate": 2.2309103081508475e-06, "loss": 0.2234, "step": 5797 }, { "epoch": 1.1, "grad_norm": 12.875, "learning_rate": 2.2301488274226426e-06, "loss": 0.2696, "step": 5798 }, { "epoch": 1.1, "grad_norm": 10.9375, "learning_rate": 2.229387372024567e-06, "loss": 0.248, "step": 5799 }, { "epoch": 1.1, "grad_norm": 11.625, "learning_rate": 2.228625942028097e-06, "loss": 0.2292, "step": 5800 }, { "epoch": 1.1, "grad_norm": 9.375, "learning_rate": 2.2278645375047066e-06, "loss": 0.222, "step": 5801 }, { "epoch": 1.1, "grad_norm": 12.75, "learning_rate": 2.227103158525867e-06, "loss": 0.2769, "step": 5802 }, { "epoch": 1.1, "grad_norm": 10.8125, "learning_rate": 2.2263418051630446e-06, "loss": 0.1938, "step": 5803 }, { "epoch": 1.1, "grad_norm": 10.0625, "learning_rate": 2.225580477487706e-06, "loss": 0.1553, "step": 5804 }, { "epoch": 1.1, "grad_norm": 10.375, "learning_rate": 2.224819175571316e-06, "loss": 0.1619, "step": 5805 }, { "epoch": 1.1, "grad_norm": 8.125, "learning_rate": 2.2240578994853358e-06, "loss": 0.1587, "step": 5806 }, { "epoch": 1.1, "grad_norm": 9.125, "learning_rate": 2.223296649301222e-06, "loss": 0.1947, "step": 5807 }, { "epoch": 1.1, "grad_norm": 13.0625, "learning_rate": 2.2225354250904326e-06, "loss": 0.1719, "step": 5808 }, { "epoch": 1.1, "grad_norm": 12.8125, "learning_rate": 2.221774226924421e-06, "loss": 0.2044, "step": 5809 }, { "epoch": 1.1, "grad_norm": 21.375, "learning_rate": 2.22101305487464e-06, "loss": 0.2278, "step": 5810 }, { "epoch": 1.1, "grad_norm": 16.625, "learning_rate": 2.2202519090125358e-06, "loss": 0.2253, "step": 5811 }, { "epoch": 1.1, "grad_norm": 10.125, "learning_rate": 2.219490789409557e-06, "loss": 0.2359, "step": 5812 }, { "epoch": 1.1, "grad_norm": 18.0, "learning_rate": 2.218729696137147e-06, "loss": 0.2296, "step": 5813 }, { "epoch": 1.1, "grad_norm": 9.3125, "learning_rate": 2.217968629266748e-06, "loss": 0.222, "step": 5814 }, { "epoch": 1.1, "grad_norm": 14.75, "learning_rate": 2.2172075888697975e-06, "loss": 0.208, "step": 5815 }, { "epoch": 1.1, "grad_norm": 10.1875, "learning_rate": 2.2164465750177336e-06, "loss": 0.2313, "step": 5816 }, { "epoch": 1.1, "grad_norm": 13.8125, "learning_rate": 2.21568558778199e-06, "loss": 0.2332, "step": 5817 }, { "epoch": 1.1, "grad_norm": 10.875, "learning_rate": 2.2149246272339973e-06, "loss": 0.2075, "step": 5818 }, { "epoch": 1.1, "grad_norm": 10.75, "learning_rate": 2.214163693445187e-06, "loss": 0.2271, "step": 5819 }, { "epoch": 1.1, "grad_norm": 10.875, "learning_rate": 2.213402786486984e-06, "loss": 0.1628, "step": 5820 }, { "epoch": 1.1, "grad_norm": 12.5, "learning_rate": 2.2126419064308128e-06, "loss": 0.2672, "step": 5821 }, { "epoch": 1.1, "grad_norm": 8.6875, "learning_rate": 2.2118810533480934e-06, "loss": 0.1973, "step": 5822 }, { "epoch": 1.1, "grad_norm": 11.5, "learning_rate": 2.211120227310248e-06, "loss": 0.2261, "step": 5823 }, { "epoch": 1.1, "grad_norm": 8.125, "learning_rate": 2.21035942838869e-06, "loss": 0.2159, "step": 5824 }, { "epoch": 1.1, "grad_norm": 8.1875, "learning_rate": 2.2095986566548363e-06, "loss": 0.1782, "step": 5825 }, { "epoch": 1.1, "grad_norm": 9.0, "learning_rate": 2.2088379121800943e-06, "loss": 0.2038, "step": 5826 }, { "epoch": 1.1, "grad_norm": 11.0, "learning_rate": 2.208077195035877e-06, "loss": 0.206, "step": 5827 }, { "epoch": 1.1, "grad_norm": 12.5, "learning_rate": 2.2073165052935885e-06, "loss": 0.2279, "step": 5828 }, { "epoch": 1.1, "grad_norm": 11.125, "learning_rate": 2.206555843024632e-06, "loss": 0.1864, "step": 5829 }, { "epoch": 1.1, "grad_norm": 12.1875, "learning_rate": 2.20579520830041e-06, "loss": 0.2381, "step": 5830 }, { "epoch": 1.1, "grad_norm": 16.625, "learning_rate": 2.205034601192321e-06, "loss": 0.2565, "step": 5831 }, { "epoch": 1.1, "grad_norm": 9.3125, "learning_rate": 2.20427402177176e-06, "loss": 0.2208, "step": 5832 }, { "epoch": 1.1, "grad_norm": 12.6875, "learning_rate": 2.2035134701101203e-06, "loss": 0.201, "step": 5833 }, { "epoch": 1.1, "grad_norm": 9.25, "learning_rate": 2.2027529462787942e-06, "loss": 0.2148, "step": 5834 }, { "epoch": 1.1, "grad_norm": 8.3125, "learning_rate": 2.201992450349168e-06, "loss": 0.1394, "step": 5835 }, { "epoch": 1.1, "grad_norm": 13.5, "learning_rate": 2.201231982392628e-06, "loss": 0.2778, "step": 5836 }, { "epoch": 1.1, "grad_norm": 9.6875, "learning_rate": 2.2004715424805564e-06, "loss": 0.2014, "step": 5837 }, { "epoch": 1.1, "grad_norm": 7.625, "learning_rate": 2.1997111306843355e-06, "loss": 0.1718, "step": 5838 }, { "epoch": 1.1, "grad_norm": 9.9375, "learning_rate": 2.1989507470753404e-06, "loss": 0.1924, "step": 5839 }, { "epoch": 1.1, "grad_norm": 11.0625, "learning_rate": 2.1981903917249467e-06, "loss": 0.2477, "step": 5840 }, { "epoch": 1.1, "grad_norm": 10.3125, "learning_rate": 2.197430064704528e-06, "loss": 0.2188, "step": 5841 }, { "epoch": 1.11, "grad_norm": 16.0, "learning_rate": 2.1966697660854532e-06, "loss": 0.214, "step": 5842 }, { "epoch": 1.11, "grad_norm": 13.5625, "learning_rate": 2.1959094959390883e-06, "loss": 0.2685, "step": 5843 }, { "epoch": 1.11, "grad_norm": 12.75, "learning_rate": 2.195149254336798e-06, "loss": 0.2945, "step": 5844 }, { "epoch": 1.11, "grad_norm": 10.5625, "learning_rate": 2.1943890413499454e-06, "loss": 0.2857, "step": 5845 }, { "epoch": 1.11, "grad_norm": 14.875, "learning_rate": 2.1936288570498886e-06, "loss": 0.2774, "step": 5846 }, { "epoch": 1.11, "grad_norm": 20.625, "learning_rate": 2.1928687015079835e-06, "loss": 0.2133, "step": 5847 }, { "epoch": 1.11, "grad_norm": 13.0625, "learning_rate": 2.1921085747955826e-06, "loss": 0.2038, "step": 5848 }, { "epoch": 1.11, "grad_norm": 14.25, "learning_rate": 2.1913484769840386e-06, "loss": 0.2645, "step": 5849 }, { "epoch": 1.11, "grad_norm": 10.5, "learning_rate": 2.1905884081447e-06, "loss": 0.2163, "step": 5850 }, { "epoch": 1.11, "grad_norm": 11.8125, "learning_rate": 2.1898283683489096e-06, "loss": 0.243, "step": 5851 }, { "epoch": 1.11, "grad_norm": 11.75, "learning_rate": 2.1890683576680128e-06, "loss": 0.1603, "step": 5852 }, { "epoch": 1.11, "grad_norm": 16.5, "learning_rate": 2.188308376173348e-06, "loss": 0.2572, "step": 5853 }, { "epoch": 1.11, "grad_norm": 14.875, "learning_rate": 2.1875484239362533e-06, "loss": 0.2214, "step": 5854 }, { "epoch": 1.11, "grad_norm": 9.125, "learning_rate": 2.186788501028062e-06, "loss": 0.1878, "step": 5855 }, { "epoch": 1.11, "grad_norm": 13.5625, "learning_rate": 2.1860286075201066e-06, "loss": 0.2019, "step": 5856 }, { "epoch": 1.11, "grad_norm": 11.9375, "learning_rate": 2.1852687434837162e-06, "loss": 0.2272, "step": 5857 }, { "epoch": 1.11, "grad_norm": 11.125, "learning_rate": 2.1845089089902172e-06, "loss": 0.1876, "step": 5858 }, { "epoch": 1.11, "grad_norm": 11.875, "learning_rate": 2.1837491041109314e-06, "loss": 0.2574, "step": 5859 }, { "epoch": 1.11, "grad_norm": 12.3125, "learning_rate": 2.182989328917181e-06, "loss": 0.1915, "step": 5860 }, { "epoch": 1.11, "grad_norm": 11.375, "learning_rate": 2.182229583480284e-06, "loss": 0.1573, "step": 5861 }, { "epoch": 1.11, "grad_norm": 9.125, "learning_rate": 2.181469867871554e-06, "loss": 0.2295, "step": 5862 }, { "epoch": 1.11, "grad_norm": 9.5, "learning_rate": 2.180710182162305e-06, "loss": 0.1643, "step": 5863 }, { "epoch": 1.11, "grad_norm": 11.3125, "learning_rate": 2.1799505264238457e-06, "loss": 0.26, "step": 5864 }, { "epoch": 1.11, "grad_norm": 11.3125, "learning_rate": 2.179190900727483e-06, "loss": 0.1984, "step": 5865 }, { "epoch": 1.11, "grad_norm": 11.5625, "learning_rate": 2.1784313051445184e-06, "loss": 0.2306, "step": 5866 }, { "epoch": 1.11, "grad_norm": 9.375, "learning_rate": 2.1776717397462564e-06, "loss": 0.2734, "step": 5867 }, { "epoch": 1.11, "grad_norm": 11.625, "learning_rate": 2.176912204603994e-06, "loss": 0.2117, "step": 5868 }, { "epoch": 1.11, "grad_norm": 18.625, "learning_rate": 2.1761526997890252e-06, "loss": 0.1942, "step": 5869 }, { "epoch": 1.11, "grad_norm": 8.4375, "learning_rate": 2.1753932253726427e-06, "loss": 0.1724, "step": 5870 }, { "epoch": 1.11, "grad_norm": 8.625, "learning_rate": 2.174633781426138e-06, "loss": 0.2312, "step": 5871 }, { "epoch": 1.11, "grad_norm": 8.5, "learning_rate": 2.173874368020796e-06, "loss": 0.1923, "step": 5872 }, { "epoch": 1.11, "grad_norm": 11.75, "learning_rate": 2.1731149852279015e-06, "loss": 0.2261, "step": 5873 }, { "epoch": 1.11, "grad_norm": 16.125, "learning_rate": 2.172355633118735e-06, "loss": 0.2479, "step": 5874 }, { "epoch": 1.11, "grad_norm": 9.8125, "learning_rate": 2.171596311764576e-06, "loss": 0.238, "step": 5875 }, { "epoch": 1.11, "grad_norm": 9.5625, "learning_rate": 2.1708370212366974e-06, "loss": 0.2025, "step": 5876 }, { "epoch": 1.11, "grad_norm": 13.9375, "learning_rate": 2.170077761606373e-06, "loss": 0.2618, "step": 5877 }, { "epoch": 1.11, "grad_norm": 9.4375, "learning_rate": 2.1693185329448725e-06, "loss": 0.1426, "step": 5878 }, { "epoch": 1.11, "grad_norm": 8.75, "learning_rate": 2.1685593353234614e-06, "loss": 0.1955, "step": 5879 }, { "epoch": 1.11, "grad_norm": 11.9375, "learning_rate": 2.1678001688134045e-06, "loss": 0.2072, "step": 5880 }, { "epoch": 1.11, "grad_norm": 9.125, "learning_rate": 2.167041033485961e-06, "loss": 0.2026, "step": 5881 }, { "epoch": 1.11, "grad_norm": 9.8125, "learning_rate": 2.1662819294123906e-06, "loss": 0.2292, "step": 5882 }, { "epoch": 1.11, "grad_norm": 8.5, "learning_rate": 2.165522856663947e-06, "loss": 0.173, "step": 5883 }, { "epoch": 1.11, "grad_norm": 10.9375, "learning_rate": 2.164763815311881e-06, "loss": 0.1785, "step": 5884 }, { "epoch": 1.11, "grad_norm": 16.875, "learning_rate": 2.164004805427444e-06, "loss": 0.2012, "step": 5885 }, { "epoch": 1.11, "grad_norm": 10.75, "learning_rate": 2.163245827081881e-06, "loss": 0.1771, "step": 5886 }, { "epoch": 1.11, "grad_norm": 9.125, "learning_rate": 2.162486880346435e-06, "loss": 0.1744, "step": 5887 }, { "epoch": 1.11, "grad_norm": 15.0, "learning_rate": 2.1617279652923446e-06, "loss": 0.2335, "step": 5888 }, { "epoch": 1.11, "grad_norm": 11.0, "learning_rate": 2.1609690819908495e-06, "loss": 0.1872, "step": 5889 }, { "epoch": 1.11, "grad_norm": 14.3125, "learning_rate": 2.1602102305131827e-06, "loss": 0.2278, "step": 5890 }, { "epoch": 1.11, "grad_norm": 10.1875, "learning_rate": 2.159451410930575e-06, "loss": 0.2049, "step": 5891 }, { "epoch": 1.11, "grad_norm": 17.125, "learning_rate": 2.158692623314254e-06, "loss": 0.2199, "step": 5892 }, { "epoch": 1.11, "grad_norm": 18.375, "learning_rate": 2.1579338677354464e-06, "loss": 0.246, "step": 5893 }, { "epoch": 1.11, "grad_norm": 9.0, "learning_rate": 2.157175144265374e-06, "loss": 0.1904, "step": 5894 }, { "epoch": 1.12, "grad_norm": 11.8125, "learning_rate": 2.1564164529752544e-06, "loss": 0.2102, "step": 5895 }, { "epoch": 1.12, "grad_norm": 9.5, "learning_rate": 2.1556577939363056e-06, "loss": 0.2665, "step": 5896 }, { "epoch": 1.12, "grad_norm": 9.1875, "learning_rate": 2.1548991672197394e-06, "loss": 0.175, "step": 5897 }, { "epoch": 1.12, "grad_norm": 9.6875, "learning_rate": 2.1541405728967674e-06, "loss": 0.2128, "step": 5898 }, { "epoch": 1.12, "grad_norm": 13.125, "learning_rate": 2.1533820110385943e-06, "loss": 0.1815, "step": 5899 }, { "epoch": 1.12, "grad_norm": 9.8125, "learning_rate": 2.1526234817164254e-06, "loss": 0.1875, "step": 5900 }, { "epoch": 1.12, "grad_norm": 10.5625, "learning_rate": 2.151864985001462e-06, "loss": 0.1846, "step": 5901 }, { "epoch": 1.12, "grad_norm": 9.5, "learning_rate": 2.151106520964902e-06, "loss": 0.1859, "step": 5902 }, { "epoch": 1.12, "grad_norm": 9.875, "learning_rate": 2.1503480896779383e-06, "loss": 0.193, "step": 5903 }, { "epoch": 1.12, "grad_norm": 7.3125, "learning_rate": 2.1495896912117643e-06, "loss": 0.1735, "step": 5904 }, { "epoch": 1.12, "grad_norm": 13.5, "learning_rate": 2.148831325637569e-06, "loss": 0.2867, "step": 5905 }, { "epoch": 1.12, "grad_norm": 12.125, "learning_rate": 2.1480729930265354e-06, "loss": 0.1548, "step": 5906 }, { "epoch": 1.12, "grad_norm": 7.875, "learning_rate": 2.1473146934498495e-06, "loss": 0.1755, "step": 5907 }, { "epoch": 1.12, "grad_norm": 13.0, "learning_rate": 2.146556426978688e-06, "loss": 0.2216, "step": 5908 }, { "epoch": 1.12, "grad_norm": 8.5, "learning_rate": 2.1457981936842285e-06, "loss": 0.2049, "step": 5909 }, { "epoch": 1.12, "grad_norm": 11.0, "learning_rate": 2.1450399936376425e-06, "loss": 0.2116, "step": 5910 }, { "epoch": 1.12, "grad_norm": 10.5, "learning_rate": 2.144281826910102e-06, "loss": 0.2266, "step": 5911 }, { "epoch": 1.12, "grad_norm": 14.125, "learning_rate": 2.1435236935727728e-06, "loss": 0.2146, "step": 5912 }, { "epoch": 1.12, "grad_norm": 13.25, "learning_rate": 2.1427655936968193e-06, "loss": 0.2285, "step": 5913 }, { "epoch": 1.12, "grad_norm": 9.3125, "learning_rate": 2.1420075273534e-06, "loss": 0.1721, "step": 5914 }, { "epoch": 1.12, "grad_norm": 13.0625, "learning_rate": 2.141249494613675e-06, "loss": 0.2347, "step": 5915 }, { "epoch": 1.12, "grad_norm": 13.375, "learning_rate": 2.140491495548797e-06, "loss": 0.1906, "step": 5916 }, { "epoch": 1.12, "grad_norm": 12.4375, "learning_rate": 2.1397335302299173e-06, "loss": 0.1656, "step": 5917 }, { "epoch": 1.12, "grad_norm": 10.0, "learning_rate": 2.1389755987281853e-06, "loss": 0.1796, "step": 5918 }, { "epoch": 1.12, "grad_norm": 33.25, "learning_rate": 2.138217701114744e-06, "loss": 0.2079, "step": 5919 }, { "epoch": 1.12, "grad_norm": 11.25, "learning_rate": 2.1374598374607357e-06, "loss": 0.1908, "step": 5920 }, { "epoch": 1.12, "grad_norm": 10.5625, "learning_rate": 2.1367020078372984e-06, "loss": 0.234, "step": 5921 }, { "epoch": 1.12, "grad_norm": 12.875, "learning_rate": 2.1359442123155685e-06, "loss": 0.2189, "step": 5922 }, { "epoch": 1.12, "grad_norm": 10.8125, "learning_rate": 2.135186450966677e-06, "loss": 0.2271, "step": 5923 }, { "epoch": 1.12, "grad_norm": 12.1875, "learning_rate": 2.1344287238617522e-06, "loss": 0.244, "step": 5924 }, { "epoch": 1.12, "grad_norm": 11.625, "learning_rate": 2.1336710310719204e-06, "loss": 0.2401, "step": 5925 }, { "epoch": 1.12, "grad_norm": 10.125, "learning_rate": 2.1329133726683054e-06, "loss": 0.1984, "step": 5926 }, { "epoch": 1.12, "grad_norm": 12.1875, "learning_rate": 2.132155748722024e-06, "loss": 0.2514, "step": 5927 }, { "epoch": 1.12, "grad_norm": 12.3125, "learning_rate": 2.1313981593041922e-06, "loss": 0.1646, "step": 5928 }, { "epoch": 1.12, "grad_norm": 12.75, "learning_rate": 2.130640604485925e-06, "loss": 0.1753, "step": 5929 }, { "epoch": 1.12, "grad_norm": 8.0625, "learning_rate": 2.12988308433833e-06, "loss": 0.2218, "step": 5930 }, { "epoch": 1.12, "grad_norm": 12.5625, "learning_rate": 2.129125598932514e-06, "loss": 0.1589, "step": 5931 }, { "epoch": 1.12, "grad_norm": 13.5, "learning_rate": 2.1283681483395784e-06, "loss": 0.2138, "step": 5932 }, { "epoch": 1.12, "grad_norm": 13.0, "learning_rate": 2.127610732630625e-06, "loss": 0.1876, "step": 5933 }, { "epoch": 1.12, "grad_norm": 14.6875, "learning_rate": 2.1268533518767496e-06, "loss": 0.2516, "step": 5934 }, { "epoch": 1.12, "grad_norm": 12.125, "learning_rate": 2.1260960061490447e-06, "loss": 0.2334, "step": 5935 }, { "epoch": 1.12, "grad_norm": 8.75, "learning_rate": 2.1253386955186e-06, "loss": 0.1754, "step": 5936 }, { "epoch": 1.12, "grad_norm": 9.9375, "learning_rate": 2.1245814200565025e-06, "loss": 0.2236, "step": 5937 }, { "epoch": 1.12, "grad_norm": 16.375, "learning_rate": 2.123824179833836e-06, "loss": 0.2096, "step": 5938 }, { "epoch": 1.12, "grad_norm": 11.875, "learning_rate": 2.1230669749216785e-06, "loss": 0.1504, "step": 5939 }, { "epoch": 1.12, "grad_norm": 12.125, "learning_rate": 2.122309805391108e-06, "loss": 0.2945, "step": 5940 }, { "epoch": 1.12, "grad_norm": 11.0625, "learning_rate": 2.1215526713131977e-06, "loss": 0.2134, "step": 5941 }, { "epoch": 1.12, "grad_norm": 10.5, "learning_rate": 2.1207955727590183e-06, "loss": 0.2032, "step": 5942 }, { "epoch": 1.12, "grad_norm": 13.6875, "learning_rate": 2.120038509799634e-06, "loss": 0.2028, "step": 5943 }, { "epoch": 1.12, "grad_norm": 10.3125, "learning_rate": 2.11928148250611e-06, "loss": 0.1805, "step": 5944 }, { "epoch": 1.12, "grad_norm": 16.0, "learning_rate": 2.118524490949506e-06, "loss": 0.2554, "step": 5945 }, { "epoch": 1.12, "grad_norm": 10.6875, "learning_rate": 2.1177675352008785e-06, "loss": 0.186, "step": 5946 }, { "epoch": 1.13, "grad_norm": 12.5, "learning_rate": 2.1170106153312795e-06, "loss": 0.1741, "step": 5947 }, { "epoch": 1.13, "grad_norm": 15.125, "learning_rate": 2.1162537314117604e-06, "loss": 0.1907, "step": 5948 }, { "epoch": 1.13, "grad_norm": 9.0, "learning_rate": 2.1154968835133677e-06, "loss": 0.2027, "step": 5949 }, { "epoch": 1.13, "grad_norm": 14.0625, "learning_rate": 2.1147400717071417e-06, "loss": 0.2342, "step": 5950 }, { "epoch": 1.13, "grad_norm": 17.625, "learning_rate": 2.1139832960641263e-06, "loss": 0.2433, "step": 5951 }, { "epoch": 1.13, "grad_norm": 12.0625, "learning_rate": 2.113226556655355e-06, "loss": 0.2052, "step": 5952 }, { "epoch": 1.13, "grad_norm": 6.3125, "learning_rate": 2.1124698535518618e-06, "loss": 0.1349, "step": 5953 }, { "epoch": 1.13, "grad_norm": 9.125, "learning_rate": 2.1117131868246745e-06, "loss": 0.1736, "step": 5954 }, { "epoch": 1.13, "grad_norm": 15.5, "learning_rate": 2.1109565565448215e-06, "loss": 0.262, "step": 5955 }, { "epoch": 1.13, "grad_norm": 10.0625, "learning_rate": 2.110199962783324e-06, "loss": 0.1766, "step": 5956 }, { "epoch": 1.13, "grad_norm": 7.6875, "learning_rate": 2.109443405611202e-06, "loss": 0.1808, "step": 5957 }, { "epoch": 1.13, "grad_norm": 11.0625, "learning_rate": 2.1086868850994696e-06, "loss": 0.2561, "step": 5958 }, { "epoch": 1.13, "grad_norm": 12.125, "learning_rate": 2.107930401319142e-06, "loss": 0.264, "step": 5959 }, { "epoch": 1.13, "grad_norm": 13.125, "learning_rate": 2.1071739543412256e-06, "loss": 0.2205, "step": 5960 }, { "epoch": 1.13, "grad_norm": 18.5, "learning_rate": 2.1064175442367275e-06, "loss": 0.1845, "step": 5961 }, { "epoch": 1.13, "grad_norm": 8.75, "learning_rate": 2.1056611710766475e-06, "loss": 0.2323, "step": 5962 }, { "epoch": 1.13, "grad_norm": 12.375, "learning_rate": 2.104904834931986e-06, "loss": 0.22, "step": 5963 }, { "epoch": 1.13, "grad_norm": 8.75, "learning_rate": 2.104148535873738e-06, "loss": 0.1741, "step": 5964 }, { "epoch": 1.13, "grad_norm": 14.8125, "learning_rate": 2.103392273972894e-06, "loss": 0.2473, "step": 5965 }, { "epoch": 1.13, "grad_norm": 14.9375, "learning_rate": 2.1026360493004436e-06, "loss": 0.193, "step": 5966 }, { "epoch": 1.13, "grad_norm": 7.71875, "learning_rate": 2.10187986192737e-06, "loss": 0.2068, "step": 5967 }, { "epoch": 1.13, "grad_norm": 15.5625, "learning_rate": 2.101123711924655e-06, "loss": 0.3092, "step": 5968 }, { "epoch": 1.13, "grad_norm": 9.5, "learning_rate": 2.1003675993632744e-06, "loss": 0.1801, "step": 5969 }, { "epoch": 1.13, "grad_norm": 13.125, "learning_rate": 2.0996115243142052e-06, "loss": 0.2988, "step": 5970 }, { "epoch": 1.13, "grad_norm": 10.875, "learning_rate": 2.098855486848416e-06, "loss": 0.2057, "step": 5971 }, { "epoch": 1.13, "grad_norm": 10.1875, "learning_rate": 2.0980994870368745e-06, "loss": 0.1755, "step": 5972 }, { "epoch": 1.13, "grad_norm": 11.5, "learning_rate": 2.097343524950543e-06, "loss": 0.2132, "step": 5973 }, { "epoch": 1.13, "grad_norm": 8.3125, "learning_rate": 2.0965876006603836e-06, "loss": 0.2045, "step": 5974 }, { "epoch": 1.13, "grad_norm": 10.25, "learning_rate": 2.095831714237351e-06, "loss": 0.2311, "step": 5975 }, { "epoch": 1.13, "grad_norm": 9.0, "learning_rate": 2.0950758657523974e-06, "loss": 0.1961, "step": 5976 }, { "epoch": 1.13, "grad_norm": 14.875, "learning_rate": 2.0943200552764737e-06, "loss": 0.2022, "step": 5977 }, { "epoch": 1.13, "grad_norm": 11.4375, "learning_rate": 2.093564282880526e-06, "loss": 0.1675, "step": 5978 }, { "epoch": 1.13, "grad_norm": 13.5, "learning_rate": 2.0928085486354943e-06, "loss": 0.1734, "step": 5979 }, { "epoch": 1.13, "grad_norm": 9.375, "learning_rate": 2.0920528526123175e-06, "loss": 0.2912, "step": 5980 }, { "epoch": 1.13, "grad_norm": 14.4375, "learning_rate": 2.0912971948819324e-06, "loss": 0.2221, "step": 5981 }, { "epoch": 1.13, "grad_norm": 13.875, "learning_rate": 2.0905415755152695e-06, "loss": 0.2356, "step": 5982 }, { "epoch": 1.13, "grad_norm": 18.625, "learning_rate": 2.0897859945832556e-06, "loss": 0.2446, "step": 5983 }, { "epoch": 1.13, "grad_norm": 11.9375, "learning_rate": 2.089030452156815e-06, "loss": 0.2243, "step": 5984 }, { "epoch": 1.13, "grad_norm": 10.375, "learning_rate": 2.0882749483068696e-06, "loss": 0.1915, "step": 5985 }, { "epoch": 1.13, "grad_norm": 8.3125, "learning_rate": 2.087519483104336e-06, "loss": 0.1781, "step": 5986 }, { "epoch": 1.13, "grad_norm": 10.875, "learning_rate": 2.0867640566201252e-06, "loss": 0.1498, "step": 5987 }, { "epoch": 1.13, "grad_norm": 13.5625, "learning_rate": 2.08600866892515e-06, "loss": 0.218, "step": 5988 }, { "epoch": 1.13, "grad_norm": 11.25, "learning_rate": 2.0852533200903147e-06, "loss": 0.2425, "step": 5989 }, { "epoch": 1.13, "grad_norm": 16.375, "learning_rate": 2.0844980101865216e-06, "loss": 0.1951, "step": 5990 }, { "epoch": 1.13, "grad_norm": 15.125, "learning_rate": 2.0837427392846698e-06, "loss": 0.2321, "step": 5991 }, { "epoch": 1.13, "grad_norm": 12.75, "learning_rate": 2.082987507455654e-06, "loss": 0.2704, "step": 5992 }, { "epoch": 1.13, "grad_norm": 10.5, "learning_rate": 2.082232314770367e-06, "loss": 0.2056, "step": 5993 }, { "epoch": 1.13, "grad_norm": 11.5625, "learning_rate": 2.0814771612996946e-06, "loss": 0.2233, "step": 5994 }, { "epoch": 1.13, "grad_norm": 8.4375, "learning_rate": 2.080722047114521e-06, "loss": 0.1989, "step": 5995 }, { "epoch": 1.13, "grad_norm": 12.125, "learning_rate": 2.0799669722857275e-06, "loss": 0.2467, "step": 5996 }, { "epoch": 1.13, "grad_norm": 11.75, "learning_rate": 2.079211936884191e-06, "loss": 0.2423, "step": 5997 }, { "epoch": 1.13, "grad_norm": 10.9375, "learning_rate": 2.0784569409807826e-06, "loss": 0.2033, "step": 5998 }, { "epoch": 1.13, "grad_norm": 9.8125, "learning_rate": 2.077701984646374e-06, "loss": 0.2224, "step": 5999 }, { "epoch": 1.14, "grad_norm": 7.84375, "learning_rate": 2.076947067951829e-06, "loss": 0.1818, "step": 6000 }, { "epoch": 1.14, "grad_norm": 9.9375, "learning_rate": 2.0761921909680104e-06, "loss": 0.1967, "step": 6001 }, { "epoch": 1.14, "grad_norm": 13.375, "learning_rate": 2.0754373537657745e-06, "loss": 0.2356, "step": 6002 }, { "epoch": 1.14, "grad_norm": 11.4375, "learning_rate": 2.074682556415977e-06, "loss": 0.2613, "step": 6003 }, { "epoch": 1.14, "grad_norm": 10.5, "learning_rate": 2.0739277989894687e-06, "loss": 0.2464, "step": 6004 }, { "epoch": 1.14, "grad_norm": 9.3125, "learning_rate": 2.0731730815570965e-06, "loss": 0.2199, "step": 6005 }, { "epoch": 1.14, "grad_norm": 10.0, "learning_rate": 2.072418404189702e-06, "loss": 0.2302, "step": 6006 }, { "epoch": 1.14, "grad_norm": 9.3125, "learning_rate": 2.0716637669581265e-06, "loss": 0.2212, "step": 6007 }, { "epoch": 1.14, "grad_norm": 12.5, "learning_rate": 2.0709091699332042e-06, "loss": 0.2506, "step": 6008 }, { "epoch": 1.14, "grad_norm": 9.875, "learning_rate": 2.070154613185767e-06, "loss": 0.1687, "step": 6009 }, { "epoch": 1.14, "grad_norm": 13.3125, "learning_rate": 2.069400096786644e-06, "loss": 0.2191, "step": 6010 }, { "epoch": 1.14, "grad_norm": 10.5625, "learning_rate": 2.0686456208066586e-06, "loss": 0.2031, "step": 6011 }, { "epoch": 1.14, "grad_norm": 7.59375, "learning_rate": 2.0678911853166313e-06, "loss": 0.167, "step": 6012 }, { "epoch": 1.14, "grad_norm": 17.75, "learning_rate": 2.0671367903873777e-06, "loss": 0.229, "step": 6013 }, { "epoch": 1.14, "grad_norm": 12.0625, "learning_rate": 2.066382436089713e-06, "loss": 0.1526, "step": 6014 }, { "epoch": 1.14, "grad_norm": 13.125, "learning_rate": 2.065628122494444e-06, "loss": 0.2698, "step": 6015 }, { "epoch": 1.14, "grad_norm": 11.0625, "learning_rate": 2.064873849672377e-06, "loss": 0.2323, "step": 6016 }, { "epoch": 1.14, "grad_norm": 9.25, "learning_rate": 2.0641196176943127e-06, "loss": 0.1431, "step": 6017 }, { "epoch": 1.14, "grad_norm": 10.75, "learning_rate": 2.06336542663105e-06, "loss": 0.2804, "step": 6018 }, { "epoch": 1.14, "grad_norm": 8.875, "learning_rate": 2.062611276553381e-06, "loss": 0.1789, "step": 6019 }, { "epoch": 1.14, "grad_norm": 12.5, "learning_rate": 2.0618571675320954e-06, "loss": 0.2234, "step": 6020 }, { "epoch": 1.14, "grad_norm": 11.0, "learning_rate": 2.0611030996379805e-06, "loss": 0.2117, "step": 6021 }, { "epoch": 1.14, "grad_norm": 11.875, "learning_rate": 2.060349072941818e-06, "loss": 0.1981, "step": 6022 }, { "epoch": 1.14, "grad_norm": 11.3125, "learning_rate": 2.059595087514386e-06, "loss": 0.2429, "step": 6023 }, { "epoch": 1.14, "grad_norm": 15.3125, "learning_rate": 2.058841143426458e-06, "loss": 0.2029, "step": 6024 }, { "epoch": 1.14, "grad_norm": 8.375, "learning_rate": 2.058087240748806e-06, "loss": 0.1447, "step": 6025 }, { "epoch": 1.14, "grad_norm": 8.6875, "learning_rate": 2.0573333795521967e-06, "loss": 0.2166, "step": 6026 }, { "epoch": 1.14, "grad_norm": 10.0, "learning_rate": 2.0565795599073912e-06, "loss": 0.1739, "step": 6027 }, { "epoch": 1.14, "grad_norm": 10.5625, "learning_rate": 2.0558257818851487e-06, "loss": 0.2152, "step": 6028 }, { "epoch": 1.14, "grad_norm": 9.8125, "learning_rate": 2.0550720455562255e-06, "loss": 0.1953, "step": 6029 }, { "epoch": 1.14, "grad_norm": 15.375, "learning_rate": 2.054318350991372e-06, "loss": 0.1338, "step": 6030 }, { "epoch": 1.14, "grad_norm": 10.625, "learning_rate": 2.0535646982613336e-06, "loss": 0.2205, "step": 6031 }, { "epoch": 1.14, "grad_norm": 11.4375, "learning_rate": 2.0528110874368557e-06, "loss": 0.1612, "step": 6032 }, { "epoch": 1.14, "grad_norm": 12.3125, "learning_rate": 2.052057518588677e-06, "loss": 0.1786, "step": 6033 }, { "epoch": 1.14, "grad_norm": 11.125, "learning_rate": 2.0513039917875318e-06, "loss": 0.2348, "step": 6034 }, { "epoch": 1.14, "grad_norm": 9.875, "learning_rate": 2.0505505071041516e-06, "loss": 0.2193, "step": 6035 }, { "epoch": 1.14, "grad_norm": 9.875, "learning_rate": 2.049797064609265e-06, "loss": 0.23, "step": 6036 }, { "epoch": 1.14, "grad_norm": 9.4375, "learning_rate": 2.049043664373595e-06, "loss": 0.1568, "step": 6037 }, { "epoch": 1.14, "grad_norm": 15.3125, "learning_rate": 2.04829030646786e-06, "loss": 0.2394, "step": 6038 }, { "epoch": 1.14, "grad_norm": 13.6875, "learning_rate": 2.0475369909627758e-06, "loss": 0.2531, "step": 6039 }, { "epoch": 1.14, "grad_norm": 11.3125, "learning_rate": 2.0467837179290547e-06, "loss": 0.2309, "step": 6040 }, { "epoch": 1.14, "grad_norm": 12.5, "learning_rate": 2.046030487437404e-06, "loss": 0.2739, "step": 6041 }, { "epoch": 1.14, "grad_norm": 10.0, "learning_rate": 2.0452772995585264e-06, "loss": 0.1419, "step": 6042 }, { "epoch": 1.14, "grad_norm": 13.375, "learning_rate": 2.044524154363123e-06, "loss": 0.215, "step": 6043 }, { "epoch": 1.14, "grad_norm": 31.75, "learning_rate": 2.0437710519218873e-06, "loss": 0.2326, "step": 6044 }, { "epoch": 1.14, "grad_norm": 8.75, "learning_rate": 2.043017992305513e-06, "loss": 0.2057, "step": 6045 }, { "epoch": 1.14, "grad_norm": 11.5, "learning_rate": 2.042264975584685e-06, "loss": 0.1913, "step": 6046 }, { "epoch": 1.14, "grad_norm": 9.625, "learning_rate": 2.041512001830089e-06, "loss": 0.1495, "step": 6047 }, { "epoch": 1.14, "grad_norm": 9.9375, "learning_rate": 2.040759071112403e-06, "loss": 0.2066, "step": 6048 }, { "epoch": 1.14, "grad_norm": 19.875, "learning_rate": 2.0400061835023037e-06, "loss": 0.2248, "step": 6049 }, { "epoch": 1.14, "grad_norm": 11.25, "learning_rate": 2.0392533390704604e-06, "loss": 0.1886, "step": 6050 }, { "epoch": 1.14, "grad_norm": 13.9375, "learning_rate": 2.0385005378875425e-06, "loss": 0.2344, "step": 6051 }, { "epoch": 1.14, "grad_norm": 13.625, "learning_rate": 2.037747780024212e-06, "loss": 0.2512, "step": 6052 }, { "epoch": 1.15, "grad_norm": 10.375, "learning_rate": 2.0369950655511283e-06, "loss": 0.2424, "step": 6053 }, { "epoch": 1.15, "grad_norm": 9.125, "learning_rate": 2.036242394538947e-06, "loss": 0.1423, "step": 6054 }, { "epoch": 1.15, "grad_norm": 8.75, "learning_rate": 2.0354897670583187e-06, "loss": 0.1387, "step": 6055 }, { "epoch": 1.15, "grad_norm": 11.75, "learning_rate": 2.03473718317989e-06, "loss": 0.2072, "step": 6056 }, { "epoch": 1.15, "grad_norm": 10.0, "learning_rate": 2.0339846429743038e-06, "loss": 0.2168, "step": 6057 }, { "epoch": 1.15, "grad_norm": 7.78125, "learning_rate": 2.0332321465121998e-06, "loss": 0.1428, "step": 6058 }, { "epoch": 1.15, "grad_norm": 12.0, "learning_rate": 2.0324796938642116e-06, "loss": 0.1898, "step": 6059 }, { "epoch": 1.15, "grad_norm": 9.8125, "learning_rate": 2.03172728510097e-06, "loss": 0.2134, "step": 6060 }, { "epoch": 1.15, "grad_norm": 12.8125, "learning_rate": 2.0309749202931e-06, "loss": 0.2567, "step": 6061 }, { "epoch": 1.15, "grad_norm": 11.6875, "learning_rate": 2.030222599511227e-06, "loss": 0.2164, "step": 6062 }, { "epoch": 1.15, "grad_norm": 10.25, "learning_rate": 2.0294703228259667e-06, "loss": 0.2077, "step": 6063 }, { "epoch": 1.15, "grad_norm": 9.0, "learning_rate": 2.028718090307933e-06, "loss": 0.2393, "step": 6064 }, { "epoch": 1.15, "grad_norm": 21.625, "learning_rate": 2.0279659020277366e-06, "loss": 0.2482, "step": 6065 }, { "epoch": 1.15, "grad_norm": 10.4375, "learning_rate": 2.027213758055984e-06, "loss": 0.1744, "step": 6066 }, { "epoch": 1.15, "grad_norm": 9.25, "learning_rate": 2.026461658463275e-06, "loss": 0.2338, "step": 6067 }, { "epoch": 1.15, "grad_norm": 11.75, "learning_rate": 2.0257096033202074e-06, "loss": 0.2017, "step": 6068 }, { "epoch": 1.15, "grad_norm": 10.1875, "learning_rate": 2.0249575926973753e-06, "loss": 0.1895, "step": 6069 }, { "epoch": 1.15, "grad_norm": 9.75, "learning_rate": 2.024205626665367e-06, "loss": 0.1809, "step": 6070 }, { "epoch": 1.15, "grad_norm": 7.78125, "learning_rate": 2.0234537052947678e-06, "loss": 0.1613, "step": 6071 }, { "epoch": 1.15, "grad_norm": 10.875, "learning_rate": 2.0227018286561566e-06, "loss": 0.1661, "step": 6072 }, { "epoch": 1.15, "grad_norm": 9.9375, "learning_rate": 2.0219499968201127e-06, "loss": 0.1754, "step": 6073 }, { "epoch": 1.15, "grad_norm": 8.1875, "learning_rate": 2.0211982098572062e-06, "loss": 0.1613, "step": 6074 }, { "epoch": 1.15, "grad_norm": 13.4375, "learning_rate": 2.0204464678380055e-06, "loss": 0.1819, "step": 6075 }, { "epoch": 1.15, "grad_norm": 12.0625, "learning_rate": 2.019694770833075e-06, "loss": 0.1669, "step": 6076 }, { "epoch": 1.15, "grad_norm": 10.8125, "learning_rate": 2.0189431189129742e-06, "loss": 0.2197, "step": 6077 }, { "epoch": 1.15, "grad_norm": 18.75, "learning_rate": 2.0181915121482576e-06, "loss": 0.2125, "step": 6078 }, { "epoch": 1.15, "grad_norm": 11.875, "learning_rate": 2.0174399506094764e-06, "loss": 0.2032, "step": 6079 }, { "epoch": 1.15, "grad_norm": 9.0625, "learning_rate": 2.0166884343671785e-06, "loss": 0.2127, "step": 6080 }, { "epoch": 1.15, "grad_norm": 7.6875, "learning_rate": 2.0159369634919063e-06, "loss": 0.127, "step": 6081 }, { "epoch": 1.15, "grad_norm": 8.0, "learning_rate": 2.0151855380541975e-06, "loss": 0.1953, "step": 6082 }, { "epoch": 1.15, "grad_norm": 10.1875, "learning_rate": 2.0144341581245856e-06, "loss": 0.1993, "step": 6083 }, { "epoch": 1.15, "grad_norm": 8.25, "learning_rate": 2.013682823773602e-06, "loss": 0.1496, "step": 6084 }, { "epoch": 1.15, "grad_norm": 9.8125, "learning_rate": 2.0129315350717716e-06, "loss": 0.2467, "step": 6085 }, { "epoch": 1.15, "grad_norm": 9.75, "learning_rate": 2.012180292089615e-06, "loss": 0.2136, "step": 6086 }, { "epoch": 1.15, "grad_norm": 10.5, "learning_rate": 2.01142909489765e-06, "loss": 0.2167, "step": 6087 }, { "epoch": 1.15, "grad_norm": 13.1875, "learning_rate": 2.0106779435663887e-06, "loss": 0.2194, "step": 6088 }, { "epoch": 1.15, "grad_norm": 11.75, "learning_rate": 2.0099268381663405e-06, "loss": 0.2606, "step": 6089 }, { "epoch": 1.15, "grad_norm": 11.25, "learning_rate": 2.0091757787680075e-06, "loss": 0.2772, "step": 6090 }, { "epoch": 1.15, "grad_norm": 10.375, "learning_rate": 2.008424765441891e-06, "loss": 0.185, "step": 6091 }, { "epoch": 1.15, "grad_norm": 14.1875, "learning_rate": 2.0076737982584855e-06, "loss": 0.1923, "step": 6092 }, { "epoch": 1.15, "grad_norm": 12.4375, "learning_rate": 2.0069228772882835e-06, "loss": 0.2153, "step": 6093 }, { "epoch": 1.15, "grad_norm": 10.3125, "learning_rate": 2.0061720026017697e-06, "loss": 0.1927, "step": 6094 }, { "epoch": 1.15, "grad_norm": 9.8125, "learning_rate": 2.005421174269428e-06, "loss": 0.2199, "step": 6095 }, { "epoch": 1.15, "grad_norm": 12.8125, "learning_rate": 2.0046703923617362e-06, "loss": 0.2512, "step": 6096 }, { "epoch": 1.15, "grad_norm": 14.9375, "learning_rate": 2.0039196569491666e-06, "loss": 0.2275, "step": 6097 }, { "epoch": 1.15, "grad_norm": 8.4375, "learning_rate": 2.003168968102191e-06, "loss": 0.1612, "step": 6098 }, { "epoch": 1.15, "grad_norm": 10.5625, "learning_rate": 2.0024183258912726e-06, "loss": 0.1977, "step": 6099 }, { "epoch": 1.15, "grad_norm": 11.0625, "learning_rate": 2.001667730386872e-06, "loss": 0.2306, "step": 6100 }, { "epoch": 1.15, "grad_norm": 11.75, "learning_rate": 2.0009171816594454e-06, "loss": 0.1817, "step": 6101 }, { "epoch": 1.15, "grad_norm": 13.5625, "learning_rate": 2.0001666797794455e-06, "loss": 0.2006, "step": 6102 }, { "epoch": 1.15, "grad_norm": 13.625, "learning_rate": 1.999416224817319e-06, "loss": 0.2171, "step": 6103 }, { "epoch": 1.15, "grad_norm": 13.5625, "learning_rate": 1.9986658168435094e-06, "loss": 0.2257, "step": 6104 }, { "epoch": 1.15, "grad_norm": 12.4375, "learning_rate": 1.9979154559284534e-06, "loss": 0.2163, "step": 6105 }, { "epoch": 1.16, "grad_norm": 15.0, "learning_rate": 1.997165142142588e-06, "loss": 0.1978, "step": 6106 }, { "epoch": 1.16, "grad_norm": 13.75, "learning_rate": 1.996414875556341e-06, "loss": 0.2428, "step": 6107 }, { "epoch": 1.16, "grad_norm": 9.9375, "learning_rate": 1.9956646562401376e-06, "loss": 0.2614, "step": 6108 }, { "epoch": 1.16, "grad_norm": 19.875, "learning_rate": 1.9949144842644e-06, "loss": 0.1715, "step": 6109 }, { "epoch": 1.16, "grad_norm": 17.875, "learning_rate": 1.994164359699544e-06, "loss": 0.2191, "step": 6110 }, { "epoch": 1.16, "grad_norm": 20.125, "learning_rate": 1.9934142826159816e-06, "loss": 0.2275, "step": 6111 }, { "epoch": 1.16, "grad_norm": 12.1875, "learning_rate": 1.9926642530841195e-06, "loss": 0.1813, "step": 6112 }, { "epoch": 1.16, "grad_norm": 9.9375, "learning_rate": 1.9919142711743614e-06, "loss": 0.161, "step": 6113 }, { "epoch": 1.16, "grad_norm": 12.25, "learning_rate": 1.9911643369571073e-06, "loss": 0.2606, "step": 6114 }, { "epoch": 1.16, "grad_norm": 10.875, "learning_rate": 1.990414450502749e-06, "loss": 0.1925, "step": 6115 }, { "epoch": 1.16, "grad_norm": 12.0625, "learning_rate": 1.989664611881677e-06, "loss": 0.2432, "step": 6116 }, { "epoch": 1.16, "grad_norm": 9.125, "learning_rate": 1.9889148211642765e-06, "loss": 0.1949, "step": 6117 }, { "epoch": 1.16, "grad_norm": 10.25, "learning_rate": 1.9881650784209284e-06, "loss": 0.2208, "step": 6118 }, { "epoch": 1.16, "grad_norm": 12.125, "learning_rate": 1.9874153837220077e-06, "loss": 0.1984, "step": 6119 }, { "epoch": 1.16, "grad_norm": 16.375, "learning_rate": 1.9866657371378877e-06, "loss": 0.1738, "step": 6120 }, { "epoch": 1.16, "grad_norm": 12.5625, "learning_rate": 1.9859161387389352e-06, "loss": 0.1457, "step": 6121 }, { "epoch": 1.16, "grad_norm": 11.4375, "learning_rate": 1.9851665885955116e-06, "loss": 0.2093, "step": 6122 }, { "epoch": 1.16, "grad_norm": 11.5, "learning_rate": 1.984417086777975e-06, "loss": 0.1934, "step": 6123 }, { "epoch": 1.16, "grad_norm": 11.3125, "learning_rate": 1.98366763335668e-06, "loss": 0.1652, "step": 6124 }, { "epoch": 1.16, "grad_norm": 13.6875, "learning_rate": 1.982918228401976e-06, "loss": 0.232, "step": 6125 }, { "epoch": 1.16, "grad_norm": 14.75, "learning_rate": 1.9821688719842054e-06, "loss": 0.2183, "step": 6126 }, { "epoch": 1.16, "grad_norm": 16.125, "learning_rate": 1.9814195641737095e-06, "loss": 0.1883, "step": 6127 }, { "epoch": 1.16, "grad_norm": 9.5625, "learning_rate": 1.980670305040823e-06, "loss": 0.1828, "step": 6128 }, { "epoch": 1.16, "grad_norm": 15.375, "learning_rate": 1.979921094655878e-06, "loss": 0.1735, "step": 6129 }, { "epoch": 1.16, "grad_norm": 10.6875, "learning_rate": 1.9791719330891983e-06, "loss": 0.1836, "step": 6130 }, { "epoch": 1.16, "grad_norm": 12.0625, "learning_rate": 1.978422820411107e-06, "loss": 0.2113, "step": 6131 }, { "epoch": 1.16, "grad_norm": 10.6875, "learning_rate": 1.977673756691921e-06, "loss": 0.2124, "step": 6132 }, { "epoch": 1.16, "grad_norm": 11.6875, "learning_rate": 1.976924742001954e-06, "loss": 0.1538, "step": 6133 }, { "epoch": 1.16, "grad_norm": 9.9375, "learning_rate": 1.9761757764115102e-06, "loss": 0.2064, "step": 6134 }, { "epoch": 1.16, "grad_norm": 9.5625, "learning_rate": 1.975426859990896e-06, "loss": 0.2225, "step": 6135 }, { "epoch": 1.16, "grad_norm": 14.4375, "learning_rate": 1.9746779928104086e-06, "loss": 0.1806, "step": 6136 }, { "epoch": 1.16, "grad_norm": 11.3125, "learning_rate": 1.973929174940343e-06, "loss": 0.2456, "step": 6137 }, { "epoch": 1.16, "grad_norm": 11.6875, "learning_rate": 1.9731804064509865e-06, "loss": 0.1906, "step": 6138 }, { "epoch": 1.16, "grad_norm": 10.1875, "learning_rate": 1.9724316874126256e-06, "loss": 0.2219, "step": 6139 }, { "epoch": 1.16, "grad_norm": 13.9375, "learning_rate": 1.97168301789554e-06, "loss": 0.1927, "step": 6140 }, { "epoch": 1.16, "grad_norm": 10.0625, "learning_rate": 1.970934397970004e-06, "loss": 0.1866, "step": 6141 }, { "epoch": 1.16, "grad_norm": 11.1875, "learning_rate": 1.97018582770629e-06, "loss": 0.1851, "step": 6142 }, { "epoch": 1.16, "grad_norm": 10.0, "learning_rate": 1.9694373071746633e-06, "loss": 0.2137, "step": 6143 }, { "epoch": 1.16, "grad_norm": 14.1875, "learning_rate": 1.9686888364453856e-06, "loss": 0.2029, "step": 6144 }, { "epoch": 1.16, "grad_norm": 10.0625, "learning_rate": 1.967940415588712e-06, "loss": 0.2052, "step": 6145 }, { "epoch": 1.16, "grad_norm": 10.75, "learning_rate": 1.967192044674897e-06, "loss": 0.2276, "step": 6146 }, { "epoch": 1.16, "grad_norm": 11.5, "learning_rate": 1.9664437237741867e-06, "loss": 0.1894, "step": 6147 }, { "epoch": 1.16, "grad_norm": 10.3125, "learning_rate": 1.965695452956824e-06, "loss": 0.2437, "step": 6148 }, { "epoch": 1.16, "grad_norm": 12.9375, "learning_rate": 1.9649472322930453e-06, "loss": 0.2613, "step": 6149 }, { "epoch": 1.16, "grad_norm": 10.9375, "learning_rate": 1.964199061853087e-06, "loss": 0.1543, "step": 6150 }, { "epoch": 1.16, "grad_norm": 11.5625, "learning_rate": 1.9634509417071757e-06, "loss": 0.1944, "step": 6151 }, { "epoch": 1.16, "grad_norm": 14.625, "learning_rate": 1.9627028719255345e-06, "loss": 0.2241, "step": 6152 }, { "epoch": 1.16, "grad_norm": 13.8125, "learning_rate": 1.961954852578384e-06, "loss": 0.216, "step": 6153 }, { "epoch": 1.16, "grad_norm": 38.25, "learning_rate": 1.9612068837359387e-06, "loss": 0.2234, "step": 6154 }, { "epoch": 1.16, "grad_norm": 9.9375, "learning_rate": 1.960458965468407e-06, "loss": 0.1909, "step": 6155 }, { "epoch": 1.16, "grad_norm": 11.3125, "learning_rate": 1.9597110978459936e-06, "loss": 0.2358, "step": 6156 }, { "epoch": 1.16, "grad_norm": 10.125, "learning_rate": 1.958963280938901e-06, "loss": 0.2026, "step": 6157 }, { "epoch": 1.16, "grad_norm": 8.125, "learning_rate": 1.958215514817322e-06, "loss": 0.1898, "step": 6158 }, { "epoch": 1.17, "grad_norm": 13.0, "learning_rate": 1.9574677995514478e-06, "loss": 0.2181, "step": 6159 }, { "epoch": 1.17, "grad_norm": 19.5, "learning_rate": 1.9567201352114644e-06, "loss": 0.2236, "step": 6160 }, { "epoch": 1.17, "grad_norm": 9.4375, "learning_rate": 1.9559725218675537e-06, "loss": 0.2127, "step": 6161 }, { "epoch": 1.17, "grad_norm": 13.3125, "learning_rate": 1.955224959589891e-06, "loss": 0.1854, "step": 6162 }, { "epoch": 1.17, "grad_norm": 11.0625, "learning_rate": 1.9544774484486475e-06, "loss": 0.2433, "step": 6163 }, { "epoch": 1.17, "grad_norm": 9.3125, "learning_rate": 1.95372998851399e-06, "loss": 0.2311, "step": 6164 }, { "epoch": 1.17, "grad_norm": 10.0625, "learning_rate": 1.952982579856082e-06, "loss": 0.1964, "step": 6165 }, { "epoch": 1.17, "grad_norm": 11.5, "learning_rate": 1.952235222545078e-06, "loss": 0.2434, "step": 6166 }, { "epoch": 1.17, "grad_norm": 9.1875, "learning_rate": 1.9514879166511317e-06, "loss": 0.2043, "step": 6167 }, { "epoch": 1.17, "grad_norm": 10.375, "learning_rate": 1.9507406622443903e-06, "loss": 0.1751, "step": 6168 }, { "epoch": 1.17, "grad_norm": 12.375, "learning_rate": 1.949993459394997e-06, "loss": 0.2468, "step": 6169 }, { "epoch": 1.17, "grad_norm": 12.9375, "learning_rate": 1.949246308173088e-06, "loss": 0.2732, "step": 6170 }, { "epoch": 1.17, "grad_norm": 10.125, "learning_rate": 1.9484992086487965e-06, "loss": 0.2325, "step": 6171 }, { "epoch": 1.17, "grad_norm": 12.6875, "learning_rate": 1.9477521608922513e-06, "loss": 0.2483, "step": 6172 }, { "epoch": 1.17, "grad_norm": 9.1875, "learning_rate": 1.947005164973576e-06, "loss": 0.2137, "step": 6173 }, { "epoch": 1.17, "grad_norm": 11.875, "learning_rate": 1.946258220962888e-06, "loss": 0.2301, "step": 6174 }, { "epoch": 1.17, "grad_norm": 7.84375, "learning_rate": 1.9455113289302995e-06, "loss": 0.1757, "step": 6175 }, { "epoch": 1.17, "grad_norm": 10.5, "learning_rate": 1.9447644889459214e-06, "loss": 0.1741, "step": 6176 }, { "epoch": 1.17, "grad_norm": 15.75, "learning_rate": 1.9440177010798567e-06, "loss": 0.2197, "step": 6177 }, { "epoch": 1.17, "grad_norm": 13.8125, "learning_rate": 1.943270965402203e-06, "loss": 0.2856, "step": 6178 }, { "epoch": 1.17, "grad_norm": 14.0625, "learning_rate": 1.9425242819830553e-06, "loss": 0.2082, "step": 6179 }, { "epoch": 1.17, "grad_norm": 8.5, "learning_rate": 1.9417776508925023e-06, "loss": 0.1544, "step": 6180 }, { "epoch": 1.17, "grad_norm": 9.875, "learning_rate": 1.941031072200629e-06, "loss": 0.1867, "step": 6181 }, { "epoch": 1.17, "grad_norm": 13.4375, "learning_rate": 1.9402845459775116e-06, "loss": 0.2021, "step": 6182 }, { "epoch": 1.17, "grad_norm": 11.3125, "learning_rate": 1.9395380722932274e-06, "loss": 0.173, "step": 6183 }, { "epoch": 1.17, "grad_norm": 11.125, "learning_rate": 1.938791651217844e-06, "loss": 0.2383, "step": 6184 }, { "epoch": 1.17, "grad_norm": 15.875, "learning_rate": 1.938045282821427e-06, "loss": 0.2385, "step": 6185 }, { "epoch": 1.17, "grad_norm": 8.5625, "learning_rate": 1.9372989671740346e-06, "loss": 0.2148, "step": 6186 }, { "epoch": 1.17, "grad_norm": 10.5625, "learning_rate": 1.9365527043457213e-06, "loss": 0.2055, "step": 6187 }, { "epoch": 1.17, "grad_norm": 9.5, "learning_rate": 1.935806494406538e-06, "loss": 0.1768, "step": 6188 }, { "epoch": 1.17, "grad_norm": 11.25, "learning_rate": 1.9350603374265263e-06, "loss": 0.1736, "step": 6189 }, { "epoch": 1.17, "grad_norm": 12.0, "learning_rate": 1.9343142334757294e-06, "loss": 0.2169, "step": 6190 }, { "epoch": 1.17, "grad_norm": 13.125, "learning_rate": 1.9335681826241794e-06, "loss": 0.2019, "step": 6191 }, { "epoch": 1.17, "grad_norm": 10.6875, "learning_rate": 1.932822184941907e-06, "loss": 0.2353, "step": 6192 }, { "epoch": 1.17, "grad_norm": 11.375, "learning_rate": 1.9320762404989353e-06, "loss": 0.2676, "step": 6193 }, { "epoch": 1.17, "grad_norm": 12.3125, "learning_rate": 1.931330349365286e-06, "loss": 0.2184, "step": 6194 }, { "epoch": 1.17, "grad_norm": 11.1875, "learning_rate": 1.930584511610972e-06, "loss": 0.166, "step": 6195 }, { "epoch": 1.17, "grad_norm": 8.875, "learning_rate": 1.9298387273060047e-06, "loss": 0.1832, "step": 6196 }, { "epoch": 1.17, "grad_norm": 12.9375, "learning_rate": 1.929092996520385e-06, "loss": 0.1821, "step": 6197 }, { "epoch": 1.17, "grad_norm": 9.5, "learning_rate": 1.9283473193241176e-06, "loss": 0.1742, "step": 6198 }, { "epoch": 1.17, "grad_norm": 14.9375, "learning_rate": 1.9276016957871933e-06, "loss": 0.2442, "step": 6199 }, { "epoch": 1.17, "grad_norm": 10.9375, "learning_rate": 1.926856125979602e-06, "loss": 0.1806, "step": 6200 }, { "epoch": 1.17, "grad_norm": 10.25, "learning_rate": 1.92611060997133e-06, "loss": 0.1851, "step": 6201 }, { "epoch": 1.17, "grad_norm": 10.5625, "learning_rate": 1.925365147832355e-06, "loss": 0.1856, "step": 6202 }, { "epoch": 1.17, "grad_norm": 12.25, "learning_rate": 1.9246197396326513e-06, "loss": 0.2401, "step": 6203 }, { "epoch": 1.17, "grad_norm": 10.75, "learning_rate": 1.923874385442188e-06, "loss": 0.2121, "step": 6204 }, { "epoch": 1.17, "grad_norm": 10.125, "learning_rate": 1.9231290853309313e-06, "loss": 0.2075, "step": 6205 }, { "epoch": 1.17, "grad_norm": 9.125, "learning_rate": 1.9223838393688387e-06, "loss": 0.1522, "step": 6206 }, { "epoch": 1.17, "grad_norm": 11.125, "learning_rate": 1.9216386476258642e-06, "loss": 0.2137, "step": 6207 }, { "epoch": 1.17, "grad_norm": 15.3125, "learning_rate": 1.9208935101719562e-06, "loss": 0.2983, "step": 6208 }, { "epoch": 1.17, "grad_norm": 11.3125, "learning_rate": 1.9201484270770606e-06, "loss": 0.1636, "step": 6209 }, { "epoch": 1.17, "grad_norm": 11.0625, "learning_rate": 1.9194033984111143e-06, "loss": 0.2357, "step": 6210 }, { "epoch": 1.17, "grad_norm": 10.1875, "learning_rate": 1.918658424244051e-06, "loss": 0.1746, "step": 6211 }, { "epoch": 1.18, "grad_norm": 8.875, "learning_rate": 1.9179135046458e-06, "loss": 0.2062, "step": 6212 }, { "epoch": 1.18, "grad_norm": 9.25, "learning_rate": 1.9171686396862855e-06, "loss": 0.2035, "step": 6213 }, { "epoch": 1.18, "grad_norm": 10.1875, "learning_rate": 1.916423829435424e-06, "loss": 0.1525, "step": 6214 }, { "epoch": 1.18, "grad_norm": 15.9375, "learning_rate": 1.9156790739631285e-06, "loss": 0.2245, "step": 6215 }, { "epoch": 1.18, "grad_norm": 8.3125, "learning_rate": 1.9149343733393084e-06, "loss": 0.1549, "step": 6216 }, { "epoch": 1.18, "grad_norm": 10.125, "learning_rate": 1.914189727633867e-06, "loss": 0.1986, "step": 6217 }, { "epoch": 1.18, "grad_norm": 8.8125, "learning_rate": 1.9134451369166997e-06, "loss": 0.2082, "step": 6218 }, { "epoch": 1.18, "grad_norm": 11.75, "learning_rate": 1.9127006012577003e-06, "loss": 0.2407, "step": 6219 }, { "epoch": 1.18, "grad_norm": 9.25, "learning_rate": 1.9119561207267565e-06, "loss": 0.2006, "step": 6220 }, { "epoch": 1.18, "grad_norm": 11.625, "learning_rate": 1.9112116953937502e-06, "loss": 0.219, "step": 6221 }, { "epoch": 1.18, "grad_norm": 7.9375, "learning_rate": 1.910467325328558e-06, "loss": 0.1868, "step": 6222 }, { "epoch": 1.18, "grad_norm": 11.125, "learning_rate": 1.9097230106010524e-06, "loss": 0.1804, "step": 6223 }, { "epoch": 1.18, "grad_norm": 11.5625, "learning_rate": 1.9089787512810993e-06, "loss": 0.2405, "step": 6224 }, { "epoch": 1.18, "grad_norm": 13.0625, "learning_rate": 1.908234547438561e-06, "loss": 0.2274, "step": 6225 }, { "epoch": 1.18, "grad_norm": 9.5625, "learning_rate": 1.9074903991432925e-06, "loss": 0.1752, "step": 6226 }, { "epoch": 1.18, "grad_norm": 8.4375, "learning_rate": 1.9067463064651455e-06, "loss": 0.2117, "step": 6227 }, { "epoch": 1.18, "grad_norm": 12.125, "learning_rate": 1.9060022694739665e-06, "loss": 0.266, "step": 6228 }, { "epoch": 1.18, "grad_norm": 11.8125, "learning_rate": 1.9052582882395948e-06, "loss": 0.2109, "step": 6229 }, { "epoch": 1.18, "grad_norm": 10.625, "learning_rate": 1.9045143628318657e-06, "loss": 0.2791, "step": 6230 }, { "epoch": 1.18, "grad_norm": 9.3125, "learning_rate": 1.9037704933206105e-06, "loss": 0.23, "step": 6231 }, { "epoch": 1.18, "grad_norm": 11.4375, "learning_rate": 1.9030266797756536e-06, "loss": 0.2173, "step": 6232 }, { "epoch": 1.18, "grad_norm": 13.375, "learning_rate": 1.9022829222668127e-06, "loss": 0.2133, "step": 6233 }, { "epoch": 1.18, "grad_norm": 13.1875, "learning_rate": 1.9015392208639055e-06, "loss": 0.2483, "step": 6234 }, { "epoch": 1.18, "grad_norm": 13.0625, "learning_rate": 1.9007955756367386e-06, "loss": 0.2249, "step": 6235 }, { "epoch": 1.18, "grad_norm": 10.375, "learning_rate": 1.9000519866551168e-06, "loss": 0.235, "step": 6236 }, { "epoch": 1.18, "grad_norm": 11.6875, "learning_rate": 1.8993084539888369e-06, "loss": 0.2007, "step": 6237 }, { "epoch": 1.18, "grad_norm": 9.8125, "learning_rate": 1.8985649777076948e-06, "loss": 0.1466, "step": 6238 }, { "epoch": 1.18, "grad_norm": 11.875, "learning_rate": 1.8978215578814765e-06, "loss": 0.2785, "step": 6239 }, { "epoch": 1.18, "grad_norm": 10.3125, "learning_rate": 1.897078194579966e-06, "loss": 0.1861, "step": 6240 }, { "epoch": 1.18, "grad_norm": 11.125, "learning_rate": 1.8963348878729384e-06, "loss": 0.173, "step": 6241 }, { "epoch": 1.18, "grad_norm": 13.4375, "learning_rate": 1.8955916378301678e-06, "loss": 0.1948, "step": 6242 }, { "epoch": 1.18, "grad_norm": 8.25, "learning_rate": 1.89484844452142e-06, "loss": 0.1803, "step": 6243 }, { "epoch": 1.18, "grad_norm": 12.6875, "learning_rate": 1.8941053080164561e-06, "loss": 0.2128, "step": 6244 }, { "epoch": 1.18, "grad_norm": 10.875, "learning_rate": 1.8933622283850336e-06, "loss": 0.1542, "step": 6245 }, { "epoch": 1.18, "grad_norm": 11.5625, "learning_rate": 1.8926192056969016e-06, "loss": 0.2162, "step": 6246 }, { "epoch": 1.18, "grad_norm": 10.3125, "learning_rate": 1.891876240021806e-06, "loss": 0.2627, "step": 6247 }, { "epoch": 1.18, "grad_norm": 8.0625, "learning_rate": 1.891133331429486e-06, "loss": 0.1623, "step": 6248 }, { "epoch": 1.18, "grad_norm": 13.5625, "learning_rate": 1.8903904799896782e-06, "loss": 0.255, "step": 6249 }, { "epoch": 1.18, "grad_norm": 11.5, "learning_rate": 1.8896476857721103e-06, "loss": 0.2094, "step": 6250 }, { "epoch": 1.18, "grad_norm": 14.1875, "learning_rate": 1.888904948846506e-06, "loss": 0.1941, "step": 6251 }, { "epoch": 1.18, "grad_norm": 8.9375, "learning_rate": 1.8881622692825844e-06, "loss": 0.193, "step": 6252 }, { "epoch": 1.18, "grad_norm": 24.5, "learning_rate": 1.8874196471500592e-06, "loss": 0.244, "step": 6253 }, { "epoch": 1.18, "grad_norm": 14.4375, "learning_rate": 1.8866770825186374e-06, "loss": 0.2204, "step": 6254 }, { "epoch": 1.18, "grad_norm": 14.0625, "learning_rate": 1.8859345754580205e-06, "loss": 0.2686, "step": 6255 }, { "epoch": 1.18, "grad_norm": 10.25, "learning_rate": 1.8851921260379073e-06, "loss": 0.2702, "step": 6256 }, { "epoch": 1.18, "grad_norm": 9.5, "learning_rate": 1.8844497343279886e-06, "loss": 0.2043, "step": 6257 }, { "epoch": 1.18, "grad_norm": 8.8125, "learning_rate": 1.88370740039795e-06, "loss": 0.2145, "step": 6258 }, { "epoch": 1.18, "grad_norm": 10.5, "learning_rate": 1.882965124317472e-06, "loss": 0.1322, "step": 6259 }, { "epoch": 1.18, "grad_norm": 11.0, "learning_rate": 1.8822229061562308e-06, "loss": 0.1989, "step": 6260 }, { "epoch": 1.18, "grad_norm": 12.625, "learning_rate": 1.8814807459838966e-06, "loss": 0.1806, "step": 6261 }, { "epoch": 1.18, "grad_norm": 9.4375, "learning_rate": 1.8807386438701325e-06, "loss": 0.1894, "step": 6262 }, { "epoch": 1.18, "grad_norm": 11.1875, "learning_rate": 1.879996599884597e-06, "loss": 0.1907, "step": 6263 }, { "epoch": 1.18, "grad_norm": 12.8125, "learning_rate": 1.8792546140969453e-06, "loss": 0.1849, "step": 6264 }, { "epoch": 1.19, "grad_norm": 10.5625, "learning_rate": 1.878512686576825e-06, "loss": 0.2137, "step": 6265 }, { "epoch": 1.19, "grad_norm": 9.1875, "learning_rate": 1.8777708173938773e-06, "loss": 0.2007, "step": 6266 }, { "epoch": 1.19, "grad_norm": 10.8125, "learning_rate": 1.8770290066177405e-06, "loss": 0.1676, "step": 6267 }, { "epoch": 1.19, "grad_norm": 17.25, "learning_rate": 1.8762872543180461e-06, "loss": 0.252, "step": 6268 }, { "epoch": 1.19, "grad_norm": 11.25, "learning_rate": 1.8755455605644204e-06, "loss": 0.1971, "step": 6269 }, { "epoch": 1.19, "grad_norm": 12.25, "learning_rate": 1.8748039254264828e-06, "loss": 0.246, "step": 6270 }, { "epoch": 1.19, "grad_norm": 10.875, "learning_rate": 1.874062348973849e-06, "loss": 0.2039, "step": 6271 }, { "epoch": 1.19, "grad_norm": 10.5, "learning_rate": 1.8733208312761298e-06, "loss": 0.1991, "step": 6272 }, { "epoch": 1.19, "grad_norm": 11.6875, "learning_rate": 1.8725793724029276e-06, "loss": 0.2012, "step": 6273 }, { "epoch": 1.19, "grad_norm": 12.125, "learning_rate": 1.8718379724238412e-06, "loss": 0.2341, "step": 6274 }, { "epoch": 1.19, "grad_norm": 15.625, "learning_rate": 1.8710966314084638e-06, "loss": 0.2213, "step": 6275 }, { "epoch": 1.19, "grad_norm": 13.1875, "learning_rate": 1.8703553494263844e-06, "loss": 0.2171, "step": 6276 }, { "epoch": 1.19, "grad_norm": 12.1875, "learning_rate": 1.8696141265471819e-06, "loss": 0.2358, "step": 6277 }, { "epoch": 1.19, "grad_norm": 12.25, "learning_rate": 1.8688729628404356e-06, "loss": 0.2476, "step": 6278 }, { "epoch": 1.19, "grad_norm": 12.375, "learning_rate": 1.8681318583757147e-06, "loss": 0.1997, "step": 6279 }, { "epoch": 1.19, "grad_norm": 9.125, "learning_rate": 1.8673908132225858e-06, "loss": 0.155, "step": 6280 }, { "epoch": 1.19, "grad_norm": 8.4375, "learning_rate": 1.866649827450606e-06, "loss": 0.1629, "step": 6281 }, { "epoch": 1.19, "grad_norm": 10.5, "learning_rate": 1.8659089011293324e-06, "loss": 0.2002, "step": 6282 }, { "epoch": 1.19, "grad_norm": 10.9375, "learning_rate": 1.8651680343283123e-06, "loss": 0.2052, "step": 6283 }, { "epoch": 1.19, "grad_norm": 8.875, "learning_rate": 1.8644272271170893e-06, "loss": 0.1845, "step": 6284 }, { "epoch": 1.19, "grad_norm": 9.8125, "learning_rate": 1.8636864795651988e-06, "loss": 0.1646, "step": 6285 }, { "epoch": 1.19, "grad_norm": 12.3125, "learning_rate": 1.8629457917421745e-06, "loss": 0.184, "step": 6286 }, { "epoch": 1.19, "grad_norm": 13.1875, "learning_rate": 1.8622051637175423e-06, "loss": 0.2692, "step": 6287 }, { "epoch": 1.19, "grad_norm": 11.125, "learning_rate": 1.8614645955608219e-06, "loss": 0.1931, "step": 6288 }, { "epoch": 1.19, "grad_norm": 12.9375, "learning_rate": 1.8607240873415299e-06, "loss": 0.2015, "step": 6289 }, { "epoch": 1.19, "grad_norm": 12.75, "learning_rate": 1.8599836391291736e-06, "loss": 0.2377, "step": 6290 }, { "epoch": 1.19, "grad_norm": 12.1875, "learning_rate": 1.8592432509932585e-06, "loss": 0.2126, "step": 6291 }, { "epoch": 1.19, "grad_norm": 16.625, "learning_rate": 1.8585029230032805e-06, "loss": 0.2542, "step": 6292 }, { "epoch": 1.19, "grad_norm": 17.875, "learning_rate": 1.8577626552287349e-06, "loss": 0.2539, "step": 6293 }, { "epoch": 1.19, "grad_norm": 12.0, "learning_rate": 1.8570224477391059e-06, "loss": 0.2677, "step": 6294 }, { "epoch": 1.19, "grad_norm": 8.9375, "learning_rate": 1.8562823006038755e-06, "loss": 0.2045, "step": 6295 }, { "epoch": 1.19, "grad_norm": 14.3125, "learning_rate": 1.8555422138925188e-06, "loss": 0.2275, "step": 6296 }, { "epoch": 1.19, "grad_norm": 9.9375, "learning_rate": 1.8548021876745067e-06, "loss": 0.2551, "step": 6297 }, { "epoch": 1.19, "grad_norm": 13.1875, "learning_rate": 1.8540622220193021e-06, "loss": 0.2228, "step": 6298 }, { "epoch": 1.19, "grad_norm": 12.5, "learning_rate": 1.8533223169963631e-06, "loss": 0.2078, "step": 6299 }, { "epoch": 1.19, "grad_norm": 22.25, "learning_rate": 1.8525824726751437e-06, "loss": 0.2643, "step": 6300 }, { "epoch": 1.19, "grad_norm": 11.875, "learning_rate": 1.8518426891250906e-06, "loss": 0.2161, "step": 6301 }, { "epoch": 1.19, "grad_norm": 9.4375, "learning_rate": 1.8511029664156443e-06, "loss": 0.233, "step": 6302 }, { "epoch": 1.19, "grad_norm": 8.625, "learning_rate": 1.8503633046162404e-06, "loss": 0.1791, "step": 6303 }, { "epoch": 1.19, "grad_norm": 13.0625, "learning_rate": 1.8496237037963094e-06, "loss": 0.2191, "step": 6304 }, { "epoch": 1.19, "grad_norm": 9.125, "learning_rate": 1.8488841640252758e-06, "loss": 0.2646, "step": 6305 }, { "epoch": 1.19, "grad_norm": 9.75, "learning_rate": 1.8481446853725568e-06, "loss": 0.251, "step": 6306 }, { "epoch": 1.19, "grad_norm": 13.625, "learning_rate": 1.8474052679075651e-06, "loss": 0.2418, "step": 6307 }, { "epoch": 1.19, "grad_norm": 8.6875, "learning_rate": 1.8466659116997089e-06, "loss": 0.2086, "step": 6308 }, { "epoch": 1.19, "grad_norm": 9.0625, "learning_rate": 1.8459266168183892e-06, "loss": 0.2261, "step": 6309 }, { "epoch": 1.19, "grad_norm": 11.625, "learning_rate": 1.8451873833329998e-06, "loss": 0.2138, "step": 6310 }, { "epoch": 1.19, "grad_norm": 8.375, "learning_rate": 1.844448211312932e-06, "loss": 0.1984, "step": 6311 }, { "epoch": 1.19, "grad_norm": 11.6875, "learning_rate": 1.8437091008275692e-06, "loss": 0.2313, "step": 6312 }, { "epoch": 1.19, "grad_norm": 12.125, "learning_rate": 1.8429700519462896e-06, "loss": 0.1902, "step": 6313 }, { "epoch": 1.19, "grad_norm": 15.8125, "learning_rate": 1.8422310647384644e-06, "loss": 0.2186, "step": 6314 }, { "epoch": 1.19, "grad_norm": 10.5, "learning_rate": 1.8414921392734619e-06, "loss": 0.1945, "step": 6315 }, { "epoch": 1.19, "grad_norm": 11.0625, "learning_rate": 1.8407532756206425e-06, "loss": 0.1831, "step": 6316 }, { "epoch": 1.2, "grad_norm": 15.25, "learning_rate": 1.8400144738493602e-06, "loss": 0.2154, "step": 6317 }, { "epoch": 1.2, "grad_norm": 12.0, "learning_rate": 1.839275734028964e-06, "loss": 0.2414, "step": 6318 }, { "epoch": 1.2, "grad_norm": 10.1875, "learning_rate": 1.8385370562287989e-06, "loss": 0.1567, "step": 6319 }, { "epoch": 1.2, "grad_norm": 9.25, "learning_rate": 1.8377984405182015e-06, "loss": 0.2199, "step": 6320 }, { "epoch": 1.2, "grad_norm": 9.9375, "learning_rate": 1.8370598869665018e-06, "loss": 0.2304, "step": 6321 }, { "epoch": 1.2, "grad_norm": 10.25, "learning_rate": 1.8363213956430288e-06, "loss": 0.231, "step": 6322 }, { "epoch": 1.2, "grad_norm": 8.625, "learning_rate": 1.8355829666170998e-06, "loss": 0.1783, "step": 6323 }, { "epoch": 1.2, "grad_norm": 10.0625, "learning_rate": 1.8348445999580312e-06, "loss": 0.1768, "step": 6324 }, { "epoch": 1.2, "grad_norm": 9.5, "learning_rate": 1.8341062957351289e-06, "loss": 0.1834, "step": 6325 }, { "epoch": 1.2, "grad_norm": 9.0, "learning_rate": 1.8333680540176969e-06, "loss": 0.1509, "step": 6326 }, { "epoch": 1.2, "grad_norm": 8.5625, "learning_rate": 1.8326298748750316e-06, "loss": 0.2102, "step": 6327 }, { "epoch": 1.2, "grad_norm": 11.9375, "learning_rate": 1.831891758376424e-06, "loss": 0.2538, "step": 6328 }, { "epoch": 1.2, "grad_norm": 14.875, "learning_rate": 1.8311537045911571e-06, "loss": 0.2296, "step": 6329 }, { "epoch": 1.2, "grad_norm": 11.25, "learning_rate": 1.8304157135885115e-06, "loss": 0.1806, "step": 6330 }, { "epoch": 1.2, "grad_norm": 8.375, "learning_rate": 1.82967778543776e-06, "loss": 0.1684, "step": 6331 }, { "epoch": 1.2, "grad_norm": 10.375, "learning_rate": 1.8289399202081693e-06, "loss": 0.1796, "step": 6332 }, { "epoch": 1.2, "grad_norm": 15.8125, "learning_rate": 1.8282021179690017e-06, "loss": 0.2054, "step": 6333 }, { "epoch": 1.2, "grad_norm": 10.6875, "learning_rate": 1.8274643787895114e-06, "loss": 0.2053, "step": 6334 }, { "epoch": 1.2, "grad_norm": 11.5, "learning_rate": 1.8267267027389478e-06, "loss": 0.2387, "step": 6335 }, { "epoch": 1.2, "grad_norm": 7.71875, "learning_rate": 1.8259890898865545e-06, "loss": 0.1826, "step": 6336 }, { "epoch": 1.2, "grad_norm": 11.4375, "learning_rate": 1.8252515403015703e-06, "loss": 0.2719, "step": 6337 }, { "epoch": 1.2, "grad_norm": 11.0625, "learning_rate": 1.8245140540532246e-06, "loss": 0.2435, "step": 6338 }, { "epoch": 1.2, "grad_norm": 10.9375, "learning_rate": 1.8237766312107446e-06, "loss": 0.2285, "step": 6339 }, { "epoch": 1.2, "grad_norm": 11.4375, "learning_rate": 1.8230392718433488e-06, "loss": 0.2225, "step": 6340 }, { "epoch": 1.2, "grad_norm": 11.5625, "learning_rate": 1.8223019760202528e-06, "loss": 0.1935, "step": 6341 }, { "epoch": 1.2, "grad_norm": 14.8125, "learning_rate": 1.8215647438106628e-06, "loss": 0.2081, "step": 6342 }, { "epoch": 1.2, "grad_norm": 12.875, "learning_rate": 1.8208275752837806e-06, "loss": 0.2159, "step": 6343 }, { "epoch": 1.2, "grad_norm": 17.625, "learning_rate": 1.8200904705088028e-06, "loss": 0.2163, "step": 6344 }, { "epoch": 1.2, "grad_norm": 7.8125, "learning_rate": 1.8193534295549197e-06, "loss": 0.1296, "step": 6345 }, { "epoch": 1.2, "grad_norm": 9.375, "learning_rate": 1.8186164524913136e-06, "loss": 0.2357, "step": 6346 }, { "epoch": 1.2, "grad_norm": 20.25, "learning_rate": 1.8178795393871629e-06, "loss": 0.2711, "step": 6347 }, { "epoch": 1.2, "grad_norm": 9.875, "learning_rate": 1.8171426903116401e-06, "loss": 0.1829, "step": 6348 }, { "epoch": 1.2, "grad_norm": 15.5, "learning_rate": 1.8164059053339112e-06, "loss": 0.2386, "step": 6349 }, { "epoch": 1.2, "grad_norm": 12.5625, "learning_rate": 1.8156691845231348e-06, "loss": 0.1662, "step": 6350 }, { "epoch": 1.2, "grad_norm": 11.9375, "learning_rate": 1.8149325279484648e-06, "loss": 0.2053, "step": 6351 }, { "epoch": 1.2, "grad_norm": 9.75, "learning_rate": 1.8141959356790501e-06, "loss": 0.1776, "step": 6352 }, { "epoch": 1.2, "grad_norm": 14.625, "learning_rate": 1.8134594077840327e-06, "loss": 0.2659, "step": 6353 }, { "epoch": 1.2, "grad_norm": 9.0, "learning_rate": 1.8127229443325458e-06, "loss": 0.1921, "step": 6354 }, { "epoch": 1.2, "grad_norm": 9.5, "learning_rate": 1.8119865453937216e-06, "loss": 0.2455, "step": 6355 }, { "epoch": 1.2, "grad_norm": 14.1875, "learning_rate": 1.8112502110366833e-06, "loss": 0.1963, "step": 6356 }, { "epoch": 1.2, "grad_norm": 20.125, "learning_rate": 1.8105139413305473e-06, "loss": 0.2044, "step": 6357 }, { "epoch": 1.2, "grad_norm": 17.125, "learning_rate": 1.8097777363444251e-06, "loss": 0.3043, "step": 6358 }, { "epoch": 1.2, "grad_norm": 10.6875, "learning_rate": 1.8090415961474234e-06, "loss": 0.1973, "step": 6359 }, { "epoch": 1.2, "grad_norm": 11.625, "learning_rate": 1.8083055208086412e-06, "loss": 0.2646, "step": 6360 }, { "epoch": 1.2, "grad_norm": 10.625, "learning_rate": 1.8075695103971706e-06, "loss": 0.213, "step": 6361 }, { "epoch": 1.2, "grad_norm": 13.9375, "learning_rate": 1.806833564982099e-06, "loss": 0.2137, "step": 6362 }, { "epoch": 1.2, "grad_norm": 10.5625, "learning_rate": 1.806097684632508e-06, "loss": 0.2081, "step": 6363 }, { "epoch": 1.2, "grad_norm": 10.625, "learning_rate": 1.8053618694174729e-06, "loss": 0.2025, "step": 6364 }, { "epoch": 1.2, "grad_norm": 9.875, "learning_rate": 1.8046261194060604e-06, "loss": 0.1684, "step": 6365 }, { "epoch": 1.2, "grad_norm": 20.375, "learning_rate": 1.8038904346673364e-06, "loss": 0.2299, "step": 6366 }, { "epoch": 1.2, "grad_norm": 11.5, "learning_rate": 1.8031548152703554e-06, "loss": 0.2088, "step": 6367 }, { "epoch": 1.2, "grad_norm": 9.8125, "learning_rate": 1.802419261284168e-06, "loss": 0.1673, "step": 6368 }, { "epoch": 1.2, "grad_norm": 8.8125, "learning_rate": 1.8016837727778181e-06, "loss": 0.21, "step": 6369 }, { "epoch": 1.21, "grad_norm": 15.9375, "learning_rate": 1.8009483498203448e-06, "loss": 0.1871, "step": 6370 }, { "epoch": 1.21, "grad_norm": 14.4375, "learning_rate": 1.8002129924807793e-06, "loss": 0.263, "step": 6371 }, { "epoch": 1.21, "grad_norm": 19.0, "learning_rate": 1.7994777008281487e-06, "loss": 0.2568, "step": 6372 }, { "epoch": 1.21, "grad_norm": 12.3125, "learning_rate": 1.7987424749314702e-06, "loss": 0.2298, "step": 6373 }, { "epoch": 1.21, "grad_norm": 15.75, "learning_rate": 1.7980073148597592e-06, "loss": 0.2577, "step": 6374 }, { "epoch": 1.21, "grad_norm": 13.3125, "learning_rate": 1.797272220682023e-06, "loss": 0.214, "step": 6375 }, { "epoch": 1.21, "grad_norm": 12.25, "learning_rate": 1.7965371924672627e-06, "loss": 0.1831, "step": 6376 }, { "epoch": 1.21, "grad_norm": 8.75, "learning_rate": 1.7958022302844716e-06, "loss": 0.1278, "step": 6377 }, { "epoch": 1.21, "grad_norm": 11.125, "learning_rate": 1.7950673342026404e-06, "loss": 0.2512, "step": 6378 }, { "epoch": 1.21, "grad_norm": 10.25, "learning_rate": 1.7943325042907505e-06, "loss": 0.1692, "step": 6379 }, { "epoch": 1.21, "grad_norm": 9.4375, "learning_rate": 1.793597740617778e-06, "loss": 0.2128, "step": 6380 }, { "epoch": 1.21, "grad_norm": 21.5, "learning_rate": 1.7928630432526944e-06, "loss": 0.2331, "step": 6381 }, { "epoch": 1.21, "grad_norm": 10.3125, "learning_rate": 1.7921284122644629e-06, "loss": 0.1858, "step": 6382 }, { "epoch": 1.21, "grad_norm": 10.1875, "learning_rate": 1.7913938477220409e-06, "loss": 0.1936, "step": 6383 }, { "epoch": 1.21, "grad_norm": 14.6875, "learning_rate": 1.7906593496943782e-06, "loss": 0.2533, "step": 6384 }, { "epoch": 1.21, "grad_norm": 14.125, "learning_rate": 1.7899249182504236e-06, "loss": 0.1929, "step": 6385 }, { "epoch": 1.21, "grad_norm": 14.3125, "learning_rate": 1.7891905534591131e-06, "loss": 0.2081, "step": 6386 }, { "epoch": 1.21, "grad_norm": 12.5, "learning_rate": 1.7884562553893808e-06, "loss": 0.1794, "step": 6387 }, { "epoch": 1.21, "grad_norm": 11.125, "learning_rate": 1.7877220241101508e-06, "loss": 0.2627, "step": 6388 }, { "epoch": 1.21, "grad_norm": 10.4375, "learning_rate": 1.7869878596903468e-06, "loss": 0.2001, "step": 6389 }, { "epoch": 1.21, "grad_norm": 10.3125, "learning_rate": 1.78625376219888e-06, "loss": 0.141, "step": 6390 }, { "epoch": 1.21, "grad_norm": 11.6875, "learning_rate": 1.7855197317046586e-06, "loss": 0.2476, "step": 6391 }, { "epoch": 1.21, "grad_norm": 11.1875, "learning_rate": 1.784785768276584e-06, "loss": 0.2156, "step": 6392 }, { "epoch": 1.21, "grad_norm": 8.4375, "learning_rate": 1.784051871983552e-06, "loss": 0.1762, "step": 6393 }, { "epoch": 1.21, "grad_norm": 10.3125, "learning_rate": 1.78331804289445e-06, "loss": 0.1719, "step": 6394 }, { "epoch": 1.21, "grad_norm": 14.125, "learning_rate": 1.7825842810781601e-06, "loss": 0.1868, "step": 6395 }, { "epoch": 1.21, "grad_norm": 12.5, "learning_rate": 1.78185058660356e-06, "loss": 0.2111, "step": 6396 }, { "epoch": 1.21, "grad_norm": 11.625, "learning_rate": 1.7811169595395183e-06, "loss": 0.2029, "step": 6397 }, { "epoch": 1.21, "grad_norm": 11.8125, "learning_rate": 1.7803833999548985e-06, "loss": 0.2411, "step": 6398 }, { "epoch": 1.21, "grad_norm": 15.3125, "learning_rate": 1.779649907918557e-06, "loss": 0.2869, "step": 6399 }, { "epoch": 1.21, "grad_norm": 7.0, "learning_rate": 1.7789164834993467e-06, "loss": 0.1494, "step": 6400 }, { "epoch": 1.21, "grad_norm": 13.3125, "learning_rate": 1.77818312676611e-06, "loss": 0.199, "step": 6401 }, { "epoch": 1.21, "grad_norm": 11.25, "learning_rate": 1.7774498377876843e-06, "loss": 0.1865, "step": 6402 }, { "epoch": 1.21, "grad_norm": 8.8125, "learning_rate": 1.7767166166329037e-06, "loss": 0.1679, "step": 6403 }, { "epoch": 1.21, "grad_norm": 11.3125, "learning_rate": 1.7759834633705924e-06, "loss": 0.2034, "step": 6404 }, { "epoch": 1.21, "grad_norm": 12.5625, "learning_rate": 1.7752503780695685e-06, "loss": 0.1989, "step": 6405 }, { "epoch": 1.21, "grad_norm": 8.5625, "learning_rate": 1.774517360798645e-06, "loss": 0.191, "step": 6406 }, { "epoch": 1.21, "grad_norm": 18.5, "learning_rate": 1.773784411626629e-06, "loss": 0.2283, "step": 6407 }, { "epoch": 1.21, "grad_norm": 11.0625, "learning_rate": 1.7730515306223195e-06, "loss": 0.2512, "step": 6408 }, { "epoch": 1.21, "grad_norm": 11.3125, "learning_rate": 1.77231871785451e-06, "loss": 0.1647, "step": 6409 }, { "epoch": 1.21, "grad_norm": 9.6875, "learning_rate": 1.7715859733919863e-06, "loss": 0.2299, "step": 6410 }, { "epoch": 1.21, "grad_norm": 15.375, "learning_rate": 1.7708532973035308e-06, "loss": 0.1787, "step": 6411 }, { "epoch": 1.21, "grad_norm": 9.8125, "learning_rate": 1.7701206896579172e-06, "loss": 0.2224, "step": 6412 }, { "epoch": 1.21, "grad_norm": 15.5, "learning_rate": 1.7693881505239119e-06, "loss": 0.2142, "step": 6413 }, { "epoch": 1.21, "grad_norm": 16.375, "learning_rate": 1.7686556799702779e-06, "loss": 0.2495, "step": 6414 }, { "epoch": 1.21, "grad_norm": 11.1875, "learning_rate": 1.767923278065769e-06, "loss": 0.2107, "step": 6415 }, { "epoch": 1.21, "grad_norm": 12.0625, "learning_rate": 1.7671909448791346e-06, "loss": 0.1673, "step": 6416 }, { "epoch": 1.21, "grad_norm": 12.875, "learning_rate": 1.766458680479115e-06, "loss": 0.2631, "step": 6417 }, { "epoch": 1.21, "grad_norm": 11.25, "learning_rate": 1.7657264849344468e-06, "loss": 0.211, "step": 6418 }, { "epoch": 1.21, "grad_norm": 16.5, "learning_rate": 1.764994358313859e-06, "loss": 0.2578, "step": 6419 }, { "epoch": 1.21, "grad_norm": 11.6875, "learning_rate": 1.764262300686075e-06, "loss": 0.2133, "step": 6420 }, { "epoch": 1.21, "grad_norm": 11.4375, "learning_rate": 1.7635303121198083e-06, "loss": 0.2153, "step": 6421 }, { "epoch": 1.21, "grad_norm": 15.9375, "learning_rate": 1.7627983926837705e-06, "loss": 0.2326, "step": 6422 }, { "epoch": 1.22, "grad_norm": 10.3125, "learning_rate": 1.7620665424466648e-06, "loss": 0.2029, "step": 6423 }, { "epoch": 1.22, "grad_norm": 7.90625, "learning_rate": 1.7613347614771864e-06, "loss": 0.1478, "step": 6424 }, { "epoch": 1.22, "grad_norm": 11.5625, "learning_rate": 1.7606030498440273e-06, "loss": 0.1935, "step": 6425 }, { "epoch": 1.22, "grad_norm": 10.625, "learning_rate": 1.7598714076158703e-06, "loss": 0.1741, "step": 6426 }, { "epoch": 1.22, "grad_norm": 17.125, "learning_rate": 1.7591398348613922e-06, "loss": 0.2891, "step": 6427 }, { "epoch": 1.22, "grad_norm": 11.4375, "learning_rate": 1.7584083316492622e-06, "loss": 0.1682, "step": 6428 }, { "epoch": 1.22, "grad_norm": 13.8125, "learning_rate": 1.7576768980481477e-06, "loss": 0.1542, "step": 6429 }, { "epoch": 1.22, "grad_norm": 9.25, "learning_rate": 1.7569455341267035e-06, "loss": 0.1477, "step": 6430 }, { "epoch": 1.22, "grad_norm": 10.8125, "learning_rate": 1.756214239953582e-06, "loss": 0.2193, "step": 6431 }, { "epoch": 1.22, "grad_norm": 15.9375, "learning_rate": 1.7554830155974255e-06, "loss": 0.2417, "step": 6432 }, { "epoch": 1.22, "grad_norm": 11.5, "learning_rate": 1.7547518611268748e-06, "loss": 0.2131, "step": 6433 }, { "epoch": 1.22, "grad_norm": 10.625, "learning_rate": 1.7540207766105596e-06, "loss": 0.2918, "step": 6434 }, { "epoch": 1.22, "grad_norm": 21.25, "learning_rate": 1.7532897621171041e-06, "loss": 0.2336, "step": 6435 }, { "epoch": 1.22, "grad_norm": 11.125, "learning_rate": 1.7525588177151279e-06, "loss": 0.1793, "step": 6436 }, { "epoch": 1.22, "grad_norm": 10.0625, "learning_rate": 1.7518279434732421e-06, "loss": 0.2304, "step": 6437 }, { "epoch": 1.22, "grad_norm": 9.0625, "learning_rate": 1.7510971394600513e-06, "loss": 0.1834, "step": 6438 }, { "epoch": 1.22, "grad_norm": 10.0, "learning_rate": 1.7503664057441534e-06, "loss": 0.1845, "step": 6439 }, { "epoch": 1.22, "grad_norm": 10.0, "learning_rate": 1.7496357423941418e-06, "loss": 0.1853, "step": 6440 }, { "epoch": 1.22, "grad_norm": 11.25, "learning_rate": 1.7489051494786003e-06, "loss": 0.1912, "step": 6441 }, { "epoch": 1.22, "grad_norm": 8.125, "learning_rate": 1.7481746270661082e-06, "loss": 0.1337, "step": 6442 }, { "epoch": 1.22, "grad_norm": 10.4375, "learning_rate": 1.7474441752252367e-06, "loss": 0.1745, "step": 6443 }, { "epoch": 1.22, "grad_norm": 10.9375, "learning_rate": 1.7467137940245527e-06, "loss": 0.2182, "step": 6444 }, { "epoch": 1.22, "grad_norm": 12.75, "learning_rate": 1.7459834835326135e-06, "loss": 0.2053, "step": 6445 }, { "epoch": 1.22, "grad_norm": 12.4375, "learning_rate": 1.745253243817971e-06, "loss": 0.22, "step": 6446 }, { "epoch": 1.22, "grad_norm": 11.5625, "learning_rate": 1.7445230749491718e-06, "loss": 0.2287, "step": 6447 }, { "epoch": 1.22, "grad_norm": 14.125, "learning_rate": 1.7437929769947547e-06, "loss": 0.2134, "step": 6448 }, { "epoch": 1.22, "grad_norm": 10.0, "learning_rate": 1.7430629500232508e-06, "loss": 0.1964, "step": 6449 }, { "epoch": 1.22, "grad_norm": 9.375, "learning_rate": 1.7423329941031852e-06, "loss": 0.2472, "step": 6450 }, { "epoch": 1.22, "grad_norm": 7.875, "learning_rate": 1.7416031093030785e-06, "loss": 0.1673, "step": 6451 }, { "epoch": 1.22, "grad_norm": 12.8125, "learning_rate": 1.7408732956914422e-06, "loss": 0.2449, "step": 6452 }, { "epoch": 1.22, "grad_norm": 13.1875, "learning_rate": 1.7401435533367811e-06, "loss": 0.2105, "step": 6453 }, { "epoch": 1.22, "grad_norm": 11.5, "learning_rate": 1.7394138823075936e-06, "loss": 0.173, "step": 6454 }, { "epoch": 1.22, "grad_norm": 10.4375, "learning_rate": 1.7386842826723732e-06, "loss": 0.1603, "step": 6455 }, { "epoch": 1.22, "grad_norm": 9.5625, "learning_rate": 1.7379547544996047e-06, "loss": 0.1848, "step": 6456 }, { "epoch": 1.22, "grad_norm": 9.6875, "learning_rate": 1.7372252978577657e-06, "loss": 0.1836, "step": 6457 }, { "epoch": 1.22, "grad_norm": 10.625, "learning_rate": 1.7364959128153294e-06, "loss": 0.1771, "step": 6458 }, { "epoch": 1.22, "grad_norm": 8.25, "learning_rate": 1.7357665994407608e-06, "loss": 0.203, "step": 6459 }, { "epoch": 1.22, "grad_norm": 16.375, "learning_rate": 1.7350373578025184e-06, "loss": 0.2309, "step": 6460 }, { "epoch": 1.22, "grad_norm": 14.6875, "learning_rate": 1.7343081879690527e-06, "loss": 0.2257, "step": 6461 }, { "epoch": 1.22, "grad_norm": 12.8125, "learning_rate": 1.7335790900088104e-06, "loss": 0.237, "step": 6462 }, { "epoch": 1.22, "grad_norm": 12.75, "learning_rate": 1.7328500639902291e-06, "loss": 0.2419, "step": 6463 }, { "epoch": 1.22, "grad_norm": 12.6875, "learning_rate": 1.732121109981741e-06, "loss": 0.229, "step": 6464 }, { "epoch": 1.22, "grad_norm": 14.6875, "learning_rate": 1.7313922280517686e-06, "loss": 0.1948, "step": 6465 }, { "epoch": 1.22, "grad_norm": 10.3125, "learning_rate": 1.7306634182687327e-06, "loss": 0.1804, "step": 6466 }, { "epoch": 1.22, "grad_norm": 10.1875, "learning_rate": 1.7299346807010434e-06, "loss": 0.251, "step": 6467 }, { "epoch": 1.22, "grad_norm": 15.5, "learning_rate": 1.7292060154171034e-06, "loss": 0.247, "step": 6468 }, { "epoch": 1.22, "grad_norm": 9.4375, "learning_rate": 1.7284774224853136e-06, "loss": 0.2167, "step": 6469 }, { "epoch": 1.22, "grad_norm": 15.3125, "learning_rate": 1.727748901974063e-06, "loss": 0.2225, "step": 6470 }, { "epoch": 1.22, "grad_norm": 11.8125, "learning_rate": 1.727020453951736e-06, "loss": 0.2025, "step": 6471 }, { "epoch": 1.22, "grad_norm": 11.125, "learning_rate": 1.7262920784867083e-06, "loss": 0.1738, "step": 6472 }, { "epoch": 1.22, "grad_norm": 12.375, "learning_rate": 1.7255637756473534e-06, "loss": 0.2185, "step": 6473 }, { "epoch": 1.22, "grad_norm": 10.625, "learning_rate": 1.7248355455020327e-06, "loss": 0.1847, "step": 6474 }, { "epoch": 1.22, "grad_norm": 14.375, "learning_rate": 1.7241073881191045e-06, "loss": 0.2706, "step": 6475 }, { "epoch": 1.23, "grad_norm": 9.375, "learning_rate": 1.723379303566916e-06, "loss": 0.1914, "step": 6476 }, { "epoch": 1.23, "grad_norm": 16.625, "learning_rate": 1.7226512919138141e-06, "loss": 0.1627, "step": 6477 }, { "epoch": 1.23, "grad_norm": 13.75, "learning_rate": 1.7219233532281325e-06, "loss": 0.1888, "step": 6478 }, { "epoch": 1.23, "grad_norm": 11.3125, "learning_rate": 1.7211954875782012e-06, "loss": 0.2193, "step": 6479 }, { "epoch": 1.23, "grad_norm": 10.3125, "learning_rate": 1.7204676950323436e-06, "loss": 0.2516, "step": 6480 }, { "epoch": 1.23, "grad_norm": 9.5, "learning_rate": 1.7197399756588746e-06, "loss": 0.1829, "step": 6481 }, { "epoch": 1.23, "grad_norm": 11.6875, "learning_rate": 1.7190123295261033e-06, "loss": 0.2015, "step": 6482 }, { "epoch": 1.23, "grad_norm": 13.6875, "learning_rate": 1.7182847567023308e-06, "loss": 0.2381, "step": 6483 }, { "epoch": 1.23, "grad_norm": 12.0625, "learning_rate": 1.7175572572558547e-06, "loss": 0.2343, "step": 6484 }, { "epoch": 1.23, "grad_norm": 11.9375, "learning_rate": 1.7168298312549609e-06, "loss": 0.2376, "step": 6485 }, { "epoch": 1.23, "grad_norm": 8.75, "learning_rate": 1.716102478767931e-06, "loss": 0.141, "step": 6486 }, { "epoch": 1.23, "grad_norm": 10.0, "learning_rate": 1.7153751998630396e-06, "loss": 0.194, "step": 6487 }, { "epoch": 1.23, "grad_norm": 11.125, "learning_rate": 1.7146479946085556e-06, "loss": 0.2063, "step": 6488 }, { "epoch": 1.23, "grad_norm": 9.625, "learning_rate": 1.7139208630727377e-06, "loss": 0.2492, "step": 6489 }, { "epoch": 1.23, "grad_norm": 13.25, "learning_rate": 1.71319380532384e-06, "loss": 0.1897, "step": 6490 }, { "epoch": 1.23, "grad_norm": 10.25, "learning_rate": 1.7124668214301104e-06, "loss": 0.2416, "step": 6491 }, { "epoch": 1.23, "grad_norm": 11.125, "learning_rate": 1.7117399114597882e-06, "loss": 0.2461, "step": 6492 }, { "epoch": 1.23, "grad_norm": 13.375, "learning_rate": 1.7110130754811055e-06, "loss": 0.2181, "step": 6493 }, { "epoch": 1.23, "grad_norm": 17.875, "learning_rate": 1.7102863135622882e-06, "loss": 0.1932, "step": 6494 }, { "epoch": 1.23, "grad_norm": 9.6875, "learning_rate": 1.7095596257715563e-06, "loss": 0.1674, "step": 6495 }, { "epoch": 1.23, "grad_norm": 11.375, "learning_rate": 1.7088330121771222e-06, "loss": 0.1928, "step": 6496 }, { "epoch": 1.23, "grad_norm": 8.8125, "learning_rate": 1.7081064728471894e-06, "loss": 0.2027, "step": 6497 }, { "epoch": 1.23, "grad_norm": 11.75, "learning_rate": 1.7073800078499564e-06, "loss": 0.1872, "step": 6498 }, { "epoch": 1.23, "grad_norm": 9.75, "learning_rate": 1.7066536172536149e-06, "loss": 0.1626, "step": 6499 }, { "epoch": 1.23, "grad_norm": 15.125, "learning_rate": 1.7059273011263494e-06, "loss": 0.2156, "step": 6500 }, { "epoch": 1.23, "grad_norm": 12.8125, "learning_rate": 1.7052010595363355e-06, "loss": 0.2073, "step": 6501 }, { "epoch": 1.23, "grad_norm": 12.125, "learning_rate": 1.704474892551745e-06, "loss": 0.1861, "step": 6502 }, { "epoch": 1.23, "grad_norm": 12.3125, "learning_rate": 1.7037488002407398e-06, "loss": 0.1938, "step": 6503 }, { "epoch": 1.23, "grad_norm": 12.75, "learning_rate": 1.7030227826714774e-06, "loss": 0.1878, "step": 6504 }, { "epoch": 1.23, "grad_norm": 9.1875, "learning_rate": 1.7022968399121055e-06, "loss": 0.1977, "step": 6505 }, { "epoch": 1.23, "grad_norm": 13.375, "learning_rate": 1.7015709720307667e-06, "loss": 0.2504, "step": 6506 }, { "epoch": 1.23, "grad_norm": 12.3125, "learning_rate": 1.700845179095596e-06, "loss": 0.2613, "step": 6507 }, { "epoch": 1.23, "grad_norm": 12.625, "learning_rate": 1.7001194611747224e-06, "loss": 0.1572, "step": 6508 }, { "epoch": 1.23, "grad_norm": 9.375, "learning_rate": 1.699393818336265e-06, "loss": 0.24, "step": 6509 }, { "epoch": 1.23, "grad_norm": 9.6875, "learning_rate": 1.6986682506483395e-06, "loss": 0.1646, "step": 6510 }, { "epoch": 1.23, "grad_norm": 9.875, "learning_rate": 1.6979427581790526e-06, "loss": 0.1656, "step": 6511 }, { "epoch": 1.23, "grad_norm": 11.4375, "learning_rate": 1.6972173409965021e-06, "loss": 0.1643, "step": 6512 }, { "epoch": 1.23, "grad_norm": 9.0625, "learning_rate": 1.6964919991687835e-06, "loss": 0.1857, "step": 6513 }, { "epoch": 1.23, "grad_norm": 11.875, "learning_rate": 1.6957667327639809e-06, "loss": 0.2313, "step": 6514 }, { "epoch": 1.23, "grad_norm": 19.0, "learning_rate": 1.6950415418501743e-06, "loss": 0.2718, "step": 6515 }, { "epoch": 1.23, "grad_norm": 11.8125, "learning_rate": 1.694316426495432e-06, "loss": 0.2096, "step": 6516 }, { "epoch": 1.23, "grad_norm": 10.4375, "learning_rate": 1.6935913867678227e-06, "loss": 0.1705, "step": 6517 }, { "epoch": 1.23, "grad_norm": 10.9375, "learning_rate": 1.6928664227354009e-06, "loss": 0.1573, "step": 6518 }, { "epoch": 1.23, "grad_norm": 13.5, "learning_rate": 1.6921415344662183e-06, "loss": 0.2203, "step": 6519 }, { "epoch": 1.23, "grad_norm": 9.8125, "learning_rate": 1.6914167220283156e-06, "loss": 0.2193, "step": 6520 }, { "epoch": 1.23, "grad_norm": 13.3125, "learning_rate": 1.6906919854897322e-06, "loss": 0.2092, "step": 6521 }, { "epoch": 1.23, "grad_norm": 14.25, "learning_rate": 1.6899673249184948e-06, "loss": 0.2167, "step": 6522 }, { "epoch": 1.23, "grad_norm": 9.0625, "learning_rate": 1.689242740382625e-06, "loss": 0.1511, "step": 6523 }, { "epoch": 1.23, "grad_norm": 14.25, "learning_rate": 1.6885182319501393e-06, "loss": 0.2323, "step": 6524 }, { "epoch": 1.23, "grad_norm": 15.875, "learning_rate": 1.6877937996890431e-06, "loss": 0.2553, "step": 6525 }, { "epoch": 1.23, "grad_norm": 12.5625, "learning_rate": 1.687069443667338e-06, "loss": 0.288, "step": 6526 }, { "epoch": 1.23, "grad_norm": 12.9375, "learning_rate": 1.6863451639530155e-06, "loss": 0.1935, "step": 6527 }, { "epoch": 1.23, "grad_norm": 10.0, "learning_rate": 1.6856209606140642e-06, "loss": 0.2121, "step": 6528 }, { "epoch": 1.24, "grad_norm": 11.9375, "learning_rate": 1.6848968337184608e-06, "loss": 0.1962, "step": 6529 }, { "epoch": 1.24, "grad_norm": 12.625, "learning_rate": 1.6841727833341775e-06, "loss": 0.2477, "step": 6530 }, { "epoch": 1.24, "grad_norm": 10.6875, "learning_rate": 1.683448809529179e-06, "loss": 0.1999, "step": 6531 }, { "epoch": 1.24, "grad_norm": 13.125, "learning_rate": 1.6827249123714228e-06, "loss": 0.2374, "step": 6532 }, { "epoch": 1.24, "grad_norm": 8.8125, "learning_rate": 1.6820010919288586e-06, "loss": 0.2058, "step": 6533 }, { "epoch": 1.24, "grad_norm": 11.5, "learning_rate": 1.6812773482694286e-06, "loss": 0.1586, "step": 6534 }, { "epoch": 1.24, "grad_norm": 8.5, "learning_rate": 1.6805536814610697e-06, "loss": 0.1863, "step": 6535 }, { "epoch": 1.24, "grad_norm": 9.625, "learning_rate": 1.6798300915717106e-06, "loss": 0.1269, "step": 6536 }, { "epoch": 1.24, "grad_norm": 10.3125, "learning_rate": 1.679106578669271e-06, "loss": 0.1931, "step": 6537 }, { "epoch": 1.24, "grad_norm": 12.5625, "learning_rate": 1.6783831428216657e-06, "loss": 0.1998, "step": 6538 }, { "epoch": 1.24, "grad_norm": 9.4375, "learning_rate": 1.6776597840968017e-06, "loss": 0.2072, "step": 6539 }, { "epoch": 1.24, "grad_norm": 11.1875, "learning_rate": 1.6769365025625788e-06, "loss": 0.1928, "step": 6540 }, { "epoch": 1.24, "grad_norm": 18.375, "learning_rate": 1.6762132982868885e-06, "loss": 0.2307, "step": 6541 }, { "epoch": 1.24, "grad_norm": 13.75, "learning_rate": 1.6754901713376154e-06, "loss": 0.2265, "step": 6542 }, { "epoch": 1.24, "grad_norm": 11.25, "learning_rate": 1.6747671217826388e-06, "loss": 0.2186, "step": 6543 }, { "epoch": 1.24, "grad_norm": 8.8125, "learning_rate": 1.674044149689829e-06, "loss": 0.171, "step": 6544 }, { "epoch": 1.24, "grad_norm": 12.0625, "learning_rate": 1.6733212551270478e-06, "loss": 0.2333, "step": 6545 }, { "epoch": 1.24, "grad_norm": 9.875, "learning_rate": 1.6725984381621523e-06, "loss": 0.1755, "step": 6546 }, { "epoch": 1.24, "grad_norm": 13.8125, "learning_rate": 1.6718756988629914e-06, "loss": 0.2176, "step": 6547 }, { "epoch": 1.24, "grad_norm": 8.5, "learning_rate": 1.6711530372974062e-06, "loss": 0.1875, "step": 6548 }, { "epoch": 1.24, "grad_norm": 9.9375, "learning_rate": 1.67043045353323e-06, "loss": 0.1645, "step": 6549 }, { "epoch": 1.24, "grad_norm": 10.6875, "learning_rate": 1.6697079476382906e-06, "loss": 0.3104, "step": 6550 }, { "epoch": 1.24, "grad_norm": 13.5625, "learning_rate": 1.6689855196804077e-06, "loss": 0.1405, "step": 6551 }, { "epoch": 1.24, "grad_norm": 12.3125, "learning_rate": 1.6682631697273926e-06, "loss": 0.1651, "step": 6552 }, { "epoch": 1.24, "grad_norm": 9.875, "learning_rate": 1.6675408978470498e-06, "loss": 0.1798, "step": 6553 }, { "epoch": 1.24, "grad_norm": 9.5, "learning_rate": 1.6668187041071782e-06, "loss": 0.227, "step": 6554 }, { "epoch": 1.24, "grad_norm": 13.0625, "learning_rate": 1.6660965885755681e-06, "loss": 0.2267, "step": 6555 }, { "epoch": 1.24, "grad_norm": 12.75, "learning_rate": 1.6653745513199996e-06, "loss": 0.198, "step": 6556 }, { "epoch": 1.24, "grad_norm": 9.3125, "learning_rate": 1.6646525924082518e-06, "loss": 0.1935, "step": 6557 }, { "epoch": 1.24, "grad_norm": 9.625, "learning_rate": 1.6639307119080909e-06, "loss": 0.2251, "step": 6558 }, { "epoch": 1.24, "grad_norm": 37.5, "learning_rate": 1.6632089098872783e-06, "loss": 0.2476, "step": 6559 }, { "epoch": 1.24, "grad_norm": 18.125, "learning_rate": 1.6624871864135656e-06, "loss": 0.2006, "step": 6560 }, { "epoch": 1.24, "grad_norm": 11.5625, "learning_rate": 1.661765541554702e-06, "loss": 0.1978, "step": 6561 }, { "epoch": 1.24, "grad_norm": 12.125, "learning_rate": 1.6610439753784239e-06, "loss": 0.2043, "step": 6562 }, { "epoch": 1.24, "grad_norm": 12.0, "learning_rate": 1.6603224879524637e-06, "loss": 0.2134, "step": 6563 }, { "epoch": 1.24, "grad_norm": 19.75, "learning_rate": 1.6596010793445433e-06, "loss": 0.2352, "step": 6564 }, { "epoch": 1.24, "grad_norm": 10.8125, "learning_rate": 1.658879749622382e-06, "loss": 0.2045, "step": 6565 }, { "epoch": 1.24, "grad_norm": 11.6875, "learning_rate": 1.6581584988536875e-06, "loss": 0.2149, "step": 6566 }, { "epoch": 1.24, "grad_norm": 16.625, "learning_rate": 1.6574373271061605e-06, "loss": 0.2142, "step": 6567 }, { "epoch": 1.24, "grad_norm": 8.375, "learning_rate": 1.656716234447498e-06, "loss": 0.1918, "step": 6568 }, { "epoch": 1.24, "grad_norm": 12.625, "learning_rate": 1.655995220945384e-06, "loss": 0.2623, "step": 6569 }, { "epoch": 1.24, "grad_norm": 9.375, "learning_rate": 1.6552742866674992e-06, "loss": 0.201, "step": 6570 }, { "epoch": 1.24, "grad_norm": 10.625, "learning_rate": 1.6545534316815148e-06, "loss": 0.1496, "step": 6571 }, { "epoch": 1.24, "grad_norm": 10.5, "learning_rate": 1.6538326560550972e-06, "loss": 0.2336, "step": 6572 }, { "epoch": 1.24, "grad_norm": 10.8125, "learning_rate": 1.6531119598559014e-06, "loss": 0.1981, "step": 6573 }, { "epoch": 1.24, "grad_norm": 13.1875, "learning_rate": 1.6523913431515781e-06, "loss": 0.2164, "step": 6574 }, { "epoch": 1.24, "grad_norm": 8.125, "learning_rate": 1.6516708060097686e-06, "loss": 0.1891, "step": 6575 }, { "epoch": 1.24, "grad_norm": 9.8125, "learning_rate": 1.650950348498109e-06, "loss": 0.2323, "step": 6576 }, { "epoch": 1.24, "grad_norm": 14.0, "learning_rate": 1.6502299706842254e-06, "loss": 0.2375, "step": 6577 }, { "epoch": 1.24, "grad_norm": 10.4375, "learning_rate": 1.6495096726357374e-06, "loss": 0.2181, "step": 6578 }, { "epoch": 1.24, "grad_norm": 10.9375, "learning_rate": 1.6487894544202576e-06, "loss": 0.2239, "step": 6579 }, { "epoch": 1.24, "grad_norm": 11.1875, "learning_rate": 1.6480693161053915e-06, "loss": 0.2654, "step": 6580 }, { "epoch": 1.24, "grad_norm": 11.0, "learning_rate": 1.6473492577587352e-06, "loss": 0.1987, "step": 6581 }, { "epoch": 1.25, "grad_norm": 11.625, "learning_rate": 1.6466292794478783e-06, "loss": 0.2492, "step": 6582 }, { "epoch": 1.25, "grad_norm": 11.5625, "learning_rate": 1.6459093812404043e-06, "loss": 0.2227, "step": 6583 }, { "epoch": 1.25, "grad_norm": 13.4375, "learning_rate": 1.6451895632038877e-06, "loss": 0.2322, "step": 6584 }, { "epoch": 1.25, "grad_norm": 12.4375, "learning_rate": 1.6444698254058946e-06, "loss": 0.2111, "step": 6585 }, { "epoch": 1.25, "grad_norm": 12.625, "learning_rate": 1.643750167913985e-06, "loss": 0.2891, "step": 6586 }, { "epoch": 1.25, "grad_norm": 9.8125, "learning_rate": 1.6430305907957116e-06, "loss": 0.2122, "step": 6587 }, { "epoch": 1.25, "grad_norm": 12.5625, "learning_rate": 1.6423110941186193e-06, "loss": 0.2315, "step": 6588 }, { "epoch": 1.25, "grad_norm": 9.3125, "learning_rate": 1.6415916779502438e-06, "loss": 0.1679, "step": 6589 }, { "epoch": 1.25, "grad_norm": 9.0, "learning_rate": 1.6408723423581146e-06, "loss": 0.2288, "step": 6590 }, { "epoch": 1.25, "grad_norm": 11.75, "learning_rate": 1.6401530874097555e-06, "loss": 0.1861, "step": 6591 }, { "epoch": 1.25, "grad_norm": 7.59375, "learning_rate": 1.6394339131726793e-06, "loss": 0.1657, "step": 6592 }, { "epoch": 1.25, "grad_norm": 10.4375, "learning_rate": 1.6387148197143924e-06, "loss": 0.224, "step": 6593 }, { "epoch": 1.25, "grad_norm": 10.625, "learning_rate": 1.637995807102395e-06, "loss": 0.2505, "step": 6594 }, { "epoch": 1.25, "grad_norm": 9.8125, "learning_rate": 1.637276875404179e-06, "loss": 0.1724, "step": 6595 }, { "epoch": 1.25, "grad_norm": 13.1875, "learning_rate": 1.636558024687227e-06, "loss": 0.1575, "step": 6596 }, { "epoch": 1.25, "grad_norm": 13.5625, "learning_rate": 1.6358392550190155e-06, "loss": 0.28, "step": 6597 }, { "epoch": 1.25, "grad_norm": 12.0, "learning_rate": 1.6351205664670145e-06, "loss": 0.2201, "step": 6598 }, { "epoch": 1.25, "grad_norm": 15.375, "learning_rate": 1.6344019590986848e-06, "loss": 0.2083, "step": 6599 }, { "epoch": 1.25, "grad_norm": 9.3125, "learning_rate": 1.6336834329814797e-06, "loss": 0.2116, "step": 6600 }, { "epoch": 1.25, "grad_norm": 10.3125, "learning_rate": 1.632964988182844e-06, "loss": 0.1699, "step": 6601 }, { "epoch": 1.25, "grad_norm": 10.5625, "learning_rate": 1.6322466247702178e-06, "loss": 0.2072, "step": 6602 }, { "epoch": 1.25, "grad_norm": 14.1875, "learning_rate": 1.6315283428110318e-06, "loss": 0.2181, "step": 6603 }, { "epoch": 1.25, "grad_norm": 11.125, "learning_rate": 1.6308101423727062e-06, "loss": 0.2347, "step": 6604 }, { "epoch": 1.25, "grad_norm": 9.3125, "learning_rate": 1.6300920235226602e-06, "loss": 0.1396, "step": 6605 }, { "epoch": 1.25, "grad_norm": 13.25, "learning_rate": 1.6293739863282993e-06, "loss": 0.2098, "step": 6606 }, { "epoch": 1.25, "grad_norm": 14.1875, "learning_rate": 1.628656030857024e-06, "loss": 0.2248, "step": 6607 }, { "epoch": 1.25, "grad_norm": 11.75, "learning_rate": 1.627938157176226e-06, "loss": 0.1916, "step": 6608 }, { "epoch": 1.25, "grad_norm": 10.25, "learning_rate": 1.6272203653532904e-06, "loss": 0.2283, "step": 6609 }, { "epoch": 1.25, "grad_norm": 12.5625, "learning_rate": 1.6265026554555947e-06, "loss": 0.2684, "step": 6610 }, { "epoch": 1.25, "grad_norm": 11.75, "learning_rate": 1.6257850275505083e-06, "loss": 0.1752, "step": 6611 }, { "epoch": 1.25, "grad_norm": 13.0625, "learning_rate": 1.6250674817053913e-06, "loss": 0.2528, "step": 6612 }, { "epoch": 1.25, "grad_norm": 11.125, "learning_rate": 1.624350017987599e-06, "loss": 0.2395, "step": 6613 }, { "epoch": 1.25, "grad_norm": 14.1875, "learning_rate": 1.6236326364644773e-06, "loss": 0.2617, "step": 6614 }, { "epoch": 1.25, "grad_norm": 15.0625, "learning_rate": 1.622915337203364e-06, "loss": 0.2273, "step": 6615 }, { "epoch": 1.25, "grad_norm": 14.6875, "learning_rate": 1.6221981202715914e-06, "loss": 0.2674, "step": 6616 }, { "epoch": 1.25, "grad_norm": 12.9375, "learning_rate": 1.6214809857364814e-06, "loss": 0.2294, "step": 6617 }, { "epoch": 1.25, "grad_norm": 10.0, "learning_rate": 1.6207639336653492e-06, "loss": 0.1791, "step": 6618 }, { "epoch": 1.25, "grad_norm": 13.0, "learning_rate": 1.620046964125502e-06, "loss": 0.2371, "step": 6619 }, { "epoch": 1.25, "grad_norm": 13.9375, "learning_rate": 1.6193300771842416e-06, "loss": 0.1911, "step": 6620 }, { "epoch": 1.25, "grad_norm": 13.0625, "learning_rate": 1.618613272908858e-06, "loss": 0.2048, "step": 6621 }, { "epoch": 1.25, "grad_norm": 11.375, "learning_rate": 1.6178965513666362e-06, "loss": 0.2255, "step": 6622 }, { "epoch": 1.25, "grad_norm": 10.0, "learning_rate": 1.6171799126248521e-06, "loss": 0.1942, "step": 6623 }, { "epoch": 1.25, "grad_norm": 10.6875, "learning_rate": 1.6164633567507763e-06, "loss": 0.2101, "step": 6624 }, { "epoch": 1.25, "grad_norm": 10.0, "learning_rate": 1.615746883811668e-06, "loss": 0.1874, "step": 6625 }, { "epoch": 1.25, "grad_norm": 13.25, "learning_rate": 1.6150304938747802e-06, "loss": 0.2376, "step": 6626 }, { "epoch": 1.25, "grad_norm": 10.625, "learning_rate": 1.6143141870073597e-06, "loss": 0.25, "step": 6627 }, { "epoch": 1.25, "grad_norm": 13.5, "learning_rate": 1.6135979632766437e-06, "loss": 0.3281, "step": 6628 }, { "epoch": 1.25, "grad_norm": 13.1875, "learning_rate": 1.6128818227498615e-06, "loss": 0.2061, "step": 6629 }, { "epoch": 1.25, "grad_norm": 10.375, "learning_rate": 1.6121657654942348e-06, "loss": 0.2512, "step": 6630 }, { "epoch": 1.25, "grad_norm": 12.0625, "learning_rate": 1.611449791576979e-06, "loss": 0.2249, "step": 6631 }, { "epoch": 1.25, "grad_norm": 11.8125, "learning_rate": 1.6107339010653e-06, "loss": 0.1609, "step": 6632 }, { "epoch": 1.25, "grad_norm": 16.125, "learning_rate": 1.6100180940263957e-06, "loss": 0.2242, "step": 6633 }, { "epoch": 1.25, "grad_norm": 11.6875, "learning_rate": 1.6093023705274566e-06, "loss": 0.2447, "step": 6634 }, { "epoch": 1.26, "grad_norm": 9.625, "learning_rate": 1.6085867306356668e-06, "loss": 0.2124, "step": 6635 }, { "epoch": 1.26, "grad_norm": 13.0, "learning_rate": 1.6078711744182006e-06, "loss": 0.2426, "step": 6636 }, { "epoch": 1.26, "grad_norm": 12.4375, "learning_rate": 1.6071557019422246e-06, "loss": 0.1909, "step": 6637 }, { "epoch": 1.26, "grad_norm": 13.6875, "learning_rate": 1.606440313274899e-06, "loss": 0.2199, "step": 6638 }, { "epoch": 1.26, "grad_norm": 13.3125, "learning_rate": 1.6057250084833757e-06, "loss": 0.191, "step": 6639 }, { "epoch": 1.26, "grad_norm": 13.8125, "learning_rate": 1.6050097876347965e-06, "loss": 0.2014, "step": 6640 }, { "epoch": 1.26, "grad_norm": 13.0, "learning_rate": 1.6042946507962975e-06, "loss": 0.2234, "step": 6641 }, { "epoch": 1.26, "grad_norm": 15.0, "learning_rate": 1.6035795980350075e-06, "loss": 0.2013, "step": 6642 }, { "epoch": 1.26, "grad_norm": 9.0625, "learning_rate": 1.6028646294180462e-06, "loss": 0.1774, "step": 6643 }, { "epoch": 1.26, "grad_norm": 11.0, "learning_rate": 1.6021497450125249e-06, "loss": 0.2131, "step": 6644 }, { "epoch": 1.26, "grad_norm": 8.1875, "learning_rate": 1.6014349448855473e-06, "loss": 0.1886, "step": 6645 }, { "epoch": 1.26, "grad_norm": 9.3125, "learning_rate": 1.6007202291042109e-06, "loss": 0.2001, "step": 6646 }, { "epoch": 1.26, "grad_norm": 12.25, "learning_rate": 1.6000055977356039e-06, "loss": 0.2298, "step": 6647 }, { "epoch": 1.26, "grad_norm": 11.4375, "learning_rate": 1.599291050846804e-06, "loss": 0.2125, "step": 6648 }, { "epoch": 1.26, "grad_norm": 12.125, "learning_rate": 1.5985765885048876e-06, "loss": 0.2923, "step": 6649 }, { "epoch": 1.26, "grad_norm": 9.5, "learning_rate": 1.5978622107769167e-06, "loss": 0.1665, "step": 6650 }, { "epoch": 1.26, "grad_norm": 9.75, "learning_rate": 1.597147917729949e-06, "loss": 0.2, "step": 6651 }, { "epoch": 1.26, "grad_norm": 9.5625, "learning_rate": 1.5964337094310312e-06, "loss": 0.1964, "step": 6652 }, { "epoch": 1.26, "grad_norm": 13.1875, "learning_rate": 1.5957195859472054e-06, "loss": 0.2329, "step": 6653 }, { "epoch": 1.26, "grad_norm": 12.25, "learning_rate": 1.5950055473455042e-06, "loss": 0.234, "step": 6654 }, { "epoch": 1.26, "grad_norm": 13.375, "learning_rate": 1.5942915936929526e-06, "loss": 0.2383, "step": 6655 }, { "epoch": 1.26, "grad_norm": 11.875, "learning_rate": 1.593577725056566e-06, "loss": 0.2118, "step": 6656 }, { "epoch": 1.26, "grad_norm": 8.125, "learning_rate": 1.5928639415033538e-06, "loss": 0.1811, "step": 6657 }, { "epoch": 1.26, "grad_norm": 11.625, "learning_rate": 1.592150243100317e-06, "loss": 0.2205, "step": 6658 }, { "epoch": 1.26, "grad_norm": 10.5625, "learning_rate": 1.5914366299144482e-06, "loss": 0.2563, "step": 6659 }, { "epoch": 1.26, "grad_norm": 11.8125, "learning_rate": 1.5907231020127328e-06, "loss": 0.1667, "step": 6660 }, { "epoch": 1.26, "grad_norm": 9.625, "learning_rate": 1.5900096594621467e-06, "loss": 0.1895, "step": 6661 }, { "epoch": 1.26, "grad_norm": 11.6875, "learning_rate": 1.5892963023296588e-06, "loss": 0.2502, "step": 6662 }, { "epoch": 1.26, "grad_norm": 12.375, "learning_rate": 1.5885830306822291e-06, "loss": 0.2238, "step": 6663 }, { "epoch": 1.26, "grad_norm": 12.0625, "learning_rate": 1.5878698445868123e-06, "loss": 0.221, "step": 6664 }, { "epoch": 1.26, "grad_norm": 11.9375, "learning_rate": 1.5871567441103513e-06, "loss": 0.1691, "step": 6665 }, { "epoch": 1.26, "grad_norm": 12.8125, "learning_rate": 1.5864437293197838e-06, "loss": 0.1876, "step": 6666 }, { "epoch": 1.26, "grad_norm": 15.8125, "learning_rate": 1.5857308002820362e-06, "loss": 0.1819, "step": 6667 }, { "epoch": 1.26, "grad_norm": 12.75, "learning_rate": 1.5850179570640322e-06, "loss": 0.2273, "step": 6668 }, { "epoch": 1.26, "grad_norm": 11.5, "learning_rate": 1.5843051997326826e-06, "loss": 0.1857, "step": 6669 }, { "epoch": 1.26, "grad_norm": 8.9375, "learning_rate": 1.5835925283548913e-06, "loss": 0.2274, "step": 6670 }, { "epoch": 1.26, "grad_norm": 10.625, "learning_rate": 1.582879942997556e-06, "loss": 0.2134, "step": 6671 }, { "epoch": 1.26, "grad_norm": 18.125, "learning_rate": 1.5821674437275647e-06, "loss": 0.3006, "step": 6672 }, { "epoch": 1.26, "grad_norm": 11.625, "learning_rate": 1.5814550306117968e-06, "loss": 0.2479, "step": 6673 }, { "epoch": 1.26, "grad_norm": 10.5625, "learning_rate": 1.580742703717124e-06, "loss": 0.2256, "step": 6674 }, { "epoch": 1.26, "grad_norm": 9.4375, "learning_rate": 1.5800304631104122e-06, "loss": 0.2173, "step": 6675 }, { "epoch": 1.26, "grad_norm": 8.0, "learning_rate": 1.5793183088585166e-06, "loss": 0.187, "step": 6676 }, { "epoch": 1.26, "grad_norm": 13.4375, "learning_rate": 1.578606241028284e-06, "loss": 0.2478, "step": 6677 }, { "epoch": 1.26, "grad_norm": 14.0625, "learning_rate": 1.5778942596865543e-06, "loss": 0.2448, "step": 6678 }, { "epoch": 1.26, "grad_norm": 10.3125, "learning_rate": 1.5771823649001609e-06, "loss": 0.1771, "step": 6679 }, { "epoch": 1.26, "grad_norm": 11.6875, "learning_rate": 1.576470556735925e-06, "loss": 0.2454, "step": 6680 }, { "epoch": 1.26, "grad_norm": 10.4375, "learning_rate": 1.5757588352606624e-06, "loss": 0.2622, "step": 6681 }, { "epoch": 1.26, "grad_norm": 10.0, "learning_rate": 1.5750472005411815e-06, "loss": 0.2094, "step": 6682 }, { "epoch": 1.26, "grad_norm": 12.875, "learning_rate": 1.574335652644281e-06, "loss": 0.2044, "step": 6683 }, { "epoch": 1.26, "grad_norm": 14.5, "learning_rate": 1.5736241916367507e-06, "loss": 0.2281, "step": 6684 }, { "epoch": 1.26, "grad_norm": 8.125, "learning_rate": 1.5729128175853737e-06, "loss": 0.162, "step": 6685 }, { "epoch": 1.26, "grad_norm": 11.3125, "learning_rate": 1.572201530556925e-06, "loss": 0.1441, "step": 6686 }, { "epoch": 1.26, "grad_norm": 9.9375, "learning_rate": 1.5714903306181716e-06, "loss": 0.1736, "step": 6687 }, { "epoch": 1.27, "grad_norm": 11.8125, "learning_rate": 1.5707792178358705e-06, "loss": 0.2648, "step": 6688 }, { "epoch": 1.27, "grad_norm": 11.8125, "learning_rate": 1.5700681922767719e-06, "loss": 0.1999, "step": 6689 }, { "epoch": 1.27, "grad_norm": 11.5625, "learning_rate": 1.5693572540076185e-06, "loss": 0.2077, "step": 6690 }, { "epoch": 1.27, "grad_norm": 10.25, "learning_rate": 1.568646403095144e-06, "loss": 0.2231, "step": 6691 }, { "epoch": 1.27, "grad_norm": 9.875, "learning_rate": 1.5679356396060725e-06, "loss": 0.2348, "step": 6692 }, { "epoch": 1.27, "grad_norm": 11.8125, "learning_rate": 1.5672249636071223e-06, "loss": 0.1924, "step": 6693 }, { "epoch": 1.27, "grad_norm": 12.3125, "learning_rate": 1.5665143751650026e-06, "loss": 0.2193, "step": 6694 }, { "epoch": 1.27, "grad_norm": 14.875, "learning_rate": 1.565803874346414e-06, "loss": 0.1801, "step": 6695 }, { "epoch": 1.27, "grad_norm": 11.9375, "learning_rate": 1.5650934612180485e-06, "loss": 0.1816, "step": 6696 }, { "epoch": 1.27, "grad_norm": 11.5625, "learning_rate": 1.5643831358465911e-06, "loss": 0.2634, "step": 6697 }, { "epoch": 1.27, "grad_norm": 10.25, "learning_rate": 1.5636728982987181e-06, "loss": 0.1762, "step": 6698 }, { "epoch": 1.27, "grad_norm": 12.5625, "learning_rate": 1.5629627486410975e-06, "loss": 0.2738, "step": 6699 }, { "epoch": 1.27, "grad_norm": 9.75, "learning_rate": 1.5622526869403876e-06, "loss": 0.2239, "step": 6700 }, { "epoch": 1.27, "grad_norm": 9.875, "learning_rate": 1.5615427132632416e-06, "loss": 0.1528, "step": 6701 }, { "epoch": 1.27, "grad_norm": 13.5, "learning_rate": 1.5608328276763013e-06, "loss": 0.2301, "step": 6702 }, { "epoch": 1.27, "grad_norm": 12.5625, "learning_rate": 1.5601230302462018e-06, "loss": 0.1938, "step": 6703 }, { "epoch": 1.27, "grad_norm": 13.75, "learning_rate": 1.559413321039571e-06, "loss": 0.194, "step": 6704 }, { "epoch": 1.27, "grad_norm": 11.25, "learning_rate": 1.5587037001230255e-06, "loss": 0.1779, "step": 6705 }, { "epoch": 1.27, "grad_norm": 11.8125, "learning_rate": 1.557994167563176e-06, "loss": 0.2442, "step": 6706 }, { "epoch": 1.27, "grad_norm": 9.125, "learning_rate": 1.5572847234266237e-06, "loss": 0.1707, "step": 6707 }, { "epoch": 1.27, "grad_norm": 9.9375, "learning_rate": 1.5565753677799634e-06, "loss": 0.184, "step": 6708 }, { "epoch": 1.27, "grad_norm": 10.0625, "learning_rate": 1.555866100689779e-06, "loss": 0.1815, "step": 6709 }, { "epoch": 1.27, "grad_norm": 9.8125, "learning_rate": 1.555156922222648e-06, "loss": 0.1874, "step": 6710 }, { "epoch": 1.27, "grad_norm": 8.875, "learning_rate": 1.5544478324451367e-06, "loss": 0.2107, "step": 6711 }, { "epoch": 1.27, "grad_norm": 12.75, "learning_rate": 1.553738831423809e-06, "loss": 0.1624, "step": 6712 }, { "epoch": 1.27, "grad_norm": 15.875, "learning_rate": 1.553029919225214e-06, "loss": 0.1733, "step": 6713 }, { "epoch": 1.27, "grad_norm": 13.0, "learning_rate": 1.5523210959158958e-06, "loss": 0.2043, "step": 6714 }, { "epoch": 1.27, "grad_norm": 19.625, "learning_rate": 1.5516123615623898e-06, "loss": 0.185, "step": 6715 }, { "epoch": 1.27, "grad_norm": 9.5, "learning_rate": 1.5509037162312235e-06, "loss": 0.1768, "step": 6716 }, { "epoch": 1.27, "grad_norm": 15.3125, "learning_rate": 1.5501951599889136e-06, "loss": 0.1761, "step": 6717 }, { "epoch": 1.27, "grad_norm": 11.0, "learning_rate": 1.549486692901971e-06, "loss": 0.1663, "step": 6718 }, { "epoch": 1.27, "grad_norm": 7.96875, "learning_rate": 1.548778315036898e-06, "loss": 0.1875, "step": 6719 }, { "epoch": 1.27, "grad_norm": 9.6875, "learning_rate": 1.548070026460188e-06, "loss": 0.1701, "step": 6720 }, { "epoch": 1.27, "grad_norm": 10.875, "learning_rate": 1.5473618272383247e-06, "loss": 0.1915, "step": 6721 }, { "epoch": 1.27, "grad_norm": 14.5625, "learning_rate": 1.5466537174377848e-06, "loss": 0.2153, "step": 6722 }, { "epoch": 1.27, "grad_norm": 8.625, "learning_rate": 1.5459456971250385e-06, "loss": 0.1864, "step": 6723 }, { "epoch": 1.27, "grad_norm": 17.0, "learning_rate": 1.5452377663665436e-06, "loss": 0.2371, "step": 6724 }, { "epoch": 1.27, "grad_norm": 10.375, "learning_rate": 1.544529925228751e-06, "loss": 0.1875, "step": 6725 }, { "epoch": 1.27, "grad_norm": 12.8125, "learning_rate": 1.5438221737781054e-06, "loss": 0.2918, "step": 6726 }, { "epoch": 1.27, "grad_norm": 13.875, "learning_rate": 1.543114512081041e-06, "loss": 0.2246, "step": 6727 }, { "epoch": 1.27, "grad_norm": 10.875, "learning_rate": 1.542406940203983e-06, "loss": 0.2335, "step": 6728 }, { "epoch": 1.27, "grad_norm": 14.75, "learning_rate": 1.5416994582133487e-06, "loss": 0.18, "step": 6729 }, { "epoch": 1.27, "grad_norm": 9.5, "learning_rate": 1.540992066175549e-06, "loss": 0.2153, "step": 6730 }, { "epoch": 1.27, "grad_norm": 10.1875, "learning_rate": 1.5402847641569845e-06, "loss": 0.1762, "step": 6731 }, { "epoch": 1.27, "grad_norm": 10.3125, "learning_rate": 1.5395775522240463e-06, "loss": 0.2929, "step": 6732 }, { "epoch": 1.27, "grad_norm": 11.9375, "learning_rate": 1.5388704304431184e-06, "loss": 0.2583, "step": 6733 }, { "epoch": 1.27, "grad_norm": 10.375, "learning_rate": 1.5381633988805773e-06, "loss": 0.2188, "step": 6734 }, { "epoch": 1.27, "grad_norm": 11.0, "learning_rate": 1.5374564576027897e-06, "loss": 0.1868, "step": 6735 }, { "epoch": 1.27, "grad_norm": 9.75, "learning_rate": 1.536749606676113e-06, "loss": 0.23, "step": 6736 }, { "epoch": 1.27, "grad_norm": 8.5625, "learning_rate": 1.5360428461668985e-06, "loss": 0.1549, "step": 6737 }, { "epoch": 1.27, "grad_norm": 10.0, "learning_rate": 1.535336176141487e-06, "loss": 0.2171, "step": 6738 }, { "epoch": 1.27, "grad_norm": 14.875, "learning_rate": 1.5346295966662123e-06, "loss": 0.2074, "step": 6739 }, { "epoch": 1.28, "grad_norm": 16.5, "learning_rate": 1.5339231078073974e-06, "loss": 0.1603, "step": 6740 }, { "epoch": 1.28, "grad_norm": 12.0625, "learning_rate": 1.5332167096313599e-06, "loss": 0.2185, "step": 6741 }, { "epoch": 1.28, "grad_norm": 9.5625, "learning_rate": 1.5325104022044062e-06, "loss": 0.1794, "step": 6742 }, { "epoch": 1.28, "grad_norm": 9.4375, "learning_rate": 1.5318041855928368e-06, "loss": 0.178, "step": 6743 }, { "epoch": 1.28, "grad_norm": 11.6875, "learning_rate": 1.53109805986294e-06, "loss": 0.2448, "step": 6744 }, { "epoch": 1.28, "grad_norm": 12.4375, "learning_rate": 1.530392025080999e-06, "loss": 0.2811, "step": 6745 }, { "epoch": 1.28, "grad_norm": 9.8125, "learning_rate": 1.5296860813132874e-06, "loss": 0.1972, "step": 6746 }, { "epoch": 1.28, "grad_norm": 8.8125, "learning_rate": 1.5289802286260691e-06, "loss": 0.1957, "step": 6747 }, { "epoch": 1.28, "grad_norm": 19.125, "learning_rate": 1.528274467085602e-06, "loss": 0.2201, "step": 6748 }, { "epoch": 1.28, "grad_norm": 12.75, "learning_rate": 1.527568796758132e-06, "loss": 0.2235, "step": 6749 }, { "epoch": 1.28, "grad_norm": 11.6875, "learning_rate": 1.5268632177099002e-06, "loss": 0.2263, "step": 6750 }, { "epoch": 1.28, "grad_norm": 11.75, "learning_rate": 1.5261577300071342e-06, "loss": 0.1987, "step": 6751 }, { "epoch": 1.28, "grad_norm": 14.625, "learning_rate": 1.5254523337160599e-06, "loss": 0.1988, "step": 6752 }, { "epoch": 1.28, "grad_norm": 9.5625, "learning_rate": 1.524747028902888e-06, "loss": 0.2416, "step": 6753 }, { "epoch": 1.28, "grad_norm": 8.8125, "learning_rate": 1.5240418156338247e-06, "loss": 0.1891, "step": 6754 }, { "epoch": 1.28, "grad_norm": 11.5, "learning_rate": 1.5233366939750644e-06, "loss": 0.1889, "step": 6755 }, { "epoch": 1.28, "grad_norm": 8.8125, "learning_rate": 1.5226316639927976e-06, "loss": 0.1727, "step": 6756 }, { "epoch": 1.28, "grad_norm": 9.8125, "learning_rate": 1.5219267257532017e-06, "loss": 0.2098, "step": 6757 }, { "epoch": 1.28, "grad_norm": 12.625, "learning_rate": 1.5212218793224464e-06, "loss": 0.2175, "step": 6758 }, { "epoch": 1.28, "grad_norm": 10.875, "learning_rate": 1.5205171247666956e-06, "loss": 0.227, "step": 6759 }, { "epoch": 1.28, "grad_norm": 10.5, "learning_rate": 1.5198124621521019e-06, "loss": 0.2824, "step": 6760 }, { "epoch": 1.28, "grad_norm": 15.375, "learning_rate": 1.5191078915448088e-06, "loss": 0.2922, "step": 6761 }, { "epoch": 1.28, "grad_norm": 10.3125, "learning_rate": 1.5184034130109527e-06, "loss": 0.1802, "step": 6762 }, { "epoch": 1.28, "grad_norm": 9.9375, "learning_rate": 1.517699026616662e-06, "loss": 0.1863, "step": 6763 }, { "epoch": 1.28, "grad_norm": 10.0625, "learning_rate": 1.5169947324280538e-06, "loss": 0.1896, "step": 6764 }, { "epoch": 1.28, "grad_norm": 7.59375, "learning_rate": 1.5162905305112392e-06, "loss": 0.1619, "step": 6765 }, { "epoch": 1.28, "grad_norm": 8.5625, "learning_rate": 1.5155864209323184e-06, "loss": 0.1315, "step": 6766 }, { "epoch": 1.28, "grad_norm": 17.125, "learning_rate": 1.5148824037573858e-06, "loss": 0.2355, "step": 6767 }, { "epoch": 1.28, "grad_norm": 10.5, "learning_rate": 1.5141784790525243e-06, "loss": 0.2579, "step": 6768 }, { "epoch": 1.28, "grad_norm": 12.125, "learning_rate": 1.5134746468838085e-06, "loss": 0.2183, "step": 6769 }, { "epoch": 1.28, "grad_norm": 10.1875, "learning_rate": 1.512770907317307e-06, "loss": 0.2246, "step": 6770 }, { "epoch": 1.28, "grad_norm": 33.25, "learning_rate": 1.5120672604190767e-06, "loss": 0.2074, "step": 6771 }, { "epoch": 1.28, "grad_norm": 11.1875, "learning_rate": 1.5113637062551662e-06, "loss": 0.2146, "step": 6772 }, { "epoch": 1.28, "grad_norm": 9.8125, "learning_rate": 1.5106602448916163e-06, "loss": 0.1883, "step": 6773 }, { "epoch": 1.28, "grad_norm": 9.375, "learning_rate": 1.50995687639446e-06, "loss": 0.1839, "step": 6774 }, { "epoch": 1.28, "grad_norm": 13.5, "learning_rate": 1.50925360082972e-06, "loss": 0.2245, "step": 6775 }, { "epoch": 1.28, "grad_norm": 11.9375, "learning_rate": 1.5085504182634099e-06, "loss": 0.2394, "step": 6776 }, { "epoch": 1.28, "grad_norm": 10.625, "learning_rate": 1.507847328761535e-06, "loss": 0.1719, "step": 6777 }, { "epoch": 1.28, "grad_norm": 7.78125, "learning_rate": 1.5071443323900936e-06, "loss": 0.1749, "step": 6778 }, { "epoch": 1.28, "grad_norm": 12.6875, "learning_rate": 1.5064414292150743e-06, "loss": 0.1938, "step": 6779 }, { "epoch": 1.28, "grad_norm": 9.9375, "learning_rate": 1.5057386193024543e-06, "loss": 0.2001, "step": 6780 }, { "epoch": 1.28, "grad_norm": 12.3125, "learning_rate": 1.5050359027182054e-06, "loss": 0.1857, "step": 6781 }, { "epoch": 1.28, "grad_norm": 10.75, "learning_rate": 1.50433327952829e-06, "loss": 0.196, "step": 6782 }, { "epoch": 1.28, "grad_norm": 12.0, "learning_rate": 1.5036307497986618e-06, "loss": 0.2356, "step": 6783 }, { "epoch": 1.28, "grad_norm": 13.5625, "learning_rate": 1.502928313595263e-06, "loss": 0.2412, "step": 6784 }, { "epoch": 1.28, "grad_norm": 14.0625, "learning_rate": 1.502225970984031e-06, "loss": 0.2053, "step": 6785 }, { "epoch": 1.28, "grad_norm": 13.875, "learning_rate": 1.501523722030892e-06, "loss": 0.2659, "step": 6786 }, { "epoch": 1.28, "grad_norm": 13.25, "learning_rate": 1.5008215668017652e-06, "loss": 0.2516, "step": 6787 }, { "epoch": 1.28, "grad_norm": 12.6875, "learning_rate": 1.5001195053625572e-06, "loss": 0.1843, "step": 6788 }, { "epoch": 1.28, "grad_norm": 14.375, "learning_rate": 1.499417537779171e-06, "loss": 0.2223, "step": 6789 }, { "epoch": 1.28, "grad_norm": 9.875, "learning_rate": 1.4987156641174966e-06, "loss": 0.2169, "step": 6790 }, { "epoch": 1.28, "grad_norm": 15.0, "learning_rate": 1.4980138844434184e-06, "loss": 0.2237, "step": 6791 }, { "epoch": 1.28, "grad_norm": 9.9375, "learning_rate": 1.497312198822808e-06, "loss": 0.1965, "step": 6792 }, { "epoch": 1.29, "grad_norm": 10.1875, "learning_rate": 1.4966106073215331e-06, "loss": 0.2116, "step": 6793 }, { "epoch": 1.29, "grad_norm": 11.5, "learning_rate": 1.4959091100054491e-06, "loss": 0.2105, "step": 6794 }, { "epoch": 1.29, "grad_norm": 15.4375, "learning_rate": 1.4952077069404015e-06, "loss": 0.2008, "step": 6795 }, { "epoch": 1.29, "grad_norm": 11.0625, "learning_rate": 1.4945063981922326e-06, "loss": 0.2396, "step": 6796 }, { "epoch": 1.29, "grad_norm": 7.59375, "learning_rate": 1.4938051838267692e-06, "loss": 0.15, "step": 6797 }, { "epoch": 1.29, "grad_norm": 11.0625, "learning_rate": 1.4931040639098343e-06, "loss": 0.2366, "step": 6798 }, { "epoch": 1.29, "grad_norm": 13.25, "learning_rate": 1.4924030385072371e-06, "loss": 0.233, "step": 6799 }, { "epoch": 1.29, "grad_norm": 14.8125, "learning_rate": 1.4917021076847844e-06, "loss": 0.2016, "step": 6800 }, { "epoch": 1.29, "grad_norm": 11.8125, "learning_rate": 1.4910012715082679e-06, "loss": 0.2556, "step": 6801 }, { "epoch": 1.29, "grad_norm": 8.875, "learning_rate": 1.4903005300434748e-06, "loss": 0.1464, "step": 6802 }, { "epoch": 1.29, "grad_norm": 10.6875, "learning_rate": 1.4895998833561786e-06, "loss": 0.2543, "step": 6803 }, { "epoch": 1.29, "grad_norm": 13.375, "learning_rate": 1.488899331512151e-06, "loss": 0.2438, "step": 6804 }, { "epoch": 1.29, "grad_norm": 7.75, "learning_rate": 1.4881988745771474e-06, "loss": 0.1849, "step": 6805 }, { "epoch": 1.29, "grad_norm": 10.875, "learning_rate": 1.487498512616919e-06, "loss": 0.1837, "step": 6806 }, { "epoch": 1.29, "grad_norm": 10.0, "learning_rate": 1.4867982456972076e-06, "loss": 0.1644, "step": 6807 }, { "epoch": 1.29, "grad_norm": 10.375, "learning_rate": 1.486098073883744e-06, "loss": 0.2259, "step": 6808 }, { "epoch": 1.29, "grad_norm": 9.75, "learning_rate": 1.4853979972422508e-06, "loss": 0.1687, "step": 6809 }, { "epoch": 1.29, "grad_norm": 23.375, "learning_rate": 1.4846980158384428e-06, "loss": 0.2178, "step": 6810 }, { "epoch": 1.29, "grad_norm": 12.6875, "learning_rate": 1.4839981297380257e-06, "loss": 0.1965, "step": 6811 }, { "epoch": 1.29, "grad_norm": 11.125, "learning_rate": 1.4832983390066952e-06, "loss": 0.2004, "step": 6812 }, { "epoch": 1.29, "grad_norm": 11.4375, "learning_rate": 1.482598643710138e-06, "loss": 0.2328, "step": 6813 }, { "epoch": 1.29, "grad_norm": 11.4375, "learning_rate": 1.481899043914033e-06, "loss": 0.2065, "step": 6814 }, { "epoch": 1.29, "grad_norm": 11.25, "learning_rate": 1.4811995396840507e-06, "loss": 0.2015, "step": 6815 }, { "epoch": 1.29, "grad_norm": 12.4375, "learning_rate": 1.4805001310858493e-06, "loss": 0.2181, "step": 6816 }, { "epoch": 1.29, "grad_norm": 14.5625, "learning_rate": 1.479800818185081e-06, "loss": 0.2111, "step": 6817 }, { "epoch": 1.29, "grad_norm": 10.75, "learning_rate": 1.479101601047389e-06, "loss": 0.2473, "step": 6818 }, { "epoch": 1.29, "grad_norm": 11.1875, "learning_rate": 1.478402479738407e-06, "loss": 0.2267, "step": 6819 }, { "epoch": 1.29, "grad_norm": 9.1875, "learning_rate": 1.477703454323758e-06, "loss": 0.1906, "step": 6820 }, { "epoch": 1.29, "grad_norm": 30.5, "learning_rate": 1.4770045248690576e-06, "loss": 0.1659, "step": 6821 }, { "epoch": 1.29, "grad_norm": 10.0625, "learning_rate": 1.4763056914399133e-06, "loss": 0.2102, "step": 6822 }, { "epoch": 1.29, "grad_norm": 14.625, "learning_rate": 1.4756069541019226e-06, "loss": 0.2323, "step": 6823 }, { "epoch": 1.29, "grad_norm": 13.0625, "learning_rate": 1.4749083129206726e-06, "loss": 0.2557, "step": 6824 }, { "epoch": 1.29, "grad_norm": 9.875, "learning_rate": 1.4742097679617428e-06, "loss": 0.1788, "step": 6825 }, { "epoch": 1.29, "grad_norm": 16.75, "learning_rate": 1.4735113192907051e-06, "loss": 0.2322, "step": 6826 }, { "epoch": 1.29, "grad_norm": 11.5625, "learning_rate": 1.47281296697312e-06, "loss": 0.2316, "step": 6827 }, { "epoch": 1.29, "grad_norm": 13.6875, "learning_rate": 1.4721147110745386e-06, "loss": 0.2023, "step": 6828 }, { "epoch": 1.29, "grad_norm": 15.625, "learning_rate": 1.471416551660506e-06, "loss": 0.2431, "step": 6829 }, { "epoch": 1.29, "grad_norm": 13.0625, "learning_rate": 1.4707184887965548e-06, "loss": 0.2059, "step": 6830 }, { "epoch": 1.29, "grad_norm": 8.8125, "learning_rate": 1.4700205225482116e-06, "loss": 0.1535, "step": 6831 }, { "epoch": 1.29, "grad_norm": 17.125, "learning_rate": 1.4693226529809906e-06, "loss": 0.1688, "step": 6832 }, { "epoch": 1.29, "grad_norm": 9.625, "learning_rate": 1.4686248801604004e-06, "loss": 0.1891, "step": 6833 }, { "epoch": 1.29, "grad_norm": 20.875, "learning_rate": 1.4679272041519386e-06, "loss": 0.3087, "step": 6834 }, { "epoch": 1.29, "grad_norm": 8.4375, "learning_rate": 1.4672296250210931e-06, "loss": 0.1555, "step": 6835 }, { "epoch": 1.29, "grad_norm": 12.0625, "learning_rate": 1.4665321428333434e-06, "loss": 0.1992, "step": 6836 }, { "epoch": 1.29, "grad_norm": 12.6875, "learning_rate": 1.4658347576541614e-06, "loss": 0.2464, "step": 6837 }, { "epoch": 1.29, "grad_norm": 8.875, "learning_rate": 1.4651374695490085e-06, "loss": 0.1832, "step": 6838 }, { "epoch": 1.29, "grad_norm": 17.0, "learning_rate": 1.464440278583335e-06, "loss": 0.2616, "step": 6839 }, { "epoch": 1.29, "grad_norm": 9.9375, "learning_rate": 1.463743184822587e-06, "loss": 0.195, "step": 6840 }, { "epoch": 1.29, "grad_norm": 11.5625, "learning_rate": 1.463046188332197e-06, "loss": 0.2268, "step": 6841 }, { "epoch": 1.29, "grad_norm": 12.5625, "learning_rate": 1.4623492891775905e-06, "loss": 0.2082, "step": 6842 }, { "epoch": 1.29, "grad_norm": 17.625, "learning_rate": 1.4616524874241819e-06, "loss": 0.1586, "step": 6843 }, { "epoch": 1.29, "grad_norm": 13.8125, "learning_rate": 1.4609557831373806e-06, "loss": 0.1811, "step": 6844 }, { "epoch": 1.29, "grad_norm": 13.5, "learning_rate": 1.4602591763825825e-06, "loss": 0.1783, "step": 6845 }, { "epoch": 1.3, "grad_norm": 15.875, "learning_rate": 1.4595626672251756e-06, "loss": 0.194, "step": 6846 }, { "epoch": 1.3, "grad_norm": 12.5625, "learning_rate": 1.4588662557305405e-06, "loss": 0.2606, "step": 6847 }, { "epoch": 1.3, "grad_norm": 8.375, "learning_rate": 1.458169941964046e-06, "loss": 0.1724, "step": 6848 }, { "epoch": 1.3, "grad_norm": 10.9375, "learning_rate": 1.4574737259910536e-06, "loss": 0.1705, "step": 6849 }, { "epoch": 1.3, "grad_norm": 9.875, "learning_rate": 1.4567776078769158e-06, "loss": 0.173, "step": 6850 }, { "epoch": 1.3, "grad_norm": 10.125, "learning_rate": 1.456081587686974e-06, "loss": 0.2164, "step": 6851 }, { "epoch": 1.3, "grad_norm": 12.0625, "learning_rate": 1.4553856654865614e-06, "loss": 0.2171, "step": 6852 }, { "epoch": 1.3, "grad_norm": 9.125, "learning_rate": 1.4546898413410033e-06, "loss": 0.1662, "step": 6853 }, { "epoch": 1.3, "grad_norm": 11.9375, "learning_rate": 1.4539941153156135e-06, "loss": 0.2265, "step": 6854 }, { "epoch": 1.3, "grad_norm": 9.0625, "learning_rate": 1.4532984874756986e-06, "loss": 0.1544, "step": 6855 }, { "epoch": 1.3, "grad_norm": 14.375, "learning_rate": 1.4526029578865548e-06, "loss": 0.2026, "step": 6856 }, { "epoch": 1.3, "grad_norm": 14.5625, "learning_rate": 1.4519075266134693e-06, "loss": 0.2318, "step": 6857 }, { "epoch": 1.3, "grad_norm": 13.0625, "learning_rate": 1.4512121937217206e-06, "loss": 0.1834, "step": 6858 }, { "epoch": 1.3, "grad_norm": 11.8125, "learning_rate": 1.4505169592765767e-06, "loss": 0.2409, "step": 6859 }, { "epoch": 1.3, "grad_norm": 16.25, "learning_rate": 1.4498218233432982e-06, "loss": 0.224, "step": 6860 }, { "epoch": 1.3, "grad_norm": 10.6875, "learning_rate": 1.4491267859871336e-06, "loss": 0.1631, "step": 6861 }, { "epoch": 1.3, "grad_norm": 13.25, "learning_rate": 1.4484318472733276e-06, "loss": 0.1994, "step": 6862 }, { "epoch": 1.3, "grad_norm": 14.4375, "learning_rate": 1.4477370072671085e-06, "loss": 0.2117, "step": 6863 }, { "epoch": 1.3, "grad_norm": 11.8125, "learning_rate": 1.4470422660337001e-06, "loss": 0.2226, "step": 6864 }, { "epoch": 1.3, "grad_norm": 11.625, "learning_rate": 1.4463476236383152e-06, "loss": 0.2212, "step": 6865 }, { "epoch": 1.3, "grad_norm": 10.5, "learning_rate": 1.4456530801461604e-06, "loss": 0.1405, "step": 6866 }, { "epoch": 1.3, "grad_norm": 13.5625, "learning_rate": 1.444958635622427e-06, "loss": 0.1649, "step": 6867 }, { "epoch": 1.3, "grad_norm": 10.625, "learning_rate": 1.4442642901323022e-06, "loss": 0.2294, "step": 6868 }, { "epoch": 1.3, "grad_norm": 13.625, "learning_rate": 1.4435700437409623e-06, "loss": 0.2861, "step": 6869 }, { "epoch": 1.3, "grad_norm": 28.25, "learning_rate": 1.442875896513573e-06, "loss": 0.205, "step": 6870 }, { "epoch": 1.3, "grad_norm": 9.6875, "learning_rate": 1.4421818485152926e-06, "loss": 0.2079, "step": 6871 }, { "epoch": 1.3, "grad_norm": 11.5, "learning_rate": 1.4414878998112686e-06, "loss": 0.2028, "step": 6872 }, { "epoch": 1.3, "grad_norm": 9.6875, "learning_rate": 1.4407940504666424e-06, "loss": 0.1802, "step": 6873 }, { "epoch": 1.3, "grad_norm": 8.625, "learning_rate": 1.4401003005465408e-06, "loss": 0.1899, "step": 6874 }, { "epoch": 1.3, "grad_norm": 10.1875, "learning_rate": 1.4394066501160854e-06, "loss": 0.1944, "step": 6875 }, { "epoch": 1.3, "grad_norm": 12.5625, "learning_rate": 1.4387130992403853e-06, "loss": 0.2608, "step": 6876 }, { "epoch": 1.3, "grad_norm": 9.4375, "learning_rate": 1.4380196479845454e-06, "loss": 0.1624, "step": 6877 }, { "epoch": 1.3, "grad_norm": 13.0625, "learning_rate": 1.4373262964136552e-06, "loss": 0.2496, "step": 6878 }, { "epoch": 1.3, "grad_norm": 13.4375, "learning_rate": 1.4366330445927985e-06, "loss": 0.2783, "step": 6879 }, { "epoch": 1.3, "grad_norm": 11.4375, "learning_rate": 1.4359398925870472e-06, "loss": 0.1804, "step": 6880 }, { "epoch": 1.3, "grad_norm": 13.8125, "learning_rate": 1.4352468404614694e-06, "loss": 0.2538, "step": 6881 }, { "epoch": 1.3, "grad_norm": 12.625, "learning_rate": 1.434553888281116e-06, "loss": 0.1938, "step": 6882 }, { "epoch": 1.3, "grad_norm": 9.875, "learning_rate": 1.4338610361110327e-06, "loss": 0.1971, "step": 6883 }, { "epoch": 1.3, "grad_norm": 13.4375, "learning_rate": 1.4331682840162578e-06, "loss": 0.3134, "step": 6884 }, { "epoch": 1.3, "grad_norm": 10.8125, "learning_rate": 1.4324756320618175e-06, "loss": 0.2198, "step": 6885 }, { "epoch": 1.3, "grad_norm": 12.125, "learning_rate": 1.431783080312727e-06, "loss": 0.2382, "step": 6886 }, { "epoch": 1.3, "grad_norm": 10.6875, "learning_rate": 1.431090628833994e-06, "loss": 0.1981, "step": 6887 }, { "epoch": 1.3, "grad_norm": 11.4375, "learning_rate": 1.430398277690619e-06, "loss": 0.2657, "step": 6888 }, { "epoch": 1.3, "grad_norm": 11.5, "learning_rate": 1.4297060269475915e-06, "loss": 0.1682, "step": 6889 }, { "epoch": 1.3, "grad_norm": 11.5, "learning_rate": 1.429013876669888e-06, "loss": 0.2186, "step": 6890 }, { "epoch": 1.3, "grad_norm": 18.5, "learning_rate": 1.4283218269224792e-06, "loss": 0.2838, "step": 6891 }, { "epoch": 1.3, "grad_norm": 15.3125, "learning_rate": 1.4276298777703277e-06, "loss": 0.1993, "step": 6892 }, { "epoch": 1.3, "grad_norm": 10.3125, "learning_rate": 1.4269380292783846e-06, "loss": 0.1908, "step": 6893 }, { "epoch": 1.3, "grad_norm": 10.625, "learning_rate": 1.4262462815115887e-06, "loss": 0.2025, "step": 6894 }, { "epoch": 1.3, "grad_norm": 10.6875, "learning_rate": 1.4255546345348756e-06, "loss": 0.265, "step": 6895 }, { "epoch": 1.3, "grad_norm": 11.875, "learning_rate": 1.4248630884131675e-06, "loss": 0.2297, "step": 6896 }, { "epoch": 1.3, "grad_norm": 10.75, "learning_rate": 1.4241716432113767e-06, "loss": 0.1529, "step": 6897 }, { "epoch": 1.3, "grad_norm": 14.9375, "learning_rate": 1.4234802989944064e-06, "loss": 0.2065, "step": 6898 }, { "epoch": 1.31, "grad_norm": 8.0625, "learning_rate": 1.4227890558271533e-06, "loss": 0.1758, "step": 6899 }, { "epoch": 1.31, "grad_norm": 13.25, "learning_rate": 1.4220979137745022e-06, "loss": 0.2054, "step": 6900 }, { "epoch": 1.31, "grad_norm": 11.1875, "learning_rate": 1.4214068729013265e-06, "loss": 0.1419, "step": 6901 }, { "epoch": 1.31, "grad_norm": 22.625, "learning_rate": 1.4207159332724923e-06, "loss": 0.2288, "step": 6902 }, { "epoch": 1.31, "grad_norm": 11.8125, "learning_rate": 1.4200250949528582e-06, "loss": 0.2852, "step": 6903 }, { "epoch": 1.31, "grad_norm": 9.9375, "learning_rate": 1.4193343580072714e-06, "loss": 0.1932, "step": 6904 }, { "epoch": 1.31, "grad_norm": 10.375, "learning_rate": 1.4186437225005656e-06, "loss": 0.1687, "step": 6905 }, { "epoch": 1.31, "grad_norm": 9.8125, "learning_rate": 1.4179531884975718e-06, "loss": 0.1759, "step": 6906 }, { "epoch": 1.31, "grad_norm": 9.25, "learning_rate": 1.417262756063108e-06, "loss": 0.1782, "step": 6907 }, { "epoch": 1.31, "grad_norm": 13.1875, "learning_rate": 1.4165724252619834e-06, "loss": 0.2383, "step": 6908 }, { "epoch": 1.31, "grad_norm": 10.5, "learning_rate": 1.4158821961589947e-06, "loss": 0.1613, "step": 6909 }, { "epoch": 1.31, "grad_norm": 19.625, "learning_rate": 1.4151920688189346e-06, "loss": 0.231, "step": 6910 }, { "epoch": 1.31, "grad_norm": 13.625, "learning_rate": 1.4145020433065826e-06, "loss": 0.2497, "step": 6911 }, { "epoch": 1.31, "grad_norm": 8.8125, "learning_rate": 1.4138121196867095e-06, "loss": 0.2075, "step": 6912 }, { "epoch": 1.31, "grad_norm": 12.375, "learning_rate": 1.4131222980240742e-06, "loss": 0.207, "step": 6913 }, { "epoch": 1.31, "grad_norm": 10.3125, "learning_rate": 1.4124325783834309e-06, "loss": 0.1901, "step": 6914 }, { "epoch": 1.31, "grad_norm": 13.6875, "learning_rate": 1.4117429608295203e-06, "loss": 0.2125, "step": 6915 }, { "epoch": 1.31, "grad_norm": 16.5, "learning_rate": 1.4110534454270752e-06, "loss": 0.2154, "step": 6916 }, { "epoch": 1.31, "grad_norm": 14.8125, "learning_rate": 1.4103640322408179e-06, "loss": 0.3136, "step": 6917 }, { "epoch": 1.31, "grad_norm": 14.875, "learning_rate": 1.4096747213354622e-06, "loss": 0.2503, "step": 6918 }, { "epoch": 1.31, "grad_norm": 12.9375, "learning_rate": 1.4089855127757112e-06, "loss": 0.2134, "step": 6919 }, { "epoch": 1.31, "grad_norm": 13.9375, "learning_rate": 1.4082964066262595e-06, "loss": 0.2014, "step": 6920 }, { "epoch": 1.31, "grad_norm": 12.8125, "learning_rate": 1.4076074029517902e-06, "loss": 0.2193, "step": 6921 }, { "epoch": 1.31, "grad_norm": 11.875, "learning_rate": 1.406918501816979e-06, "loss": 0.2278, "step": 6922 }, { "epoch": 1.31, "grad_norm": 11.0, "learning_rate": 1.4062297032864912e-06, "loss": 0.2281, "step": 6923 }, { "epoch": 1.31, "grad_norm": 11.125, "learning_rate": 1.4055410074249815e-06, "loss": 0.1789, "step": 6924 }, { "epoch": 1.31, "grad_norm": 15.4375, "learning_rate": 1.404852414297096e-06, "loss": 0.2091, "step": 6925 }, { "epoch": 1.31, "grad_norm": 13.9375, "learning_rate": 1.4041639239674712e-06, "loss": 0.1949, "step": 6926 }, { "epoch": 1.31, "grad_norm": 15.875, "learning_rate": 1.4034755365007335e-06, "loss": 0.2257, "step": 6927 }, { "epoch": 1.31, "grad_norm": 14.3125, "learning_rate": 1.4027872519614997e-06, "loss": 0.2247, "step": 6928 }, { "epoch": 1.31, "grad_norm": 10.25, "learning_rate": 1.4020990704143773e-06, "loss": 0.2237, "step": 6929 }, { "epoch": 1.31, "grad_norm": 15.8125, "learning_rate": 1.4014109919239632e-06, "loss": 0.26, "step": 6930 }, { "epoch": 1.31, "grad_norm": 10.6875, "learning_rate": 1.4007230165548458e-06, "loss": 0.1495, "step": 6931 }, { "epoch": 1.31, "grad_norm": 9.9375, "learning_rate": 1.4000351443716037e-06, "loss": 0.2111, "step": 6932 }, { "epoch": 1.31, "grad_norm": 16.75, "learning_rate": 1.3993473754388045e-06, "loss": 0.2169, "step": 6933 }, { "epoch": 1.31, "grad_norm": 18.75, "learning_rate": 1.3986597098210076e-06, "loss": 0.314, "step": 6934 }, { "epoch": 1.31, "grad_norm": 10.3125, "learning_rate": 1.3979721475827623e-06, "loss": 0.1941, "step": 6935 }, { "epoch": 1.31, "grad_norm": 8.1875, "learning_rate": 1.3972846887886077e-06, "loss": 0.1424, "step": 6936 }, { "epoch": 1.31, "grad_norm": 10.3125, "learning_rate": 1.3965973335030733e-06, "loss": 0.1554, "step": 6937 }, { "epoch": 1.31, "grad_norm": 9.5, "learning_rate": 1.3959100817906795e-06, "loss": 0.1776, "step": 6938 }, { "epoch": 1.31, "grad_norm": 10.5, "learning_rate": 1.3952229337159367e-06, "loss": 0.1622, "step": 6939 }, { "epoch": 1.31, "grad_norm": 11.6875, "learning_rate": 1.3945358893433452e-06, "loss": 0.2007, "step": 6940 }, { "epoch": 1.31, "grad_norm": 8.25, "learning_rate": 1.3938489487373958e-06, "loss": 0.1838, "step": 6941 }, { "epoch": 1.31, "grad_norm": 11.75, "learning_rate": 1.3931621119625692e-06, "loss": 0.1852, "step": 6942 }, { "epoch": 1.31, "grad_norm": 11.0625, "learning_rate": 1.3924753790833378e-06, "loss": 0.2132, "step": 6943 }, { "epoch": 1.31, "grad_norm": 11.4375, "learning_rate": 1.3917887501641624e-06, "loss": 0.141, "step": 6944 }, { "epoch": 1.31, "grad_norm": 10.8125, "learning_rate": 1.391102225269495e-06, "loss": 0.2112, "step": 6945 }, { "epoch": 1.31, "grad_norm": 9.0625, "learning_rate": 1.3904158044637772e-06, "loss": 0.2131, "step": 6946 }, { "epoch": 1.31, "grad_norm": 10.375, "learning_rate": 1.3897294878114422e-06, "loss": 0.1898, "step": 6947 }, { "epoch": 1.31, "grad_norm": 13.625, "learning_rate": 1.3890432753769118e-06, "loss": 0.1866, "step": 6948 }, { "epoch": 1.31, "grad_norm": 12.0, "learning_rate": 1.3883571672245983e-06, "loss": 0.1821, "step": 6949 }, { "epoch": 1.31, "grad_norm": 12.1875, "learning_rate": 1.3876711634189069e-06, "loss": 0.2114, "step": 6950 }, { "epoch": 1.31, "grad_norm": 9.125, "learning_rate": 1.3869852640242282e-06, "loss": 0.1996, "step": 6951 }, { "epoch": 1.32, "grad_norm": 11.1875, "learning_rate": 1.3862994691049466e-06, "loss": 0.2259, "step": 6952 }, { "epoch": 1.32, "grad_norm": 11.625, "learning_rate": 1.3856137787254356e-06, "loss": 0.1921, "step": 6953 }, { "epoch": 1.32, "grad_norm": 9.0, "learning_rate": 1.384928192950059e-06, "loss": 0.2009, "step": 6954 }, { "epoch": 1.32, "grad_norm": 11.375, "learning_rate": 1.3842427118431701e-06, "loss": 0.1982, "step": 6955 }, { "epoch": 1.32, "grad_norm": 12.25, "learning_rate": 1.3835573354691136e-06, "loss": 0.2177, "step": 6956 }, { "epoch": 1.32, "grad_norm": 10.6875, "learning_rate": 1.3828720638922238e-06, "loss": 0.2148, "step": 6957 }, { "epoch": 1.32, "grad_norm": 13.875, "learning_rate": 1.3821868971768248e-06, "loss": 0.2083, "step": 6958 }, { "epoch": 1.32, "grad_norm": 10.6875, "learning_rate": 1.381501835387231e-06, "loss": 0.1922, "step": 6959 }, { "epoch": 1.32, "grad_norm": 10.1875, "learning_rate": 1.3808168785877468e-06, "loss": 0.1923, "step": 6960 }, { "epoch": 1.32, "grad_norm": 10.6875, "learning_rate": 1.38013202684267e-06, "loss": 0.2034, "step": 6961 }, { "epoch": 1.32, "grad_norm": 10.6875, "learning_rate": 1.3794472802162815e-06, "loss": 0.2361, "step": 6962 }, { "epoch": 1.32, "grad_norm": 13.75, "learning_rate": 1.3787626387728587e-06, "loss": 0.273, "step": 6963 }, { "epoch": 1.32, "grad_norm": 11.4375, "learning_rate": 1.3780781025766649e-06, "loss": 0.2388, "step": 6964 }, { "epoch": 1.32, "grad_norm": 11.8125, "learning_rate": 1.3773936716919594e-06, "loss": 0.2185, "step": 6965 }, { "epoch": 1.32, "grad_norm": 12.125, "learning_rate": 1.376709346182984e-06, "loss": 0.2574, "step": 6966 }, { "epoch": 1.32, "grad_norm": 10.375, "learning_rate": 1.3760251261139756e-06, "loss": 0.2161, "step": 6967 }, { "epoch": 1.32, "grad_norm": 13.5, "learning_rate": 1.375341011549159e-06, "loss": 0.2754, "step": 6968 }, { "epoch": 1.32, "grad_norm": 12.625, "learning_rate": 1.3746570025527528e-06, "loss": 0.2211, "step": 6969 }, { "epoch": 1.32, "grad_norm": 9.5625, "learning_rate": 1.37397309918896e-06, "loss": 0.2256, "step": 6970 }, { "epoch": 1.32, "grad_norm": 10.4375, "learning_rate": 1.3732893015219767e-06, "loss": 0.2135, "step": 6971 }, { "epoch": 1.32, "grad_norm": 10.5625, "learning_rate": 1.3726056096159906e-06, "loss": 0.2101, "step": 6972 }, { "epoch": 1.32, "grad_norm": 8.1875, "learning_rate": 1.3719220235351782e-06, "loss": 0.1631, "step": 6973 }, { "epoch": 1.32, "grad_norm": 12.4375, "learning_rate": 1.3712385433437036e-06, "loss": 0.2908, "step": 6974 }, { "epoch": 1.32, "grad_norm": 14.5625, "learning_rate": 1.370555169105723e-06, "loss": 0.1959, "step": 6975 }, { "epoch": 1.32, "grad_norm": 10.25, "learning_rate": 1.369871900885385e-06, "loss": 0.1966, "step": 6976 }, { "epoch": 1.32, "grad_norm": 10.9375, "learning_rate": 1.3691887387468253e-06, "loss": 0.2438, "step": 6977 }, { "epoch": 1.32, "grad_norm": 12.25, "learning_rate": 1.3685056827541687e-06, "loss": 0.1429, "step": 6978 }, { "epoch": 1.32, "grad_norm": 10.6875, "learning_rate": 1.3678227329715321e-06, "loss": 0.2056, "step": 6979 }, { "epoch": 1.32, "grad_norm": 11.5, "learning_rate": 1.3671398894630231e-06, "loss": 0.2242, "step": 6980 }, { "epoch": 1.32, "grad_norm": 9.8125, "learning_rate": 1.366457152292739e-06, "loss": 0.1599, "step": 6981 }, { "epoch": 1.32, "grad_norm": 11.1875, "learning_rate": 1.3657745215247624e-06, "loss": 0.197, "step": 6982 }, { "epoch": 1.32, "grad_norm": 15.0625, "learning_rate": 1.3650919972231739e-06, "loss": 0.1989, "step": 6983 }, { "epoch": 1.32, "grad_norm": 11.9375, "learning_rate": 1.364409579452039e-06, "loss": 0.1859, "step": 6984 }, { "epoch": 1.32, "grad_norm": 10.0, "learning_rate": 1.3637272682754132e-06, "loss": 0.19, "step": 6985 }, { "epoch": 1.32, "grad_norm": 15.6875, "learning_rate": 1.3630450637573423e-06, "loss": 0.1963, "step": 6986 }, { "epoch": 1.32, "grad_norm": 13.0, "learning_rate": 1.3623629659618653e-06, "loss": 0.2365, "step": 6987 }, { "epoch": 1.32, "grad_norm": 11.125, "learning_rate": 1.3616809749530085e-06, "loss": 0.1716, "step": 6988 }, { "epoch": 1.32, "grad_norm": 21.375, "learning_rate": 1.3609990907947861e-06, "loss": 0.2196, "step": 6989 }, { "epoch": 1.32, "grad_norm": 18.5, "learning_rate": 1.3603173135512051e-06, "loss": 0.2126, "step": 6990 }, { "epoch": 1.32, "grad_norm": 10.125, "learning_rate": 1.3596356432862633e-06, "loss": 0.1908, "step": 6991 }, { "epoch": 1.32, "grad_norm": 12.375, "learning_rate": 1.3589540800639477e-06, "loss": 0.1812, "step": 6992 }, { "epoch": 1.32, "grad_norm": 8.9375, "learning_rate": 1.3582726239482325e-06, "loss": 0.1824, "step": 6993 }, { "epoch": 1.32, "grad_norm": 8.75, "learning_rate": 1.3575912750030833e-06, "loss": 0.15, "step": 6994 }, { "epoch": 1.32, "grad_norm": 13.75, "learning_rate": 1.356910033292459e-06, "loss": 0.2629, "step": 6995 }, { "epoch": 1.32, "grad_norm": 11.0, "learning_rate": 1.3562288988803057e-06, "loss": 0.2162, "step": 6996 }, { "epoch": 1.32, "grad_norm": 8.6875, "learning_rate": 1.355547871830556e-06, "loss": 0.1617, "step": 6997 }, { "epoch": 1.32, "grad_norm": 8.9375, "learning_rate": 1.3548669522071395e-06, "loss": 0.1973, "step": 6998 }, { "epoch": 1.32, "grad_norm": 11.875, "learning_rate": 1.3541861400739703e-06, "loss": 0.2235, "step": 6999 }, { "epoch": 1.32, "grad_norm": 11.0, "learning_rate": 1.3535054354949563e-06, "loss": 0.1797, "step": 7000 }, { "epoch": 1.32, "grad_norm": 14.4375, "learning_rate": 1.3528248385339894e-06, "loss": 0.215, "step": 7001 }, { "epoch": 1.32, "grad_norm": 11.0625, "learning_rate": 1.3521443492549586e-06, "loss": 0.1947, "step": 7002 }, { "epoch": 1.32, "grad_norm": 14.125, "learning_rate": 1.3514639677217379e-06, "loss": 0.2579, "step": 7003 }, { "epoch": 1.32, "grad_norm": 12.125, "learning_rate": 1.3507836939981943e-06, "loss": 0.2485, "step": 7004 }, { "epoch": 1.33, "grad_norm": 12.8125, "learning_rate": 1.35010352814818e-06, "loss": 0.2146, "step": 7005 }, { "epoch": 1.33, "grad_norm": 7.53125, "learning_rate": 1.3494234702355427e-06, "loss": 0.1707, "step": 7006 }, { "epoch": 1.33, "grad_norm": 10.8125, "learning_rate": 1.348743520324117e-06, "loss": 0.163, "step": 7007 }, { "epoch": 1.33, "grad_norm": 16.375, "learning_rate": 1.3480636784777274e-06, "loss": 0.2153, "step": 7008 }, { "epoch": 1.33, "grad_norm": 9.5, "learning_rate": 1.3473839447601888e-06, "loss": 0.184, "step": 7009 }, { "epoch": 1.33, "grad_norm": 11.8125, "learning_rate": 1.3467043192353058e-06, "loss": 0.2648, "step": 7010 }, { "epoch": 1.33, "grad_norm": 10.875, "learning_rate": 1.3460248019668737e-06, "loss": 0.1983, "step": 7011 }, { "epoch": 1.33, "grad_norm": 12.625, "learning_rate": 1.3453453930186738e-06, "loss": 0.2588, "step": 7012 }, { "epoch": 1.33, "grad_norm": 8.5625, "learning_rate": 1.3446660924544836e-06, "loss": 0.1763, "step": 7013 }, { "epoch": 1.33, "grad_norm": 12.1875, "learning_rate": 1.3439869003380651e-06, "loss": 0.2333, "step": 7014 }, { "epoch": 1.33, "grad_norm": 15.0, "learning_rate": 1.3433078167331742e-06, "loss": 0.2432, "step": 7015 }, { "epoch": 1.33, "grad_norm": 13.5625, "learning_rate": 1.3426288417035504e-06, "loss": 0.2202, "step": 7016 }, { "epoch": 1.33, "grad_norm": 10.3125, "learning_rate": 1.3419499753129312e-06, "loss": 0.1778, "step": 7017 }, { "epoch": 1.33, "grad_norm": 12.125, "learning_rate": 1.3412712176250377e-06, "loss": 0.1772, "step": 7018 }, { "epoch": 1.33, "grad_norm": 10.875, "learning_rate": 1.3405925687035837e-06, "loss": 0.1929, "step": 7019 }, { "epoch": 1.33, "grad_norm": 17.5, "learning_rate": 1.3399140286122713e-06, "loss": 0.2592, "step": 7020 }, { "epoch": 1.33, "grad_norm": 14.25, "learning_rate": 1.3392355974147936e-06, "loss": 0.2082, "step": 7021 }, { "epoch": 1.33, "grad_norm": 11.875, "learning_rate": 1.3385572751748324e-06, "loss": 0.2325, "step": 7022 }, { "epoch": 1.33, "grad_norm": 12.8125, "learning_rate": 1.3378790619560602e-06, "loss": 0.2111, "step": 7023 }, { "epoch": 1.33, "grad_norm": 18.625, "learning_rate": 1.337200957822139e-06, "loss": 0.2223, "step": 7024 }, { "epoch": 1.33, "grad_norm": 12.4375, "learning_rate": 1.3365229628367198e-06, "loss": 0.1994, "step": 7025 }, { "epoch": 1.33, "grad_norm": 10.4375, "learning_rate": 1.3358450770634446e-06, "loss": 0.2067, "step": 7026 }, { "epoch": 1.33, "grad_norm": 12.6875, "learning_rate": 1.3351673005659443e-06, "loss": 0.2475, "step": 7027 }, { "epoch": 1.33, "grad_norm": 10.9375, "learning_rate": 1.3344896334078395e-06, "loss": 0.1757, "step": 7028 }, { "epoch": 1.33, "grad_norm": 10.0, "learning_rate": 1.333812075652741e-06, "loss": 0.2392, "step": 7029 }, { "epoch": 1.33, "grad_norm": 11.5, "learning_rate": 1.3331346273642497e-06, "loss": 0.2266, "step": 7030 }, { "epoch": 1.33, "grad_norm": 11.25, "learning_rate": 1.332457288605955e-06, "loss": 0.2467, "step": 7031 }, { "epoch": 1.33, "grad_norm": 10.1875, "learning_rate": 1.3317800594414368e-06, "loss": 0.1495, "step": 7032 }, { "epoch": 1.33, "grad_norm": 10.9375, "learning_rate": 1.3311029399342644e-06, "loss": 0.1871, "step": 7033 }, { "epoch": 1.33, "grad_norm": 9.3125, "learning_rate": 1.3304259301479977e-06, "loss": 0.1899, "step": 7034 }, { "epoch": 1.33, "grad_norm": 12.9375, "learning_rate": 1.3297490301461851e-06, "loss": 0.1938, "step": 7035 }, { "epoch": 1.33, "grad_norm": 15.75, "learning_rate": 1.329072239992365e-06, "loss": 0.2177, "step": 7036 }, { "epoch": 1.33, "grad_norm": 12.625, "learning_rate": 1.328395559750066e-06, "loss": 0.2332, "step": 7037 }, { "epoch": 1.33, "grad_norm": 11.25, "learning_rate": 1.3277189894828062e-06, "loss": 0.2285, "step": 7038 }, { "epoch": 1.33, "grad_norm": 11.375, "learning_rate": 1.327042529254093e-06, "loss": 0.2398, "step": 7039 }, { "epoch": 1.33, "grad_norm": 11.125, "learning_rate": 1.3263661791274236e-06, "loss": 0.1821, "step": 7040 }, { "epoch": 1.33, "grad_norm": 10.875, "learning_rate": 1.3256899391662854e-06, "loss": 0.1944, "step": 7041 }, { "epoch": 1.33, "grad_norm": 11.3125, "learning_rate": 1.3250138094341546e-06, "loss": 0.2009, "step": 7042 }, { "epoch": 1.33, "grad_norm": 16.125, "learning_rate": 1.324337789994498e-06, "loss": 0.2486, "step": 7043 }, { "epoch": 1.33, "grad_norm": 11.3125, "learning_rate": 1.323661880910771e-06, "loss": 0.1977, "step": 7044 }, { "epoch": 1.33, "grad_norm": 10.5625, "learning_rate": 1.322986082246419e-06, "loss": 0.1907, "step": 7045 }, { "epoch": 1.33, "grad_norm": 13.1875, "learning_rate": 1.3223103940648785e-06, "loss": 0.3027, "step": 7046 }, { "epoch": 1.33, "grad_norm": 13.5, "learning_rate": 1.321634816429573e-06, "loss": 0.215, "step": 7047 }, { "epoch": 1.33, "grad_norm": 11.625, "learning_rate": 1.3209593494039174e-06, "loss": 0.1911, "step": 7048 }, { "epoch": 1.33, "grad_norm": 8.1875, "learning_rate": 1.3202839930513162e-06, "loss": 0.1312, "step": 7049 }, { "epoch": 1.33, "grad_norm": 11.8125, "learning_rate": 1.3196087474351627e-06, "loss": 0.2282, "step": 7050 }, { "epoch": 1.33, "grad_norm": 11.3125, "learning_rate": 1.31893361261884e-06, "loss": 0.2297, "step": 7051 }, { "epoch": 1.33, "grad_norm": 12.3125, "learning_rate": 1.3182585886657207e-06, "loss": 0.2007, "step": 7052 }, { "epoch": 1.33, "grad_norm": 11.4375, "learning_rate": 1.3175836756391694e-06, "loss": 0.2053, "step": 7053 }, { "epoch": 1.33, "grad_norm": 13.25, "learning_rate": 1.316908873602536e-06, "loss": 0.1944, "step": 7054 }, { "epoch": 1.33, "grad_norm": 13.6875, "learning_rate": 1.3162341826191628e-06, "loss": 0.1846, "step": 7055 }, { "epoch": 1.33, "grad_norm": 12.6875, "learning_rate": 1.31555960275238e-06, "loss": 0.2152, "step": 7056 }, { "epoch": 1.33, "grad_norm": 12.6875, "learning_rate": 1.3148851340655112e-06, "loss": 0.1688, "step": 7057 }, { "epoch": 1.34, "grad_norm": 11.9375, "learning_rate": 1.3142107766218642e-06, "loss": 0.1954, "step": 7058 }, { "epoch": 1.34, "grad_norm": 8.25, "learning_rate": 1.31353653048474e-06, "loss": 0.2012, "step": 7059 }, { "epoch": 1.34, "grad_norm": 11.0625, "learning_rate": 1.312862395717427e-06, "loss": 0.1744, "step": 7060 }, { "epoch": 1.34, "grad_norm": 9.6875, "learning_rate": 1.3121883723832069e-06, "loss": 0.2291, "step": 7061 }, { "epoch": 1.34, "grad_norm": 11.0, "learning_rate": 1.3115144605453452e-06, "loss": 0.1997, "step": 7062 }, { "epoch": 1.34, "grad_norm": 13.5, "learning_rate": 1.310840660267101e-06, "loss": 0.1779, "step": 7063 }, { "epoch": 1.34, "grad_norm": 10.5625, "learning_rate": 1.3101669716117227e-06, "loss": 0.178, "step": 7064 }, { "epoch": 1.34, "grad_norm": 10.0, "learning_rate": 1.3094933946424481e-06, "loss": 0.2417, "step": 7065 }, { "epoch": 1.34, "grad_norm": 15.5625, "learning_rate": 1.3088199294225018e-06, "loss": 0.2216, "step": 7066 }, { "epoch": 1.34, "grad_norm": 7.40625, "learning_rate": 1.3081465760150998e-06, "loss": 0.1806, "step": 7067 }, { "epoch": 1.34, "grad_norm": 13.375, "learning_rate": 1.307473334483451e-06, "loss": 0.2502, "step": 7068 }, { "epoch": 1.34, "grad_norm": 15.8125, "learning_rate": 1.306800204890748e-06, "loss": 0.2271, "step": 7069 }, { "epoch": 1.34, "grad_norm": 9.0, "learning_rate": 1.3061271873001757e-06, "loss": 0.1661, "step": 7070 }, { "epoch": 1.34, "grad_norm": 7.75, "learning_rate": 1.3054542817749076e-06, "loss": 0.1543, "step": 7071 }, { "epoch": 1.34, "grad_norm": 11.375, "learning_rate": 1.3047814883781102e-06, "loss": 0.2129, "step": 7072 }, { "epoch": 1.34, "grad_norm": 11.125, "learning_rate": 1.3041088071729339e-06, "loss": 0.2502, "step": 7073 }, { "epoch": 1.34, "grad_norm": 13.6875, "learning_rate": 1.303436238222521e-06, "loss": 0.3044, "step": 7074 }, { "epoch": 1.34, "grad_norm": 10.4375, "learning_rate": 1.3027637815900057e-06, "loss": 0.2118, "step": 7075 }, { "epoch": 1.34, "grad_norm": 17.0, "learning_rate": 1.30209143733851e-06, "loss": 0.1985, "step": 7076 }, { "epoch": 1.34, "grad_norm": 12.5, "learning_rate": 1.301419205531142e-06, "loss": 0.2312, "step": 7077 }, { "epoch": 1.34, "grad_norm": 12.25, "learning_rate": 1.3007470862310024e-06, "loss": 0.266, "step": 7078 }, { "epoch": 1.34, "grad_norm": 9.125, "learning_rate": 1.3000750795011834e-06, "loss": 0.2243, "step": 7079 }, { "epoch": 1.34, "grad_norm": 15.25, "learning_rate": 1.299403185404764e-06, "loss": 0.2923, "step": 7080 }, { "epoch": 1.34, "grad_norm": 11.5, "learning_rate": 1.2987314040048105e-06, "loss": 0.1977, "step": 7081 }, { "epoch": 1.34, "grad_norm": 11.25, "learning_rate": 1.298059735364382e-06, "loss": 0.2464, "step": 7082 }, { "epoch": 1.34, "grad_norm": 18.375, "learning_rate": 1.2973881795465271e-06, "loss": 0.1951, "step": 7083 }, { "epoch": 1.34, "grad_norm": 20.375, "learning_rate": 1.2967167366142835e-06, "loss": 0.1925, "step": 7084 }, { "epoch": 1.34, "grad_norm": 9.1875, "learning_rate": 1.296045406630674e-06, "loss": 0.1765, "step": 7085 }, { "epoch": 1.34, "grad_norm": 12.4375, "learning_rate": 1.295374189658718e-06, "loss": 0.1929, "step": 7086 }, { "epoch": 1.34, "grad_norm": 9.5, "learning_rate": 1.2947030857614188e-06, "loss": 0.2211, "step": 7087 }, { "epoch": 1.34, "grad_norm": 9.5625, "learning_rate": 1.2940320950017723e-06, "loss": 0.2022, "step": 7088 }, { "epoch": 1.34, "grad_norm": 8.3125, "learning_rate": 1.2933612174427596e-06, "loss": 0.1683, "step": 7089 }, { "epoch": 1.34, "grad_norm": 10.3125, "learning_rate": 1.2926904531473572e-06, "loss": 0.2444, "step": 7090 }, { "epoch": 1.34, "grad_norm": 15.25, "learning_rate": 1.292019802178526e-06, "loss": 0.2429, "step": 7091 }, { "epoch": 1.34, "grad_norm": 11.75, "learning_rate": 1.2913492645992195e-06, "loss": 0.1928, "step": 7092 }, { "epoch": 1.34, "grad_norm": 11.875, "learning_rate": 1.2906788404723764e-06, "loss": 0.2066, "step": 7093 }, { "epoch": 1.34, "grad_norm": 12.3125, "learning_rate": 1.2900085298609296e-06, "loss": 0.1802, "step": 7094 }, { "epoch": 1.34, "grad_norm": 9.4375, "learning_rate": 1.2893383328277998e-06, "loss": 0.1677, "step": 7095 }, { "epoch": 1.34, "grad_norm": 18.125, "learning_rate": 1.2886682494358933e-06, "loss": 0.234, "step": 7096 }, { "epoch": 1.34, "grad_norm": 12.0625, "learning_rate": 1.2879982797481116e-06, "loss": 0.2322, "step": 7097 }, { "epoch": 1.34, "grad_norm": 9.0625, "learning_rate": 1.287328423827342e-06, "loss": 0.1682, "step": 7098 }, { "epoch": 1.34, "grad_norm": 10.9375, "learning_rate": 1.286658681736463e-06, "loss": 0.2451, "step": 7099 }, { "epoch": 1.34, "grad_norm": 17.25, "learning_rate": 1.2859890535383382e-06, "loss": 0.219, "step": 7100 }, { "epoch": 1.34, "grad_norm": 10.5, "learning_rate": 1.2853195392958263e-06, "loss": 0.1786, "step": 7101 }, { "epoch": 1.34, "grad_norm": 11.3125, "learning_rate": 1.2846501390717719e-06, "loss": 0.165, "step": 7102 }, { "epoch": 1.34, "grad_norm": 11.0625, "learning_rate": 1.283980852929011e-06, "loss": 0.26, "step": 7103 }, { "epoch": 1.34, "grad_norm": 12.25, "learning_rate": 1.2833116809303637e-06, "loss": 0.1923, "step": 7104 }, { "epoch": 1.34, "grad_norm": 12.375, "learning_rate": 1.282642623138647e-06, "loss": 0.1955, "step": 7105 }, { "epoch": 1.34, "grad_norm": 11.375, "learning_rate": 1.2819736796166615e-06, "loss": 0.2145, "step": 7106 }, { "epoch": 1.34, "grad_norm": 12.375, "learning_rate": 1.2813048504271997e-06, "loss": 0.2168, "step": 7107 }, { "epoch": 1.34, "grad_norm": 9.375, "learning_rate": 1.2806361356330421e-06, "loss": 0.2155, "step": 7108 }, { "epoch": 1.34, "grad_norm": 10.4375, "learning_rate": 1.2799675352969593e-06, "loss": 0.2087, "step": 7109 }, { "epoch": 1.35, "grad_norm": 11.3125, "learning_rate": 1.2792990494817108e-06, "loss": 0.2524, "step": 7110 }, { "epoch": 1.35, "grad_norm": 14.3125, "learning_rate": 1.2786306782500454e-06, "loss": 0.2424, "step": 7111 }, { "epoch": 1.35, "grad_norm": 15.8125, "learning_rate": 1.277962421664701e-06, "loss": 0.2071, "step": 7112 }, { "epoch": 1.35, "grad_norm": 11.0625, "learning_rate": 1.277294279788405e-06, "loss": 0.2332, "step": 7113 }, { "epoch": 1.35, "grad_norm": 10.1875, "learning_rate": 1.2766262526838736e-06, "loss": 0.1652, "step": 7114 }, { "epoch": 1.35, "grad_norm": 12.0625, "learning_rate": 1.275958340413813e-06, "loss": 0.2661, "step": 7115 }, { "epoch": 1.35, "grad_norm": 11.4375, "learning_rate": 1.2752905430409178e-06, "loss": 0.269, "step": 7116 }, { "epoch": 1.35, "grad_norm": 13.9375, "learning_rate": 1.274622860627872e-06, "loss": 0.2608, "step": 7117 }, { "epoch": 1.35, "grad_norm": 11.375, "learning_rate": 1.2739552932373498e-06, "loss": 0.2543, "step": 7118 }, { "epoch": 1.35, "grad_norm": 14.125, "learning_rate": 1.2732878409320129e-06, "loss": 0.1848, "step": 7119 }, { "epoch": 1.35, "grad_norm": 10.9375, "learning_rate": 1.2726205037745132e-06, "loss": 0.1996, "step": 7120 }, { "epoch": 1.35, "grad_norm": 13.3125, "learning_rate": 1.2719532818274922e-06, "loss": 0.2492, "step": 7121 }, { "epoch": 1.35, "grad_norm": 8.0, "learning_rate": 1.2712861751535793e-06, "loss": 0.1748, "step": 7122 }, { "epoch": 1.35, "grad_norm": 12.625, "learning_rate": 1.2706191838153948e-06, "loss": 0.2017, "step": 7123 }, { "epoch": 1.35, "grad_norm": 11.3125, "learning_rate": 1.269952307875546e-06, "loss": 0.1947, "step": 7124 }, { "epoch": 1.35, "grad_norm": 15.0625, "learning_rate": 1.2692855473966315e-06, "loss": 0.2246, "step": 7125 }, { "epoch": 1.35, "grad_norm": 8.9375, "learning_rate": 1.2686189024412376e-06, "loss": 0.1722, "step": 7126 }, { "epoch": 1.35, "grad_norm": 10.25, "learning_rate": 1.2679523730719407e-06, "loss": 0.1425, "step": 7127 }, { "epoch": 1.35, "grad_norm": 12.375, "learning_rate": 1.2672859593513056e-06, "loss": 0.237, "step": 7128 }, { "epoch": 1.35, "grad_norm": 11.8125, "learning_rate": 1.2666196613418868e-06, "loss": 0.2287, "step": 7129 }, { "epoch": 1.35, "grad_norm": 14.75, "learning_rate": 1.2659534791062278e-06, "loss": 0.198, "step": 7130 }, { "epoch": 1.35, "grad_norm": 11.9375, "learning_rate": 1.265287412706861e-06, "loss": 0.2322, "step": 7131 }, { "epoch": 1.35, "grad_norm": 13.5625, "learning_rate": 1.2646214622063076e-06, "loss": 0.2315, "step": 7132 }, { "epoch": 1.35, "grad_norm": 13.5, "learning_rate": 1.2639556276670795e-06, "loss": 0.2336, "step": 7133 }, { "epoch": 1.35, "grad_norm": 12.4375, "learning_rate": 1.2632899091516754e-06, "loss": 0.212, "step": 7134 }, { "epoch": 1.35, "grad_norm": 10.8125, "learning_rate": 1.2626243067225852e-06, "loss": 0.1639, "step": 7135 }, { "epoch": 1.35, "grad_norm": 8.875, "learning_rate": 1.261958820442287e-06, "loss": 0.1551, "step": 7136 }, { "epoch": 1.35, "grad_norm": 11.9375, "learning_rate": 1.2612934503732477e-06, "loss": 0.2011, "step": 7137 }, { "epoch": 1.35, "grad_norm": 11.0, "learning_rate": 1.2606281965779232e-06, "loss": 0.1862, "step": 7138 }, { "epoch": 1.35, "grad_norm": 9.0625, "learning_rate": 1.25996305911876e-06, "loss": 0.1909, "step": 7139 }, { "epoch": 1.35, "grad_norm": 8.9375, "learning_rate": 1.2592980380581908e-06, "loss": 0.2321, "step": 7140 }, { "epoch": 1.35, "grad_norm": 11.5, "learning_rate": 1.2586331334586422e-06, "loss": 0.1807, "step": 7141 }, { "epoch": 1.35, "grad_norm": 9.9375, "learning_rate": 1.257968345382524e-06, "loss": 0.2113, "step": 7142 }, { "epoch": 1.35, "grad_norm": 11.1875, "learning_rate": 1.257303673892239e-06, "loss": 0.1744, "step": 7143 }, { "epoch": 1.35, "grad_norm": 10.9375, "learning_rate": 1.2566391190501764e-06, "loss": 0.2355, "step": 7144 }, { "epoch": 1.35, "grad_norm": 11.4375, "learning_rate": 1.2559746809187199e-06, "loss": 0.194, "step": 7145 }, { "epoch": 1.35, "grad_norm": 10.875, "learning_rate": 1.2553103595602344e-06, "loss": 0.2015, "step": 7146 }, { "epoch": 1.35, "grad_norm": 11.3125, "learning_rate": 1.2546461550370796e-06, "loss": 0.1768, "step": 7147 }, { "epoch": 1.35, "grad_norm": 10.875, "learning_rate": 1.253982067411601e-06, "loss": 0.1919, "step": 7148 }, { "epoch": 1.35, "grad_norm": 12.25, "learning_rate": 1.2533180967461373e-06, "loss": 0.2652, "step": 7149 }, { "epoch": 1.35, "grad_norm": 11.75, "learning_rate": 1.2526542431030109e-06, "loss": 0.1987, "step": 7150 }, { "epoch": 1.35, "grad_norm": 15.375, "learning_rate": 1.2519905065445354e-06, "loss": 0.2417, "step": 7151 }, { "epoch": 1.35, "grad_norm": 9.0625, "learning_rate": 1.251326887133017e-06, "loss": 0.1829, "step": 7152 }, { "epoch": 1.35, "grad_norm": 9.75, "learning_rate": 1.2506633849307449e-06, "loss": 0.1983, "step": 7153 }, { "epoch": 1.35, "grad_norm": 12.8125, "learning_rate": 1.2500000000000007e-06, "loss": 0.1994, "step": 7154 }, { "epoch": 1.35, "grad_norm": 12.8125, "learning_rate": 1.2493367324030533e-06, "loss": 0.2435, "step": 7155 }, { "epoch": 1.35, "grad_norm": 13.0625, "learning_rate": 1.2486735822021653e-06, "loss": 0.1889, "step": 7156 }, { "epoch": 1.35, "grad_norm": 14.8125, "learning_rate": 1.2480105494595812e-06, "loss": 0.2113, "step": 7157 }, { "epoch": 1.35, "grad_norm": 15.375, "learning_rate": 1.2473476342375386e-06, "loss": 0.2544, "step": 7158 }, { "epoch": 1.35, "grad_norm": 14.4375, "learning_rate": 1.2466848365982633e-06, "loss": 0.2032, "step": 7159 }, { "epoch": 1.35, "grad_norm": 13.75, "learning_rate": 1.2460221566039723e-06, "loss": 0.2587, "step": 7160 }, { "epoch": 1.35, "grad_norm": 12.9375, "learning_rate": 1.2453595943168667e-06, "loss": 0.203, "step": 7161 }, { "epoch": 1.35, "grad_norm": 11.5, "learning_rate": 1.2446971497991394e-06, "loss": 0.1926, "step": 7162 }, { "epoch": 1.36, "grad_norm": 13.5, "learning_rate": 1.244034823112974e-06, "loss": 0.197, "step": 7163 }, { "epoch": 1.36, "grad_norm": 13.0, "learning_rate": 1.2433726143205408e-06, "loss": 0.2264, "step": 7164 }, { "epoch": 1.36, "grad_norm": 10.8125, "learning_rate": 1.242710523483998e-06, "loss": 0.187, "step": 7165 }, { "epoch": 1.36, "grad_norm": 18.0, "learning_rate": 1.2420485506654935e-06, "loss": 0.2708, "step": 7166 }, { "epoch": 1.36, "grad_norm": 14.875, "learning_rate": 1.2413866959271676e-06, "loss": 0.2005, "step": 7167 }, { "epoch": 1.36, "grad_norm": 23.5, "learning_rate": 1.2407249593311457e-06, "loss": 0.2, "step": 7168 }, { "epoch": 1.36, "grad_norm": 13.8125, "learning_rate": 1.2400633409395412e-06, "loss": 0.1769, "step": 7169 }, { "epoch": 1.36, "grad_norm": 13.25, "learning_rate": 1.2394018408144589e-06, "loss": 0.2048, "step": 7170 }, { "epoch": 1.36, "grad_norm": 13.25, "learning_rate": 1.2387404590179935e-06, "loss": 0.2036, "step": 7171 }, { "epoch": 1.36, "grad_norm": 10.875, "learning_rate": 1.2380791956122266e-06, "loss": 0.2196, "step": 7172 }, { "epoch": 1.36, "grad_norm": 8.875, "learning_rate": 1.2374180506592265e-06, "loss": 0.1427, "step": 7173 }, { "epoch": 1.36, "grad_norm": 9.25, "learning_rate": 1.2367570242210558e-06, "loss": 0.1792, "step": 7174 }, { "epoch": 1.36, "grad_norm": 9.1875, "learning_rate": 1.236096116359762e-06, "loss": 0.1406, "step": 7175 }, { "epoch": 1.36, "grad_norm": 12.125, "learning_rate": 1.2354353271373834e-06, "loss": 0.2137, "step": 7176 }, { "epoch": 1.36, "grad_norm": 10.5, "learning_rate": 1.2347746566159437e-06, "loss": 0.2171, "step": 7177 }, { "epoch": 1.36, "grad_norm": 10.9375, "learning_rate": 1.2341141048574613e-06, "loss": 0.1775, "step": 7178 }, { "epoch": 1.36, "grad_norm": 15.1875, "learning_rate": 1.2334536719239392e-06, "loss": 0.2394, "step": 7179 }, { "epoch": 1.36, "grad_norm": 9.5, "learning_rate": 1.2327933578773692e-06, "loss": 0.1756, "step": 7180 }, { "epoch": 1.36, "grad_norm": 11.1875, "learning_rate": 1.232133162779733e-06, "loss": 0.2005, "step": 7181 }, { "epoch": 1.36, "grad_norm": 14.8125, "learning_rate": 1.2314730866930027e-06, "loss": 0.1924, "step": 7182 }, { "epoch": 1.36, "grad_norm": 11.1875, "learning_rate": 1.2308131296791381e-06, "loss": 0.1855, "step": 7183 }, { "epoch": 1.36, "grad_norm": 11.875, "learning_rate": 1.2301532918000842e-06, "loss": 0.1967, "step": 7184 }, { "epoch": 1.36, "grad_norm": 13.0, "learning_rate": 1.229493573117781e-06, "loss": 0.226, "step": 7185 }, { "epoch": 1.36, "grad_norm": 12.8125, "learning_rate": 1.2288339736941535e-06, "loss": 0.2346, "step": 7186 }, { "epoch": 1.36, "grad_norm": 13.5, "learning_rate": 1.2281744935911168e-06, "loss": 0.1653, "step": 7187 }, { "epoch": 1.36, "grad_norm": 10.75, "learning_rate": 1.2275151328705724e-06, "loss": 0.2533, "step": 7188 }, { "epoch": 1.36, "grad_norm": 12.375, "learning_rate": 1.2268558915944145e-06, "loss": 0.1963, "step": 7189 }, { "epoch": 1.36, "grad_norm": 11.0625, "learning_rate": 1.2261967698245237e-06, "loss": 0.2013, "step": 7190 }, { "epoch": 1.36, "grad_norm": 13.375, "learning_rate": 1.2255377676227707e-06, "loss": 0.2081, "step": 7191 }, { "epoch": 1.36, "grad_norm": 25.125, "learning_rate": 1.2248788850510107e-06, "loss": 0.1945, "step": 7192 }, { "epoch": 1.36, "grad_norm": 14.3125, "learning_rate": 1.2242201221710947e-06, "loss": 0.2373, "step": 7193 }, { "epoch": 1.36, "grad_norm": 14.625, "learning_rate": 1.2235614790448578e-06, "loss": 0.1826, "step": 7194 }, { "epoch": 1.36, "grad_norm": 9.5625, "learning_rate": 1.2229029557341251e-06, "loss": 0.2675, "step": 7195 }, { "epoch": 1.36, "grad_norm": 9.125, "learning_rate": 1.2222445523007079e-06, "loss": 0.1805, "step": 7196 }, { "epoch": 1.36, "grad_norm": 12.0, "learning_rate": 1.2215862688064114e-06, "loss": 0.2261, "step": 7197 }, { "epoch": 1.36, "grad_norm": 10.9375, "learning_rate": 1.220928105313026e-06, "loss": 0.255, "step": 7198 }, { "epoch": 1.36, "grad_norm": 11.25, "learning_rate": 1.2202700618823306e-06, "loss": 0.2008, "step": 7199 }, { "epoch": 1.36, "grad_norm": 10.375, "learning_rate": 1.219612138576095e-06, "loss": 0.1798, "step": 7200 }, { "epoch": 1.36, "grad_norm": 13.1875, "learning_rate": 1.218954335456076e-06, "loss": 0.2194, "step": 7201 }, { "epoch": 1.36, "grad_norm": 17.125, "learning_rate": 1.2182966525840195e-06, "loss": 0.1842, "step": 7202 }, { "epoch": 1.36, "grad_norm": 9.5, "learning_rate": 1.2176390900216607e-06, "loss": 0.1666, "step": 7203 }, { "epoch": 1.36, "grad_norm": 9.625, "learning_rate": 1.2169816478307222e-06, "loss": 0.201, "step": 7204 }, { "epoch": 1.36, "grad_norm": 11.6875, "learning_rate": 1.2163243260729169e-06, "loss": 0.1434, "step": 7205 }, { "epoch": 1.36, "grad_norm": 18.25, "learning_rate": 1.2156671248099454e-06, "loss": 0.2418, "step": 7206 }, { "epoch": 1.36, "grad_norm": 12.8125, "learning_rate": 1.2150100441034973e-06, "loss": 0.1845, "step": 7207 }, { "epoch": 1.36, "grad_norm": 11.5625, "learning_rate": 1.2143530840152507e-06, "loss": 0.1952, "step": 7208 }, { "epoch": 1.36, "grad_norm": 17.25, "learning_rate": 1.2136962446068725e-06, "loss": 0.2235, "step": 7209 }, { "epoch": 1.36, "grad_norm": 11.75, "learning_rate": 1.2130395259400185e-06, "loss": 0.2212, "step": 7210 }, { "epoch": 1.36, "grad_norm": 10.9375, "learning_rate": 1.2123829280763326e-06, "loss": 0.2685, "step": 7211 }, { "epoch": 1.36, "grad_norm": 9.75, "learning_rate": 1.2117264510774481e-06, "loss": 0.1749, "step": 7212 }, { "epoch": 1.36, "grad_norm": 11.5, "learning_rate": 1.211070095004986e-06, "loss": 0.2341, "step": 7213 }, { "epoch": 1.36, "grad_norm": 10.75, "learning_rate": 1.2104138599205572e-06, "loss": 0.2684, "step": 7214 }, { "epoch": 1.36, "grad_norm": 24.125, "learning_rate": 1.2097577458857601e-06, "loss": 0.2839, "step": 7215 }, { "epoch": 1.37, "grad_norm": 14.3125, "learning_rate": 1.209101752962182e-06, "loss": 0.1912, "step": 7216 }, { "epoch": 1.37, "grad_norm": 8.5, "learning_rate": 1.2084458812113993e-06, "loss": 0.1616, "step": 7217 }, { "epoch": 1.37, "grad_norm": 21.0, "learning_rate": 1.2077901306949769e-06, "loss": 0.2074, "step": 7218 }, { "epoch": 1.37, "grad_norm": 14.875, "learning_rate": 1.2071345014744678e-06, "loss": 0.1667, "step": 7219 }, { "epoch": 1.37, "grad_norm": 9.8125, "learning_rate": 1.2064789936114144e-06, "loss": 0.1162, "step": 7220 }, { "epoch": 1.37, "grad_norm": 9.625, "learning_rate": 1.2058236071673468e-06, "loss": 0.1564, "step": 7221 }, { "epoch": 1.37, "grad_norm": 11.9375, "learning_rate": 1.2051683422037843e-06, "loss": 0.2641, "step": 7222 }, { "epoch": 1.37, "grad_norm": 14.9375, "learning_rate": 1.204513198782235e-06, "loss": 0.2069, "step": 7223 }, { "epoch": 1.37, "grad_norm": 9.0, "learning_rate": 1.203858176964195e-06, "loss": 0.2102, "step": 7224 }, { "epoch": 1.37, "grad_norm": 9.75, "learning_rate": 1.203203276811149e-06, "loss": 0.1831, "step": 7225 }, { "epoch": 1.37, "grad_norm": 10.8125, "learning_rate": 1.2025484983845713e-06, "loss": 0.2045, "step": 7226 }, { "epoch": 1.37, "grad_norm": 14.25, "learning_rate": 1.2018938417459233e-06, "loss": 0.2158, "step": 7227 }, { "epoch": 1.37, "grad_norm": 12.125, "learning_rate": 1.201239306956656e-06, "loss": 0.2335, "step": 7228 }, { "epoch": 1.37, "grad_norm": 13.875, "learning_rate": 1.2005848940782087e-06, "loss": 0.2755, "step": 7229 }, { "epoch": 1.37, "grad_norm": 11.4375, "learning_rate": 1.1999306031720088e-06, "loss": 0.244, "step": 7230 }, { "epoch": 1.37, "grad_norm": 13.6875, "learning_rate": 1.1992764342994731e-06, "loss": 0.2399, "step": 7231 }, { "epoch": 1.37, "grad_norm": 14.625, "learning_rate": 1.1986223875220055e-06, "loss": 0.211, "step": 7232 }, { "epoch": 1.37, "grad_norm": 13.375, "learning_rate": 1.1979684629010018e-06, "loss": 0.2514, "step": 7233 }, { "epoch": 1.37, "grad_norm": 9.25, "learning_rate": 1.197314660497842e-06, "loss": 0.1572, "step": 7234 }, { "epoch": 1.37, "grad_norm": 10.625, "learning_rate": 1.1966609803738965e-06, "loss": 0.2123, "step": 7235 }, { "epoch": 1.37, "grad_norm": 12.6875, "learning_rate": 1.1960074225905252e-06, "loss": 0.2735, "step": 7236 }, { "epoch": 1.37, "grad_norm": 11.875, "learning_rate": 1.195353987209075e-06, "loss": 0.1527, "step": 7237 }, { "epoch": 1.37, "grad_norm": 8.125, "learning_rate": 1.1947006742908824e-06, "loss": 0.1583, "step": 7238 }, { "epoch": 1.37, "grad_norm": 12.8125, "learning_rate": 1.1940474838972719e-06, "loss": 0.2732, "step": 7239 }, { "epoch": 1.37, "grad_norm": 12.3125, "learning_rate": 1.193394416089556e-06, "loss": 0.2189, "step": 7240 }, { "epoch": 1.37, "grad_norm": 11.0625, "learning_rate": 1.1927414709290369e-06, "loss": 0.2446, "step": 7241 }, { "epoch": 1.37, "grad_norm": 9.625, "learning_rate": 1.1920886484770044e-06, "loss": 0.2102, "step": 7242 }, { "epoch": 1.37, "grad_norm": 11.1875, "learning_rate": 1.191435948794736e-06, "loss": 0.241, "step": 7243 }, { "epoch": 1.37, "grad_norm": 12.3125, "learning_rate": 1.1907833719435016e-06, "loss": 0.2077, "step": 7244 }, { "epoch": 1.37, "grad_norm": 9.4375, "learning_rate": 1.190130917984554e-06, "loss": 0.1853, "step": 7245 }, { "epoch": 1.37, "grad_norm": 13.25, "learning_rate": 1.1894785869791375e-06, "loss": 0.2266, "step": 7246 }, { "epoch": 1.37, "grad_norm": 9.375, "learning_rate": 1.188826378988484e-06, "loss": 0.1861, "step": 7247 }, { "epoch": 1.37, "grad_norm": 8.125, "learning_rate": 1.1881742940738173e-06, "loss": 0.1569, "step": 7248 }, { "epoch": 1.37, "grad_norm": 13.375, "learning_rate": 1.1875223322963433e-06, "loss": 0.1921, "step": 7249 }, { "epoch": 1.37, "grad_norm": 10.8125, "learning_rate": 1.1868704937172612e-06, "loss": 0.1969, "step": 7250 }, { "epoch": 1.37, "grad_norm": 11.4375, "learning_rate": 1.1862187783977558e-06, "loss": 0.2198, "step": 7251 }, { "epoch": 1.37, "grad_norm": 10.6875, "learning_rate": 1.1855671863990047e-06, "loss": 0.2251, "step": 7252 }, { "epoch": 1.37, "grad_norm": 14.125, "learning_rate": 1.1849157177821682e-06, "loss": 0.1994, "step": 7253 }, { "epoch": 1.37, "grad_norm": 10.0, "learning_rate": 1.1842643726083975e-06, "loss": 0.1596, "step": 7254 }, { "epoch": 1.37, "grad_norm": 13.6875, "learning_rate": 1.1836131509388346e-06, "loss": 0.2181, "step": 7255 }, { "epoch": 1.37, "grad_norm": 13.125, "learning_rate": 1.1829620528346075e-06, "loss": 0.1942, "step": 7256 }, { "epoch": 1.37, "grad_norm": 16.875, "learning_rate": 1.1823110783568309e-06, "loss": 0.2592, "step": 7257 }, { "epoch": 1.37, "grad_norm": 15.5, "learning_rate": 1.1816602275666103e-06, "loss": 0.1718, "step": 7258 }, { "epoch": 1.37, "grad_norm": 13.5625, "learning_rate": 1.1810095005250405e-06, "loss": 0.1956, "step": 7259 }, { "epoch": 1.37, "grad_norm": 15.4375, "learning_rate": 1.1803588972932037e-06, "loss": 0.2273, "step": 7260 }, { "epoch": 1.37, "grad_norm": 12.1875, "learning_rate": 1.179708417932168e-06, "loss": 0.2055, "step": 7261 }, { "epoch": 1.37, "grad_norm": 17.5, "learning_rate": 1.1790580625029919e-06, "loss": 0.2151, "step": 7262 }, { "epoch": 1.37, "grad_norm": 11.5, "learning_rate": 1.1784078310667254e-06, "loss": 0.2149, "step": 7263 }, { "epoch": 1.37, "grad_norm": 9.8125, "learning_rate": 1.177757723684401e-06, "loss": 0.1787, "step": 7264 }, { "epoch": 1.37, "grad_norm": 8.1875, "learning_rate": 1.177107740417042e-06, "loss": 0.1546, "step": 7265 }, { "epoch": 1.37, "grad_norm": 8.625, "learning_rate": 1.1764578813256624e-06, "loss": 0.1913, "step": 7266 }, { "epoch": 1.37, "grad_norm": 8.875, "learning_rate": 1.1758081464712628e-06, "loss": 0.1755, "step": 7267 }, { "epoch": 1.37, "grad_norm": 18.375, "learning_rate": 1.1751585359148301e-06, "loss": 0.2326, "step": 7268 }, { "epoch": 1.38, "grad_norm": 12.5, "learning_rate": 1.1745090497173408e-06, "loss": 0.1826, "step": 7269 }, { "epoch": 1.38, "grad_norm": 11.9375, "learning_rate": 1.1738596879397626e-06, "loss": 0.166, "step": 7270 }, { "epoch": 1.38, "grad_norm": 14.5, "learning_rate": 1.1732104506430488e-06, "loss": 0.229, "step": 7271 }, { "epoch": 1.38, "grad_norm": 14.5625, "learning_rate": 1.1725613378881397e-06, "loss": 0.2083, "step": 7272 }, { "epoch": 1.38, "grad_norm": 9.25, "learning_rate": 1.1719123497359655e-06, "loss": 0.1811, "step": 7273 }, { "epoch": 1.38, "grad_norm": 11.0625, "learning_rate": 1.171263486247447e-06, "loss": 0.2061, "step": 7274 }, { "epoch": 1.38, "grad_norm": 12.625, "learning_rate": 1.1706147474834905e-06, "loss": 0.1546, "step": 7275 }, { "epoch": 1.38, "grad_norm": 9.8125, "learning_rate": 1.169966133504989e-06, "loss": 0.2047, "step": 7276 }, { "epoch": 1.38, "grad_norm": 9.1875, "learning_rate": 1.1693176443728286e-06, "loss": 0.1578, "step": 7277 }, { "epoch": 1.38, "grad_norm": 10.25, "learning_rate": 1.1686692801478798e-06, "loss": 0.2057, "step": 7278 }, { "epoch": 1.38, "grad_norm": 12.25, "learning_rate": 1.1680210408910037e-06, "loss": 0.2006, "step": 7279 }, { "epoch": 1.38, "grad_norm": 10.9375, "learning_rate": 1.167372926663046e-06, "loss": 0.1852, "step": 7280 }, { "epoch": 1.38, "grad_norm": 11.875, "learning_rate": 1.1667249375248457e-06, "loss": 0.2107, "step": 7281 }, { "epoch": 1.38, "grad_norm": 11.375, "learning_rate": 1.1660770735372271e-06, "loss": 0.1516, "step": 7282 }, { "epoch": 1.38, "grad_norm": 15.6875, "learning_rate": 1.165429334761004e-06, "loss": 0.2648, "step": 7283 }, { "epoch": 1.38, "grad_norm": 12.5, "learning_rate": 1.1647817212569751e-06, "loss": 0.2257, "step": 7284 }, { "epoch": 1.38, "grad_norm": 10.0625, "learning_rate": 1.1641342330859325e-06, "loss": 0.1522, "step": 7285 }, { "epoch": 1.38, "grad_norm": 19.375, "learning_rate": 1.1634868703086533e-06, "loss": 0.1967, "step": 7286 }, { "epoch": 1.38, "grad_norm": 7.71875, "learning_rate": 1.1628396329859032e-06, "loss": 0.1794, "step": 7287 }, { "epoch": 1.38, "grad_norm": 10.8125, "learning_rate": 1.1621925211784365e-06, "loss": 0.2114, "step": 7288 }, { "epoch": 1.38, "grad_norm": 12.0625, "learning_rate": 1.1615455349469962e-06, "loss": 0.2042, "step": 7289 }, { "epoch": 1.38, "grad_norm": 10.8125, "learning_rate": 1.1608986743523126e-06, "loss": 0.2087, "step": 7290 }, { "epoch": 1.38, "grad_norm": 14.6875, "learning_rate": 1.1602519394551042e-06, "loss": 0.2202, "step": 7291 }, { "epoch": 1.38, "grad_norm": 11.0, "learning_rate": 1.159605330316079e-06, "loss": 0.1684, "step": 7292 }, { "epoch": 1.38, "grad_norm": 11.4375, "learning_rate": 1.1589588469959314e-06, "loss": 0.2319, "step": 7293 }, { "epoch": 1.38, "grad_norm": 11.375, "learning_rate": 1.1583124895553462e-06, "loss": 0.2173, "step": 7294 }, { "epoch": 1.38, "grad_norm": 8.1875, "learning_rate": 1.1576662580549925e-06, "loss": 0.1778, "step": 7295 }, { "epoch": 1.38, "grad_norm": 11.9375, "learning_rate": 1.1570201525555324e-06, "loss": 0.2413, "step": 7296 }, { "epoch": 1.38, "grad_norm": 10.6875, "learning_rate": 1.1563741731176134e-06, "loss": 0.2437, "step": 7297 }, { "epoch": 1.38, "grad_norm": 17.75, "learning_rate": 1.1557283198018714e-06, "loss": 0.2476, "step": 7298 }, { "epoch": 1.38, "grad_norm": 9.3125, "learning_rate": 1.155082592668931e-06, "loss": 0.2161, "step": 7299 }, { "epoch": 1.38, "grad_norm": 12.625, "learning_rate": 1.1544369917794045e-06, "loss": 0.1834, "step": 7300 }, { "epoch": 1.38, "grad_norm": 12.0625, "learning_rate": 1.1537915171938927e-06, "loss": 0.2437, "step": 7301 }, { "epoch": 1.38, "grad_norm": 10.25, "learning_rate": 1.1531461689729842e-06, "loss": 0.1514, "step": 7302 }, { "epoch": 1.38, "grad_norm": 10.6875, "learning_rate": 1.1525009471772558e-06, "loss": 0.2071, "step": 7303 }, { "epoch": 1.38, "grad_norm": 10.125, "learning_rate": 1.1518558518672732e-06, "loss": 0.1831, "step": 7304 }, { "epoch": 1.38, "grad_norm": 7.34375, "learning_rate": 1.1512108831035886e-06, "loss": 0.1714, "step": 7305 }, { "epoch": 1.38, "grad_norm": 11.6875, "learning_rate": 1.1505660409467445e-06, "loss": 0.2163, "step": 7306 }, { "epoch": 1.38, "grad_norm": 16.625, "learning_rate": 1.1499213254572694e-06, "loss": 0.1706, "step": 7307 }, { "epoch": 1.38, "grad_norm": 9.3125, "learning_rate": 1.1492767366956811e-06, "loss": 0.1673, "step": 7308 }, { "epoch": 1.38, "grad_norm": 12.5625, "learning_rate": 1.148632274722485e-06, "loss": 0.2033, "step": 7309 }, { "epoch": 1.38, "grad_norm": 10.375, "learning_rate": 1.1479879395981756e-06, "loss": 0.2108, "step": 7310 }, { "epoch": 1.38, "grad_norm": 10.9375, "learning_rate": 1.1473437313832338e-06, "loss": 0.2155, "step": 7311 }, { "epoch": 1.38, "grad_norm": 17.75, "learning_rate": 1.1466996501381303e-06, "loss": 0.2202, "step": 7312 }, { "epoch": 1.38, "grad_norm": 11.625, "learning_rate": 1.1460556959233227e-06, "loss": 0.1731, "step": 7313 }, { "epoch": 1.38, "grad_norm": 16.125, "learning_rate": 1.1454118687992568e-06, "loss": 0.2288, "step": 7314 }, { "epoch": 1.38, "grad_norm": 13.875, "learning_rate": 1.1447681688263672e-06, "loss": 0.2392, "step": 7315 }, { "epoch": 1.38, "grad_norm": 12.375, "learning_rate": 1.1441245960650763e-06, "loss": 0.2013, "step": 7316 }, { "epoch": 1.38, "grad_norm": 10.0, "learning_rate": 1.1434811505757937e-06, "loss": 0.2074, "step": 7317 }, { "epoch": 1.38, "grad_norm": 11.0625, "learning_rate": 1.1428378324189183e-06, "loss": 0.2037, "step": 7318 }, { "epoch": 1.38, "grad_norm": 11.1875, "learning_rate": 1.142194641654836e-06, "loss": 0.2496, "step": 7319 }, { "epoch": 1.38, "grad_norm": 14.1875, "learning_rate": 1.1415515783439215e-06, "loss": 0.2163, "step": 7320 }, { "epoch": 1.38, "grad_norm": 11.4375, "learning_rate": 1.1409086425465374e-06, "loss": 0.2937, "step": 7321 }, { "epoch": 1.39, "grad_norm": 11.625, "learning_rate": 1.1402658343230339e-06, "loss": 0.1976, "step": 7322 }, { "epoch": 1.39, "grad_norm": 12.75, "learning_rate": 1.1396231537337494e-06, "loss": 0.2032, "step": 7323 }, { "epoch": 1.39, "grad_norm": 9.8125, "learning_rate": 1.1389806008390109e-06, "loss": 0.2136, "step": 7324 }, { "epoch": 1.39, "grad_norm": 10.4375, "learning_rate": 1.1383381756991324e-06, "loss": 0.2013, "step": 7325 }, { "epoch": 1.39, "grad_norm": 14.8125, "learning_rate": 1.1376958783744165e-06, "loss": 0.2334, "step": 7326 }, { "epoch": 1.39, "grad_norm": 10.6875, "learning_rate": 1.1370537089251543e-06, "loss": 0.1812, "step": 7327 }, { "epoch": 1.39, "grad_norm": 15.5, "learning_rate": 1.1364116674116238e-06, "loss": 0.2452, "step": 7328 }, { "epoch": 1.39, "grad_norm": 6.28125, "learning_rate": 1.1357697538940918e-06, "loss": 0.122, "step": 7329 }, { "epoch": 1.39, "grad_norm": 15.6875, "learning_rate": 1.1351279684328128e-06, "loss": 0.2539, "step": 7330 }, { "epoch": 1.39, "grad_norm": 10.0, "learning_rate": 1.1344863110880282e-06, "loss": 0.1812, "step": 7331 }, { "epoch": 1.39, "grad_norm": 10.25, "learning_rate": 1.1338447819199714e-06, "loss": 0.1657, "step": 7332 }, { "epoch": 1.39, "grad_norm": 10.3125, "learning_rate": 1.1332033809888578e-06, "loss": 0.2041, "step": 7333 }, { "epoch": 1.39, "grad_norm": 9.25, "learning_rate": 1.1325621083548955e-06, "loss": 0.2061, "step": 7334 }, { "epoch": 1.39, "grad_norm": 14.375, "learning_rate": 1.1319209640782768e-06, "loss": 0.2766, "step": 7335 }, { "epoch": 1.39, "grad_norm": 10.5625, "learning_rate": 1.1312799482191877e-06, "loss": 0.1809, "step": 7336 }, { "epoch": 1.39, "grad_norm": 11.0625, "learning_rate": 1.1306390608377955e-06, "loss": 0.1929, "step": 7337 }, { "epoch": 1.39, "grad_norm": 11.6875, "learning_rate": 1.129998301994259e-06, "loss": 0.21, "step": 7338 }, { "epoch": 1.39, "grad_norm": 12.9375, "learning_rate": 1.1293576717487238e-06, "loss": 0.217, "step": 7339 }, { "epoch": 1.39, "grad_norm": 9.75, "learning_rate": 1.1287171701613262e-06, "loss": 0.2494, "step": 7340 }, { "epoch": 1.39, "grad_norm": 11.75, "learning_rate": 1.128076797292186e-06, "loss": 0.2664, "step": 7341 }, { "epoch": 1.39, "grad_norm": 12.9375, "learning_rate": 1.1274365532014127e-06, "loss": 0.2153, "step": 7342 }, { "epoch": 1.39, "grad_norm": 16.625, "learning_rate": 1.1267964379491064e-06, "loss": 0.2177, "step": 7343 }, { "epoch": 1.39, "grad_norm": 9.4375, "learning_rate": 1.1261564515953524e-06, "loss": 0.1835, "step": 7344 }, { "epoch": 1.39, "grad_norm": 12.4375, "learning_rate": 1.1255165942002228e-06, "loss": 0.1999, "step": 7345 }, { "epoch": 1.39, "grad_norm": 13.5625, "learning_rate": 1.1248768658237786e-06, "loss": 0.208, "step": 7346 }, { "epoch": 1.39, "grad_norm": 17.0, "learning_rate": 1.1242372665260725e-06, "loss": 0.2236, "step": 7347 }, { "epoch": 1.39, "grad_norm": 9.0, "learning_rate": 1.1235977963671388e-06, "loss": 0.2108, "step": 7348 }, { "epoch": 1.39, "grad_norm": 12.5625, "learning_rate": 1.1229584554070038e-06, "loss": 0.1866, "step": 7349 }, { "epoch": 1.39, "grad_norm": 9.9375, "learning_rate": 1.1223192437056793e-06, "loss": 0.1982, "step": 7350 }, { "epoch": 1.39, "grad_norm": 9.3125, "learning_rate": 1.1216801613231692e-06, "loss": 0.1715, "step": 7351 }, { "epoch": 1.39, "grad_norm": 11.625, "learning_rate": 1.1210412083194595e-06, "loss": 0.2193, "step": 7352 }, { "epoch": 1.39, "grad_norm": 11.9375, "learning_rate": 1.1204023847545266e-06, "loss": 0.2283, "step": 7353 }, { "epoch": 1.39, "grad_norm": 12.9375, "learning_rate": 1.1197636906883372e-06, "loss": 0.209, "step": 7354 }, { "epoch": 1.39, "grad_norm": 10.125, "learning_rate": 1.1191251261808431e-06, "loss": 0.1795, "step": 7355 }, { "epoch": 1.39, "grad_norm": 13.375, "learning_rate": 1.1184866912919831e-06, "loss": 0.245, "step": 7356 }, { "epoch": 1.39, "grad_norm": 9.0, "learning_rate": 1.1178483860816852e-06, "loss": 0.1566, "step": 7357 }, { "epoch": 1.39, "grad_norm": 14.875, "learning_rate": 1.1172102106098667e-06, "loss": 0.2186, "step": 7358 }, { "epoch": 1.39, "grad_norm": 11.1875, "learning_rate": 1.1165721649364314e-06, "loss": 0.2061, "step": 7359 }, { "epoch": 1.39, "grad_norm": 16.375, "learning_rate": 1.115934249121269e-06, "loss": 0.2246, "step": 7360 }, { "epoch": 1.39, "grad_norm": 11.875, "learning_rate": 1.1152964632242585e-06, "loss": 0.2264, "step": 7361 }, { "epoch": 1.39, "grad_norm": 19.875, "learning_rate": 1.1146588073052695e-06, "loss": 0.2222, "step": 7362 }, { "epoch": 1.39, "grad_norm": 11.875, "learning_rate": 1.1140212814241558e-06, "loss": 0.1891, "step": 7363 }, { "epoch": 1.39, "grad_norm": 13.0, "learning_rate": 1.1133838856407583e-06, "loss": 0.192, "step": 7364 }, { "epoch": 1.39, "grad_norm": 12.6875, "learning_rate": 1.1127466200149098e-06, "loss": 0.1781, "step": 7365 }, { "epoch": 1.39, "grad_norm": 12.4375, "learning_rate": 1.1121094846064273e-06, "loss": 0.251, "step": 7366 }, { "epoch": 1.39, "grad_norm": 12.375, "learning_rate": 1.1114724794751182e-06, "loss": 0.2186, "step": 7367 }, { "epoch": 1.39, "grad_norm": 12.4375, "learning_rate": 1.1108356046807737e-06, "loss": 0.195, "step": 7368 }, { "epoch": 1.39, "grad_norm": 12.4375, "learning_rate": 1.1101988602831776e-06, "loss": 0.2277, "step": 7369 }, { "epoch": 1.39, "grad_norm": 13.5, "learning_rate": 1.1095622463420982e-06, "loss": 0.2805, "step": 7370 }, { "epoch": 1.39, "grad_norm": 9.0, "learning_rate": 1.1089257629172943e-06, "loss": 0.1711, "step": 7371 }, { "epoch": 1.39, "grad_norm": 14.3125, "learning_rate": 1.1082894100685071e-06, "loss": 0.2086, "step": 7372 }, { "epoch": 1.39, "grad_norm": 10.6875, "learning_rate": 1.1076531878554723e-06, "loss": 0.195, "step": 7373 }, { "epoch": 1.39, "grad_norm": 17.875, "learning_rate": 1.1070170963379093e-06, "loss": 0.2229, "step": 7374 }, { "epoch": 1.4, "grad_norm": 27.625, "learning_rate": 1.1063811355755258e-06, "loss": 0.2206, "step": 7375 }, { "epoch": 1.4, "grad_norm": 9.625, "learning_rate": 1.1057453056280182e-06, "loss": 0.1964, "step": 7376 }, { "epoch": 1.4, "grad_norm": 11.75, "learning_rate": 1.1051096065550693e-06, "loss": 0.1906, "step": 7377 }, { "epoch": 1.4, "grad_norm": 10.5625, "learning_rate": 1.1044740384163512e-06, "loss": 0.1932, "step": 7378 }, { "epoch": 1.4, "grad_norm": 13.5625, "learning_rate": 1.1038386012715207e-06, "loss": 0.2211, "step": 7379 }, { "epoch": 1.4, "grad_norm": 12.5625, "learning_rate": 1.1032032951802269e-06, "loss": 0.2619, "step": 7380 }, { "epoch": 1.4, "grad_norm": 13.75, "learning_rate": 1.102568120202103e-06, "loss": 0.2359, "step": 7381 }, { "epoch": 1.4, "grad_norm": 11.0, "learning_rate": 1.101933076396772e-06, "loss": 0.1924, "step": 7382 }, { "epoch": 1.4, "grad_norm": 12.625, "learning_rate": 1.1012981638238407e-06, "loss": 0.2162, "step": 7383 }, { "epoch": 1.4, "grad_norm": 9.0, "learning_rate": 1.1006633825429092e-06, "loss": 0.1803, "step": 7384 }, { "epoch": 1.4, "grad_norm": 9.875, "learning_rate": 1.100028732613562e-06, "loss": 0.1868, "step": 7385 }, { "epoch": 1.4, "grad_norm": 10.875, "learning_rate": 1.0993942140953712e-06, "loss": 0.1835, "step": 7386 }, { "epoch": 1.4, "grad_norm": 13.6875, "learning_rate": 1.0987598270478978e-06, "loss": 0.1918, "step": 7387 }, { "epoch": 1.4, "grad_norm": 11.375, "learning_rate": 1.0981255715306897e-06, "loss": 0.1995, "step": 7388 }, { "epoch": 1.4, "grad_norm": 9.9375, "learning_rate": 1.0974914476032821e-06, "loss": 0.1887, "step": 7389 }, { "epoch": 1.4, "grad_norm": 45.0, "learning_rate": 1.0968574553251988e-06, "loss": 0.2217, "step": 7390 }, { "epoch": 1.4, "grad_norm": 8.875, "learning_rate": 1.0962235947559507e-06, "loss": 0.2128, "step": 7391 }, { "epoch": 1.4, "grad_norm": 14.125, "learning_rate": 1.0955898659550365e-06, "loss": 0.2408, "step": 7392 }, { "epoch": 1.4, "grad_norm": 10.625, "learning_rate": 1.094956268981942e-06, "loss": 0.1999, "step": 7393 }, { "epoch": 1.4, "grad_norm": 9.625, "learning_rate": 1.0943228038961417e-06, "loss": 0.1515, "step": 7394 }, { "epoch": 1.4, "grad_norm": 10.3125, "learning_rate": 1.0936894707570966e-06, "loss": 0.2131, "step": 7395 }, { "epoch": 1.4, "grad_norm": 10.5, "learning_rate": 1.0930562696242561e-06, "loss": 0.1643, "step": 7396 }, { "epoch": 1.4, "grad_norm": 13.8125, "learning_rate": 1.0924232005570567e-06, "loss": 0.2164, "step": 7397 }, { "epoch": 1.4, "grad_norm": 12.375, "learning_rate": 1.091790263614923e-06, "loss": 0.2496, "step": 7398 }, { "epoch": 1.4, "grad_norm": 10.5, "learning_rate": 1.0911574588572669e-06, "loss": 0.1828, "step": 7399 }, { "epoch": 1.4, "grad_norm": 14.375, "learning_rate": 1.0905247863434872e-06, "loss": 0.2426, "step": 7400 }, { "epoch": 1.4, "grad_norm": 12.125, "learning_rate": 1.0898922461329719e-06, "loss": 0.2328, "step": 7401 }, { "epoch": 1.4, "grad_norm": 10.9375, "learning_rate": 1.0892598382850953e-06, "loss": 0.175, "step": 7402 }, { "epoch": 1.4, "grad_norm": 10.3125, "learning_rate": 1.0886275628592197e-06, "loss": 0.2195, "step": 7403 }, { "epoch": 1.4, "grad_norm": 13.0625, "learning_rate": 1.0879954199146948e-06, "loss": 0.1601, "step": 7404 }, { "epoch": 1.4, "grad_norm": 14.4375, "learning_rate": 1.0873634095108582e-06, "loss": 0.1672, "step": 7405 }, { "epoch": 1.4, "grad_norm": 12.1875, "learning_rate": 1.0867315317070345e-06, "loss": 0.2024, "step": 7406 }, { "epoch": 1.4, "grad_norm": 10.75, "learning_rate": 1.0860997865625363e-06, "loss": 0.1507, "step": 7407 }, { "epoch": 1.4, "grad_norm": 9.5625, "learning_rate": 1.0854681741366643e-06, "loss": 0.158, "step": 7408 }, { "epoch": 1.4, "grad_norm": 8.0, "learning_rate": 1.084836694488705e-06, "loss": 0.1667, "step": 7409 }, { "epoch": 1.4, "grad_norm": 8.75, "learning_rate": 1.0842053476779346e-06, "loss": 0.205, "step": 7410 }, { "epoch": 1.4, "grad_norm": 10.9375, "learning_rate": 1.083574133763615e-06, "loss": 0.2781, "step": 7411 }, { "epoch": 1.4, "grad_norm": 11.375, "learning_rate": 1.0829430528049968e-06, "loss": 0.1679, "step": 7412 }, { "epoch": 1.4, "grad_norm": 8.6875, "learning_rate": 1.0823121048613172e-06, "loss": 0.1758, "step": 7413 }, { "epoch": 1.4, "grad_norm": 10.5, "learning_rate": 1.0816812899918016e-06, "loss": 0.209, "step": 7414 }, { "epoch": 1.4, "grad_norm": 11.5625, "learning_rate": 1.0810506082556632e-06, "loss": 0.2201, "step": 7415 }, { "epoch": 1.4, "grad_norm": 13.75, "learning_rate": 1.0804200597121017e-06, "loss": 0.2189, "step": 7416 }, { "epoch": 1.4, "grad_norm": 14.375, "learning_rate": 1.079789644420305e-06, "loss": 0.2483, "step": 7417 }, { "epoch": 1.4, "grad_norm": 9.4375, "learning_rate": 1.0791593624394487e-06, "loss": 0.1619, "step": 7418 }, { "epoch": 1.4, "grad_norm": 10.625, "learning_rate": 1.0785292138286948e-06, "loss": 0.2166, "step": 7419 }, { "epoch": 1.4, "grad_norm": 17.25, "learning_rate": 1.077899198647194e-06, "loss": 0.1954, "step": 7420 }, { "epoch": 1.4, "grad_norm": 16.5, "learning_rate": 1.0772693169540834e-06, "loss": 0.2434, "step": 7421 }, { "epoch": 1.4, "grad_norm": 13.125, "learning_rate": 1.0766395688084889e-06, "loss": 0.1845, "step": 7422 }, { "epoch": 1.4, "grad_norm": 9.1875, "learning_rate": 1.0760099542695215e-06, "loss": 0.1486, "step": 7423 }, { "epoch": 1.4, "grad_norm": 11.75, "learning_rate": 1.0753804733962845e-06, "loss": 0.2884, "step": 7424 }, { "epoch": 1.4, "grad_norm": 12.25, "learning_rate": 1.0747511262478622e-06, "loss": 0.1687, "step": 7425 }, { "epoch": 1.4, "grad_norm": 12.375, "learning_rate": 1.0741219128833307e-06, "loss": 0.2127, "step": 7426 }, { "epoch": 1.4, "grad_norm": 12.9375, "learning_rate": 1.0734928333617516e-06, "loss": 0.2677, "step": 7427 }, { "epoch": 1.41, "grad_norm": 7.3125, "learning_rate": 1.0728638877421774e-06, "loss": 0.1324, "step": 7428 }, { "epoch": 1.41, "grad_norm": 10.75, "learning_rate": 1.0722350760836424e-06, "loss": 0.1911, "step": 7429 }, { "epoch": 1.41, "grad_norm": 9.4375, "learning_rate": 1.0716063984451727e-06, "loss": 0.1537, "step": 7430 }, { "epoch": 1.41, "grad_norm": 11.0625, "learning_rate": 1.07097785488578e-06, "loss": 0.1944, "step": 7431 }, { "epoch": 1.41, "grad_norm": 12.25, "learning_rate": 1.0703494454644636e-06, "loss": 0.2029, "step": 7432 }, { "epoch": 1.41, "grad_norm": 10.0625, "learning_rate": 1.069721170240211e-06, "loss": 0.2087, "step": 7433 }, { "epoch": 1.41, "grad_norm": 15.125, "learning_rate": 1.0690930292719953e-06, "loss": 0.1795, "step": 7434 }, { "epoch": 1.41, "grad_norm": 16.625, "learning_rate": 1.0684650226187813e-06, "loss": 0.1868, "step": 7435 }, { "epoch": 1.41, "grad_norm": 11.75, "learning_rate": 1.0678371503395146e-06, "loss": 0.2278, "step": 7436 }, { "epoch": 1.41, "grad_norm": 10.1875, "learning_rate": 1.0672094124931334e-06, "loss": 0.2191, "step": 7437 }, { "epoch": 1.41, "grad_norm": 13.125, "learning_rate": 1.0665818091385605e-06, "loss": 0.1698, "step": 7438 }, { "epoch": 1.41, "grad_norm": 10.25, "learning_rate": 1.0659543403347097e-06, "loss": 0.1885, "step": 7439 }, { "epoch": 1.41, "grad_norm": 10.625, "learning_rate": 1.0653270061404771e-06, "loss": 0.1638, "step": 7440 }, { "epoch": 1.41, "grad_norm": 10.1875, "learning_rate": 1.0646998066147495e-06, "loss": 0.2333, "step": 7441 }, { "epoch": 1.41, "grad_norm": 10.375, "learning_rate": 1.0640727418163996e-06, "loss": 0.2143, "step": 7442 }, { "epoch": 1.41, "grad_norm": 13.0, "learning_rate": 1.0634458118042907e-06, "loss": 0.2208, "step": 7443 }, { "epoch": 1.41, "grad_norm": 10.125, "learning_rate": 1.0628190166372675e-06, "loss": 0.2464, "step": 7444 }, { "epoch": 1.41, "grad_norm": 10.625, "learning_rate": 1.0621923563741664e-06, "loss": 0.1687, "step": 7445 }, { "epoch": 1.41, "grad_norm": 8.5625, "learning_rate": 1.0615658310738117e-06, "loss": 0.1843, "step": 7446 }, { "epoch": 1.41, "grad_norm": 18.25, "learning_rate": 1.060939440795013e-06, "loss": 0.2516, "step": 7447 }, { "epoch": 1.41, "grad_norm": 9.6875, "learning_rate": 1.0603131855965665e-06, "loss": 0.1632, "step": 7448 }, { "epoch": 1.41, "grad_norm": 10.1875, "learning_rate": 1.0596870655372562e-06, "loss": 0.1643, "step": 7449 }, { "epoch": 1.41, "grad_norm": 11.25, "learning_rate": 1.0590610806758567e-06, "loss": 0.2028, "step": 7450 }, { "epoch": 1.41, "grad_norm": 14.0625, "learning_rate": 1.0584352310711267e-06, "loss": 0.3094, "step": 7451 }, { "epoch": 1.41, "grad_norm": 13.4375, "learning_rate": 1.057809516781812e-06, "loss": 0.2944, "step": 7452 }, { "epoch": 1.41, "grad_norm": 12.0625, "learning_rate": 1.0571839378666454e-06, "loss": 0.231, "step": 7453 }, { "epoch": 1.41, "grad_norm": 12.8125, "learning_rate": 1.0565584943843505e-06, "loss": 0.1372, "step": 7454 }, { "epoch": 1.41, "grad_norm": 9.5, "learning_rate": 1.055933186393636e-06, "loss": 0.1963, "step": 7455 }, { "epoch": 1.41, "grad_norm": 15.75, "learning_rate": 1.0553080139531944e-06, "loss": 0.2185, "step": 7456 }, { "epoch": 1.41, "grad_norm": 11.0, "learning_rate": 1.054682977121712e-06, "loss": 0.2381, "step": 7457 }, { "epoch": 1.41, "grad_norm": 12.4375, "learning_rate": 1.0540580759578584e-06, "loss": 0.1966, "step": 7458 }, { "epoch": 1.41, "grad_norm": 11.0, "learning_rate": 1.0534333105202915e-06, "loss": 0.2464, "step": 7459 }, { "epoch": 1.41, "grad_norm": 11.0, "learning_rate": 1.052808680867654e-06, "loss": 0.2097, "step": 7460 }, { "epoch": 1.41, "grad_norm": 13.125, "learning_rate": 1.0521841870585806e-06, "loss": 0.2745, "step": 7461 }, { "epoch": 1.41, "grad_norm": 14.3125, "learning_rate": 1.0515598291516908e-06, "loss": 0.1957, "step": 7462 }, { "epoch": 1.41, "grad_norm": 9.9375, "learning_rate": 1.050935607205589e-06, "loss": 0.2241, "step": 7463 }, { "epoch": 1.41, "grad_norm": 11.6875, "learning_rate": 1.0503115212788694e-06, "loss": 0.2306, "step": 7464 }, { "epoch": 1.41, "grad_norm": 13.9375, "learning_rate": 1.0496875714301149e-06, "loss": 0.1869, "step": 7465 }, { "epoch": 1.41, "grad_norm": 12.1875, "learning_rate": 1.0490637577178936e-06, "loss": 0.1994, "step": 7466 }, { "epoch": 1.41, "grad_norm": 13.875, "learning_rate": 1.0484400802007588e-06, "loss": 0.2668, "step": 7467 }, { "epoch": 1.41, "grad_norm": 14.25, "learning_rate": 1.0478165389372555e-06, "loss": 0.2056, "step": 7468 }, { "epoch": 1.41, "grad_norm": 9.9375, "learning_rate": 1.047193133985913e-06, "loss": 0.1369, "step": 7469 }, { "epoch": 1.41, "grad_norm": 10.1875, "learning_rate": 1.0465698654052493e-06, "loss": 0.2597, "step": 7470 }, { "epoch": 1.41, "grad_norm": 12.1875, "learning_rate": 1.045946733253766e-06, "loss": 0.2336, "step": 7471 }, { "epoch": 1.41, "grad_norm": 10.25, "learning_rate": 1.0453237375899574e-06, "loss": 0.2251, "step": 7472 }, { "epoch": 1.41, "grad_norm": 10.3125, "learning_rate": 1.0447008784723017e-06, "loss": 0.1763, "step": 7473 }, { "epoch": 1.41, "grad_norm": 13.125, "learning_rate": 1.0440781559592652e-06, "loss": 0.1996, "step": 7474 }, { "epoch": 1.41, "grad_norm": 13.1875, "learning_rate": 1.0434555701092988e-06, "loss": 0.1775, "step": 7475 }, { "epoch": 1.41, "grad_norm": 8.6875, "learning_rate": 1.0428331209808454e-06, "loss": 0.1759, "step": 7476 }, { "epoch": 1.41, "grad_norm": 12.625, "learning_rate": 1.042210808632331e-06, "loss": 0.1847, "step": 7477 }, { "epoch": 1.41, "grad_norm": 10.5625, "learning_rate": 1.0415886331221708e-06, "loss": 0.2359, "step": 7478 }, { "epoch": 1.41, "grad_norm": 11.1875, "learning_rate": 1.0409665945087665e-06, "loss": 0.2223, "step": 7479 }, { "epoch": 1.42, "grad_norm": 8.9375, "learning_rate": 1.040344692850507e-06, "loss": 0.1892, "step": 7480 }, { "epoch": 1.42, "grad_norm": 13.3125, "learning_rate": 1.0397229282057684e-06, "loss": 0.2782, "step": 7481 }, { "epoch": 1.42, "grad_norm": 11.1875, "learning_rate": 1.0391013006329134e-06, "loss": 0.175, "step": 7482 }, { "epoch": 1.42, "grad_norm": 12.0625, "learning_rate": 1.0384798101902933e-06, "loss": 0.2277, "step": 7483 }, { "epoch": 1.42, "grad_norm": 10.8125, "learning_rate": 1.0378584569362447e-06, "loss": 0.2154, "step": 7484 }, { "epoch": 1.42, "grad_norm": 10.3125, "learning_rate": 1.0372372409290924e-06, "loss": 0.1739, "step": 7485 }, { "epoch": 1.42, "grad_norm": 9.625, "learning_rate": 1.0366161622271483e-06, "loss": 0.1482, "step": 7486 }, { "epoch": 1.42, "grad_norm": 11.75, "learning_rate": 1.0359952208887111e-06, "loss": 0.1749, "step": 7487 }, { "epoch": 1.42, "grad_norm": 8.4375, "learning_rate": 1.0353744169720669e-06, "loss": 0.1973, "step": 7488 }, { "epoch": 1.42, "grad_norm": 12.125, "learning_rate": 1.0347537505354887e-06, "loss": 0.2297, "step": 7489 }, { "epoch": 1.42, "grad_norm": 13.0625, "learning_rate": 1.0341332216372363e-06, "loss": 0.2167, "step": 7490 }, { "epoch": 1.42, "grad_norm": 9.3125, "learning_rate": 1.033512830335557e-06, "loss": 0.1735, "step": 7491 }, { "epoch": 1.42, "grad_norm": 12.125, "learning_rate": 1.0328925766886855e-06, "loss": 0.2287, "step": 7492 }, { "epoch": 1.42, "grad_norm": 11.5, "learning_rate": 1.0322724607548426e-06, "loss": 0.2134, "step": 7493 }, { "epoch": 1.42, "grad_norm": 8.625, "learning_rate": 1.0316524825922372e-06, "loss": 0.2152, "step": 7494 }, { "epoch": 1.42, "grad_norm": 10.0, "learning_rate": 1.0310326422590646e-06, "loss": 0.2337, "step": 7495 }, { "epoch": 1.42, "grad_norm": 9.4375, "learning_rate": 1.0304129398135077e-06, "loss": 0.2004, "step": 7496 }, { "epoch": 1.42, "grad_norm": 12.375, "learning_rate": 1.0297933753137354e-06, "loss": 0.2382, "step": 7497 }, { "epoch": 1.42, "grad_norm": 14.1875, "learning_rate": 1.0291739488179051e-06, "loss": 0.1932, "step": 7498 }, { "epoch": 1.42, "grad_norm": 15.1875, "learning_rate": 1.02855466038416e-06, "loss": 0.2194, "step": 7499 }, { "epoch": 1.42, "grad_norm": 11.1875, "learning_rate": 1.0279355100706315e-06, "loss": 0.2043, "step": 7500 }, { "epoch": 1.42, "grad_norm": 14.3125, "learning_rate": 1.0273164979354368e-06, "loss": 0.1687, "step": 7501 }, { "epoch": 1.42, "grad_norm": 8.5625, "learning_rate": 1.026697624036681e-06, "loss": 0.1497, "step": 7502 }, { "epoch": 1.42, "grad_norm": 11.4375, "learning_rate": 1.026078888432456e-06, "loss": 0.2271, "step": 7503 }, { "epoch": 1.42, "grad_norm": 10.875, "learning_rate": 1.0254602911808407e-06, "loss": 0.2011, "step": 7504 }, { "epoch": 1.42, "grad_norm": 15.1875, "learning_rate": 1.024841832339901e-06, "loss": 0.2607, "step": 7505 }, { "epoch": 1.42, "grad_norm": 10.8125, "learning_rate": 1.0242235119676893e-06, "loss": 0.1876, "step": 7506 }, { "epoch": 1.42, "grad_norm": 10.8125, "learning_rate": 1.0236053301222462e-06, "loss": 0.1932, "step": 7507 }, { "epoch": 1.42, "grad_norm": 9.5, "learning_rate": 1.0229872868615983e-06, "loss": 0.2262, "step": 7508 }, { "epoch": 1.42, "grad_norm": 13.25, "learning_rate": 1.0223693822437594e-06, "loss": 0.2225, "step": 7509 }, { "epoch": 1.42, "grad_norm": 9.9375, "learning_rate": 1.0217516163267308e-06, "loss": 0.1915, "step": 7510 }, { "epoch": 1.42, "grad_norm": 9.5625, "learning_rate": 1.0211339891684988e-06, "loss": 0.2085, "step": 7511 }, { "epoch": 1.42, "grad_norm": 14.375, "learning_rate": 1.0205165008270412e-06, "loss": 0.2118, "step": 7512 }, { "epoch": 1.42, "grad_norm": 12.8125, "learning_rate": 1.0198991513603174e-06, "loss": 0.1872, "step": 7513 }, { "epoch": 1.42, "grad_norm": 9.1875, "learning_rate": 1.0192819408262767e-06, "loss": 0.2045, "step": 7514 }, { "epoch": 1.42, "grad_norm": 9.1875, "learning_rate": 1.0186648692828552e-06, "loss": 0.2115, "step": 7515 }, { "epoch": 1.42, "grad_norm": 9.9375, "learning_rate": 1.0180479367879751e-06, "loss": 0.1836, "step": 7516 }, { "epoch": 1.42, "grad_norm": 22.875, "learning_rate": 1.0174311433995463e-06, "loss": 0.207, "step": 7517 }, { "epoch": 1.42, "grad_norm": 12.625, "learning_rate": 1.0168144891754653e-06, "loss": 0.2708, "step": 7518 }, { "epoch": 1.42, "grad_norm": 9.25, "learning_rate": 1.0161979741736159e-06, "loss": 0.225, "step": 7519 }, { "epoch": 1.42, "grad_norm": 10.6875, "learning_rate": 1.015581598451868e-06, "loss": 0.2349, "step": 7520 }, { "epoch": 1.42, "grad_norm": 18.375, "learning_rate": 1.0149653620680794e-06, "loss": 0.2095, "step": 7521 }, { "epoch": 1.42, "grad_norm": 7.84375, "learning_rate": 1.0143492650800932e-06, "loss": 0.136, "step": 7522 }, { "epoch": 1.42, "grad_norm": 12.5625, "learning_rate": 1.0137333075457435e-06, "loss": 0.1981, "step": 7523 }, { "epoch": 1.42, "grad_norm": 9.3125, "learning_rate": 1.0131174895228454e-06, "loss": 0.1717, "step": 7524 }, { "epoch": 1.42, "grad_norm": 9.9375, "learning_rate": 1.0125018110692053e-06, "loss": 0.214, "step": 7525 }, { "epoch": 1.42, "grad_norm": 13.4375, "learning_rate": 1.011886272242614e-06, "loss": 0.2774, "step": 7526 }, { "epoch": 1.42, "grad_norm": 10.6875, "learning_rate": 1.0112708731008527e-06, "loss": 0.202, "step": 7527 }, { "epoch": 1.42, "grad_norm": 10.4375, "learning_rate": 1.0106556137016852e-06, "loss": 0.1894, "step": 7528 }, { "epoch": 1.42, "grad_norm": 15.8125, "learning_rate": 1.010040494102864e-06, "loss": 0.2073, "step": 7529 }, { "epoch": 1.42, "grad_norm": 9.6875, "learning_rate": 1.0094255143621282e-06, "loss": 0.2157, "step": 7530 }, { "epoch": 1.42, "grad_norm": 13.75, "learning_rate": 1.0088106745372067e-06, "loss": 0.2188, "step": 7531 }, { "epoch": 1.42, "grad_norm": 12.375, "learning_rate": 1.00819597468581e-06, "loss": 0.2169, "step": 7532 }, { "epoch": 1.43, "grad_norm": 14.25, "learning_rate": 1.0075814148656383e-06, "loss": 0.2021, "step": 7533 }, { "epoch": 1.43, "grad_norm": 13.375, "learning_rate": 1.00696699513438e-06, "loss": 0.1964, "step": 7534 }, { "epoch": 1.43, "grad_norm": 13.3125, "learning_rate": 1.006352715549709e-06, "loss": 0.1661, "step": 7535 }, { "epoch": 1.43, "grad_norm": 11.0, "learning_rate": 1.0057385761692841e-06, "loss": 0.187, "step": 7536 }, { "epoch": 1.43, "grad_norm": 14.375, "learning_rate": 1.0051245770507532e-06, "loss": 0.1658, "step": 7537 }, { "epoch": 1.43, "grad_norm": 15.875, "learning_rate": 1.004510718251752e-06, "loss": 0.1765, "step": 7538 }, { "epoch": 1.43, "grad_norm": 9.625, "learning_rate": 1.0038969998299011e-06, "loss": 0.1932, "step": 7539 }, { "epoch": 1.43, "grad_norm": 9.8125, "learning_rate": 1.0032834218428072e-06, "loss": 0.1642, "step": 7540 }, { "epoch": 1.43, "grad_norm": 10.5, "learning_rate": 1.0026699843480652e-06, "loss": 0.1671, "step": 7541 }, { "epoch": 1.43, "grad_norm": 10.25, "learning_rate": 1.002056687403258e-06, "loss": 0.2056, "step": 7542 }, { "epoch": 1.43, "grad_norm": 12.625, "learning_rate": 1.0014435310659543e-06, "loss": 0.2177, "step": 7543 }, { "epoch": 1.43, "grad_norm": 11.75, "learning_rate": 1.0008305153937065e-06, "loss": 0.2077, "step": 7544 }, { "epoch": 1.43, "grad_norm": 13.0625, "learning_rate": 1.0002176404440591e-06, "loss": 0.2609, "step": 7545 }, { "epoch": 1.43, "grad_norm": 15.875, "learning_rate": 9.99604906274541e-07, "loss": 0.2232, "step": 7546 }, { "epoch": 1.43, "grad_norm": 9.8125, "learning_rate": 9.989923129426663e-07, "loss": 0.1777, "step": 7547 }, { "epoch": 1.43, "grad_norm": 10.3125, "learning_rate": 9.983798605059364e-07, "loss": 0.2027, "step": 7548 }, { "epoch": 1.43, "grad_norm": 13.125, "learning_rate": 9.97767549021843e-07, "loss": 0.2293, "step": 7549 }, { "epoch": 1.43, "grad_norm": 10.375, "learning_rate": 9.971553785478616e-07, "loss": 0.1962, "step": 7550 }, { "epoch": 1.43, "grad_norm": 10.625, "learning_rate": 9.965433491414534e-07, "loss": 0.1932, "step": 7551 }, { "epoch": 1.43, "grad_norm": 13.0, "learning_rate": 9.95931460860067e-07, "loss": 0.2127, "step": 7552 }, { "epoch": 1.43, "grad_norm": 11.6875, "learning_rate": 9.953197137611415e-07, "loss": 0.1773, "step": 7553 }, { "epoch": 1.43, "grad_norm": 10.875, "learning_rate": 9.947081079020986e-07, "loss": 0.2355, "step": 7554 }, { "epoch": 1.43, "grad_norm": 11.8125, "learning_rate": 9.94096643340346e-07, "loss": 0.2019, "step": 7555 }, { "epoch": 1.43, "grad_norm": 17.125, "learning_rate": 9.934853201332825e-07, "loss": 0.2014, "step": 7556 }, { "epoch": 1.43, "grad_norm": 10.5, "learning_rate": 9.928741383382906e-07, "loss": 0.2356, "step": 7557 }, { "epoch": 1.43, "grad_norm": 14.25, "learning_rate": 9.922630980127404e-07, "loss": 0.2165, "step": 7558 }, { "epoch": 1.43, "grad_norm": 17.875, "learning_rate": 9.916521992139862e-07, "loss": 0.197, "step": 7559 }, { "epoch": 1.43, "grad_norm": 8.6875, "learning_rate": 9.910414419993742e-07, "loss": 0.2029, "step": 7560 }, { "epoch": 1.43, "grad_norm": 9.9375, "learning_rate": 9.904308264262328e-07, "loss": 0.2457, "step": 7561 }, { "epoch": 1.43, "grad_norm": 12.375, "learning_rate": 9.8982035255188e-07, "loss": 0.2338, "step": 7562 }, { "epoch": 1.43, "grad_norm": 12.5, "learning_rate": 9.892100204336166e-07, "loss": 0.2203, "step": 7563 }, { "epoch": 1.43, "grad_norm": 15.75, "learning_rate": 9.88599830128735e-07, "loss": 0.185, "step": 7564 }, { "epoch": 1.43, "grad_norm": 20.625, "learning_rate": 9.879897816945111e-07, "loss": 0.2261, "step": 7565 }, { "epoch": 1.43, "grad_norm": 11.9375, "learning_rate": 9.873798751882088e-07, "loss": 0.2167, "step": 7566 }, { "epoch": 1.43, "grad_norm": 13.125, "learning_rate": 9.867701106670776e-07, "loss": 0.1958, "step": 7567 }, { "epoch": 1.43, "grad_norm": 11.375, "learning_rate": 9.861604881883547e-07, "loss": 0.193, "step": 7568 }, { "epoch": 1.43, "grad_norm": 10.6875, "learning_rate": 9.855510078092633e-07, "loss": 0.1951, "step": 7569 }, { "epoch": 1.43, "grad_norm": 12.5, "learning_rate": 9.849416695870137e-07, "loss": 0.2208, "step": 7570 }, { "epoch": 1.43, "grad_norm": 11.875, "learning_rate": 9.843324735788026e-07, "loss": 0.1917, "step": 7571 }, { "epoch": 1.43, "grad_norm": 13.4375, "learning_rate": 9.837234198418137e-07, "loss": 0.198, "step": 7572 }, { "epoch": 1.43, "grad_norm": 11.875, "learning_rate": 9.831145084332174e-07, "loss": 0.268, "step": 7573 }, { "epoch": 1.43, "grad_norm": 15.9375, "learning_rate": 9.825057394101684e-07, "loss": 0.2036, "step": 7574 }, { "epoch": 1.43, "grad_norm": 9.9375, "learning_rate": 9.81897112829812e-07, "loss": 0.1714, "step": 7575 }, { "epoch": 1.43, "grad_norm": 13.125, "learning_rate": 9.812886287492778e-07, "loss": 0.2472, "step": 7576 }, { "epoch": 1.43, "grad_norm": 11.8125, "learning_rate": 9.806802872256821e-07, "loss": 0.1437, "step": 7577 }, { "epoch": 1.43, "grad_norm": 11.9375, "learning_rate": 9.800720883161285e-07, "loss": 0.2266, "step": 7578 }, { "epoch": 1.43, "grad_norm": 18.5, "learning_rate": 9.794640320777067e-07, "loss": 0.1519, "step": 7579 }, { "epoch": 1.43, "grad_norm": 12.1875, "learning_rate": 9.788561185674925e-07, "loss": 0.2136, "step": 7580 }, { "epoch": 1.43, "grad_norm": 10.5, "learning_rate": 9.782483478425498e-07, "loss": 0.1932, "step": 7581 }, { "epoch": 1.43, "grad_norm": 12.1875, "learning_rate": 9.77640719959928e-07, "loss": 0.2211, "step": 7582 }, { "epoch": 1.43, "grad_norm": 11.8125, "learning_rate": 9.77033234976663e-07, "loss": 0.1855, "step": 7583 }, { "epoch": 1.43, "grad_norm": 10.875, "learning_rate": 9.76425892949778e-07, "loss": 0.2115, "step": 7584 }, { "epoch": 1.43, "grad_norm": 12.5, "learning_rate": 9.758186939362825e-07, "loss": 0.2739, "step": 7585 }, { "epoch": 1.44, "grad_norm": 10.5, "learning_rate": 9.752116379931717e-07, "loss": 0.1919, "step": 7586 }, { "epoch": 1.44, "grad_norm": 9.375, "learning_rate": 9.74604725177429e-07, "loss": 0.2474, "step": 7587 }, { "epoch": 1.44, "grad_norm": 13.0, "learning_rate": 9.739979555460233e-07, "loss": 0.2646, "step": 7588 }, { "epoch": 1.44, "grad_norm": 12.125, "learning_rate": 9.7339132915591e-07, "loss": 0.2105, "step": 7589 }, { "epoch": 1.44, "grad_norm": 10.9375, "learning_rate": 9.727848460640314e-07, "loss": 0.1773, "step": 7590 }, { "epoch": 1.44, "grad_norm": 10.25, "learning_rate": 9.721785063273161e-07, "loss": 0.1437, "step": 7591 }, { "epoch": 1.44, "grad_norm": 12.5625, "learning_rate": 9.715723100026803e-07, "loss": 0.1433, "step": 7592 }, { "epoch": 1.44, "grad_norm": 10.4375, "learning_rate": 9.709662571470246e-07, "loss": 0.2356, "step": 7593 }, { "epoch": 1.44, "grad_norm": 13.25, "learning_rate": 9.703603478172382e-07, "loss": 0.2728, "step": 7594 }, { "epoch": 1.44, "grad_norm": 8.5625, "learning_rate": 9.69754582070196e-07, "loss": 0.174, "step": 7595 }, { "epoch": 1.44, "grad_norm": 11.875, "learning_rate": 9.69148959962759e-07, "loss": 0.1895, "step": 7596 }, { "epoch": 1.44, "grad_norm": 9.75, "learning_rate": 9.685434815517759e-07, "loss": 0.1508, "step": 7597 }, { "epoch": 1.44, "grad_norm": 12.5, "learning_rate": 9.6793814689408e-07, "loss": 0.2138, "step": 7598 }, { "epoch": 1.44, "grad_norm": 13.0, "learning_rate": 9.673329560464934e-07, "loss": 0.207, "step": 7599 }, { "epoch": 1.44, "grad_norm": 10.8125, "learning_rate": 9.667279090658226e-07, "loss": 0.1655, "step": 7600 }, { "epoch": 1.44, "grad_norm": 13.5, "learning_rate": 9.661230060088627e-07, "loss": 0.2064, "step": 7601 }, { "epoch": 1.44, "grad_norm": 11.6875, "learning_rate": 9.655182469323931e-07, "loss": 0.2094, "step": 7602 }, { "epoch": 1.44, "grad_norm": 11.0, "learning_rate": 9.649136318931815e-07, "loss": 0.1968, "step": 7603 }, { "epoch": 1.44, "grad_norm": 12.4375, "learning_rate": 9.643091609479806e-07, "loss": 0.1731, "step": 7604 }, { "epoch": 1.44, "grad_norm": 17.0, "learning_rate": 9.637048341535312e-07, "loss": 0.2535, "step": 7605 }, { "epoch": 1.44, "grad_norm": 8.25, "learning_rate": 9.631006515665588e-07, "loss": 0.1884, "step": 7606 }, { "epoch": 1.44, "grad_norm": 9.5, "learning_rate": 9.62496613243777e-07, "loss": 0.1578, "step": 7607 }, { "epoch": 1.44, "grad_norm": 8.8125, "learning_rate": 9.618927192418842e-07, "loss": 0.1601, "step": 7608 }, { "epoch": 1.44, "grad_norm": 10.6875, "learning_rate": 9.61288969617567e-07, "loss": 0.21, "step": 7609 }, { "epoch": 1.44, "grad_norm": 12.1875, "learning_rate": 9.60685364427497e-07, "loss": 0.215, "step": 7610 }, { "epoch": 1.44, "grad_norm": 42.75, "learning_rate": 9.60081903728333e-07, "loss": 0.2068, "step": 7611 }, { "epoch": 1.44, "grad_norm": 10.6875, "learning_rate": 9.594785875767204e-07, "loss": 0.2443, "step": 7612 }, { "epoch": 1.44, "grad_norm": 10.875, "learning_rate": 9.588754160292902e-07, "loss": 0.2286, "step": 7613 }, { "epoch": 1.44, "grad_norm": 10.4375, "learning_rate": 9.582723891426595e-07, "loss": 0.2047, "step": 7614 }, { "epoch": 1.44, "grad_norm": 9.125, "learning_rate": 9.576695069734358e-07, "loss": 0.1985, "step": 7615 }, { "epoch": 1.44, "grad_norm": 10.5, "learning_rate": 9.570667695782069e-07, "loss": 0.266, "step": 7616 }, { "epoch": 1.44, "grad_norm": 11.1875, "learning_rate": 9.564641770135505e-07, "loss": 0.1986, "step": 7617 }, { "epoch": 1.44, "grad_norm": 17.0, "learning_rate": 9.558617293360298e-07, "loss": 0.2741, "step": 7618 }, { "epoch": 1.44, "grad_norm": 10.125, "learning_rate": 9.552594266021972e-07, "loss": 0.2269, "step": 7619 }, { "epoch": 1.44, "grad_norm": 15.5625, "learning_rate": 9.546572688685867e-07, "loss": 0.2378, "step": 7620 }, { "epoch": 1.44, "grad_norm": 12.0, "learning_rate": 9.540552561917215e-07, "loss": 0.2853, "step": 7621 }, { "epoch": 1.44, "grad_norm": 12.125, "learning_rate": 9.534533886281103e-07, "loss": 0.2025, "step": 7622 }, { "epoch": 1.44, "grad_norm": 12.6875, "learning_rate": 9.528516662342507e-07, "loss": 0.2239, "step": 7623 }, { "epoch": 1.44, "grad_norm": 11.5625, "learning_rate": 9.522500890666228e-07, "loss": 0.2059, "step": 7624 }, { "epoch": 1.44, "grad_norm": 12.5, "learning_rate": 9.516486571816941e-07, "loss": 0.231, "step": 7625 }, { "epoch": 1.44, "grad_norm": 11.8125, "learning_rate": 9.510473706359214e-07, "loss": 0.1619, "step": 7626 }, { "epoch": 1.44, "grad_norm": 15.6875, "learning_rate": 9.504462294857453e-07, "loss": 0.2298, "step": 7627 }, { "epoch": 1.44, "grad_norm": 9.8125, "learning_rate": 9.498452337875918e-07, "loss": 0.2012, "step": 7628 }, { "epoch": 1.44, "grad_norm": 14.375, "learning_rate": 9.492443835978746e-07, "loss": 0.2607, "step": 7629 }, { "epoch": 1.44, "grad_norm": 10.375, "learning_rate": 9.486436789729961e-07, "loss": 0.2354, "step": 7630 }, { "epoch": 1.44, "grad_norm": 12.3125, "learning_rate": 9.4804311996934e-07, "loss": 0.2, "step": 7631 }, { "epoch": 1.44, "grad_norm": 12.4375, "learning_rate": 9.474427066432801e-07, "loss": 0.1614, "step": 7632 }, { "epoch": 1.44, "grad_norm": 11.4375, "learning_rate": 9.468424390511743e-07, "loss": 0.2036, "step": 7633 }, { "epoch": 1.44, "grad_norm": 29.875, "learning_rate": 9.46242317249371e-07, "loss": 0.244, "step": 7634 }, { "epoch": 1.44, "grad_norm": 12.25, "learning_rate": 9.456423412941986e-07, "loss": 0.2452, "step": 7635 }, { "epoch": 1.44, "grad_norm": 11.9375, "learning_rate": 9.450425112419756e-07, "loss": 0.2527, "step": 7636 }, { "epoch": 1.44, "grad_norm": 8.3125, "learning_rate": 9.444428271490077e-07, "loss": 0.2235, "step": 7637 }, { "epoch": 1.44, "grad_norm": 10.375, "learning_rate": 9.438432890715857e-07, "loss": 0.1781, "step": 7638 }, { "epoch": 1.45, "grad_norm": 10.5, "learning_rate": 9.432438970659846e-07, "loss": 0.174, "step": 7639 }, { "epoch": 1.45, "grad_norm": 11.75, "learning_rate": 9.426446511884676e-07, "loss": 0.246, "step": 7640 }, { "epoch": 1.45, "grad_norm": 20.0, "learning_rate": 9.420455514952859e-07, "loss": 0.2226, "step": 7641 }, { "epoch": 1.45, "grad_norm": 10.1875, "learning_rate": 9.414465980426748e-07, "loss": 0.1861, "step": 7642 }, { "epoch": 1.45, "grad_norm": 10.4375, "learning_rate": 9.408477908868554e-07, "loss": 0.1751, "step": 7643 }, { "epoch": 1.45, "grad_norm": 8.1875, "learning_rate": 9.402491300840352e-07, "loss": 0.1747, "step": 7644 }, { "epoch": 1.45, "grad_norm": 11.8125, "learning_rate": 9.396506156904106e-07, "loss": 0.274, "step": 7645 }, { "epoch": 1.45, "grad_norm": 12.75, "learning_rate": 9.390522477621627e-07, "loss": 0.2592, "step": 7646 }, { "epoch": 1.45, "grad_norm": 12.0625, "learning_rate": 9.384540263554557e-07, "loss": 0.2102, "step": 7647 }, { "epoch": 1.45, "grad_norm": 10.75, "learning_rate": 9.378559515264454e-07, "loss": 0.2098, "step": 7648 }, { "epoch": 1.45, "grad_norm": 13.5625, "learning_rate": 9.372580233312706e-07, "loss": 0.1644, "step": 7649 }, { "epoch": 1.45, "grad_norm": 13.125, "learning_rate": 9.366602418260578e-07, "loss": 0.2539, "step": 7650 }, { "epoch": 1.45, "grad_norm": 15.75, "learning_rate": 9.360626070669163e-07, "loss": 0.2631, "step": 7651 }, { "epoch": 1.45, "grad_norm": 19.875, "learning_rate": 9.354651191099473e-07, "loss": 0.185, "step": 7652 }, { "epoch": 1.45, "grad_norm": 11.875, "learning_rate": 9.348677780112339e-07, "loss": 0.2079, "step": 7653 }, { "epoch": 1.45, "grad_norm": 14.0, "learning_rate": 9.342705838268478e-07, "loss": 0.2177, "step": 7654 }, { "epoch": 1.45, "grad_norm": 11.0, "learning_rate": 9.336735366128433e-07, "loss": 0.1794, "step": 7655 }, { "epoch": 1.45, "grad_norm": 11.75, "learning_rate": 9.330766364252658e-07, "loss": 0.2176, "step": 7656 }, { "epoch": 1.45, "grad_norm": 11.4375, "learning_rate": 9.324798833201452e-07, "loss": 0.1996, "step": 7657 }, { "epoch": 1.45, "grad_norm": 12.5, "learning_rate": 9.318832773534934e-07, "loss": 0.2301, "step": 7658 }, { "epoch": 1.45, "grad_norm": 12.5625, "learning_rate": 9.312868185813153e-07, "loss": 0.2125, "step": 7659 }, { "epoch": 1.45, "grad_norm": 10.3125, "learning_rate": 9.306905070595979e-07, "loss": 0.2367, "step": 7660 }, { "epoch": 1.45, "grad_norm": 11.6875, "learning_rate": 9.300943428443155e-07, "loss": 0.2371, "step": 7661 }, { "epoch": 1.45, "grad_norm": 11.0, "learning_rate": 9.294983259914261e-07, "loss": 0.2121, "step": 7662 }, { "epoch": 1.45, "grad_norm": 8.4375, "learning_rate": 9.289024565568789e-07, "loss": 0.1975, "step": 7663 }, { "epoch": 1.45, "grad_norm": 8.6875, "learning_rate": 9.283067345966048e-07, "loss": 0.1572, "step": 7664 }, { "epoch": 1.45, "grad_norm": 12.625, "learning_rate": 9.277111601665237e-07, "loss": 0.192, "step": 7665 }, { "epoch": 1.45, "grad_norm": 11.6875, "learning_rate": 9.271157333225381e-07, "loss": 0.1984, "step": 7666 }, { "epoch": 1.45, "grad_norm": 15.0625, "learning_rate": 9.265204541205411e-07, "loss": 0.1755, "step": 7667 }, { "epoch": 1.45, "grad_norm": 14.4375, "learning_rate": 9.259253226164091e-07, "loss": 0.2405, "step": 7668 }, { "epoch": 1.45, "grad_norm": 14.875, "learning_rate": 9.253303388660054e-07, "loss": 0.2413, "step": 7669 }, { "epoch": 1.45, "grad_norm": 11.4375, "learning_rate": 9.247355029251792e-07, "loss": 0.2283, "step": 7670 }, { "epoch": 1.45, "grad_norm": 9.8125, "learning_rate": 9.24140814849766e-07, "loss": 0.172, "step": 7671 }, { "epoch": 1.45, "grad_norm": 9.375, "learning_rate": 9.235462746955876e-07, "loss": 0.2014, "step": 7672 }, { "epoch": 1.45, "grad_norm": 11.375, "learning_rate": 9.229518825184516e-07, "loss": 0.159, "step": 7673 }, { "epoch": 1.45, "grad_norm": 8.9375, "learning_rate": 9.223576383741519e-07, "loss": 0.1986, "step": 7674 }, { "epoch": 1.45, "grad_norm": 18.625, "learning_rate": 9.217635423184684e-07, "loss": 0.2063, "step": 7675 }, { "epoch": 1.45, "grad_norm": 12.3125, "learning_rate": 9.211695944071671e-07, "loss": 0.1784, "step": 7676 }, { "epoch": 1.45, "grad_norm": 13.875, "learning_rate": 9.205757946960001e-07, "loss": 0.2044, "step": 7677 }, { "epoch": 1.45, "grad_norm": 13.5, "learning_rate": 9.199821432407058e-07, "loss": 0.2445, "step": 7678 }, { "epoch": 1.45, "grad_norm": 13.375, "learning_rate": 9.193886400970082e-07, "loss": 0.244, "step": 7679 }, { "epoch": 1.45, "grad_norm": 11.75, "learning_rate": 9.187952853206181e-07, "loss": 0.191, "step": 7680 }, { "epoch": 1.45, "grad_norm": 11.875, "learning_rate": 9.182020789672313e-07, "loss": 0.2139, "step": 7681 }, { "epoch": 1.45, "grad_norm": 11.0625, "learning_rate": 9.176090210925312e-07, "loss": 0.1851, "step": 7682 }, { "epoch": 1.45, "grad_norm": 10.6875, "learning_rate": 9.170161117521855e-07, "loss": 0.2638, "step": 7683 }, { "epoch": 1.45, "grad_norm": 12.4375, "learning_rate": 9.164233510018497e-07, "loss": 0.2385, "step": 7684 }, { "epoch": 1.45, "grad_norm": 10.75, "learning_rate": 9.158307388971637e-07, "loss": 0.1698, "step": 7685 }, { "epoch": 1.45, "grad_norm": 12.625, "learning_rate": 9.15238275493755e-07, "loss": 0.2196, "step": 7686 }, { "epoch": 1.45, "grad_norm": 13.125, "learning_rate": 9.146459608472355e-07, "loss": 0.2109, "step": 7687 }, { "epoch": 1.45, "grad_norm": 8.625, "learning_rate": 9.14053795013205e-07, "loss": 0.1764, "step": 7688 }, { "epoch": 1.45, "grad_norm": 14.6875, "learning_rate": 9.134617780472477e-07, "loss": 0.19, "step": 7689 }, { "epoch": 1.45, "grad_norm": 12.25, "learning_rate": 9.128699100049343e-07, "loss": 0.2711, "step": 7690 }, { "epoch": 1.45, "grad_norm": 13.125, "learning_rate": 9.122781909418224e-07, "loss": 0.2382, "step": 7691 }, { "epoch": 1.46, "grad_norm": 15.5, "learning_rate": 9.116866209134545e-07, "loss": 0.2193, "step": 7692 }, { "epoch": 1.46, "grad_norm": 11.6875, "learning_rate": 9.110951999753598e-07, "loss": 0.1637, "step": 7693 }, { "epoch": 1.46, "grad_norm": 14.375, "learning_rate": 9.105039281830527e-07, "loss": 0.2365, "step": 7694 }, { "epoch": 1.46, "grad_norm": 10.375, "learning_rate": 9.099128055920345e-07, "loss": 0.2183, "step": 7695 }, { "epoch": 1.46, "grad_norm": 7.15625, "learning_rate": 9.093218322577921e-07, "loss": 0.1487, "step": 7696 }, { "epoch": 1.46, "grad_norm": 8.6875, "learning_rate": 9.087310082357986e-07, "loss": 0.1813, "step": 7697 }, { "epoch": 1.46, "grad_norm": 11.125, "learning_rate": 9.081403335815126e-07, "loss": 0.2391, "step": 7698 }, { "epoch": 1.46, "grad_norm": 11.5, "learning_rate": 9.075498083503789e-07, "loss": 0.1616, "step": 7699 }, { "epoch": 1.46, "grad_norm": 10.6875, "learning_rate": 9.069594325978287e-07, "loss": 0.2285, "step": 7700 }, { "epoch": 1.46, "grad_norm": 14.75, "learning_rate": 9.063692063792787e-07, "loss": 0.1959, "step": 7701 }, { "epoch": 1.46, "grad_norm": 21.75, "learning_rate": 9.057791297501306e-07, "loss": 0.2919, "step": 7702 }, { "epoch": 1.46, "grad_norm": 10.375, "learning_rate": 9.051892027657761e-07, "loss": 0.2012, "step": 7703 }, { "epoch": 1.46, "grad_norm": 12.125, "learning_rate": 9.045994254815871e-07, "loss": 0.2103, "step": 7704 }, { "epoch": 1.46, "grad_norm": 19.25, "learning_rate": 9.04009797952925e-07, "loss": 0.2185, "step": 7705 }, { "epoch": 1.46, "grad_norm": 11.5, "learning_rate": 9.034203202351357e-07, "loss": 0.2234, "step": 7706 }, { "epoch": 1.46, "grad_norm": 10.4375, "learning_rate": 9.028309923835543e-07, "loss": 0.2054, "step": 7707 }, { "epoch": 1.46, "grad_norm": 9.0, "learning_rate": 9.022418144534966e-07, "loss": 0.1679, "step": 7708 }, { "epoch": 1.46, "grad_norm": 9.125, "learning_rate": 9.016527865002681e-07, "loss": 0.1725, "step": 7709 }, { "epoch": 1.46, "grad_norm": 10.125, "learning_rate": 9.010639085791581e-07, "loss": 0.1841, "step": 7710 }, { "epoch": 1.46, "grad_norm": 10.9375, "learning_rate": 9.004751807454451e-07, "loss": 0.1714, "step": 7711 }, { "epoch": 1.46, "grad_norm": 12.125, "learning_rate": 8.99886603054389e-07, "loss": 0.2285, "step": 7712 }, { "epoch": 1.46, "grad_norm": 10.9375, "learning_rate": 8.992981755612378e-07, "loss": 0.2091, "step": 7713 }, { "epoch": 1.46, "grad_norm": 13.3125, "learning_rate": 8.987098983212278e-07, "loss": 0.189, "step": 7714 }, { "epoch": 1.46, "grad_norm": 11.9375, "learning_rate": 8.981217713895768e-07, "loss": 0.2037, "step": 7715 }, { "epoch": 1.46, "grad_norm": 10.8125, "learning_rate": 8.975337948214907e-07, "loss": 0.1798, "step": 7716 }, { "epoch": 1.46, "grad_norm": 13.8125, "learning_rate": 8.969459686721607e-07, "loss": 0.1918, "step": 7717 }, { "epoch": 1.46, "grad_norm": 10.125, "learning_rate": 8.963582929967668e-07, "loss": 0.2326, "step": 7718 }, { "epoch": 1.46, "grad_norm": 8.5, "learning_rate": 8.957707678504699e-07, "loss": 0.1848, "step": 7719 }, { "epoch": 1.46, "grad_norm": 9.0625, "learning_rate": 8.951833932884199e-07, "loss": 0.2365, "step": 7720 }, { "epoch": 1.46, "grad_norm": 8.6875, "learning_rate": 8.945961693657512e-07, "loss": 0.1712, "step": 7721 }, { "epoch": 1.46, "grad_norm": 10.25, "learning_rate": 8.940090961375872e-07, "loss": 0.2717, "step": 7722 }, { "epoch": 1.46, "grad_norm": 10.375, "learning_rate": 8.93422173659032e-07, "loss": 0.2012, "step": 7723 }, { "epoch": 1.46, "grad_norm": 10.25, "learning_rate": 8.92835401985179e-07, "loss": 0.2239, "step": 7724 }, { "epoch": 1.46, "grad_norm": 13.25, "learning_rate": 8.922487811711076e-07, "loss": 0.1751, "step": 7725 }, { "epoch": 1.46, "grad_norm": 16.75, "learning_rate": 8.916623112718825e-07, "loss": 0.209, "step": 7726 }, { "epoch": 1.46, "grad_norm": 14.9375, "learning_rate": 8.910759923425524e-07, "loss": 0.2542, "step": 7727 }, { "epoch": 1.46, "grad_norm": 11.6875, "learning_rate": 8.904898244381529e-07, "loss": 0.2266, "step": 7728 }, { "epoch": 1.46, "grad_norm": 14.25, "learning_rate": 8.899038076137079e-07, "loss": 0.1828, "step": 7729 }, { "epoch": 1.46, "grad_norm": 15.25, "learning_rate": 8.893179419242245e-07, "loss": 0.2089, "step": 7730 }, { "epoch": 1.46, "grad_norm": 11.1875, "learning_rate": 8.887322274246954e-07, "loss": 0.1805, "step": 7731 }, { "epoch": 1.46, "grad_norm": 12.25, "learning_rate": 8.881466641700992e-07, "loss": 0.2302, "step": 7732 }, { "epoch": 1.46, "grad_norm": 16.375, "learning_rate": 8.875612522154028e-07, "loss": 0.258, "step": 7733 }, { "epoch": 1.46, "grad_norm": 8.5625, "learning_rate": 8.869759916155574e-07, "loss": 0.1528, "step": 7734 }, { "epoch": 1.46, "grad_norm": 10.5, "learning_rate": 8.863908824254966e-07, "loss": 0.1959, "step": 7735 }, { "epoch": 1.46, "grad_norm": 14.875, "learning_rate": 8.858059247001457e-07, "loss": 0.2012, "step": 7736 }, { "epoch": 1.46, "grad_norm": 12.9375, "learning_rate": 8.85221118494412e-07, "loss": 0.1562, "step": 7737 }, { "epoch": 1.46, "grad_norm": 14.625, "learning_rate": 8.846364638631907e-07, "loss": 0.2079, "step": 7738 }, { "epoch": 1.46, "grad_norm": 12.1875, "learning_rate": 8.840519608613591e-07, "loss": 0.2267, "step": 7739 }, { "epoch": 1.46, "grad_norm": 18.875, "learning_rate": 8.834676095437844e-07, "loss": 0.1674, "step": 7740 }, { "epoch": 1.46, "grad_norm": 10.25, "learning_rate": 8.82883409965318e-07, "loss": 0.1954, "step": 7741 }, { "epoch": 1.46, "grad_norm": 12.0, "learning_rate": 8.822993621807976e-07, "loss": 0.2291, "step": 7742 }, { "epoch": 1.46, "grad_norm": 10.125, "learning_rate": 8.817154662450436e-07, "loss": 0.2231, "step": 7743 }, { "epoch": 1.46, "grad_norm": 10.4375, "learning_rate": 8.811317222128668e-07, "loss": 0.1869, "step": 7744 }, { "epoch": 1.47, "grad_norm": 12.3125, "learning_rate": 8.805481301390617e-07, "loss": 0.196, "step": 7745 }, { "epoch": 1.47, "grad_norm": 8.8125, "learning_rate": 8.799646900784057e-07, "loss": 0.2046, "step": 7746 }, { "epoch": 1.47, "grad_norm": 11.1875, "learning_rate": 8.793814020856672e-07, "loss": 0.2186, "step": 7747 }, { "epoch": 1.47, "grad_norm": 8.5625, "learning_rate": 8.787982662155969e-07, "loss": 0.1745, "step": 7748 }, { "epoch": 1.47, "grad_norm": 11.375, "learning_rate": 8.782152825229331e-07, "loss": 0.1883, "step": 7749 }, { "epoch": 1.47, "grad_norm": 10.125, "learning_rate": 8.77632451062396e-07, "loss": 0.1919, "step": 7750 }, { "epoch": 1.47, "grad_norm": 11.625, "learning_rate": 8.770497718886966e-07, "loss": 0.1985, "step": 7751 }, { "epoch": 1.47, "grad_norm": 9.6875, "learning_rate": 8.764672450565289e-07, "loss": 0.1918, "step": 7752 }, { "epoch": 1.47, "grad_norm": 13.8125, "learning_rate": 8.758848706205733e-07, "loss": 0.2273, "step": 7753 }, { "epoch": 1.47, "grad_norm": 9.1875, "learning_rate": 8.753026486354934e-07, "loss": 0.1875, "step": 7754 }, { "epoch": 1.47, "grad_norm": 12.0, "learning_rate": 8.74720579155943e-07, "loss": 0.2298, "step": 7755 }, { "epoch": 1.47, "grad_norm": 10.625, "learning_rate": 8.741386622365586e-07, "loss": 0.1637, "step": 7756 }, { "epoch": 1.47, "grad_norm": 13.0625, "learning_rate": 8.735568979319628e-07, "loss": 0.1862, "step": 7757 }, { "epoch": 1.47, "grad_norm": 9.9375, "learning_rate": 8.729752862967641e-07, "loss": 0.1294, "step": 7758 }, { "epoch": 1.47, "grad_norm": 14.5625, "learning_rate": 8.723938273855567e-07, "loss": 0.2019, "step": 7759 }, { "epoch": 1.47, "grad_norm": 10.875, "learning_rate": 8.718125212529205e-07, "loss": 0.1815, "step": 7760 }, { "epoch": 1.47, "grad_norm": 13.0, "learning_rate": 8.712313679534209e-07, "loss": 0.2998, "step": 7761 }, { "epoch": 1.47, "grad_norm": 11.375, "learning_rate": 8.706503675416092e-07, "loss": 0.2016, "step": 7762 }, { "epoch": 1.47, "grad_norm": 15.4375, "learning_rate": 8.700695200720218e-07, "loss": 0.224, "step": 7763 }, { "epoch": 1.47, "grad_norm": 9.75, "learning_rate": 8.694888255991815e-07, "loss": 0.1393, "step": 7764 }, { "epoch": 1.47, "grad_norm": 13.3125, "learning_rate": 8.689082841775964e-07, "loss": 0.2127, "step": 7765 }, { "epoch": 1.47, "grad_norm": 10.1875, "learning_rate": 8.6832789586176e-07, "loss": 0.2066, "step": 7766 }, { "epoch": 1.47, "grad_norm": 14.9375, "learning_rate": 8.677476607061516e-07, "loss": 0.2306, "step": 7767 }, { "epoch": 1.47, "grad_norm": 9.6875, "learning_rate": 8.671675787652364e-07, "loss": 0.2187, "step": 7768 }, { "epoch": 1.47, "grad_norm": 17.625, "learning_rate": 8.665876500934648e-07, "loss": 0.2456, "step": 7769 }, { "epoch": 1.47, "grad_norm": 10.3125, "learning_rate": 8.66007874745273e-07, "loss": 0.1816, "step": 7770 }, { "epoch": 1.47, "grad_norm": 12.875, "learning_rate": 8.654282527750829e-07, "loss": 0.188, "step": 7771 }, { "epoch": 1.47, "grad_norm": 11.125, "learning_rate": 8.648487842373016e-07, "loss": 0.2316, "step": 7772 }, { "epoch": 1.47, "grad_norm": 13.375, "learning_rate": 8.642694691863224e-07, "loss": 0.2095, "step": 7773 }, { "epoch": 1.47, "grad_norm": 18.75, "learning_rate": 8.636903076765238e-07, "loss": 0.2142, "step": 7774 }, { "epoch": 1.47, "grad_norm": 14.5, "learning_rate": 8.631112997622701e-07, "loss": 0.2072, "step": 7775 }, { "epoch": 1.47, "grad_norm": 11.0, "learning_rate": 8.625324454979109e-07, "loss": 0.2211, "step": 7776 }, { "epoch": 1.47, "grad_norm": 13.25, "learning_rate": 8.619537449377813e-07, "loss": 0.2316, "step": 7777 }, { "epoch": 1.47, "grad_norm": 12.5, "learning_rate": 8.613751981362028e-07, "loss": 0.1819, "step": 7778 }, { "epoch": 1.47, "grad_norm": 11.6875, "learning_rate": 8.607968051474813e-07, "loss": 0.289, "step": 7779 }, { "epoch": 1.47, "grad_norm": 11.6875, "learning_rate": 8.602185660259094e-07, "loss": 0.2176, "step": 7780 }, { "epoch": 1.47, "grad_norm": 13.5625, "learning_rate": 8.596404808257641e-07, "loss": 0.2107, "step": 7781 }, { "epoch": 1.47, "grad_norm": 8.8125, "learning_rate": 8.590625496013088e-07, "loss": 0.1447, "step": 7782 }, { "epoch": 1.47, "grad_norm": 10.375, "learning_rate": 8.584847724067924e-07, "loss": 0.2055, "step": 7783 }, { "epoch": 1.47, "grad_norm": 13.625, "learning_rate": 8.579071492964486e-07, "loss": 0.2129, "step": 7784 }, { "epoch": 1.47, "grad_norm": 13.375, "learning_rate": 8.573296803244979e-07, "loss": 0.2395, "step": 7785 }, { "epoch": 1.47, "grad_norm": 10.5, "learning_rate": 8.567523655451451e-07, "loss": 0.2527, "step": 7786 }, { "epoch": 1.47, "grad_norm": 10.625, "learning_rate": 8.56175205012581e-07, "loss": 0.2071, "step": 7787 }, { "epoch": 1.47, "grad_norm": 11.375, "learning_rate": 8.555981987809822e-07, "loss": 0.2828, "step": 7788 }, { "epoch": 1.47, "grad_norm": 12.0, "learning_rate": 8.550213469045107e-07, "loss": 0.1801, "step": 7789 }, { "epoch": 1.47, "grad_norm": 7.90625, "learning_rate": 8.544446494373126e-07, "loss": 0.16, "step": 7790 }, { "epoch": 1.47, "grad_norm": 10.5625, "learning_rate": 8.538681064335236e-07, "loss": 0.1797, "step": 7791 }, { "epoch": 1.47, "grad_norm": 9.3125, "learning_rate": 8.532917179472591e-07, "loss": 0.1582, "step": 7792 }, { "epoch": 1.47, "grad_norm": 10.5, "learning_rate": 8.527154840326246e-07, "loss": 0.1962, "step": 7793 }, { "epoch": 1.47, "grad_norm": 13.0, "learning_rate": 8.521394047437081e-07, "loss": 0.2075, "step": 7794 }, { "epoch": 1.47, "grad_norm": 12.6875, "learning_rate": 8.515634801345871e-07, "loss": 0.228, "step": 7795 }, { "epoch": 1.47, "grad_norm": 12.375, "learning_rate": 8.509877102593194e-07, "loss": 0.2485, "step": 7796 }, { "epoch": 1.47, "grad_norm": 10.375, "learning_rate": 8.504120951719516e-07, "loss": 0.2328, "step": 7797 }, { "epoch": 1.48, "grad_norm": 9.75, "learning_rate": 8.498366349265152e-07, "loss": 0.2064, "step": 7798 }, { "epoch": 1.48, "grad_norm": 11.8125, "learning_rate": 8.492613295770269e-07, "loss": 0.2355, "step": 7799 }, { "epoch": 1.48, "grad_norm": 12.75, "learning_rate": 8.486861791774886e-07, "loss": 0.1833, "step": 7800 }, { "epoch": 1.48, "grad_norm": 11.875, "learning_rate": 8.481111837818873e-07, "loss": 0.2453, "step": 7801 }, { "epoch": 1.48, "grad_norm": 10.9375, "learning_rate": 8.475363434441988e-07, "loss": 0.2673, "step": 7802 }, { "epoch": 1.48, "grad_norm": 13.9375, "learning_rate": 8.46961658218379e-07, "loss": 0.2345, "step": 7803 }, { "epoch": 1.48, "grad_norm": 12.5, "learning_rate": 8.463871281583727e-07, "loss": 0.209, "step": 7804 }, { "epoch": 1.48, "grad_norm": 11.375, "learning_rate": 8.458127533181084e-07, "loss": 0.2411, "step": 7805 }, { "epoch": 1.48, "grad_norm": 10.5, "learning_rate": 8.45238533751504e-07, "loss": 0.1709, "step": 7806 }, { "epoch": 1.48, "grad_norm": 9.1875, "learning_rate": 8.446644695124567e-07, "loss": 0.1929, "step": 7807 }, { "epoch": 1.48, "grad_norm": 13.1875, "learning_rate": 8.440905606548533e-07, "loss": 0.2623, "step": 7808 }, { "epoch": 1.48, "grad_norm": 10.1875, "learning_rate": 8.435168072325639e-07, "loss": 0.2277, "step": 7809 }, { "epoch": 1.48, "grad_norm": 8.625, "learning_rate": 8.429432092994477e-07, "loss": 0.1797, "step": 7810 }, { "epoch": 1.48, "grad_norm": 13.9375, "learning_rate": 8.423697669093442e-07, "loss": 0.1761, "step": 7811 }, { "epoch": 1.48, "grad_norm": 12.5625, "learning_rate": 8.417964801160816e-07, "loss": 0.2258, "step": 7812 }, { "epoch": 1.48, "grad_norm": 11.4375, "learning_rate": 8.412233489734717e-07, "loss": 0.2024, "step": 7813 }, { "epoch": 1.48, "grad_norm": 9.5625, "learning_rate": 8.40650373535315e-07, "loss": 0.1969, "step": 7814 }, { "epoch": 1.48, "grad_norm": 13.125, "learning_rate": 8.400775538553923e-07, "loss": 0.1857, "step": 7815 }, { "epoch": 1.48, "grad_norm": 11.3125, "learning_rate": 8.395048899874734e-07, "loss": 0.2109, "step": 7816 }, { "epoch": 1.48, "grad_norm": 17.5, "learning_rate": 8.389323819853132e-07, "loss": 0.2454, "step": 7817 }, { "epoch": 1.48, "grad_norm": 15.25, "learning_rate": 8.38360029902652e-07, "loss": 0.2007, "step": 7818 }, { "epoch": 1.48, "grad_norm": 13.6875, "learning_rate": 8.377878337932128e-07, "loss": 0.1609, "step": 7819 }, { "epoch": 1.48, "grad_norm": 19.375, "learning_rate": 8.372157937107059e-07, "loss": 0.2139, "step": 7820 }, { "epoch": 1.48, "grad_norm": 13.0, "learning_rate": 8.366439097088291e-07, "loss": 0.2045, "step": 7821 }, { "epoch": 1.48, "grad_norm": 15.25, "learning_rate": 8.360721818412629e-07, "loss": 0.2334, "step": 7822 }, { "epoch": 1.48, "grad_norm": 14.3125, "learning_rate": 8.355006101616725e-07, "loss": 0.2308, "step": 7823 }, { "epoch": 1.48, "grad_norm": 13.0625, "learning_rate": 8.349291947237096e-07, "loss": 0.2346, "step": 7824 }, { "epoch": 1.48, "grad_norm": 11.9375, "learning_rate": 8.343579355810125e-07, "loss": 0.2496, "step": 7825 }, { "epoch": 1.48, "grad_norm": 9.375, "learning_rate": 8.33786832787204e-07, "loss": 0.2105, "step": 7826 }, { "epoch": 1.48, "grad_norm": 27.625, "learning_rate": 8.332158863958892e-07, "loss": 0.21, "step": 7827 }, { "epoch": 1.48, "grad_norm": 9.8125, "learning_rate": 8.326450964606639e-07, "loss": 0.2538, "step": 7828 }, { "epoch": 1.48, "grad_norm": 9.0, "learning_rate": 8.320744630351061e-07, "loss": 0.1674, "step": 7829 }, { "epoch": 1.48, "grad_norm": 9.375, "learning_rate": 8.315039861727777e-07, "loss": 0.1727, "step": 7830 }, { "epoch": 1.48, "grad_norm": 13.3125, "learning_rate": 8.309336659272282e-07, "loss": 0.215, "step": 7831 }, { "epoch": 1.48, "grad_norm": 13.5, "learning_rate": 8.303635023519932e-07, "loss": 0.1916, "step": 7832 }, { "epoch": 1.48, "grad_norm": 14.1875, "learning_rate": 8.297934955005923e-07, "loss": 0.2956, "step": 7833 }, { "epoch": 1.48, "grad_norm": 12.625, "learning_rate": 8.292236454265287e-07, "loss": 0.2011, "step": 7834 }, { "epoch": 1.48, "grad_norm": 12.125, "learning_rate": 8.286539521832926e-07, "loss": 0.2003, "step": 7835 }, { "epoch": 1.48, "grad_norm": 23.0, "learning_rate": 8.280844158243609e-07, "loss": 0.2221, "step": 7836 }, { "epoch": 1.48, "grad_norm": 13.0625, "learning_rate": 8.275150364031945e-07, "loss": 0.1888, "step": 7837 }, { "epoch": 1.48, "grad_norm": 12.5625, "learning_rate": 8.26945813973237e-07, "loss": 0.2215, "step": 7838 }, { "epoch": 1.48, "grad_norm": 9.5, "learning_rate": 8.263767485879215e-07, "loss": 0.1707, "step": 7839 }, { "epoch": 1.48, "grad_norm": 12.25, "learning_rate": 8.258078403006641e-07, "loss": 0.1794, "step": 7840 }, { "epoch": 1.48, "grad_norm": 9.375, "learning_rate": 8.252390891648676e-07, "loss": 0.2226, "step": 7841 }, { "epoch": 1.48, "grad_norm": 10.75, "learning_rate": 8.24670495233916e-07, "loss": 0.203, "step": 7842 }, { "epoch": 1.48, "grad_norm": 14.25, "learning_rate": 8.241020585611844e-07, "loss": 0.2635, "step": 7843 }, { "epoch": 1.48, "grad_norm": 10.9375, "learning_rate": 8.235337792000292e-07, "loss": 0.1948, "step": 7844 }, { "epoch": 1.48, "grad_norm": 12.125, "learning_rate": 8.229656572037939e-07, "loss": 0.2162, "step": 7845 }, { "epoch": 1.48, "grad_norm": 16.5, "learning_rate": 8.223976926258043e-07, "loss": 0.1687, "step": 7846 }, { "epoch": 1.48, "grad_norm": 11.375, "learning_rate": 8.218298855193754e-07, "loss": 0.2281, "step": 7847 }, { "epoch": 1.48, "grad_norm": 10.8125, "learning_rate": 8.212622359378055e-07, "loss": 0.1882, "step": 7848 }, { "epoch": 1.48, "grad_norm": 11.3125, "learning_rate": 8.206947439343776e-07, "loss": 0.2682, "step": 7849 }, { "epoch": 1.49, "grad_norm": 11.1875, "learning_rate": 8.201274095623607e-07, "loss": 0.2116, "step": 7850 }, { "epoch": 1.49, "grad_norm": 15.9375, "learning_rate": 8.195602328750088e-07, "loss": 0.2693, "step": 7851 }, { "epoch": 1.49, "grad_norm": 8.625, "learning_rate": 8.189932139255613e-07, "loss": 0.2037, "step": 7852 }, { "epoch": 1.49, "grad_norm": 12.9375, "learning_rate": 8.184263527672423e-07, "loss": 0.2166, "step": 7853 }, { "epoch": 1.49, "grad_norm": 13.5625, "learning_rate": 8.178596494532612e-07, "loss": 0.1912, "step": 7854 }, { "epoch": 1.49, "grad_norm": 23.875, "learning_rate": 8.172931040368134e-07, "loss": 0.2283, "step": 7855 }, { "epoch": 1.49, "grad_norm": 13.0625, "learning_rate": 8.16726716571079e-07, "loss": 0.2608, "step": 7856 }, { "epoch": 1.49, "grad_norm": 12.375, "learning_rate": 8.16160487109221e-07, "loss": 0.198, "step": 7857 }, { "epoch": 1.49, "grad_norm": 15.4375, "learning_rate": 8.155944157043921e-07, "loss": 0.2417, "step": 7858 }, { "epoch": 1.49, "grad_norm": 13.5, "learning_rate": 8.150285024097271e-07, "loss": 0.2682, "step": 7859 }, { "epoch": 1.49, "grad_norm": 14.8125, "learning_rate": 8.144627472783462e-07, "loss": 0.18, "step": 7860 }, { "epoch": 1.49, "grad_norm": 7.96875, "learning_rate": 8.138971503633555e-07, "loss": 0.1597, "step": 7861 }, { "epoch": 1.49, "grad_norm": 8.75, "learning_rate": 8.133317117178458e-07, "loss": 0.197, "step": 7862 }, { "epoch": 1.49, "grad_norm": 11.375, "learning_rate": 8.12766431394893e-07, "loss": 0.1988, "step": 7863 }, { "epoch": 1.49, "grad_norm": 15.25, "learning_rate": 8.122013094475584e-07, "loss": 0.255, "step": 7864 }, { "epoch": 1.49, "grad_norm": 9.4375, "learning_rate": 8.116363459288884e-07, "loss": 0.134, "step": 7865 }, { "epoch": 1.49, "grad_norm": 11.625, "learning_rate": 8.110715408919145e-07, "loss": 0.1836, "step": 7866 }, { "epoch": 1.49, "grad_norm": 11.0625, "learning_rate": 8.10506894389653e-07, "loss": 0.1631, "step": 7867 }, { "epoch": 1.49, "grad_norm": 19.125, "learning_rate": 8.099424064751057e-07, "loss": 0.266, "step": 7868 }, { "epoch": 1.49, "grad_norm": 8.9375, "learning_rate": 8.093780772012599e-07, "loss": 0.1967, "step": 7869 }, { "epoch": 1.49, "grad_norm": 13.0, "learning_rate": 8.088139066210867e-07, "loss": 0.2523, "step": 7870 }, { "epoch": 1.49, "grad_norm": 12.3125, "learning_rate": 8.082498947875436e-07, "loss": 0.2153, "step": 7871 }, { "epoch": 1.49, "grad_norm": 11.6875, "learning_rate": 8.076860417535728e-07, "loss": 0.2355, "step": 7872 }, { "epoch": 1.49, "grad_norm": 13.0625, "learning_rate": 8.071223475721013e-07, "loss": 0.1461, "step": 7873 }, { "epoch": 1.49, "grad_norm": 9.6875, "learning_rate": 8.065588122960413e-07, "loss": 0.1608, "step": 7874 }, { "epoch": 1.49, "grad_norm": 8.9375, "learning_rate": 8.059954359782904e-07, "loss": 0.1838, "step": 7875 }, { "epoch": 1.49, "grad_norm": 8.9375, "learning_rate": 8.054322186717311e-07, "loss": 0.2105, "step": 7876 }, { "epoch": 1.49, "grad_norm": 19.5, "learning_rate": 8.048691604292308e-07, "loss": 0.2156, "step": 7877 }, { "epoch": 1.49, "grad_norm": 13.4375, "learning_rate": 8.043062613036421e-07, "loss": 0.1631, "step": 7878 }, { "epoch": 1.49, "grad_norm": 19.375, "learning_rate": 8.037435213478029e-07, "loss": 0.1587, "step": 7879 }, { "epoch": 1.49, "grad_norm": 9.625, "learning_rate": 8.031809406145355e-07, "loss": 0.173, "step": 7880 }, { "epoch": 1.49, "grad_norm": 10.1875, "learning_rate": 8.026185191566479e-07, "loss": 0.2757, "step": 7881 }, { "epoch": 1.49, "grad_norm": 12.5, "learning_rate": 8.020562570269333e-07, "loss": 0.1552, "step": 7882 }, { "epoch": 1.49, "grad_norm": 9.875, "learning_rate": 8.014941542781693e-07, "loss": 0.1892, "step": 7883 }, { "epoch": 1.49, "grad_norm": 16.375, "learning_rate": 8.009322109631185e-07, "loss": 0.2485, "step": 7884 }, { "epoch": 1.49, "grad_norm": 13.5625, "learning_rate": 8.003704271345295e-07, "loss": 0.2305, "step": 7885 }, { "epoch": 1.49, "grad_norm": 10.75, "learning_rate": 7.998088028451345e-07, "loss": 0.2247, "step": 7886 }, { "epoch": 1.49, "grad_norm": 9.3125, "learning_rate": 7.992473381476523e-07, "loss": 0.2045, "step": 7887 }, { "epoch": 1.49, "grad_norm": 11.375, "learning_rate": 7.986860330947854e-07, "loss": 0.2645, "step": 7888 }, { "epoch": 1.49, "grad_norm": 14.5625, "learning_rate": 7.98124887739222e-07, "loss": 0.2494, "step": 7889 }, { "epoch": 1.49, "grad_norm": 12.1875, "learning_rate": 7.975639021336351e-07, "loss": 0.175, "step": 7890 }, { "epoch": 1.49, "grad_norm": 11.1875, "learning_rate": 7.970030763306829e-07, "loss": 0.1722, "step": 7891 }, { "epoch": 1.49, "grad_norm": 10.0, "learning_rate": 7.964424103830085e-07, "loss": 0.1427, "step": 7892 }, { "epoch": 1.49, "grad_norm": 14.625, "learning_rate": 7.958819043432389e-07, "loss": 0.1623, "step": 7893 }, { "epoch": 1.49, "grad_norm": 13.5625, "learning_rate": 7.953215582639898e-07, "loss": 0.1989, "step": 7894 }, { "epoch": 1.49, "grad_norm": 16.5, "learning_rate": 7.947613721978565e-07, "loss": 0.1904, "step": 7895 }, { "epoch": 1.49, "grad_norm": 9.8125, "learning_rate": 7.942013461974232e-07, "loss": 0.1743, "step": 7896 }, { "epoch": 1.49, "grad_norm": 11.6875, "learning_rate": 7.936414803152567e-07, "loss": 0.1924, "step": 7897 }, { "epoch": 1.49, "grad_norm": 8.4375, "learning_rate": 7.930817746039127e-07, "loss": 0.1869, "step": 7898 }, { "epoch": 1.49, "grad_norm": 11.0, "learning_rate": 7.925222291159265e-07, "loss": 0.1995, "step": 7899 }, { "epoch": 1.49, "grad_norm": 10.9375, "learning_rate": 7.919628439038218e-07, "loss": 0.1866, "step": 7900 }, { "epoch": 1.49, "grad_norm": 16.0, "learning_rate": 7.914036190201058e-07, "loss": 0.2063, "step": 7901 }, { "epoch": 1.49, "grad_norm": 10.5, "learning_rate": 7.908445545172732e-07, "loss": 0.2025, "step": 7902 }, { "epoch": 1.5, "grad_norm": 10.75, "learning_rate": 7.902856504477999e-07, "loss": 0.2379, "step": 7903 }, { "epoch": 1.5, "grad_norm": 9.9375, "learning_rate": 7.897269068641486e-07, "loss": 0.1653, "step": 7904 }, { "epoch": 1.5, "grad_norm": 8.1875, "learning_rate": 7.891683238187677e-07, "loss": 0.2163, "step": 7905 }, { "epoch": 1.5, "grad_norm": 11.6875, "learning_rate": 7.886099013640907e-07, "loss": 0.1872, "step": 7906 }, { "epoch": 1.5, "grad_norm": 9.875, "learning_rate": 7.880516395525331e-07, "loss": 0.255, "step": 7907 }, { "epoch": 1.5, "grad_norm": 10.75, "learning_rate": 7.874935384364971e-07, "loss": 0.1942, "step": 7908 }, { "epoch": 1.5, "grad_norm": 13.25, "learning_rate": 7.869355980683719e-07, "loss": 0.2116, "step": 7909 }, { "epoch": 1.5, "grad_norm": 12.5625, "learning_rate": 7.863778185005291e-07, "loss": 0.2237, "step": 7910 }, { "epoch": 1.5, "grad_norm": 13.0, "learning_rate": 7.858201997853252e-07, "loss": 0.2425, "step": 7911 }, { "epoch": 1.5, "grad_norm": 22.5, "learning_rate": 7.852627419751013e-07, "loss": 0.2411, "step": 7912 }, { "epoch": 1.5, "grad_norm": 7.40625, "learning_rate": 7.847054451221872e-07, "loss": 0.1522, "step": 7913 }, { "epoch": 1.5, "grad_norm": 10.8125, "learning_rate": 7.841483092788923e-07, "loss": 0.2293, "step": 7914 }, { "epoch": 1.5, "grad_norm": 11.9375, "learning_rate": 7.835913344975133e-07, "loss": 0.2004, "step": 7915 }, { "epoch": 1.5, "grad_norm": 15.875, "learning_rate": 7.830345208303328e-07, "loss": 0.2855, "step": 7916 }, { "epoch": 1.5, "grad_norm": 10.375, "learning_rate": 7.82477868329618e-07, "loss": 0.2417, "step": 7917 }, { "epoch": 1.5, "grad_norm": 15.8125, "learning_rate": 7.819213770476183e-07, "loss": 0.2112, "step": 7918 }, { "epoch": 1.5, "grad_norm": 14.1875, "learning_rate": 7.8136504703657e-07, "loss": 0.2538, "step": 7919 }, { "epoch": 1.5, "grad_norm": 12.4375, "learning_rate": 7.808088783486955e-07, "loss": 0.1904, "step": 7920 }, { "epoch": 1.5, "grad_norm": 10.4375, "learning_rate": 7.802528710362011e-07, "loss": 0.2258, "step": 7921 }, { "epoch": 1.5, "grad_norm": 10.0, "learning_rate": 7.796970251512756e-07, "loss": 0.1762, "step": 7922 }, { "epoch": 1.5, "grad_norm": 9.1875, "learning_rate": 7.791413407460949e-07, "loss": 0.1893, "step": 7923 }, { "epoch": 1.5, "grad_norm": 11.1875, "learning_rate": 7.785858178728209e-07, "loss": 0.2262, "step": 7924 }, { "epoch": 1.5, "grad_norm": 10.6875, "learning_rate": 7.780304565835986e-07, "loss": 0.1915, "step": 7925 }, { "epoch": 1.5, "grad_norm": 11.6875, "learning_rate": 7.774752569305563e-07, "loss": 0.1981, "step": 7926 }, { "epoch": 1.5, "grad_norm": 11.375, "learning_rate": 7.769202189658109e-07, "loss": 0.2101, "step": 7927 }, { "epoch": 1.5, "grad_norm": 9.5, "learning_rate": 7.763653427414617e-07, "loss": 0.1991, "step": 7928 }, { "epoch": 1.5, "grad_norm": 15.1875, "learning_rate": 7.758106283095937e-07, "loss": 0.2275, "step": 7929 }, { "epoch": 1.5, "grad_norm": 14.0, "learning_rate": 7.75256075722274e-07, "loss": 0.2628, "step": 7930 }, { "epoch": 1.5, "grad_norm": 12.8125, "learning_rate": 7.747016850315594e-07, "loss": 0.2416, "step": 7931 }, { "epoch": 1.5, "grad_norm": 8.1875, "learning_rate": 7.741474562894882e-07, "loss": 0.1926, "step": 7932 }, { "epoch": 1.5, "grad_norm": 13.8125, "learning_rate": 7.735933895480849e-07, "loss": 0.2121, "step": 7933 }, { "epoch": 1.5, "grad_norm": 10.125, "learning_rate": 7.730394848593556e-07, "loss": 0.2133, "step": 7934 }, { "epoch": 1.5, "grad_norm": 9.3125, "learning_rate": 7.724857422752959e-07, "loss": 0.2099, "step": 7935 }, { "epoch": 1.5, "grad_norm": 11.375, "learning_rate": 7.719321618478836e-07, "loss": 0.2093, "step": 7936 }, { "epoch": 1.5, "grad_norm": 10.8125, "learning_rate": 7.713787436290815e-07, "loss": 0.16, "step": 7937 }, { "epoch": 1.5, "grad_norm": 10.25, "learning_rate": 7.708254876708371e-07, "loss": 0.1562, "step": 7938 }, { "epoch": 1.5, "grad_norm": 12.125, "learning_rate": 7.702723940250834e-07, "loss": 0.1608, "step": 7939 }, { "epoch": 1.5, "grad_norm": 10.25, "learning_rate": 7.697194627437379e-07, "loss": 0.1797, "step": 7940 }, { "epoch": 1.5, "grad_norm": 23.625, "learning_rate": 7.691666938787007e-07, "loss": 0.1882, "step": 7941 }, { "epoch": 1.5, "grad_norm": 10.75, "learning_rate": 7.686140874818604e-07, "loss": 0.176, "step": 7942 }, { "epoch": 1.5, "grad_norm": 12.9375, "learning_rate": 7.680616436050883e-07, "loss": 0.2325, "step": 7943 }, { "epoch": 1.5, "grad_norm": 10.9375, "learning_rate": 7.675093623002413e-07, "loss": 0.1744, "step": 7944 }, { "epoch": 1.5, "grad_norm": 12.5625, "learning_rate": 7.66957243619158e-07, "loss": 0.2397, "step": 7945 }, { "epoch": 1.5, "grad_norm": 10.4375, "learning_rate": 7.664052876136666e-07, "loss": 0.2173, "step": 7946 }, { "epoch": 1.5, "grad_norm": 10.125, "learning_rate": 7.658534943355764e-07, "loss": 0.1997, "step": 7947 }, { "epoch": 1.5, "grad_norm": 9.5625, "learning_rate": 7.653018638366829e-07, "loss": 0.1981, "step": 7948 }, { "epoch": 1.5, "grad_norm": 10.625, "learning_rate": 7.647503961687664e-07, "loss": 0.2003, "step": 7949 }, { "epoch": 1.5, "grad_norm": 15.9375, "learning_rate": 7.641990913835908e-07, "loss": 0.1774, "step": 7950 }, { "epoch": 1.5, "grad_norm": 9.8125, "learning_rate": 7.636479495329058e-07, "loss": 0.2009, "step": 7951 }, { "epoch": 1.5, "grad_norm": 11.1875, "learning_rate": 7.630969706684454e-07, "loss": 0.1838, "step": 7952 }, { "epoch": 1.5, "grad_norm": 11.375, "learning_rate": 7.625461548419285e-07, "loss": 0.2346, "step": 7953 }, { "epoch": 1.5, "grad_norm": 11.9375, "learning_rate": 7.619955021050587e-07, "loss": 0.2111, "step": 7954 }, { "epoch": 1.5, "grad_norm": 12.3125, "learning_rate": 7.614450125095235e-07, "loss": 0.1404, "step": 7955 }, { "epoch": 1.51, "grad_norm": 10.8125, "learning_rate": 7.608946861069963e-07, "loss": 0.2128, "step": 7956 }, { "epoch": 1.51, "grad_norm": 9.25, "learning_rate": 7.603445229491346e-07, "loss": 0.1864, "step": 7957 }, { "epoch": 1.51, "grad_norm": 9.1875, "learning_rate": 7.5979452308758e-07, "loss": 0.1735, "step": 7958 }, { "epoch": 1.51, "grad_norm": 11.8125, "learning_rate": 7.5924468657396e-07, "loss": 0.2343, "step": 7959 }, { "epoch": 1.51, "grad_norm": 15.0, "learning_rate": 7.586950134598861e-07, "loss": 0.2355, "step": 7960 }, { "epoch": 1.51, "grad_norm": 15.375, "learning_rate": 7.58145503796954e-07, "loss": 0.2248, "step": 7961 }, { "epoch": 1.51, "grad_norm": 11.25, "learning_rate": 7.575961576367449e-07, "loss": 0.182, "step": 7962 }, { "epoch": 1.51, "grad_norm": 11.8125, "learning_rate": 7.570469750308243e-07, "loss": 0.194, "step": 7963 }, { "epoch": 1.51, "grad_norm": 10.375, "learning_rate": 7.564979560307423e-07, "loss": 0.177, "step": 7964 }, { "epoch": 1.51, "grad_norm": 9.1875, "learning_rate": 7.559491006880335e-07, "loss": 0.2066, "step": 7965 }, { "epoch": 1.51, "grad_norm": 10.5, "learning_rate": 7.554004090542174e-07, "loss": 0.2013, "step": 7966 }, { "epoch": 1.51, "grad_norm": 10.9375, "learning_rate": 7.548518811807983e-07, "loss": 0.1693, "step": 7967 }, { "epoch": 1.51, "grad_norm": 11.5625, "learning_rate": 7.543035171192651e-07, "loss": 0.2068, "step": 7968 }, { "epoch": 1.51, "grad_norm": 9.75, "learning_rate": 7.537553169210901e-07, "loss": 0.142, "step": 7969 }, { "epoch": 1.51, "grad_norm": 14.875, "learning_rate": 7.532072806377322e-07, "loss": 0.2674, "step": 7970 }, { "epoch": 1.51, "grad_norm": 12.6875, "learning_rate": 7.526594083206338e-07, "loss": 0.2227, "step": 7971 }, { "epoch": 1.51, "grad_norm": 10.4375, "learning_rate": 7.521117000212214e-07, "loss": 0.1891, "step": 7972 }, { "epoch": 1.51, "grad_norm": 12.0, "learning_rate": 7.515641557909076e-07, "loss": 0.1951, "step": 7973 }, { "epoch": 1.51, "grad_norm": 12.6875, "learning_rate": 7.510167756810882e-07, "loss": 0.2597, "step": 7974 }, { "epoch": 1.51, "grad_norm": 11.6875, "learning_rate": 7.504695597431444e-07, "loss": 0.218, "step": 7975 }, { "epoch": 1.51, "grad_norm": 15.0, "learning_rate": 7.499225080284417e-07, "loss": 0.2332, "step": 7976 }, { "epoch": 1.51, "grad_norm": 14.125, "learning_rate": 7.493756205883299e-07, "loss": 0.2204, "step": 7977 }, { "epoch": 1.51, "grad_norm": 9.875, "learning_rate": 7.488288974741442e-07, "loss": 0.2394, "step": 7978 }, { "epoch": 1.51, "grad_norm": 12.75, "learning_rate": 7.482823387372038e-07, "loss": 0.2051, "step": 7979 }, { "epoch": 1.51, "grad_norm": 8.9375, "learning_rate": 7.477359444288123e-07, "loss": 0.1896, "step": 7980 }, { "epoch": 1.51, "grad_norm": 8.625, "learning_rate": 7.471897146002574e-07, "loss": 0.1579, "step": 7981 }, { "epoch": 1.51, "grad_norm": 11.625, "learning_rate": 7.466436493028148e-07, "loss": 0.165, "step": 7982 }, { "epoch": 1.51, "grad_norm": 9.5625, "learning_rate": 7.460977485877391e-07, "loss": 0.2582, "step": 7983 }, { "epoch": 1.51, "grad_norm": 10.625, "learning_rate": 7.455520125062735e-07, "loss": 0.2366, "step": 7984 }, { "epoch": 1.51, "grad_norm": 11.875, "learning_rate": 7.450064411096436e-07, "loss": 0.2342, "step": 7985 }, { "epoch": 1.51, "grad_norm": 13.8125, "learning_rate": 7.444610344490632e-07, "loss": 0.1675, "step": 7986 }, { "epoch": 1.51, "grad_norm": 14.6875, "learning_rate": 7.439157925757256e-07, "loss": 0.2145, "step": 7987 }, { "epoch": 1.51, "grad_norm": 8.25, "learning_rate": 7.433707155408118e-07, "loss": 0.182, "step": 7988 }, { "epoch": 1.51, "grad_norm": 11.75, "learning_rate": 7.428258033954857e-07, "loss": 0.1601, "step": 7989 }, { "epoch": 1.51, "grad_norm": 10.9375, "learning_rate": 7.422810561908992e-07, "loss": 0.2092, "step": 7990 }, { "epoch": 1.51, "grad_norm": 10.9375, "learning_rate": 7.417364739781832e-07, "loss": 0.1848, "step": 7991 }, { "epoch": 1.51, "grad_norm": 15.25, "learning_rate": 7.411920568084566e-07, "loss": 0.2099, "step": 7992 }, { "epoch": 1.51, "grad_norm": 9.5, "learning_rate": 7.406478047328236e-07, "loss": 0.2284, "step": 7993 }, { "epoch": 1.51, "grad_norm": 9.375, "learning_rate": 7.401037178023715e-07, "loss": 0.1912, "step": 7994 }, { "epoch": 1.51, "grad_norm": 10.125, "learning_rate": 7.395597960681705e-07, "loss": 0.1584, "step": 7995 }, { "epoch": 1.51, "grad_norm": 12.0, "learning_rate": 7.390160395812771e-07, "loss": 0.2353, "step": 7996 }, { "epoch": 1.51, "grad_norm": 10.375, "learning_rate": 7.384724483927345e-07, "loss": 0.2119, "step": 7997 }, { "epoch": 1.51, "grad_norm": 10.875, "learning_rate": 7.379290225535651e-07, "loss": 0.2109, "step": 7998 }, { "epoch": 1.51, "grad_norm": 8.5625, "learning_rate": 7.373857621147801e-07, "loss": 0.1746, "step": 7999 }, { "epoch": 1.51, "grad_norm": 13.5, "learning_rate": 7.368426671273726e-07, "loss": 0.2312, "step": 8000 }, { "epoch": 1.51, "grad_norm": 9.6875, "learning_rate": 7.362997376423237e-07, "loss": 0.1942, "step": 8001 }, { "epoch": 1.51, "grad_norm": 10.0, "learning_rate": 7.357569737105947e-07, "loss": 0.1568, "step": 8002 }, { "epoch": 1.51, "grad_norm": 13.25, "learning_rate": 7.352143753831326e-07, "loss": 0.2423, "step": 8003 }, { "epoch": 1.51, "grad_norm": 10.5625, "learning_rate": 7.346719427108711e-07, "loss": 0.218, "step": 8004 }, { "epoch": 1.51, "grad_norm": 17.5, "learning_rate": 7.341296757447272e-07, "loss": 0.238, "step": 8005 }, { "epoch": 1.51, "grad_norm": 9.875, "learning_rate": 7.335875745356003e-07, "loss": 0.1824, "step": 8006 }, { "epoch": 1.51, "grad_norm": 17.375, "learning_rate": 7.330456391343754e-07, "loss": 0.1982, "step": 8007 }, { "epoch": 1.51, "grad_norm": 12.1875, "learning_rate": 7.325038695919242e-07, "loss": 0.156, "step": 8008 }, { "epoch": 1.52, "grad_norm": 9.8125, "learning_rate": 7.319622659591011e-07, "loss": 0.2038, "step": 8009 }, { "epoch": 1.52, "grad_norm": 9.8125, "learning_rate": 7.314208282867429e-07, "loss": 0.1808, "step": 8010 }, { "epoch": 1.52, "grad_norm": 15.75, "learning_rate": 7.308795566256732e-07, "loss": 0.2272, "step": 8011 }, { "epoch": 1.52, "grad_norm": 14.5, "learning_rate": 7.303384510267009e-07, "loss": 0.2434, "step": 8012 }, { "epoch": 1.52, "grad_norm": 8.25, "learning_rate": 7.297975115406181e-07, "loss": 0.1385, "step": 8013 }, { "epoch": 1.52, "grad_norm": 9.5, "learning_rate": 7.292567382181992e-07, "loss": 0.1884, "step": 8014 }, { "epoch": 1.52, "grad_norm": 11.0625, "learning_rate": 7.287161311102056e-07, "loss": 0.1512, "step": 8015 }, { "epoch": 1.52, "grad_norm": 11.875, "learning_rate": 7.281756902673836e-07, "loss": 0.2119, "step": 8016 }, { "epoch": 1.52, "grad_norm": 13.0, "learning_rate": 7.27635415740463e-07, "loss": 0.2056, "step": 8017 }, { "epoch": 1.52, "grad_norm": 9.625, "learning_rate": 7.270953075801554e-07, "loss": 0.205, "step": 8018 }, { "epoch": 1.52, "grad_norm": 10.6875, "learning_rate": 7.265553658371616e-07, "loss": 0.2096, "step": 8019 }, { "epoch": 1.52, "grad_norm": 11.375, "learning_rate": 7.260155905621633e-07, "loss": 0.2586, "step": 8020 }, { "epoch": 1.52, "grad_norm": 15.0625, "learning_rate": 7.254759818058285e-07, "loss": 0.1724, "step": 8021 }, { "epoch": 1.52, "grad_norm": 8.875, "learning_rate": 7.249365396188063e-07, "loss": 0.178, "step": 8022 }, { "epoch": 1.52, "grad_norm": 10.5625, "learning_rate": 7.243972640517352e-07, "loss": 0.1617, "step": 8023 }, { "epoch": 1.52, "grad_norm": 10.75, "learning_rate": 7.238581551552349e-07, "loss": 0.1914, "step": 8024 }, { "epoch": 1.52, "grad_norm": 10.625, "learning_rate": 7.233192129799085e-07, "loss": 0.2593, "step": 8025 }, { "epoch": 1.52, "grad_norm": 10.25, "learning_rate": 7.227804375763453e-07, "loss": 0.188, "step": 8026 }, { "epoch": 1.52, "grad_norm": 9.8125, "learning_rate": 7.222418289951197e-07, "loss": 0.1407, "step": 8027 }, { "epoch": 1.52, "grad_norm": 15.875, "learning_rate": 7.217033872867896e-07, "loss": 0.2108, "step": 8028 }, { "epoch": 1.52, "grad_norm": 11.25, "learning_rate": 7.211651125018943e-07, "loss": 0.1814, "step": 8029 }, { "epoch": 1.52, "grad_norm": 17.25, "learning_rate": 7.206270046909625e-07, "loss": 0.2354, "step": 8030 }, { "epoch": 1.52, "grad_norm": 16.125, "learning_rate": 7.200890639045038e-07, "loss": 0.209, "step": 8031 }, { "epoch": 1.52, "grad_norm": 15.125, "learning_rate": 7.195512901930143e-07, "loss": 0.1777, "step": 8032 }, { "epoch": 1.52, "grad_norm": 9.125, "learning_rate": 7.190136836069708e-07, "loss": 0.1449, "step": 8033 }, { "epoch": 1.52, "grad_norm": 14.25, "learning_rate": 7.18476244196839e-07, "loss": 0.2377, "step": 8034 }, { "epoch": 1.52, "grad_norm": 13.25, "learning_rate": 7.179389720130661e-07, "loss": 0.2308, "step": 8035 }, { "epoch": 1.52, "grad_norm": 11.375, "learning_rate": 7.174018671060851e-07, "loss": 0.1926, "step": 8036 }, { "epoch": 1.52, "grad_norm": 14.6875, "learning_rate": 7.168649295263097e-07, "loss": 0.2497, "step": 8037 }, { "epoch": 1.52, "grad_norm": 11.625, "learning_rate": 7.163281593241436e-07, "loss": 0.2585, "step": 8038 }, { "epoch": 1.52, "grad_norm": 12.1875, "learning_rate": 7.157915565499704e-07, "loss": 0.1978, "step": 8039 }, { "epoch": 1.52, "grad_norm": 11.375, "learning_rate": 7.152551212541598e-07, "loss": 0.2117, "step": 8040 }, { "epoch": 1.52, "grad_norm": 13.0, "learning_rate": 7.147188534870652e-07, "loss": 0.2025, "step": 8041 }, { "epoch": 1.52, "grad_norm": 10.625, "learning_rate": 7.141827532990251e-07, "loss": 0.2101, "step": 8042 }, { "epoch": 1.52, "grad_norm": 10.1875, "learning_rate": 7.136468207403607e-07, "loss": 0.1864, "step": 8043 }, { "epoch": 1.52, "grad_norm": 8.375, "learning_rate": 7.131110558613788e-07, "loss": 0.1449, "step": 8044 }, { "epoch": 1.52, "grad_norm": 9.4375, "learning_rate": 7.125754587123701e-07, "loss": 0.2122, "step": 8045 }, { "epoch": 1.52, "grad_norm": 12.125, "learning_rate": 7.120400293436097e-07, "loss": 0.2076, "step": 8046 }, { "epoch": 1.52, "grad_norm": 10.9375, "learning_rate": 7.115047678053563e-07, "loss": 0.1753, "step": 8047 }, { "epoch": 1.52, "grad_norm": 12.25, "learning_rate": 7.109696741478539e-07, "loss": 0.231, "step": 8048 }, { "epoch": 1.52, "grad_norm": 22.0, "learning_rate": 7.104347484213297e-07, "loss": 0.2227, "step": 8049 }, { "epoch": 1.52, "grad_norm": 31.875, "learning_rate": 7.098999906759956e-07, "loss": 0.2004, "step": 8050 }, { "epoch": 1.52, "grad_norm": 13.9375, "learning_rate": 7.093654009620479e-07, "loss": 0.1901, "step": 8051 }, { "epoch": 1.52, "grad_norm": 18.25, "learning_rate": 7.088309793296672e-07, "loss": 0.2303, "step": 8052 }, { "epoch": 1.52, "grad_norm": 12.8125, "learning_rate": 7.082967258290174e-07, "loss": 0.2144, "step": 8053 }, { "epoch": 1.52, "grad_norm": 9.6875, "learning_rate": 7.07762640510248e-07, "loss": 0.1503, "step": 8054 }, { "epoch": 1.52, "grad_norm": 14.875, "learning_rate": 7.072287234234918e-07, "loss": 0.1779, "step": 8055 }, { "epoch": 1.52, "grad_norm": 12.0625, "learning_rate": 7.066949746188656e-07, "loss": 0.1961, "step": 8056 }, { "epoch": 1.52, "grad_norm": 8.5, "learning_rate": 7.061613941464712e-07, "loss": 0.1779, "step": 8057 }, { "epoch": 1.52, "grad_norm": 10.875, "learning_rate": 7.056279820563941e-07, "loss": 0.2156, "step": 8058 }, { "epoch": 1.52, "grad_norm": 11.6875, "learning_rate": 7.050947383987041e-07, "loss": 0.1998, "step": 8059 }, { "epoch": 1.52, "grad_norm": 11.125, "learning_rate": 7.045616632234556e-07, "loss": 0.1587, "step": 8060 }, { "epoch": 1.52, "grad_norm": 9.6875, "learning_rate": 7.040287565806864e-07, "loss": 0.1761, "step": 8061 }, { "epoch": 1.53, "grad_norm": 15.125, "learning_rate": 7.034960185204187e-07, "loss": 0.2695, "step": 8062 }, { "epoch": 1.53, "grad_norm": 12.75, "learning_rate": 7.029634490926593e-07, "loss": 0.2427, "step": 8063 }, { "epoch": 1.53, "grad_norm": 11.9375, "learning_rate": 7.024310483473989e-07, "loss": 0.2376, "step": 8064 }, { "epoch": 1.53, "grad_norm": 14.4375, "learning_rate": 7.018988163346125e-07, "loss": 0.2588, "step": 8065 }, { "epoch": 1.53, "grad_norm": 13.4375, "learning_rate": 7.013667531042589e-07, "loss": 0.2135, "step": 8066 }, { "epoch": 1.53, "grad_norm": 14.125, "learning_rate": 7.008348587062813e-07, "loss": 0.2347, "step": 8067 }, { "epoch": 1.53, "grad_norm": 12.3125, "learning_rate": 7.003031331906074e-07, "loss": 0.2314, "step": 8068 }, { "epoch": 1.53, "grad_norm": 12.25, "learning_rate": 6.997715766071483e-07, "loss": 0.1904, "step": 8069 }, { "epoch": 1.53, "grad_norm": 11.25, "learning_rate": 6.992401890057999e-07, "loss": 0.2089, "step": 8070 }, { "epoch": 1.53, "grad_norm": 17.5, "learning_rate": 6.987089704364417e-07, "loss": 0.3044, "step": 8071 }, { "epoch": 1.53, "grad_norm": 10.8125, "learning_rate": 6.981779209489378e-07, "loss": 0.2545, "step": 8072 }, { "epoch": 1.53, "grad_norm": 12.6875, "learning_rate": 6.976470405931357e-07, "loss": 0.1928, "step": 8073 }, { "epoch": 1.53, "grad_norm": 11.75, "learning_rate": 6.971163294188693e-07, "loss": 0.2614, "step": 8074 }, { "epoch": 1.53, "grad_norm": 8.0625, "learning_rate": 6.965857874759532e-07, "loss": 0.149, "step": 8075 }, { "epoch": 1.53, "grad_norm": 10.3125, "learning_rate": 6.960554148141882e-07, "loss": 0.2128, "step": 8076 }, { "epoch": 1.53, "grad_norm": 8.625, "learning_rate": 6.95525211483358e-07, "loss": 0.2149, "step": 8077 }, { "epoch": 1.53, "grad_norm": 9.875, "learning_rate": 6.949951775332339e-07, "loss": 0.2047, "step": 8078 }, { "epoch": 1.53, "grad_norm": 9.75, "learning_rate": 6.944653130135659e-07, "loss": 0.1647, "step": 8079 }, { "epoch": 1.53, "grad_norm": 8.9375, "learning_rate": 6.939356179740919e-07, "loss": 0.1774, "step": 8080 }, { "epoch": 1.53, "grad_norm": 11.6875, "learning_rate": 6.934060924645325e-07, "loss": 0.246, "step": 8081 }, { "epoch": 1.53, "grad_norm": 15.75, "learning_rate": 6.928767365345928e-07, "loss": 0.2447, "step": 8082 }, { "epoch": 1.53, "grad_norm": 13.3125, "learning_rate": 6.923475502339622e-07, "loss": 0.2628, "step": 8083 }, { "epoch": 1.53, "grad_norm": 11.9375, "learning_rate": 6.918185336123126e-07, "loss": 0.2019, "step": 8084 }, { "epoch": 1.53, "grad_norm": 11.3125, "learning_rate": 6.912896867193037e-07, "loss": 0.2396, "step": 8085 }, { "epoch": 1.53, "grad_norm": 11.625, "learning_rate": 6.907610096045745e-07, "loss": 0.1973, "step": 8086 }, { "epoch": 1.53, "grad_norm": 12.0625, "learning_rate": 6.902325023177511e-07, "loss": 0.1885, "step": 8087 }, { "epoch": 1.53, "grad_norm": 10.75, "learning_rate": 6.897041649084424e-07, "loss": 0.1787, "step": 8088 }, { "epoch": 1.53, "grad_norm": 10.5625, "learning_rate": 6.891759974262438e-07, "loss": 0.1863, "step": 8089 }, { "epoch": 1.53, "grad_norm": 14.5, "learning_rate": 6.886479999207308e-07, "loss": 0.2552, "step": 8090 }, { "epoch": 1.53, "grad_norm": 11.0625, "learning_rate": 6.881201724414655e-07, "loss": 0.1894, "step": 8091 }, { "epoch": 1.53, "grad_norm": 10.5, "learning_rate": 6.875925150379928e-07, "loss": 0.1597, "step": 8092 }, { "epoch": 1.53, "grad_norm": 14.875, "learning_rate": 6.870650277598445e-07, "loss": 0.2345, "step": 8093 }, { "epoch": 1.53, "grad_norm": 9.625, "learning_rate": 6.865377106565324e-07, "loss": 0.1783, "step": 8094 }, { "epoch": 1.53, "grad_norm": 13.4375, "learning_rate": 6.860105637775539e-07, "loss": 0.1928, "step": 8095 }, { "epoch": 1.53, "grad_norm": 8.6875, "learning_rate": 6.854835871723922e-07, "loss": 0.1699, "step": 8096 }, { "epoch": 1.53, "grad_norm": 13.8125, "learning_rate": 6.849567808905133e-07, "loss": 0.2147, "step": 8097 }, { "epoch": 1.53, "grad_norm": 11.375, "learning_rate": 6.844301449813651e-07, "loss": 0.1697, "step": 8098 }, { "epoch": 1.53, "grad_norm": 9.625, "learning_rate": 6.839036794943813e-07, "loss": 0.1842, "step": 8099 }, { "epoch": 1.53, "grad_norm": 11.1875, "learning_rate": 6.833773844789818e-07, "loss": 0.1986, "step": 8100 }, { "epoch": 1.53, "grad_norm": 10.4375, "learning_rate": 6.828512599845675e-07, "loss": 0.235, "step": 8101 }, { "epoch": 1.53, "grad_norm": 11.8125, "learning_rate": 6.823253060605234e-07, "loss": 0.2305, "step": 8102 }, { "epoch": 1.53, "grad_norm": 11.625, "learning_rate": 6.817995227562186e-07, "loss": 0.1959, "step": 8103 }, { "epoch": 1.53, "grad_norm": 8.125, "learning_rate": 6.812739101210087e-07, "loss": 0.1526, "step": 8104 }, { "epoch": 1.53, "grad_norm": 12.3125, "learning_rate": 6.807484682042315e-07, "loss": 0.1912, "step": 8105 }, { "epoch": 1.53, "grad_norm": 10.1875, "learning_rate": 6.802231970552059e-07, "loss": 0.1391, "step": 8106 }, { "epoch": 1.53, "grad_norm": 14.5, "learning_rate": 6.796980967232406e-07, "loss": 0.2184, "step": 8107 }, { "epoch": 1.53, "grad_norm": 11.4375, "learning_rate": 6.791731672576243e-07, "loss": 0.2022, "step": 8108 }, { "epoch": 1.53, "grad_norm": 12.6875, "learning_rate": 6.786484087076297e-07, "loss": 0.2244, "step": 8109 }, { "epoch": 1.53, "grad_norm": 11.5625, "learning_rate": 6.781238211225145e-07, "loss": 0.197, "step": 8110 }, { "epoch": 1.53, "grad_norm": 10.875, "learning_rate": 6.775994045515208e-07, "loss": 0.2427, "step": 8111 }, { "epoch": 1.53, "grad_norm": 12.0, "learning_rate": 6.770751590438749e-07, "loss": 0.2138, "step": 8112 }, { "epoch": 1.53, "grad_norm": 10.9375, "learning_rate": 6.765510846487841e-07, "loss": 0.1917, "step": 8113 }, { "epoch": 1.53, "grad_norm": 11.5, "learning_rate": 6.760271814154423e-07, "loss": 0.2316, "step": 8114 }, { "epoch": 1.54, "grad_norm": 9.0625, "learning_rate": 6.755034493930277e-07, "loss": 0.1713, "step": 8115 }, { "epoch": 1.54, "grad_norm": 11.125, "learning_rate": 6.749798886307016e-07, "loss": 0.1928, "step": 8116 }, { "epoch": 1.54, "grad_norm": 15.5, "learning_rate": 6.744564991776071e-07, "loss": 0.2872, "step": 8117 }, { "epoch": 1.54, "grad_norm": 12.4375, "learning_rate": 6.739332810828752e-07, "loss": 0.1732, "step": 8118 }, { "epoch": 1.54, "grad_norm": 12.6875, "learning_rate": 6.734102343956181e-07, "loss": 0.2674, "step": 8119 }, { "epoch": 1.54, "grad_norm": 10.5625, "learning_rate": 6.728873591649338e-07, "loss": 0.2137, "step": 8120 }, { "epoch": 1.54, "grad_norm": 15.6875, "learning_rate": 6.723646554399002e-07, "loss": 0.243, "step": 8121 }, { "epoch": 1.54, "grad_norm": 12.1875, "learning_rate": 6.718421232695843e-07, "loss": 0.1704, "step": 8122 }, { "epoch": 1.54, "grad_norm": 12.125, "learning_rate": 6.713197627030343e-07, "loss": 0.2196, "step": 8123 }, { "epoch": 1.54, "grad_norm": 10.0625, "learning_rate": 6.707975737892833e-07, "loss": 0.1988, "step": 8124 }, { "epoch": 1.54, "grad_norm": 14.8125, "learning_rate": 6.70275556577345e-07, "loss": 0.2173, "step": 8125 }, { "epoch": 1.54, "grad_norm": 14.0625, "learning_rate": 6.697537111162223e-07, "loss": 0.257, "step": 8126 }, { "epoch": 1.54, "grad_norm": 15.0625, "learning_rate": 6.69232037454898e-07, "loss": 0.2519, "step": 8127 }, { "epoch": 1.54, "grad_norm": 14.5, "learning_rate": 6.687105356423407e-07, "loss": 0.2462, "step": 8128 }, { "epoch": 1.54, "grad_norm": 22.25, "learning_rate": 6.681892057275019e-07, "loss": 0.2444, "step": 8129 }, { "epoch": 1.54, "grad_norm": 16.875, "learning_rate": 6.676680477593173e-07, "loss": 0.2106, "step": 8130 }, { "epoch": 1.54, "grad_norm": 10.375, "learning_rate": 6.671470617867063e-07, "loss": 0.1784, "step": 8131 }, { "epoch": 1.54, "grad_norm": 10.4375, "learning_rate": 6.666262478585725e-07, "loss": 0.226, "step": 8132 }, { "epoch": 1.54, "grad_norm": 16.0, "learning_rate": 6.661056060238033e-07, "loss": 0.3203, "step": 8133 }, { "epoch": 1.54, "grad_norm": 12.8125, "learning_rate": 6.655851363312696e-07, "loss": 0.2298, "step": 8134 }, { "epoch": 1.54, "grad_norm": 9.375, "learning_rate": 6.650648388298265e-07, "loss": 0.1564, "step": 8135 }, { "epoch": 1.54, "grad_norm": 9.5, "learning_rate": 6.645447135683126e-07, "loss": 0.2198, "step": 8136 }, { "epoch": 1.54, "grad_norm": 12.875, "learning_rate": 6.640247605955505e-07, "loss": 0.2445, "step": 8137 }, { "epoch": 1.54, "grad_norm": 13.875, "learning_rate": 6.635049799603468e-07, "loss": 0.238, "step": 8138 }, { "epoch": 1.54, "grad_norm": 12.625, "learning_rate": 6.629853717114917e-07, "loss": 0.1882, "step": 8139 }, { "epoch": 1.54, "grad_norm": 11.375, "learning_rate": 6.624659358977595e-07, "loss": 0.1724, "step": 8140 }, { "epoch": 1.54, "grad_norm": 14.1875, "learning_rate": 6.619466725679077e-07, "loss": 0.25, "step": 8141 }, { "epoch": 1.54, "grad_norm": 10.75, "learning_rate": 6.614275817706784e-07, "loss": 0.171, "step": 8142 }, { "epoch": 1.54, "grad_norm": 14.375, "learning_rate": 6.609086635547968e-07, "loss": 0.206, "step": 8143 }, { "epoch": 1.54, "grad_norm": 9.4375, "learning_rate": 6.603899179689721e-07, "loss": 0.2047, "step": 8144 }, { "epoch": 1.54, "grad_norm": 10.875, "learning_rate": 6.598713450618983e-07, "loss": 0.2249, "step": 8145 }, { "epoch": 1.54, "grad_norm": 9.5625, "learning_rate": 6.593529448822511e-07, "loss": 0.1728, "step": 8146 }, { "epoch": 1.54, "grad_norm": 14.625, "learning_rate": 6.588347174786919e-07, "loss": 0.256, "step": 8147 }, { "epoch": 1.54, "grad_norm": 15.0, "learning_rate": 6.583166628998652e-07, "loss": 0.2533, "step": 8148 }, { "epoch": 1.54, "grad_norm": 10.125, "learning_rate": 6.577987811943989e-07, "loss": 0.1665, "step": 8149 }, { "epoch": 1.54, "grad_norm": 13.875, "learning_rate": 6.572810724109052e-07, "loss": 0.1752, "step": 8150 }, { "epoch": 1.54, "grad_norm": 11.4375, "learning_rate": 6.567635365979799e-07, "loss": 0.2528, "step": 8151 }, { "epoch": 1.54, "grad_norm": 11.6875, "learning_rate": 6.562461738042028e-07, "loss": 0.2206, "step": 8152 }, { "epoch": 1.54, "grad_norm": 12.1875, "learning_rate": 6.557289840781367e-07, "loss": 0.2056, "step": 8153 }, { "epoch": 1.54, "grad_norm": 8.375, "learning_rate": 6.552119674683288e-07, "loss": 0.1889, "step": 8154 }, { "epoch": 1.54, "grad_norm": 7.5625, "learning_rate": 6.546951240233104e-07, "loss": 0.1499, "step": 8155 }, { "epoch": 1.54, "grad_norm": 12.3125, "learning_rate": 6.541784537915955e-07, "loss": 0.2301, "step": 8156 }, { "epoch": 1.54, "grad_norm": 11.0625, "learning_rate": 6.536619568216826e-07, "loss": 0.2217, "step": 8157 }, { "epoch": 1.54, "grad_norm": 12.5, "learning_rate": 6.531456331620537e-07, "loss": 0.1684, "step": 8158 }, { "epoch": 1.54, "grad_norm": 10.6875, "learning_rate": 6.526294828611748e-07, "loss": 0.1856, "step": 8159 }, { "epoch": 1.54, "grad_norm": 14.375, "learning_rate": 6.521135059674952e-07, "loss": 0.2436, "step": 8160 }, { "epoch": 1.54, "grad_norm": 9.1875, "learning_rate": 6.515977025294473e-07, "loss": 0.1947, "step": 8161 }, { "epoch": 1.54, "grad_norm": 8.5625, "learning_rate": 6.510820725954503e-07, "loss": 0.1738, "step": 8162 }, { "epoch": 1.54, "grad_norm": 9.625, "learning_rate": 6.505666162139027e-07, "loss": 0.1532, "step": 8163 }, { "epoch": 1.54, "grad_norm": 10.8125, "learning_rate": 6.500513334331898e-07, "loss": 0.2161, "step": 8164 }, { "epoch": 1.54, "grad_norm": 14.25, "learning_rate": 6.495362243016795e-07, "loss": 0.2329, "step": 8165 }, { "epoch": 1.54, "grad_norm": 11.8125, "learning_rate": 6.490212888677233e-07, "loss": 0.1754, "step": 8166 }, { "epoch": 1.54, "grad_norm": 10.4375, "learning_rate": 6.48506527179657e-07, "loss": 0.2003, "step": 8167 }, { "epoch": 1.55, "grad_norm": 14.25, "learning_rate": 6.479919392857998e-07, "loss": 0.2569, "step": 8168 }, { "epoch": 1.55, "grad_norm": 17.125, "learning_rate": 6.474775252344542e-07, "loss": 0.248, "step": 8169 }, { "epoch": 1.55, "grad_norm": 13.1875, "learning_rate": 6.469632850739071e-07, "loss": 0.1478, "step": 8170 }, { "epoch": 1.55, "grad_norm": 12.125, "learning_rate": 6.464492188524285e-07, "loss": 0.2374, "step": 8171 }, { "epoch": 1.55, "grad_norm": 11.3125, "learning_rate": 6.459353266182713e-07, "loss": 0.207, "step": 8172 }, { "epoch": 1.55, "grad_norm": 10.8125, "learning_rate": 6.45421608419676e-07, "loss": 0.2415, "step": 8173 }, { "epoch": 1.55, "grad_norm": 12.3125, "learning_rate": 6.449080643048608e-07, "loss": 0.2101, "step": 8174 }, { "epoch": 1.55, "grad_norm": 13.375, "learning_rate": 6.443946943220317e-07, "loss": 0.1791, "step": 8175 }, { "epoch": 1.55, "grad_norm": 13.75, "learning_rate": 6.438814985193761e-07, "loss": 0.1834, "step": 8176 }, { "epoch": 1.55, "grad_norm": 8.75, "learning_rate": 6.43368476945069e-07, "loss": 0.183, "step": 8177 }, { "epoch": 1.55, "grad_norm": 19.5, "learning_rate": 6.428556296472638e-07, "loss": 0.2664, "step": 8178 }, { "epoch": 1.55, "grad_norm": 13.125, "learning_rate": 6.423429566741004e-07, "loss": 0.2088, "step": 8179 }, { "epoch": 1.55, "grad_norm": 11.3125, "learning_rate": 6.418304580737014e-07, "loss": 0.2173, "step": 8180 }, { "epoch": 1.55, "grad_norm": 11.875, "learning_rate": 6.413181338941754e-07, "loss": 0.2187, "step": 8181 }, { "epoch": 1.55, "grad_norm": 8.75, "learning_rate": 6.408059841836112e-07, "loss": 0.1846, "step": 8182 }, { "epoch": 1.55, "grad_norm": 11.4375, "learning_rate": 6.402940089900822e-07, "loss": 0.2008, "step": 8183 }, { "epoch": 1.55, "grad_norm": 15.5, "learning_rate": 6.397822083616475e-07, "loss": 0.2222, "step": 8184 }, { "epoch": 1.55, "grad_norm": 11.25, "learning_rate": 6.392705823463488e-07, "loss": 0.232, "step": 8185 }, { "epoch": 1.55, "grad_norm": 12.125, "learning_rate": 6.387591309922089e-07, "loss": 0.2176, "step": 8186 }, { "epoch": 1.55, "grad_norm": 10.1875, "learning_rate": 6.382478543472364e-07, "loss": 0.2155, "step": 8187 }, { "epoch": 1.55, "grad_norm": 14.9375, "learning_rate": 6.377367524594252e-07, "loss": 0.2454, "step": 8188 }, { "epoch": 1.55, "grad_norm": 11.3125, "learning_rate": 6.372258253767502e-07, "loss": 0.1996, "step": 8189 }, { "epoch": 1.55, "grad_norm": 8.5, "learning_rate": 6.367150731471697e-07, "loss": 0.1733, "step": 8190 }, { "epoch": 1.55, "grad_norm": 11.25, "learning_rate": 6.362044958186262e-07, "loss": 0.1801, "step": 8191 }, { "epoch": 1.55, "grad_norm": 8.5, "learning_rate": 6.356940934390482e-07, "loss": 0.1637, "step": 8192 }, { "epoch": 1.55, "grad_norm": 10.3125, "learning_rate": 6.351838660563439e-07, "loss": 0.2389, "step": 8193 }, { "epoch": 1.55, "grad_norm": 14.625, "learning_rate": 6.346738137184064e-07, "loss": 0.2646, "step": 8194 }, { "epoch": 1.55, "grad_norm": 11.875, "learning_rate": 6.341639364731147e-07, "loss": 0.2269, "step": 8195 }, { "epoch": 1.55, "grad_norm": 16.25, "learning_rate": 6.336542343683289e-07, "loss": 0.2694, "step": 8196 }, { "epoch": 1.55, "grad_norm": 10.625, "learning_rate": 6.331447074518923e-07, "loss": 0.2141, "step": 8197 }, { "epoch": 1.55, "grad_norm": 17.25, "learning_rate": 6.326353557716322e-07, "loss": 0.194, "step": 8198 }, { "epoch": 1.55, "grad_norm": 11.4375, "learning_rate": 6.32126179375362e-07, "loss": 0.1896, "step": 8199 }, { "epoch": 1.55, "grad_norm": 9.0625, "learning_rate": 6.31617178310876e-07, "loss": 0.1747, "step": 8200 }, { "epoch": 1.55, "grad_norm": 17.625, "learning_rate": 6.311083526259515e-07, "loss": 0.2509, "step": 8201 }, { "epoch": 1.55, "grad_norm": 10.6875, "learning_rate": 6.305997023683502e-07, "loss": 0.2515, "step": 8202 }, { "epoch": 1.55, "grad_norm": 11.375, "learning_rate": 6.300912275858193e-07, "loss": 0.236, "step": 8203 }, { "epoch": 1.55, "grad_norm": 10.4375, "learning_rate": 6.295829283260877e-07, "loss": 0.1547, "step": 8204 }, { "epoch": 1.55, "grad_norm": 9.375, "learning_rate": 6.290748046368658e-07, "loss": 0.1424, "step": 8205 }, { "epoch": 1.55, "grad_norm": 11.5625, "learning_rate": 6.285668565658517e-07, "loss": 0.1869, "step": 8206 }, { "epoch": 1.55, "grad_norm": 14.0, "learning_rate": 6.280590841607248e-07, "loss": 0.1569, "step": 8207 }, { "epoch": 1.55, "grad_norm": 13.375, "learning_rate": 6.275514874691483e-07, "loss": 0.1742, "step": 8208 }, { "epoch": 1.55, "grad_norm": 11.6875, "learning_rate": 6.270440665387667e-07, "loss": 0.2198, "step": 8209 }, { "epoch": 1.55, "grad_norm": 9.5625, "learning_rate": 6.265368214172124e-07, "loss": 0.167, "step": 8210 }, { "epoch": 1.55, "grad_norm": 11.8125, "learning_rate": 6.260297521520984e-07, "loss": 0.2037, "step": 8211 }, { "epoch": 1.55, "grad_norm": 12.6875, "learning_rate": 6.255228587910223e-07, "loss": 0.2299, "step": 8212 }, { "epoch": 1.55, "grad_norm": 15.25, "learning_rate": 6.250161413815626e-07, "loss": 0.2401, "step": 8213 }, { "epoch": 1.55, "grad_norm": 11.4375, "learning_rate": 6.245095999712855e-07, "loss": 0.1526, "step": 8214 }, { "epoch": 1.55, "grad_norm": 12.5, "learning_rate": 6.240032346077379e-07, "loss": 0.1961, "step": 8215 }, { "epoch": 1.55, "grad_norm": 24.5, "learning_rate": 6.234970453384517e-07, "loss": 0.2281, "step": 8216 }, { "epoch": 1.55, "grad_norm": 10.1875, "learning_rate": 6.229910322109386e-07, "loss": 0.2552, "step": 8217 }, { "epoch": 1.55, "grad_norm": 9.6875, "learning_rate": 6.224851952726996e-07, "loss": 0.2065, "step": 8218 }, { "epoch": 1.55, "grad_norm": 11.5, "learning_rate": 6.219795345712146e-07, "loss": 0.1697, "step": 8219 }, { "epoch": 1.55, "grad_norm": 9.5, "learning_rate": 6.21474050153949e-07, "loss": 0.14, "step": 8220 }, { "epoch": 1.56, "grad_norm": 9.25, "learning_rate": 6.209687420683508e-07, "loss": 0.2396, "step": 8221 }, { "epoch": 1.56, "grad_norm": 13.9375, "learning_rate": 6.204636103618519e-07, "loss": 0.2606, "step": 8222 }, { "epoch": 1.56, "grad_norm": 9.8125, "learning_rate": 6.199586550818681e-07, "loss": 0.2053, "step": 8223 }, { "epoch": 1.56, "grad_norm": 15.9375, "learning_rate": 6.194538762757962e-07, "loss": 0.1771, "step": 8224 }, { "epoch": 1.56, "grad_norm": 8.8125, "learning_rate": 6.189492739910202e-07, "loss": 0.1763, "step": 8225 }, { "epoch": 1.56, "grad_norm": 9.75, "learning_rate": 6.18444848274905e-07, "loss": 0.1717, "step": 8226 }, { "epoch": 1.56, "grad_norm": 8.125, "learning_rate": 6.179405991748005e-07, "loss": 0.1473, "step": 8227 }, { "epoch": 1.56, "grad_norm": 12.4375, "learning_rate": 6.174365267380364e-07, "loss": 0.1983, "step": 8228 }, { "epoch": 1.56, "grad_norm": 12.375, "learning_rate": 6.169326310119309e-07, "loss": 0.1719, "step": 8229 }, { "epoch": 1.56, "grad_norm": 12.1875, "learning_rate": 6.164289120437825e-07, "loss": 0.2107, "step": 8230 }, { "epoch": 1.56, "grad_norm": 10.8125, "learning_rate": 6.159253698808739e-07, "loss": 0.201, "step": 8231 }, { "epoch": 1.56, "grad_norm": 13.9375, "learning_rate": 6.154220045704709e-07, "loss": 0.2213, "step": 8232 }, { "epoch": 1.56, "grad_norm": 9.125, "learning_rate": 6.149188161598229e-07, "loss": 0.1749, "step": 8233 }, { "epoch": 1.56, "grad_norm": 16.0, "learning_rate": 6.144158046961632e-07, "loss": 0.2225, "step": 8234 }, { "epoch": 1.56, "grad_norm": 11.9375, "learning_rate": 6.139129702267074e-07, "loss": 0.2202, "step": 8235 }, { "epoch": 1.56, "grad_norm": 9.3125, "learning_rate": 6.134103127986551e-07, "loss": 0.24, "step": 8236 }, { "epoch": 1.56, "grad_norm": 13.0, "learning_rate": 6.129078324591897e-07, "loss": 0.2098, "step": 8237 }, { "epoch": 1.56, "grad_norm": 20.625, "learning_rate": 6.124055292554773e-07, "loss": 0.2218, "step": 8238 }, { "epoch": 1.56, "grad_norm": 10.0, "learning_rate": 6.119034032346677e-07, "loss": 0.1754, "step": 8239 }, { "epoch": 1.56, "grad_norm": 12.0, "learning_rate": 6.114014544438939e-07, "loss": 0.206, "step": 8240 }, { "epoch": 1.56, "grad_norm": 12.3125, "learning_rate": 6.108996829302721e-07, "loss": 0.227, "step": 8241 }, { "epoch": 1.56, "grad_norm": 9.0, "learning_rate": 6.103980887409025e-07, "loss": 0.216, "step": 8242 }, { "epoch": 1.56, "grad_norm": 11.75, "learning_rate": 6.098966719228683e-07, "loss": 0.1614, "step": 8243 }, { "epoch": 1.56, "grad_norm": 12.125, "learning_rate": 6.093954325232356e-07, "loss": 0.2195, "step": 8244 }, { "epoch": 1.56, "grad_norm": 13.6875, "learning_rate": 6.088943705890546e-07, "loss": 0.2151, "step": 8245 }, { "epoch": 1.56, "grad_norm": 14.75, "learning_rate": 6.083934861673582e-07, "loss": 0.2744, "step": 8246 }, { "epoch": 1.56, "grad_norm": 9.8125, "learning_rate": 6.078927793051634e-07, "loss": 0.1386, "step": 8247 }, { "epoch": 1.56, "grad_norm": 11.8125, "learning_rate": 6.073922500494697e-07, "loss": 0.2285, "step": 8248 }, { "epoch": 1.56, "grad_norm": 10.0625, "learning_rate": 6.068918984472602e-07, "loss": 0.1712, "step": 8249 }, { "epoch": 1.56, "grad_norm": 17.25, "learning_rate": 6.06391724545502e-07, "loss": 0.2969, "step": 8250 }, { "epoch": 1.56, "grad_norm": 8.75, "learning_rate": 6.058917283911442e-07, "loss": 0.1802, "step": 8251 }, { "epoch": 1.56, "grad_norm": 11.0, "learning_rate": 6.053919100311204e-07, "loss": 0.2, "step": 8252 }, { "epoch": 1.56, "grad_norm": 10.0, "learning_rate": 6.048922695123472e-07, "loss": 0.1696, "step": 8253 }, { "epoch": 1.56, "grad_norm": 12.625, "learning_rate": 6.04392806881724e-07, "loss": 0.2354, "step": 8254 }, { "epoch": 1.56, "grad_norm": 18.875, "learning_rate": 6.038935221861339e-07, "loss": 0.1697, "step": 8255 }, { "epoch": 1.56, "grad_norm": 10.4375, "learning_rate": 6.033944154724436e-07, "loss": 0.2089, "step": 8256 }, { "epoch": 1.56, "grad_norm": 11.8125, "learning_rate": 6.028954867875025e-07, "loss": 0.23, "step": 8257 }, { "epoch": 1.56, "grad_norm": 15.375, "learning_rate": 6.023967361781438e-07, "loss": 0.2145, "step": 8258 }, { "epoch": 1.56, "grad_norm": 9.1875, "learning_rate": 6.018981636911833e-07, "loss": 0.1885, "step": 8259 }, { "epoch": 1.56, "grad_norm": 13.0, "learning_rate": 6.013997693734211e-07, "loss": 0.2604, "step": 8260 }, { "epoch": 1.56, "grad_norm": 13.5, "learning_rate": 6.009015532716397e-07, "loss": 0.167, "step": 8261 }, { "epoch": 1.56, "grad_norm": 10.6875, "learning_rate": 6.004035154326049e-07, "loss": 0.1804, "step": 8262 }, { "epoch": 1.56, "grad_norm": 11.4375, "learning_rate": 5.999056559030667e-07, "loss": 0.1585, "step": 8263 }, { "epoch": 1.56, "grad_norm": 11.875, "learning_rate": 5.994079747297562e-07, "loss": 0.2333, "step": 8264 }, { "epoch": 1.56, "grad_norm": 16.375, "learning_rate": 5.989104719593922e-07, "loss": 0.1547, "step": 8265 }, { "epoch": 1.56, "grad_norm": 11.5, "learning_rate": 5.984131476386709e-07, "loss": 0.2243, "step": 8266 }, { "epoch": 1.56, "grad_norm": 15.5, "learning_rate": 5.97916001814276e-07, "loss": 0.2518, "step": 8267 }, { "epoch": 1.56, "grad_norm": 15.375, "learning_rate": 5.97419034532872e-07, "loss": 0.2502, "step": 8268 }, { "epoch": 1.56, "grad_norm": 14.5, "learning_rate": 5.9692224584111e-07, "loss": 0.2029, "step": 8269 }, { "epoch": 1.56, "grad_norm": 8.625, "learning_rate": 5.964256357856201e-07, "loss": 0.1841, "step": 8270 }, { "epoch": 1.56, "grad_norm": 11.3125, "learning_rate": 5.95929204413018e-07, "loss": 0.1926, "step": 8271 }, { "epoch": 1.56, "grad_norm": 16.125, "learning_rate": 5.954329517699021e-07, "loss": 0.2288, "step": 8272 }, { "epoch": 1.57, "grad_norm": 9.375, "learning_rate": 5.949368779028555e-07, "loss": 0.1652, "step": 8273 }, { "epoch": 1.57, "grad_norm": 14.125, "learning_rate": 5.944409828584416e-07, "loss": 0.2654, "step": 8274 }, { "epoch": 1.57, "grad_norm": 7.75, "learning_rate": 5.939452666832088e-07, "loss": 0.1863, "step": 8275 }, { "epoch": 1.57, "grad_norm": 14.75, "learning_rate": 5.934497294236899e-07, "loss": 0.2302, "step": 8276 }, { "epoch": 1.57, "grad_norm": 12.3125, "learning_rate": 5.92954371126398e-07, "loss": 0.2579, "step": 8277 }, { "epoch": 1.57, "grad_norm": 11.75, "learning_rate": 5.924591918378314e-07, "loss": 0.2303, "step": 8278 }, { "epoch": 1.57, "grad_norm": 9.1875, "learning_rate": 5.919641916044705e-07, "loss": 0.1958, "step": 8279 }, { "epoch": 1.57, "grad_norm": 9.8125, "learning_rate": 5.914693704727814e-07, "loss": 0.2129, "step": 8280 }, { "epoch": 1.57, "grad_norm": 14.4375, "learning_rate": 5.909747284892098e-07, "loss": 0.2356, "step": 8281 }, { "epoch": 1.57, "grad_norm": 14.8125, "learning_rate": 5.904802657001865e-07, "loss": 0.1954, "step": 8282 }, { "epoch": 1.57, "grad_norm": 10.125, "learning_rate": 5.899859821521247e-07, "loss": 0.215, "step": 8283 }, { "epoch": 1.57, "grad_norm": 10.0625, "learning_rate": 5.894918778914238e-07, "loss": 0.2158, "step": 8284 }, { "epoch": 1.57, "grad_norm": 15.0, "learning_rate": 5.889979529644612e-07, "loss": 0.2163, "step": 8285 }, { "epoch": 1.57, "grad_norm": 11.3125, "learning_rate": 5.885042074176007e-07, "loss": 0.2276, "step": 8286 }, { "epoch": 1.57, "grad_norm": 13.625, "learning_rate": 5.880106412971897e-07, "loss": 0.201, "step": 8287 }, { "epoch": 1.57, "grad_norm": 7.6875, "learning_rate": 5.875172546495583e-07, "loss": 0.1665, "step": 8288 }, { "epoch": 1.57, "grad_norm": 13.375, "learning_rate": 5.870240475210173e-07, "loss": 0.1776, "step": 8289 }, { "epoch": 1.57, "grad_norm": 10.5, "learning_rate": 5.865310199578631e-07, "loss": 0.2148, "step": 8290 }, { "epoch": 1.57, "grad_norm": 9.5, "learning_rate": 5.860381720063757e-07, "loss": 0.1957, "step": 8291 }, { "epoch": 1.57, "grad_norm": 26.0, "learning_rate": 5.855455037128177e-07, "loss": 0.2106, "step": 8292 }, { "epoch": 1.57, "grad_norm": 12.0, "learning_rate": 5.850530151234323e-07, "loss": 0.177, "step": 8293 }, { "epoch": 1.57, "grad_norm": 16.125, "learning_rate": 5.845607062844489e-07, "loss": 0.257, "step": 8294 }, { "epoch": 1.57, "grad_norm": 12.4375, "learning_rate": 5.840685772420798e-07, "loss": 0.2059, "step": 8295 }, { "epoch": 1.57, "grad_norm": 11.375, "learning_rate": 5.835766280425198e-07, "loss": 0.2065, "step": 8296 }, { "epoch": 1.57, "grad_norm": 11.875, "learning_rate": 5.830848587319449e-07, "loss": 0.2133, "step": 8297 }, { "epoch": 1.57, "grad_norm": 11.6875, "learning_rate": 5.825932693565175e-07, "loss": 0.2152, "step": 8298 }, { "epoch": 1.57, "grad_norm": 13.4375, "learning_rate": 5.821018599623818e-07, "loss": 0.2906, "step": 8299 }, { "epoch": 1.57, "grad_norm": 13.5, "learning_rate": 5.816106305956651e-07, "loss": 0.2103, "step": 8300 }, { "epoch": 1.57, "grad_norm": 10.125, "learning_rate": 5.811195813024756e-07, "loss": 0.1847, "step": 8301 }, { "epoch": 1.57, "grad_norm": 10.25, "learning_rate": 5.80628712128909e-07, "loss": 0.2158, "step": 8302 }, { "epoch": 1.57, "grad_norm": 11.5, "learning_rate": 5.801380231210404e-07, "loss": 0.2476, "step": 8303 }, { "epoch": 1.57, "grad_norm": 11.75, "learning_rate": 5.796475143249309e-07, "loss": 0.2476, "step": 8304 }, { "epoch": 1.57, "grad_norm": 14.3125, "learning_rate": 5.791571857866205e-07, "loss": 0.1834, "step": 8305 }, { "epoch": 1.57, "grad_norm": 15.4375, "learning_rate": 5.78667037552137e-07, "loss": 0.1962, "step": 8306 }, { "epoch": 1.57, "grad_norm": 13.1875, "learning_rate": 5.78177069667489e-07, "loss": 0.2409, "step": 8307 }, { "epoch": 1.57, "grad_norm": 12.0, "learning_rate": 5.776872821786667e-07, "loss": 0.2236, "step": 8308 }, { "epoch": 1.57, "grad_norm": 12.0625, "learning_rate": 5.771976751316469e-07, "loss": 0.2606, "step": 8309 }, { "epoch": 1.57, "grad_norm": 8.625, "learning_rate": 5.767082485723865e-07, "loss": 0.2009, "step": 8310 }, { "epoch": 1.57, "grad_norm": 14.125, "learning_rate": 5.762190025468278e-07, "loss": 0.2672, "step": 8311 }, { "epoch": 1.57, "grad_norm": 9.0625, "learning_rate": 5.757299371008925e-07, "loss": 0.195, "step": 8312 }, { "epoch": 1.57, "grad_norm": 15.125, "learning_rate": 5.752410522804896e-07, "loss": 0.1703, "step": 8313 }, { "epoch": 1.57, "grad_norm": 11.0, "learning_rate": 5.747523481315093e-07, "loss": 0.2199, "step": 8314 }, { "epoch": 1.57, "grad_norm": 11.8125, "learning_rate": 5.742638246998248e-07, "loss": 0.1746, "step": 8315 }, { "epoch": 1.57, "grad_norm": 8.25, "learning_rate": 5.737754820312902e-07, "loss": 0.1837, "step": 8316 }, { "epoch": 1.57, "grad_norm": 11.375, "learning_rate": 5.732873201717473e-07, "loss": 0.1405, "step": 8317 }, { "epoch": 1.57, "grad_norm": 9.3125, "learning_rate": 5.727993391670178e-07, "loss": 0.1765, "step": 8318 }, { "epoch": 1.57, "grad_norm": 10.3125, "learning_rate": 5.723115390629063e-07, "loss": 0.1655, "step": 8319 }, { "epoch": 1.57, "grad_norm": 11.1875, "learning_rate": 5.718239199052022e-07, "loss": 0.2261, "step": 8320 }, { "epoch": 1.57, "grad_norm": 11.875, "learning_rate": 5.71336481739676e-07, "loss": 0.1851, "step": 8321 }, { "epoch": 1.57, "grad_norm": 14.8125, "learning_rate": 5.708492246120826e-07, "loss": 0.1965, "step": 8322 }, { "epoch": 1.57, "grad_norm": 11.1875, "learning_rate": 5.703621485681588e-07, "loss": 0.2296, "step": 8323 }, { "epoch": 1.57, "grad_norm": 11.1875, "learning_rate": 5.698752536536256e-07, "loss": 0.1648, "step": 8324 }, { "epoch": 1.57, "grad_norm": 13.625, "learning_rate": 5.69388539914186e-07, "loss": 0.2794, "step": 8325 }, { "epoch": 1.58, "grad_norm": 9.9375, "learning_rate": 5.689020073955267e-07, "loss": 0.2312, "step": 8326 }, { "epoch": 1.58, "grad_norm": 12.875, "learning_rate": 5.684156561433169e-07, "loss": 0.2261, "step": 8327 }, { "epoch": 1.58, "grad_norm": 12.0625, "learning_rate": 5.679294862032086e-07, "loss": 0.1992, "step": 8328 }, { "epoch": 1.58, "grad_norm": 11.0625, "learning_rate": 5.674434976208377e-07, "loss": 0.1601, "step": 8329 }, { "epoch": 1.58, "grad_norm": 12.25, "learning_rate": 5.669576904418222e-07, "loss": 0.2524, "step": 8330 }, { "epoch": 1.58, "grad_norm": 9.5, "learning_rate": 5.664720647117636e-07, "loss": 0.1936, "step": 8331 }, { "epoch": 1.58, "grad_norm": 11.25, "learning_rate": 5.65986620476246e-07, "loss": 0.1859, "step": 8332 }, { "epoch": 1.58, "grad_norm": 18.875, "learning_rate": 5.655013577808363e-07, "loss": 0.2248, "step": 8333 }, { "epoch": 1.58, "grad_norm": 8.375, "learning_rate": 5.650162766710854e-07, "loss": 0.2015, "step": 8334 }, { "epoch": 1.58, "grad_norm": 10.8125, "learning_rate": 5.645313771925259e-07, "loss": 0.2155, "step": 8335 }, { "epoch": 1.58, "grad_norm": 8.4375, "learning_rate": 5.640466593906738e-07, "loss": 0.1558, "step": 8336 }, { "epoch": 1.58, "grad_norm": 13.75, "learning_rate": 5.635621233110283e-07, "loss": 0.2505, "step": 8337 }, { "epoch": 1.58, "grad_norm": 9.5, "learning_rate": 5.630777689990715e-07, "loss": 0.2323, "step": 8338 }, { "epoch": 1.58, "grad_norm": 11.3125, "learning_rate": 5.625935965002685e-07, "loss": 0.2192, "step": 8339 }, { "epoch": 1.58, "grad_norm": 17.875, "learning_rate": 5.621096058600664e-07, "loss": 0.1748, "step": 8340 }, { "epoch": 1.58, "grad_norm": 15.4375, "learning_rate": 5.616257971238967e-07, "loss": 0.2439, "step": 8341 }, { "epoch": 1.58, "grad_norm": 14.0625, "learning_rate": 5.611421703371725e-07, "loss": 0.2897, "step": 8342 }, { "epoch": 1.58, "grad_norm": 12.1875, "learning_rate": 5.606587255452909e-07, "loss": 0.246, "step": 8343 }, { "epoch": 1.58, "grad_norm": 11.0625, "learning_rate": 5.601754627936312e-07, "loss": 0.2341, "step": 8344 }, { "epoch": 1.58, "grad_norm": 12.6875, "learning_rate": 5.596923821275558e-07, "loss": 0.2674, "step": 8345 }, { "epoch": 1.58, "grad_norm": 12.25, "learning_rate": 5.592094835924103e-07, "loss": 0.2325, "step": 8346 }, { "epoch": 1.58, "grad_norm": 12.1875, "learning_rate": 5.587267672335226e-07, "loss": 0.2714, "step": 8347 }, { "epoch": 1.58, "grad_norm": 10.0625, "learning_rate": 5.58244233096204e-07, "loss": 0.21, "step": 8348 }, { "epoch": 1.58, "grad_norm": 10.1875, "learning_rate": 5.577618812257488e-07, "loss": 0.1826, "step": 8349 }, { "epoch": 1.58, "grad_norm": 12.8125, "learning_rate": 5.572797116674336e-07, "loss": 0.1752, "step": 8350 }, { "epoch": 1.58, "grad_norm": 16.875, "learning_rate": 5.567977244665182e-07, "loss": 0.2503, "step": 8351 }, { "epoch": 1.58, "grad_norm": 12.375, "learning_rate": 5.563159196682453e-07, "loss": 0.2716, "step": 8352 }, { "epoch": 1.58, "grad_norm": 10.5625, "learning_rate": 5.558342973178415e-07, "loss": 0.1895, "step": 8353 }, { "epoch": 1.58, "grad_norm": 10.0, "learning_rate": 5.553528574605138e-07, "loss": 0.1887, "step": 8354 }, { "epoch": 1.58, "grad_norm": 12.4375, "learning_rate": 5.548716001414545e-07, "loss": 0.166, "step": 8355 }, { "epoch": 1.58, "grad_norm": 15.0, "learning_rate": 5.543905254058365e-07, "loss": 0.205, "step": 8356 }, { "epoch": 1.58, "grad_norm": 11.9375, "learning_rate": 5.539096332988191e-07, "loss": 0.216, "step": 8357 }, { "epoch": 1.58, "grad_norm": 14.75, "learning_rate": 5.534289238655405e-07, "loss": 0.2799, "step": 8358 }, { "epoch": 1.58, "grad_norm": 11.3125, "learning_rate": 5.529483971511241e-07, "loss": 0.1915, "step": 8359 }, { "epoch": 1.58, "grad_norm": 11.375, "learning_rate": 5.52468053200674e-07, "loss": 0.175, "step": 8360 }, { "epoch": 1.58, "grad_norm": 12.5, "learning_rate": 5.519878920592819e-07, "loss": 0.2183, "step": 8361 }, { "epoch": 1.58, "grad_norm": 16.0, "learning_rate": 5.515079137720164e-07, "loss": 0.1622, "step": 8362 }, { "epoch": 1.58, "grad_norm": 9.0, "learning_rate": 5.510281183839317e-07, "loss": 0.1842, "step": 8363 }, { "epoch": 1.58, "grad_norm": 13.0, "learning_rate": 5.505485059400668e-07, "loss": 0.1917, "step": 8364 }, { "epoch": 1.58, "grad_norm": 10.1875, "learning_rate": 5.500690764854394e-07, "loss": 0.2267, "step": 8365 }, { "epoch": 1.58, "grad_norm": 14.4375, "learning_rate": 5.495898300650531e-07, "loss": 0.1998, "step": 8366 }, { "epoch": 1.58, "grad_norm": 13.0625, "learning_rate": 5.491107667238926e-07, "loss": 0.1878, "step": 8367 }, { "epoch": 1.58, "grad_norm": 11.75, "learning_rate": 5.486318865069279e-07, "loss": 0.1804, "step": 8368 }, { "epoch": 1.58, "grad_norm": 10.6875, "learning_rate": 5.481531894591083e-07, "loss": 0.1743, "step": 8369 }, { "epoch": 1.58, "grad_norm": 16.375, "learning_rate": 5.476746756253684e-07, "loss": 0.1982, "step": 8370 }, { "epoch": 1.58, "grad_norm": 13.4375, "learning_rate": 5.47196345050624e-07, "loss": 0.1838, "step": 8371 }, { "epoch": 1.58, "grad_norm": 13.6875, "learning_rate": 5.467181977797767e-07, "loss": 0.196, "step": 8372 }, { "epoch": 1.58, "grad_norm": 10.1875, "learning_rate": 5.462402338577066e-07, "loss": 0.2214, "step": 8373 }, { "epoch": 1.58, "grad_norm": 14.0, "learning_rate": 5.457624533292788e-07, "loss": 0.1763, "step": 8374 }, { "epoch": 1.58, "grad_norm": 11.4375, "learning_rate": 5.452848562393426e-07, "loss": 0.2025, "step": 8375 }, { "epoch": 1.58, "grad_norm": 13.0, "learning_rate": 5.448074426327288e-07, "loss": 0.2086, "step": 8376 }, { "epoch": 1.58, "grad_norm": 8.5, "learning_rate": 5.443302125542491e-07, "loss": 0.1667, "step": 8377 }, { "epoch": 1.58, "grad_norm": 10.375, "learning_rate": 5.438531660486998e-07, "loss": 0.157, "step": 8378 }, { "epoch": 1.59, "grad_norm": 12.0, "learning_rate": 5.433763031608613e-07, "loss": 0.2389, "step": 8379 }, { "epoch": 1.59, "grad_norm": 15.1875, "learning_rate": 5.42899623935495e-07, "loss": 0.2431, "step": 8380 }, { "epoch": 1.59, "grad_norm": 16.0, "learning_rate": 5.424231284173445e-07, "loss": 0.1972, "step": 8381 }, { "epoch": 1.59, "grad_norm": 13.375, "learning_rate": 5.419468166511366e-07, "loss": 0.188, "step": 8382 }, { "epoch": 1.59, "grad_norm": 9.8125, "learning_rate": 5.414706886815826e-07, "loss": 0.2051, "step": 8383 }, { "epoch": 1.59, "grad_norm": 11.3125, "learning_rate": 5.409947445533758e-07, "loss": 0.1816, "step": 8384 }, { "epoch": 1.59, "grad_norm": 9.5, "learning_rate": 5.405189843111891e-07, "loss": 0.1891, "step": 8385 }, { "epoch": 1.59, "grad_norm": 8.9375, "learning_rate": 5.400434079996827e-07, "loss": 0.1983, "step": 8386 }, { "epoch": 1.59, "grad_norm": 12.875, "learning_rate": 5.395680156634972e-07, "loss": 0.2443, "step": 8387 }, { "epoch": 1.59, "grad_norm": 8.8125, "learning_rate": 5.390928073472567e-07, "loss": 0.1719, "step": 8388 }, { "epoch": 1.59, "grad_norm": 11.875, "learning_rate": 5.386177830955658e-07, "loss": 0.2346, "step": 8389 }, { "epoch": 1.59, "grad_norm": 8.5, "learning_rate": 5.381429429530155e-07, "loss": 0.2048, "step": 8390 }, { "epoch": 1.59, "grad_norm": 11.1875, "learning_rate": 5.376682869641775e-07, "loss": 0.2186, "step": 8391 }, { "epoch": 1.59, "grad_norm": 12.125, "learning_rate": 5.371938151736053e-07, "loss": 0.2202, "step": 8392 }, { "epoch": 1.59, "grad_norm": 13.0, "learning_rate": 5.367195276258358e-07, "loss": 0.1933, "step": 8393 }, { "epoch": 1.59, "grad_norm": 9.5625, "learning_rate": 5.362454243653906e-07, "loss": 0.2057, "step": 8394 }, { "epoch": 1.59, "grad_norm": 12.9375, "learning_rate": 5.357715054367723e-07, "loss": 0.2374, "step": 8395 }, { "epoch": 1.59, "grad_norm": 11.0625, "learning_rate": 5.352977708844642e-07, "loss": 0.2033, "step": 8396 }, { "epoch": 1.59, "grad_norm": 16.125, "learning_rate": 5.348242207529366e-07, "loss": 0.1819, "step": 8397 }, { "epoch": 1.59, "grad_norm": 11.0625, "learning_rate": 5.343508550866391e-07, "loss": 0.2133, "step": 8398 }, { "epoch": 1.59, "grad_norm": 11.5625, "learning_rate": 5.338776739300064e-07, "loss": 0.2157, "step": 8399 }, { "epoch": 1.59, "grad_norm": 14.375, "learning_rate": 5.33404677327452e-07, "loss": 0.1928, "step": 8400 }, { "epoch": 1.59, "grad_norm": 12.875, "learning_rate": 5.329318653233773e-07, "loss": 0.2192, "step": 8401 }, { "epoch": 1.59, "grad_norm": 12.625, "learning_rate": 5.324592379621626e-07, "loss": 0.2845, "step": 8402 }, { "epoch": 1.59, "grad_norm": 13.8125, "learning_rate": 5.31986795288173e-07, "loss": 0.2465, "step": 8403 }, { "epoch": 1.59, "grad_norm": 17.25, "learning_rate": 5.31514537345753e-07, "loss": 0.1778, "step": 8404 }, { "epoch": 1.59, "grad_norm": 12.625, "learning_rate": 5.310424641792345e-07, "loss": 0.2272, "step": 8405 }, { "epoch": 1.59, "grad_norm": 9.125, "learning_rate": 5.305705758329283e-07, "loss": 0.231, "step": 8406 }, { "epoch": 1.59, "grad_norm": 10.6875, "learning_rate": 5.300988723511296e-07, "loss": 0.2183, "step": 8407 }, { "epoch": 1.59, "grad_norm": 11.625, "learning_rate": 5.29627353778116e-07, "loss": 0.193, "step": 8408 }, { "epoch": 1.59, "grad_norm": 15.625, "learning_rate": 5.291560201581472e-07, "loss": 0.2244, "step": 8409 }, { "epoch": 1.59, "grad_norm": 16.0, "learning_rate": 5.286848715354658e-07, "loss": 0.3109, "step": 8410 }, { "epoch": 1.59, "grad_norm": 11.25, "learning_rate": 5.282139079542977e-07, "loss": 0.2299, "step": 8411 }, { "epoch": 1.59, "grad_norm": 13.5625, "learning_rate": 5.277431294588504e-07, "loss": 0.2243, "step": 8412 }, { "epoch": 1.59, "grad_norm": 10.0, "learning_rate": 5.272725360933143e-07, "loss": 0.1553, "step": 8413 }, { "epoch": 1.59, "grad_norm": 11.5625, "learning_rate": 5.268021279018632e-07, "loss": 0.2405, "step": 8414 }, { "epoch": 1.59, "grad_norm": 9.3125, "learning_rate": 5.263319049286522e-07, "loss": 0.1706, "step": 8415 }, { "epoch": 1.59, "grad_norm": 11.5625, "learning_rate": 5.258618672178203e-07, "loss": 0.2021, "step": 8416 }, { "epoch": 1.59, "grad_norm": 16.375, "learning_rate": 5.253920148134883e-07, "loss": 0.2408, "step": 8417 }, { "epoch": 1.59, "grad_norm": 11.25, "learning_rate": 5.249223477597601e-07, "loss": 0.1748, "step": 8418 }, { "epoch": 1.59, "grad_norm": 11.3125, "learning_rate": 5.244528661007217e-07, "loss": 0.2226, "step": 8419 }, { "epoch": 1.59, "grad_norm": 32.5, "learning_rate": 5.239835698804421e-07, "loss": 0.1549, "step": 8420 }, { "epoch": 1.59, "grad_norm": 12.6875, "learning_rate": 5.235144591429725e-07, "loss": 0.2093, "step": 8421 }, { "epoch": 1.59, "grad_norm": 13.8125, "learning_rate": 5.230455339323473e-07, "loss": 0.2097, "step": 8422 }, { "epoch": 1.59, "grad_norm": 9.75, "learning_rate": 5.225767942925827e-07, "loss": 0.1801, "step": 8423 }, { "epoch": 1.59, "grad_norm": 10.625, "learning_rate": 5.221082402676783e-07, "loss": 0.2071, "step": 8424 }, { "epoch": 1.59, "grad_norm": 11.3125, "learning_rate": 5.216398719016158e-07, "loss": 0.2019, "step": 8425 }, { "epoch": 1.59, "grad_norm": 12.25, "learning_rate": 5.211716892383595e-07, "loss": 0.2216, "step": 8426 }, { "epoch": 1.59, "grad_norm": 10.6875, "learning_rate": 5.207036923218561e-07, "loss": 0.2035, "step": 8427 }, { "epoch": 1.59, "grad_norm": 11.4375, "learning_rate": 5.202358811960353e-07, "loss": 0.2399, "step": 8428 }, { "epoch": 1.59, "grad_norm": 11.0, "learning_rate": 5.197682559048092e-07, "loss": 0.2171, "step": 8429 }, { "epoch": 1.59, "grad_norm": 10.375, "learning_rate": 5.193008164920724e-07, "loss": 0.2027, "step": 8430 }, { "epoch": 1.59, "grad_norm": 9.0625, "learning_rate": 5.188335630017016e-07, "loss": 0.1957, "step": 8431 }, { "epoch": 1.6, "grad_norm": 11.625, "learning_rate": 5.183664954775572e-07, "loss": 0.2339, "step": 8432 }, { "epoch": 1.6, "grad_norm": 12.75, "learning_rate": 5.178996139634812e-07, "loss": 0.225, "step": 8433 }, { "epoch": 1.6, "grad_norm": 12.75, "learning_rate": 5.174329185032981e-07, "loss": 0.1975, "step": 8434 }, { "epoch": 1.6, "grad_norm": 9.4375, "learning_rate": 5.169664091408155e-07, "loss": 0.1974, "step": 8435 }, { "epoch": 1.6, "grad_norm": 11.1875, "learning_rate": 5.16500085919823e-07, "loss": 0.2064, "step": 8436 }, { "epoch": 1.6, "grad_norm": 10.0, "learning_rate": 5.160339488840935e-07, "loss": 0.1716, "step": 8437 }, { "epoch": 1.6, "grad_norm": 9.375, "learning_rate": 5.155679980773812e-07, "loss": 0.2272, "step": 8438 }, { "epoch": 1.6, "grad_norm": 9.5625, "learning_rate": 5.151022335434242e-07, "loss": 0.1754, "step": 8439 }, { "epoch": 1.6, "grad_norm": 10.3125, "learning_rate": 5.146366553259419e-07, "loss": 0.2377, "step": 8440 }, { "epoch": 1.6, "grad_norm": 14.8125, "learning_rate": 5.141712634686372e-07, "loss": 0.2193, "step": 8441 }, { "epoch": 1.6, "grad_norm": 17.25, "learning_rate": 5.137060580151945e-07, "loss": 0.229, "step": 8442 }, { "epoch": 1.6, "grad_norm": 11.1875, "learning_rate": 5.132410390092818e-07, "loss": 0.2632, "step": 8443 }, { "epoch": 1.6, "grad_norm": 11.4375, "learning_rate": 5.12776206494548e-07, "loss": 0.1551, "step": 8444 }, { "epoch": 1.6, "grad_norm": 13.9375, "learning_rate": 5.123115605146278e-07, "loss": 0.1947, "step": 8445 }, { "epoch": 1.6, "grad_norm": 8.625, "learning_rate": 5.118471011131343e-07, "loss": 0.157, "step": 8446 }, { "epoch": 1.6, "grad_norm": 9.75, "learning_rate": 5.11382828333665e-07, "loss": 0.1469, "step": 8447 }, { "epoch": 1.6, "grad_norm": 25.75, "learning_rate": 5.109187422198006e-07, "loss": 0.2388, "step": 8448 }, { "epoch": 1.6, "grad_norm": 13.4375, "learning_rate": 5.104548428151027e-07, "loss": 0.2164, "step": 8449 }, { "epoch": 1.6, "grad_norm": 11.6875, "learning_rate": 5.099911301631166e-07, "loss": 0.1976, "step": 8450 }, { "epoch": 1.6, "grad_norm": 10.5625, "learning_rate": 5.095276043073696e-07, "loss": 0.1743, "step": 8451 }, { "epoch": 1.6, "grad_norm": 9.9375, "learning_rate": 5.090642652913713e-07, "loss": 0.2036, "step": 8452 }, { "epoch": 1.6, "grad_norm": 7.0, "learning_rate": 5.086011131586143e-07, "loss": 0.1516, "step": 8453 }, { "epoch": 1.6, "grad_norm": 16.5, "learning_rate": 5.081381479525732e-07, "loss": 0.1887, "step": 8454 }, { "epoch": 1.6, "grad_norm": 12.125, "learning_rate": 5.076753697167045e-07, "loss": 0.2046, "step": 8455 }, { "epoch": 1.6, "grad_norm": 11.75, "learning_rate": 5.072127784944494e-07, "loss": 0.1757, "step": 8456 }, { "epoch": 1.6, "grad_norm": 19.625, "learning_rate": 5.067503743292285e-07, "loss": 0.2509, "step": 8457 }, { "epoch": 1.6, "grad_norm": 10.75, "learning_rate": 5.062881572644471e-07, "loss": 0.2401, "step": 8458 }, { "epoch": 1.6, "grad_norm": 8.9375, "learning_rate": 5.058261273434908e-07, "loss": 0.1938, "step": 8459 }, { "epoch": 1.6, "grad_norm": 12.5, "learning_rate": 5.053642846097315e-07, "loss": 0.2531, "step": 8460 }, { "epoch": 1.6, "grad_norm": 15.8125, "learning_rate": 5.04902629106519e-07, "loss": 0.257, "step": 8461 }, { "epoch": 1.6, "grad_norm": 12.9375, "learning_rate": 5.04441160877188e-07, "loss": 0.2351, "step": 8462 }, { "epoch": 1.6, "grad_norm": 21.5, "learning_rate": 5.039798799650544e-07, "loss": 0.2284, "step": 8463 }, { "epoch": 1.6, "grad_norm": 14.4375, "learning_rate": 5.035187864134197e-07, "loss": 0.2538, "step": 8464 }, { "epoch": 1.6, "grad_norm": 14.75, "learning_rate": 5.030578802655628e-07, "loss": 0.1698, "step": 8465 }, { "epoch": 1.6, "grad_norm": 11.6875, "learning_rate": 5.025971615647482e-07, "loss": 0.1438, "step": 8466 }, { "epoch": 1.6, "grad_norm": 10.875, "learning_rate": 5.021366303542233e-07, "loss": 0.1912, "step": 8467 }, { "epoch": 1.6, "grad_norm": 10.25, "learning_rate": 5.016762866772166e-07, "loss": 0.2182, "step": 8468 }, { "epoch": 1.6, "grad_norm": 13.5, "learning_rate": 5.012161305769381e-07, "loss": 0.2329, "step": 8469 }, { "epoch": 1.6, "grad_norm": 12.75, "learning_rate": 5.007561620965812e-07, "loss": 0.2148, "step": 8470 }, { "epoch": 1.6, "grad_norm": 17.125, "learning_rate": 5.002963812793232e-07, "loss": 0.248, "step": 8471 }, { "epoch": 1.6, "grad_norm": 12.875, "learning_rate": 4.998367881683222e-07, "loss": 0.2511, "step": 8472 }, { "epoch": 1.6, "grad_norm": 11.5, "learning_rate": 4.993773828067175e-07, "loss": 0.2745, "step": 8473 }, { "epoch": 1.6, "grad_norm": 11.125, "learning_rate": 4.989181652376324e-07, "loss": 0.1868, "step": 8474 }, { "epoch": 1.6, "grad_norm": 13.8125, "learning_rate": 4.98459135504174e-07, "loss": 0.2601, "step": 8475 }, { "epoch": 1.6, "grad_norm": 12.125, "learning_rate": 4.980002936494283e-07, "loss": 0.2392, "step": 8476 }, { "epoch": 1.6, "grad_norm": 9.25, "learning_rate": 4.975416397164653e-07, "loss": 0.1766, "step": 8477 }, { "epoch": 1.6, "grad_norm": 14.75, "learning_rate": 4.970831737483386e-07, "loss": 0.2436, "step": 8478 }, { "epoch": 1.6, "grad_norm": 10.875, "learning_rate": 4.966248957880834e-07, "loss": 0.1827, "step": 8479 }, { "epoch": 1.6, "grad_norm": 25.0, "learning_rate": 4.961668058787156e-07, "loss": 0.1318, "step": 8480 }, { "epoch": 1.6, "grad_norm": 11.625, "learning_rate": 4.957089040632343e-07, "loss": 0.2118, "step": 8481 }, { "epoch": 1.6, "grad_norm": 12.125, "learning_rate": 4.952511903846233e-07, "loss": 0.22, "step": 8482 }, { "epoch": 1.6, "grad_norm": 10.5625, "learning_rate": 4.947936648858462e-07, "loss": 0.2199, "step": 8483 }, { "epoch": 1.6, "grad_norm": 15.75, "learning_rate": 4.943363276098487e-07, "loss": 0.1969, "step": 8484 }, { "epoch": 1.61, "grad_norm": 9.125, "learning_rate": 4.938791785995595e-07, "loss": 0.1535, "step": 8485 }, { "epoch": 1.61, "grad_norm": 10.9375, "learning_rate": 4.934222178978912e-07, "loss": 0.1464, "step": 8486 }, { "epoch": 1.61, "grad_norm": 11.1875, "learning_rate": 4.929654455477372e-07, "loss": 0.1473, "step": 8487 }, { "epoch": 1.61, "grad_norm": 9.8125, "learning_rate": 4.925088615919719e-07, "loss": 0.2212, "step": 8488 }, { "epoch": 1.61, "grad_norm": 10.1875, "learning_rate": 4.920524660734549e-07, "loss": 0.1729, "step": 8489 }, { "epoch": 1.61, "grad_norm": 12.0, "learning_rate": 4.915962590350263e-07, "loss": 0.2022, "step": 8490 }, { "epoch": 1.61, "grad_norm": 24.75, "learning_rate": 4.911402405195092e-07, "loss": 0.2258, "step": 8491 }, { "epoch": 1.61, "grad_norm": 10.0, "learning_rate": 4.906844105697073e-07, "loss": 0.1864, "step": 8492 }, { "epoch": 1.61, "grad_norm": 10.875, "learning_rate": 4.902287692284097e-07, "loss": 0.1857, "step": 8493 }, { "epoch": 1.61, "grad_norm": 9.75, "learning_rate": 4.897733165383853e-07, "loss": 0.2027, "step": 8494 }, { "epoch": 1.61, "grad_norm": 9.75, "learning_rate": 4.893180525423871e-07, "loss": 0.1761, "step": 8495 }, { "epoch": 1.61, "grad_norm": 12.8125, "learning_rate": 4.888629772831472e-07, "loss": 0.23, "step": 8496 }, { "epoch": 1.61, "grad_norm": 9.125, "learning_rate": 4.884080908033839e-07, "loss": 0.1959, "step": 8497 }, { "epoch": 1.61, "grad_norm": 16.25, "learning_rate": 4.879533931457958e-07, "loss": 0.1943, "step": 8498 }, { "epoch": 1.61, "grad_norm": 14.375, "learning_rate": 4.874988843530639e-07, "loss": 0.2094, "step": 8499 }, { "epoch": 1.61, "grad_norm": 11.75, "learning_rate": 4.870445644678515e-07, "loss": 0.1879, "step": 8500 }, { "epoch": 1.61, "grad_norm": 12.75, "learning_rate": 4.865904335328042e-07, "loss": 0.1806, "step": 8501 }, { "epoch": 1.61, "grad_norm": 12.0, "learning_rate": 4.861364915905503e-07, "loss": 0.1481, "step": 8502 }, { "epoch": 1.61, "grad_norm": 12.75, "learning_rate": 4.856827386836996e-07, "loss": 0.2456, "step": 8503 }, { "epoch": 1.61, "grad_norm": 8.9375, "learning_rate": 4.852291748548446e-07, "loss": 0.193, "step": 8504 }, { "epoch": 1.61, "grad_norm": 12.4375, "learning_rate": 4.847758001465602e-07, "loss": 0.2092, "step": 8505 }, { "epoch": 1.61, "grad_norm": 8.5, "learning_rate": 4.843226146014038e-07, "loss": 0.1995, "step": 8506 }, { "epoch": 1.61, "grad_norm": 12.5625, "learning_rate": 4.838696182619129e-07, "loss": 0.2237, "step": 8507 }, { "epoch": 1.61, "grad_norm": 15.625, "learning_rate": 4.834168111706103e-07, "loss": 0.2423, "step": 8508 }, { "epoch": 1.61, "grad_norm": 9.875, "learning_rate": 4.829641933699997e-07, "loss": 0.1459, "step": 8509 }, { "epoch": 1.61, "grad_norm": 8.9375, "learning_rate": 4.825117649025665e-07, "loss": 0.1899, "step": 8510 }, { "epoch": 1.61, "grad_norm": 10.3125, "learning_rate": 4.820595258107791e-07, "loss": 0.1655, "step": 8511 }, { "epoch": 1.61, "grad_norm": 10.125, "learning_rate": 4.816074761370876e-07, "loss": 0.2021, "step": 8512 }, { "epoch": 1.61, "grad_norm": 10.3125, "learning_rate": 4.811556159239248e-07, "loss": 0.1982, "step": 8513 }, { "epoch": 1.61, "grad_norm": 10.625, "learning_rate": 4.807039452137053e-07, "loss": 0.1764, "step": 8514 }, { "epoch": 1.61, "grad_norm": 13.3125, "learning_rate": 4.802524640488263e-07, "loss": 0.1917, "step": 8515 }, { "epoch": 1.61, "grad_norm": 14.9375, "learning_rate": 4.798011724716667e-07, "loss": 0.2274, "step": 8516 }, { "epoch": 1.61, "grad_norm": 11.875, "learning_rate": 4.793500705245882e-07, "loss": 0.2174, "step": 8517 }, { "epoch": 1.61, "grad_norm": 13.3125, "learning_rate": 4.788991582499341e-07, "loss": 0.1918, "step": 8518 }, { "epoch": 1.61, "grad_norm": 15.375, "learning_rate": 4.784484356900307e-07, "loss": 0.2474, "step": 8519 }, { "epoch": 1.61, "grad_norm": 16.5, "learning_rate": 4.779979028871856e-07, "loss": 0.196, "step": 8520 }, { "epoch": 1.61, "grad_norm": 10.9375, "learning_rate": 4.775475598836887e-07, "loss": 0.1822, "step": 8521 }, { "epoch": 1.61, "grad_norm": 8.5625, "learning_rate": 4.770974067218132e-07, "loss": 0.1857, "step": 8522 }, { "epoch": 1.61, "grad_norm": 8.5625, "learning_rate": 4.766474434438129e-07, "loss": 0.1774, "step": 8523 }, { "epoch": 1.61, "grad_norm": 10.8125, "learning_rate": 4.7619767009192463e-07, "loss": 0.1971, "step": 8524 }, { "epoch": 1.61, "grad_norm": 10.375, "learning_rate": 4.757480867083677e-07, "loss": 0.1565, "step": 8525 }, { "epoch": 1.61, "grad_norm": 11.375, "learning_rate": 4.752986933353429e-07, "loss": 0.2056, "step": 8526 }, { "epoch": 1.61, "grad_norm": 12.875, "learning_rate": 4.748494900150333e-07, "loss": 0.2047, "step": 8527 }, { "epoch": 1.61, "grad_norm": 13.5, "learning_rate": 4.744004767896046e-07, "loss": 0.2382, "step": 8528 }, { "epoch": 1.61, "grad_norm": 11.5625, "learning_rate": 4.73951653701204e-07, "loss": 0.1507, "step": 8529 }, { "epoch": 1.61, "grad_norm": 8.25, "learning_rate": 4.7350302079196167e-07, "loss": 0.1882, "step": 8530 }, { "epoch": 1.61, "grad_norm": 10.3125, "learning_rate": 4.7305457810398886e-07, "loss": 0.2065, "step": 8531 }, { "epoch": 1.61, "grad_norm": 14.4375, "learning_rate": 4.7260632567938004e-07, "loss": 0.2014, "step": 8532 }, { "epoch": 1.61, "grad_norm": 8.25, "learning_rate": 4.7215826356021093e-07, "loss": 0.1711, "step": 8533 }, { "epoch": 1.61, "grad_norm": 8.375, "learning_rate": 4.7171039178854016e-07, "loss": 0.1835, "step": 8534 }, { "epoch": 1.61, "grad_norm": 10.75, "learning_rate": 4.71262710406408e-07, "loss": 0.2524, "step": 8535 }, { "epoch": 1.61, "grad_norm": 14.875, "learning_rate": 4.7081521945583713e-07, "loss": 0.237, "step": 8536 }, { "epoch": 1.61, "grad_norm": 15.125, "learning_rate": 4.703679189788318e-07, "loss": 0.2277, "step": 8537 }, { "epoch": 1.62, "grad_norm": 9.9375, "learning_rate": 4.6992080901737904e-07, "loss": 0.222, "step": 8538 }, { "epoch": 1.62, "grad_norm": 11.375, "learning_rate": 4.694738896134479e-07, "loss": 0.1897, "step": 8539 }, { "epoch": 1.62, "grad_norm": 15.0, "learning_rate": 4.6902716080898887e-07, "loss": 0.2185, "step": 8540 }, { "epoch": 1.62, "grad_norm": 13.8125, "learning_rate": 4.6858062264593553e-07, "loss": 0.1898, "step": 8541 }, { "epoch": 1.62, "grad_norm": 16.75, "learning_rate": 4.6813427516620295e-07, "loss": 0.1778, "step": 8542 }, { "epoch": 1.62, "grad_norm": 10.4375, "learning_rate": 4.67688118411688e-07, "loss": 0.2488, "step": 8543 }, { "epoch": 1.62, "grad_norm": 11.625, "learning_rate": 4.672421524242715e-07, "loss": 0.1781, "step": 8544 }, { "epoch": 1.62, "grad_norm": 9.75, "learning_rate": 4.667963772458137e-07, "loss": 0.2195, "step": 8545 }, { "epoch": 1.62, "grad_norm": 8.5625, "learning_rate": 4.663507929181582e-07, "loss": 0.1884, "step": 8546 }, { "epoch": 1.62, "grad_norm": 9.25, "learning_rate": 4.6590539948313055e-07, "loss": 0.2388, "step": 8547 }, { "epoch": 1.62, "grad_norm": 16.375, "learning_rate": 4.6546019698254007e-07, "loss": 0.1982, "step": 8548 }, { "epoch": 1.62, "grad_norm": 11.875, "learning_rate": 4.650151854581747e-07, "loss": 0.2451, "step": 8549 }, { "epoch": 1.62, "grad_norm": 9.1875, "learning_rate": 4.645703649518074e-07, "loss": 0.1623, "step": 8550 }, { "epoch": 1.62, "grad_norm": 19.375, "learning_rate": 4.641257355051909e-07, "loss": 0.2219, "step": 8551 }, { "epoch": 1.62, "grad_norm": 15.375, "learning_rate": 4.636812971600638e-07, "loss": 0.202, "step": 8552 }, { "epoch": 1.62, "grad_norm": 13.875, "learning_rate": 4.632370499581418e-07, "loss": 0.1851, "step": 8553 }, { "epoch": 1.62, "grad_norm": 11.125, "learning_rate": 4.627929939411255e-07, "loss": 0.1696, "step": 8554 }, { "epoch": 1.62, "grad_norm": 12.0, "learning_rate": 4.6234912915069793e-07, "loss": 0.222, "step": 8555 }, { "epoch": 1.62, "grad_norm": 11.0, "learning_rate": 4.6190545562852377e-07, "loss": 0.2025, "step": 8556 }, { "epoch": 1.62, "grad_norm": 13.75, "learning_rate": 4.614619734162476e-07, "loss": 0.2514, "step": 8557 }, { "epoch": 1.62, "grad_norm": 13.75, "learning_rate": 4.6101868255549825e-07, "loss": 0.2488, "step": 8558 }, { "epoch": 1.62, "grad_norm": 10.125, "learning_rate": 4.60575583087888e-07, "loss": 0.1246, "step": 8559 }, { "epoch": 1.62, "grad_norm": 15.375, "learning_rate": 4.6013267505500695e-07, "loss": 0.209, "step": 8560 }, { "epoch": 1.62, "grad_norm": 14.5, "learning_rate": 4.5968995849843073e-07, "loss": 0.2085, "step": 8561 }, { "epoch": 1.62, "grad_norm": 10.9375, "learning_rate": 4.5924743345971495e-07, "loss": 0.1877, "step": 8562 }, { "epoch": 1.62, "grad_norm": 13.8125, "learning_rate": 4.5880509998039974e-07, "loss": 0.211, "step": 8563 }, { "epoch": 1.62, "grad_norm": 13.875, "learning_rate": 4.583629581020044e-07, "loss": 0.2299, "step": 8564 }, { "epoch": 1.62, "grad_norm": 7.5, "learning_rate": 4.5792100786603064e-07, "loss": 0.1309, "step": 8565 }, { "epoch": 1.62, "grad_norm": 19.625, "learning_rate": 4.5747924931396517e-07, "loss": 0.2218, "step": 8566 }, { "epoch": 1.62, "grad_norm": 11.5625, "learning_rate": 4.5703768248727384e-07, "loss": 0.1937, "step": 8567 }, { "epoch": 1.62, "grad_norm": 14.125, "learning_rate": 4.5659630742740435e-07, "loss": 0.2022, "step": 8568 }, { "epoch": 1.62, "grad_norm": 10.0625, "learning_rate": 4.561551241757872e-07, "loss": 0.2002, "step": 8569 }, { "epoch": 1.62, "grad_norm": 10.5, "learning_rate": 4.5571413277383623e-07, "loss": 0.166, "step": 8570 }, { "epoch": 1.62, "grad_norm": 10.4375, "learning_rate": 4.552733332629458e-07, "loss": 0.2345, "step": 8571 }, { "epoch": 1.62, "grad_norm": 12.875, "learning_rate": 4.5483272568449144e-07, "loss": 0.1763, "step": 8572 }, { "epoch": 1.62, "grad_norm": 13.25, "learning_rate": 4.543923100798314e-07, "loss": 0.1646, "step": 8573 }, { "epoch": 1.62, "grad_norm": 9.375, "learning_rate": 4.53952086490308e-07, "loss": 0.2351, "step": 8574 }, { "epoch": 1.62, "grad_norm": 13.8125, "learning_rate": 4.5351205495724303e-07, "loss": 0.1915, "step": 8575 }, { "epoch": 1.62, "grad_norm": 12.9375, "learning_rate": 4.5307221552193943e-07, "loss": 0.2192, "step": 8576 }, { "epoch": 1.62, "grad_norm": 12.0, "learning_rate": 4.5263256822568547e-07, "loss": 0.2158, "step": 8577 }, { "epoch": 1.62, "grad_norm": 10.375, "learning_rate": 4.52193113109749e-07, "loss": 0.1922, "step": 8578 }, { "epoch": 1.62, "grad_norm": 14.5625, "learning_rate": 4.5175385021538094e-07, "loss": 0.2093, "step": 8579 }, { "epoch": 1.62, "grad_norm": 22.875, "learning_rate": 4.513147795838116e-07, "loss": 0.2142, "step": 8580 }, { "epoch": 1.62, "grad_norm": 14.5625, "learning_rate": 4.508759012562572e-07, "loss": 0.1922, "step": 8581 }, { "epoch": 1.62, "grad_norm": 8.625, "learning_rate": 4.504372152739131e-07, "loss": 0.174, "step": 8582 }, { "epoch": 1.62, "grad_norm": 12.125, "learning_rate": 4.499987216779586e-07, "loss": 0.198, "step": 8583 }, { "epoch": 1.62, "grad_norm": 11.625, "learning_rate": 4.495604205095516e-07, "loss": 0.2318, "step": 8584 }, { "epoch": 1.62, "grad_norm": 13.375, "learning_rate": 4.4912231180983587e-07, "loss": 0.2135, "step": 8585 }, { "epoch": 1.62, "grad_norm": 14.9375, "learning_rate": 4.486843956199349e-07, "loss": 0.2979, "step": 8586 }, { "epoch": 1.62, "grad_norm": 12.9375, "learning_rate": 4.4824667198095473e-07, "loss": 0.202, "step": 8587 }, { "epoch": 1.62, "grad_norm": 11.6875, "learning_rate": 4.47809140933983e-07, "loss": 0.2205, "step": 8588 }, { "epoch": 1.62, "grad_norm": 12.6875, "learning_rate": 4.473718025200896e-07, "loss": 0.2181, "step": 8589 }, { "epoch": 1.62, "grad_norm": 12.3125, "learning_rate": 4.469346567803265e-07, "loss": 0.2, "step": 8590 }, { "epoch": 1.63, "grad_norm": 11.875, "learning_rate": 4.4649770375572583e-07, "loss": 0.1695, "step": 8591 }, { "epoch": 1.63, "grad_norm": 14.0625, "learning_rate": 4.460609434873045e-07, "loss": 0.2219, "step": 8592 }, { "epoch": 1.63, "grad_norm": 14.4375, "learning_rate": 4.456243760160597e-07, "loss": 0.1995, "step": 8593 }, { "epoch": 1.63, "grad_norm": 10.3125, "learning_rate": 4.4518800138297126e-07, "loss": 0.2414, "step": 8594 }, { "epoch": 1.63, "grad_norm": 10.25, "learning_rate": 4.4475181962899824e-07, "loss": 0.2035, "step": 8595 }, { "epoch": 1.63, "grad_norm": 12.0625, "learning_rate": 4.4431583079508565e-07, "loss": 0.1844, "step": 8596 }, { "epoch": 1.63, "grad_norm": 10.25, "learning_rate": 4.4388003492215836e-07, "loss": 0.1565, "step": 8597 }, { "epoch": 1.63, "grad_norm": 12.25, "learning_rate": 4.4344443205112245e-07, "loss": 0.1823, "step": 8598 }, { "epoch": 1.63, "grad_norm": 14.75, "learning_rate": 4.430090222228672e-07, "loss": 0.2712, "step": 8599 }, { "epoch": 1.63, "grad_norm": 15.625, "learning_rate": 4.425738054782633e-07, "loss": 0.2199, "step": 8600 }, { "epoch": 1.63, "grad_norm": 10.1875, "learning_rate": 4.4213878185816295e-07, "loss": 0.2749, "step": 8601 }, { "epoch": 1.63, "grad_norm": 9.3125, "learning_rate": 4.417039514034005e-07, "loss": 0.1948, "step": 8602 }, { "epoch": 1.63, "grad_norm": 14.25, "learning_rate": 4.4126931415479245e-07, "loss": 0.2514, "step": 8603 }, { "epoch": 1.63, "grad_norm": 9.9375, "learning_rate": 4.408348701531368e-07, "loss": 0.204, "step": 8604 }, { "epoch": 1.63, "grad_norm": 12.375, "learning_rate": 4.4040061943921355e-07, "loss": 0.1882, "step": 8605 }, { "epoch": 1.63, "grad_norm": 8.0625, "learning_rate": 4.3996656205378433e-07, "loss": 0.1941, "step": 8606 }, { "epoch": 1.63, "grad_norm": 14.0, "learning_rate": 4.3953269803759303e-07, "loss": 0.1774, "step": 8607 }, { "epoch": 1.63, "grad_norm": 10.875, "learning_rate": 4.3909902743136526e-07, "loss": 0.2024, "step": 8608 }, { "epoch": 1.63, "grad_norm": 12.0625, "learning_rate": 4.3866555027580833e-07, "loss": 0.1728, "step": 8609 }, { "epoch": 1.63, "grad_norm": 11.125, "learning_rate": 4.3823226661161125e-07, "loss": 0.2237, "step": 8610 }, { "epoch": 1.63, "grad_norm": 12.5625, "learning_rate": 4.3779917647944536e-07, "loss": 0.1839, "step": 8611 }, { "epoch": 1.63, "grad_norm": 15.625, "learning_rate": 4.373662799199635e-07, "loss": 0.2852, "step": 8612 }, { "epoch": 1.63, "grad_norm": 10.0, "learning_rate": 4.3693357697380004e-07, "loss": 0.1939, "step": 8613 }, { "epoch": 1.63, "grad_norm": 13.5, "learning_rate": 4.36501067681572e-07, "loss": 0.2187, "step": 8614 }, { "epoch": 1.63, "grad_norm": 8.875, "learning_rate": 4.360687520838774e-07, "loss": 0.1666, "step": 8615 }, { "epoch": 1.63, "grad_norm": 13.625, "learning_rate": 4.3563663022129663e-07, "loss": 0.2157, "step": 8616 }, { "epoch": 1.63, "grad_norm": 14.4375, "learning_rate": 4.3520470213439167e-07, "loss": 0.216, "step": 8617 }, { "epoch": 1.63, "grad_norm": 11.375, "learning_rate": 4.347729678637061e-07, "loss": 0.1708, "step": 8618 }, { "epoch": 1.63, "grad_norm": 10.3125, "learning_rate": 4.343414274497659e-07, "loss": 0.2177, "step": 8619 }, { "epoch": 1.63, "grad_norm": 13.875, "learning_rate": 4.339100809330782e-07, "loss": 0.2415, "step": 8620 }, { "epoch": 1.63, "grad_norm": 15.8125, "learning_rate": 4.334789283541324e-07, "loss": 0.1544, "step": 8621 }, { "epoch": 1.63, "grad_norm": 10.1875, "learning_rate": 4.3304796975339956e-07, "loss": 0.1886, "step": 8622 }, { "epoch": 1.63, "grad_norm": 12.125, "learning_rate": 4.3261720517133205e-07, "loss": 0.2215, "step": 8623 }, { "epoch": 1.63, "grad_norm": 8.8125, "learning_rate": 4.3218663464836494e-07, "loss": 0.176, "step": 8624 }, { "epoch": 1.63, "grad_norm": 8.625, "learning_rate": 4.3175625822491464e-07, "loss": 0.1494, "step": 8625 }, { "epoch": 1.63, "grad_norm": 13.1875, "learning_rate": 4.31326075941379e-07, "loss": 0.2278, "step": 8626 }, { "epoch": 1.63, "grad_norm": 9.1875, "learning_rate": 4.3089608783813796e-07, "loss": 0.2171, "step": 8627 }, { "epoch": 1.63, "grad_norm": 22.625, "learning_rate": 4.3046629395555345e-07, "loss": 0.1382, "step": 8628 }, { "epoch": 1.63, "grad_norm": 9.625, "learning_rate": 4.30036694333969e-07, "loss": 0.2158, "step": 8629 }, { "epoch": 1.63, "grad_norm": 12.5625, "learning_rate": 4.2960728901370955e-07, "loss": 0.16, "step": 8630 }, { "epoch": 1.63, "grad_norm": 8.875, "learning_rate": 4.2917807803508236e-07, "loss": 0.2165, "step": 8631 }, { "epoch": 1.63, "grad_norm": 8.4375, "learning_rate": 4.2874906143837603e-07, "loss": 0.1387, "step": 8632 }, { "epoch": 1.63, "grad_norm": 13.3125, "learning_rate": 4.2832023926386107e-07, "loss": 0.1701, "step": 8633 }, { "epoch": 1.63, "grad_norm": 9.125, "learning_rate": 4.278916115517898e-07, "loss": 0.222, "step": 8634 }, { "epoch": 1.63, "grad_norm": 11.25, "learning_rate": 4.2746317834239574e-07, "loss": 0.1741, "step": 8635 }, { "epoch": 1.63, "grad_norm": 11.5625, "learning_rate": 4.270349396758963e-07, "loss": 0.236, "step": 8636 }, { "epoch": 1.63, "grad_norm": 11.625, "learning_rate": 4.2660689559248736e-07, "loss": 0.2154, "step": 8637 }, { "epoch": 1.63, "grad_norm": 15.6875, "learning_rate": 4.261790461323487e-07, "loss": 0.2553, "step": 8638 }, { "epoch": 1.63, "grad_norm": 17.5, "learning_rate": 4.257513913356404e-07, "loss": 0.178, "step": 8639 }, { "epoch": 1.63, "grad_norm": 9.125, "learning_rate": 4.2532393124250725e-07, "loss": 0.1772, "step": 8640 }, { "epoch": 1.63, "grad_norm": 12.6875, "learning_rate": 4.248966658930717e-07, "loss": 0.2653, "step": 8641 }, { "epoch": 1.63, "grad_norm": 14.875, "learning_rate": 4.244695953274408e-07, "loss": 0.2095, "step": 8642 }, { "epoch": 1.64, "grad_norm": 18.875, "learning_rate": 4.2404271958570194e-07, "loss": 0.2879, "step": 8643 }, { "epoch": 1.64, "grad_norm": 8.875, "learning_rate": 4.2361603870792494e-07, "loss": 0.1793, "step": 8644 }, { "epoch": 1.64, "grad_norm": 11.8125, "learning_rate": 4.231895527341609e-07, "loss": 0.1891, "step": 8645 }, { "epoch": 1.64, "grad_norm": 10.0, "learning_rate": 4.227632617044425e-07, "loss": 0.2218, "step": 8646 }, { "epoch": 1.64, "grad_norm": 12.0625, "learning_rate": 4.22337165658786e-07, "loss": 0.1953, "step": 8647 }, { "epoch": 1.64, "grad_norm": 21.375, "learning_rate": 4.219112646371859e-07, "loss": 0.1997, "step": 8648 }, { "epoch": 1.64, "grad_norm": 8.4375, "learning_rate": 4.21485558679621e-07, "loss": 0.164, "step": 8649 }, { "epoch": 1.64, "grad_norm": 7.96875, "learning_rate": 4.210600478260507e-07, "loss": 0.1748, "step": 8650 }, { "epoch": 1.64, "grad_norm": 12.625, "learning_rate": 4.206347321164178e-07, "loss": 0.2317, "step": 8651 }, { "epoch": 1.64, "grad_norm": 10.1875, "learning_rate": 4.2020961159064374e-07, "loss": 0.1836, "step": 8652 }, { "epoch": 1.64, "grad_norm": 9.375, "learning_rate": 4.19784686288634e-07, "loss": 0.1864, "step": 8653 }, { "epoch": 1.64, "grad_norm": 13.375, "learning_rate": 4.1935995625027454e-07, "loss": 0.1638, "step": 8654 }, { "epoch": 1.64, "grad_norm": 11.25, "learning_rate": 4.189354215154351e-07, "loss": 0.212, "step": 8655 }, { "epoch": 1.64, "grad_norm": 7.125, "learning_rate": 4.185110821239641e-07, "loss": 0.1602, "step": 8656 }, { "epoch": 1.64, "grad_norm": 17.75, "learning_rate": 4.180869381156924e-07, "loss": 0.1495, "step": 8657 }, { "epoch": 1.64, "grad_norm": 14.375, "learning_rate": 4.176629895304349e-07, "loss": 0.2209, "step": 8658 }, { "epoch": 1.64, "grad_norm": 12.6875, "learning_rate": 4.172392364079861e-07, "loss": 0.2088, "step": 8659 }, { "epoch": 1.64, "grad_norm": 9.5, "learning_rate": 4.168156787881211e-07, "loss": 0.1488, "step": 8660 }, { "epoch": 1.64, "grad_norm": 16.5, "learning_rate": 4.163923167105985e-07, "loss": 0.2365, "step": 8661 }, { "epoch": 1.64, "grad_norm": 11.9375, "learning_rate": 4.1596915021515915e-07, "loss": 0.1929, "step": 8662 }, { "epoch": 1.64, "grad_norm": 10.1875, "learning_rate": 4.1554617934152384e-07, "loss": 0.1803, "step": 8663 }, { "epoch": 1.64, "grad_norm": 15.0625, "learning_rate": 4.151234041293953e-07, "loss": 0.2313, "step": 8664 }, { "epoch": 1.64, "grad_norm": 9.6875, "learning_rate": 4.1470082461845724e-07, "loss": 0.1546, "step": 8665 }, { "epoch": 1.64, "grad_norm": 12.125, "learning_rate": 4.1427844084837796e-07, "loss": 0.2477, "step": 8666 }, { "epoch": 1.64, "grad_norm": 12.25, "learning_rate": 4.1385625285880473e-07, "loss": 0.2176, "step": 8667 }, { "epoch": 1.64, "grad_norm": 11.6875, "learning_rate": 4.134342606893657e-07, "loss": 0.201, "step": 8668 }, { "epoch": 1.64, "grad_norm": 10.4375, "learning_rate": 4.1301246437967374e-07, "loss": 0.171, "step": 8669 }, { "epoch": 1.64, "grad_norm": 11.8125, "learning_rate": 4.12590863969321e-07, "loss": 0.2068, "step": 8670 }, { "epoch": 1.64, "grad_norm": 10.8125, "learning_rate": 4.1216945949788254e-07, "loss": 0.1529, "step": 8671 }, { "epoch": 1.64, "grad_norm": 9.8125, "learning_rate": 4.117482510049123e-07, "loss": 0.1744, "step": 8672 }, { "epoch": 1.64, "grad_norm": 13.875, "learning_rate": 4.1132723852994966e-07, "loss": 0.2637, "step": 8673 }, { "epoch": 1.64, "grad_norm": 23.5, "learning_rate": 4.1090642211251447e-07, "loss": 0.2223, "step": 8674 }, { "epoch": 1.64, "grad_norm": 8.5625, "learning_rate": 4.104858017921054e-07, "loss": 0.2116, "step": 8675 }, { "epoch": 1.64, "grad_norm": 13.0, "learning_rate": 4.100653776082053e-07, "loss": 0.1315, "step": 8676 }, { "epoch": 1.64, "grad_norm": 11.75, "learning_rate": 4.0964514960027954e-07, "loss": 0.1795, "step": 8677 }, { "epoch": 1.64, "grad_norm": 8.375, "learning_rate": 4.092251178077733e-07, "loss": 0.1656, "step": 8678 }, { "epoch": 1.64, "grad_norm": 11.125, "learning_rate": 4.088052822701122e-07, "loss": 0.2311, "step": 8679 }, { "epoch": 1.64, "grad_norm": 15.1875, "learning_rate": 4.083856430267066e-07, "loss": 0.207, "step": 8680 }, { "epoch": 1.64, "grad_norm": 13.4375, "learning_rate": 4.07966200116946e-07, "loss": 0.1952, "step": 8681 }, { "epoch": 1.64, "grad_norm": 13.8125, "learning_rate": 4.0754695358020347e-07, "loss": 0.2147, "step": 8682 }, { "epoch": 1.64, "grad_norm": 13.4375, "learning_rate": 4.071279034558301e-07, "loss": 0.2392, "step": 8683 }, { "epoch": 1.64, "grad_norm": 9.625, "learning_rate": 4.067090497831627e-07, "loss": 0.1934, "step": 8684 }, { "epoch": 1.64, "grad_norm": 9.8125, "learning_rate": 4.0629039260151744e-07, "loss": 0.1699, "step": 8685 }, { "epoch": 1.64, "grad_norm": 10.5625, "learning_rate": 4.05871931950193e-07, "loss": 0.1969, "step": 8686 }, { "epoch": 1.64, "grad_norm": 11.25, "learning_rate": 4.05453667868467e-07, "loss": 0.2139, "step": 8687 }, { "epoch": 1.64, "grad_norm": 11.25, "learning_rate": 4.05035600395603e-07, "loss": 0.2095, "step": 8688 }, { "epoch": 1.64, "grad_norm": 12.625, "learning_rate": 4.0461772957084256e-07, "loss": 0.2274, "step": 8689 }, { "epoch": 1.64, "grad_norm": 12.6875, "learning_rate": 4.042000554334102e-07, "loss": 0.2148, "step": 8690 }, { "epoch": 1.64, "grad_norm": 10.75, "learning_rate": 4.037825780225119e-07, "loss": 0.1789, "step": 8691 }, { "epoch": 1.64, "grad_norm": 10.4375, "learning_rate": 4.03365297377335e-07, "loss": 0.1884, "step": 8692 }, { "epoch": 1.64, "grad_norm": 10.75, "learning_rate": 4.0294821353704797e-07, "loss": 0.2045, "step": 8693 }, { "epoch": 1.64, "grad_norm": 10.6875, "learning_rate": 4.025313265408018e-07, "loss": 0.2146, "step": 8694 }, { "epoch": 1.64, "grad_norm": 9.8125, "learning_rate": 4.021146364277281e-07, "loss": 0.1827, "step": 8695 }, { "epoch": 1.65, "grad_norm": 10.25, "learning_rate": 4.0169814323694046e-07, "loss": 0.2015, "step": 8696 }, { "epoch": 1.65, "grad_norm": 10.25, "learning_rate": 4.0128184700753386e-07, "loss": 0.205, "step": 8697 }, { "epoch": 1.65, "grad_norm": 14.5625, "learning_rate": 4.008657477785849e-07, "loss": 0.2502, "step": 8698 }, { "epoch": 1.65, "grad_norm": 11.3125, "learning_rate": 4.0044984558915147e-07, "loss": 0.176, "step": 8699 }, { "epoch": 1.65, "grad_norm": 11.5625, "learning_rate": 4.0003414047827286e-07, "loss": 0.162, "step": 8700 }, { "epoch": 1.65, "grad_norm": 11.1875, "learning_rate": 3.9961863248497045e-07, "loss": 0.2772, "step": 8701 }, { "epoch": 1.65, "grad_norm": 12.6875, "learning_rate": 3.9920332164824634e-07, "loss": 0.2279, "step": 8702 }, { "epoch": 1.65, "grad_norm": 9.8125, "learning_rate": 3.987882080070851e-07, "loss": 0.2032, "step": 8703 }, { "epoch": 1.65, "grad_norm": 12.1875, "learning_rate": 3.983732916004515e-07, "loss": 0.2202, "step": 8704 }, { "epoch": 1.65, "grad_norm": 11.375, "learning_rate": 3.9795857246729326e-07, "loss": 0.2158, "step": 8705 }, { "epoch": 1.65, "grad_norm": 11.875, "learning_rate": 3.975440506465383e-07, "loss": 0.2007, "step": 8706 }, { "epoch": 1.65, "grad_norm": 14.0625, "learning_rate": 3.9712972617709677e-07, "loss": 0.2458, "step": 8707 }, { "epoch": 1.65, "grad_norm": 9.375, "learning_rate": 3.9671559909785993e-07, "loss": 0.2029, "step": 8708 }, { "epoch": 1.65, "grad_norm": 10.4375, "learning_rate": 3.9630166944770095e-07, "loss": 0.1938, "step": 8709 }, { "epoch": 1.65, "grad_norm": 13.25, "learning_rate": 3.958879372654739e-07, "loss": 0.218, "step": 8710 }, { "epoch": 1.65, "grad_norm": 10.875, "learning_rate": 3.954744025900148e-07, "loss": 0.208, "step": 8711 }, { "epoch": 1.65, "grad_norm": 10.3125, "learning_rate": 3.9506106546014085e-07, "loss": 0.2163, "step": 8712 }, { "epoch": 1.65, "grad_norm": 10.8125, "learning_rate": 3.9464792591465047e-07, "loss": 0.1978, "step": 8713 }, { "epoch": 1.65, "grad_norm": 11.1875, "learning_rate": 3.9423498399232427e-07, "loss": 0.1782, "step": 8714 }, { "epoch": 1.65, "grad_norm": 12.5625, "learning_rate": 3.9382223973192385e-07, "loss": 0.2169, "step": 8715 }, { "epoch": 1.65, "grad_norm": 13.875, "learning_rate": 3.934096931721923e-07, "loss": 0.2692, "step": 8716 }, { "epoch": 1.65, "grad_norm": 10.0, "learning_rate": 3.929973443518539e-07, "loss": 0.1746, "step": 8717 }, { "epoch": 1.65, "grad_norm": 11.6875, "learning_rate": 3.9258519330961477e-07, "loss": 0.24, "step": 8718 }, { "epoch": 1.65, "grad_norm": 13.1875, "learning_rate": 3.9217324008416214e-07, "loss": 0.2549, "step": 8719 }, { "epoch": 1.65, "grad_norm": 10.9375, "learning_rate": 3.9176148471416513e-07, "loss": 0.2139, "step": 8720 }, { "epoch": 1.65, "grad_norm": 11.8125, "learning_rate": 3.9134992723827367e-07, "loss": 0.214, "step": 8721 }, { "epoch": 1.65, "grad_norm": 15.375, "learning_rate": 3.9093856769511994e-07, "loss": 0.295, "step": 8722 }, { "epoch": 1.65, "grad_norm": 19.5, "learning_rate": 3.905274061233158e-07, "loss": 0.2289, "step": 8723 }, { "epoch": 1.65, "grad_norm": 12.0, "learning_rate": 3.9011644256145813e-07, "loss": 0.2508, "step": 8724 }, { "epoch": 1.65, "grad_norm": 10.5625, "learning_rate": 3.897056770481206e-07, "loss": 0.2038, "step": 8725 }, { "epoch": 1.65, "grad_norm": 11.9375, "learning_rate": 3.892951096218617e-07, "loss": 0.2374, "step": 8726 }, { "epoch": 1.65, "grad_norm": 10.0625, "learning_rate": 3.888847403212195e-07, "loss": 0.1859, "step": 8727 }, { "epoch": 1.65, "grad_norm": 12.25, "learning_rate": 3.884745691847147e-07, "loss": 0.1795, "step": 8728 }, { "epoch": 1.65, "grad_norm": 8.625, "learning_rate": 3.8806459625084864e-07, "loss": 0.1805, "step": 8729 }, { "epoch": 1.65, "grad_norm": 15.0625, "learning_rate": 3.8765482155810416e-07, "loss": 0.1261, "step": 8730 }, { "epoch": 1.65, "grad_norm": 8.375, "learning_rate": 3.87245245144946e-07, "loss": 0.1459, "step": 8731 }, { "epoch": 1.65, "grad_norm": 12.75, "learning_rate": 3.868358670498193e-07, "loss": 0.2442, "step": 8732 }, { "epoch": 1.65, "grad_norm": 14.0, "learning_rate": 3.864266873111516e-07, "loss": 0.1719, "step": 8733 }, { "epoch": 1.65, "grad_norm": 13.125, "learning_rate": 3.860177059673506e-07, "loss": 0.2025, "step": 8734 }, { "epoch": 1.65, "grad_norm": 9.375, "learning_rate": 3.856089230568083e-07, "loss": 0.229, "step": 8735 }, { "epoch": 1.65, "grad_norm": 12.0, "learning_rate": 3.852003386178935e-07, "loss": 0.229, "step": 8736 }, { "epoch": 1.65, "grad_norm": 12.9375, "learning_rate": 3.847919526889601e-07, "loss": 0.2586, "step": 8737 }, { "epoch": 1.65, "grad_norm": 17.375, "learning_rate": 3.8438376530834096e-07, "loss": 0.2025, "step": 8738 }, { "epoch": 1.65, "grad_norm": 12.1875, "learning_rate": 3.839757765143534e-07, "loss": 0.176, "step": 8739 }, { "epoch": 1.65, "grad_norm": 9.8125, "learning_rate": 3.8356798634529233e-07, "loss": 0.2371, "step": 8740 }, { "epoch": 1.65, "grad_norm": 12.9375, "learning_rate": 3.8316039483943665e-07, "loss": 0.239, "step": 8741 }, { "epoch": 1.65, "grad_norm": 16.875, "learning_rate": 3.8275300203504476e-07, "loss": 0.2137, "step": 8742 }, { "epoch": 1.65, "grad_norm": 13.625, "learning_rate": 3.823458079703593e-07, "loss": 0.1937, "step": 8743 }, { "epoch": 1.65, "grad_norm": 8.625, "learning_rate": 3.819388126836007e-07, "loss": 0.1804, "step": 8744 }, { "epoch": 1.65, "grad_norm": 16.875, "learning_rate": 3.8153201621297263e-07, "loss": 0.2264, "step": 8745 }, { "epoch": 1.65, "grad_norm": 8.1875, "learning_rate": 3.811254185966606e-07, "loss": 0.1844, "step": 8746 }, { "epoch": 1.65, "grad_norm": 15.75, "learning_rate": 3.8071901987283084e-07, "loss": 0.2333, "step": 8747 }, { "epoch": 1.65, "grad_norm": 12.125, "learning_rate": 3.8031282007962995e-07, "loss": 0.1522, "step": 8748 }, { "epoch": 1.66, "grad_norm": 14.875, "learning_rate": 3.7990681925518606e-07, "loss": 0.2484, "step": 8749 }, { "epoch": 1.66, "grad_norm": 12.125, "learning_rate": 3.795010174376113e-07, "loss": 0.2177, "step": 8750 }, { "epoch": 1.66, "grad_norm": 16.125, "learning_rate": 3.7909541466499613e-07, "loss": 0.2332, "step": 8751 }, { "epoch": 1.66, "grad_norm": 12.1875, "learning_rate": 3.7869001097541296e-07, "loss": 0.2076, "step": 8752 }, { "epoch": 1.66, "grad_norm": 11.9375, "learning_rate": 3.782848064069153e-07, "loss": 0.1946, "step": 8753 }, { "epoch": 1.66, "grad_norm": 11.5, "learning_rate": 3.7787980099753986e-07, "loss": 0.1579, "step": 8754 }, { "epoch": 1.66, "grad_norm": 13.625, "learning_rate": 3.774749947853032e-07, "loss": 0.2275, "step": 8755 }, { "epoch": 1.66, "grad_norm": 9.625, "learning_rate": 3.770703878082016e-07, "loss": 0.1954, "step": 8756 }, { "epoch": 1.66, "grad_norm": 8.375, "learning_rate": 3.7666598010421605e-07, "loss": 0.1877, "step": 8757 }, { "epoch": 1.66, "grad_norm": 10.875, "learning_rate": 3.7626177171130733e-07, "loss": 0.1783, "step": 8758 }, { "epoch": 1.66, "grad_norm": 10.875, "learning_rate": 3.7585776266741565e-07, "loss": 0.2204, "step": 8759 }, { "epoch": 1.66, "grad_norm": 17.0, "learning_rate": 3.754539530104645e-07, "loss": 0.2248, "step": 8760 }, { "epoch": 1.66, "grad_norm": 13.9375, "learning_rate": 3.7505034277835947e-07, "loss": 0.2235, "step": 8761 }, { "epoch": 1.66, "grad_norm": 13.8125, "learning_rate": 3.7464693200898625e-07, "loss": 0.2121, "step": 8762 }, { "epoch": 1.66, "grad_norm": 11.6875, "learning_rate": 3.742437207402103e-07, "loss": 0.2005, "step": 8763 }, { "epoch": 1.66, "grad_norm": 10.4375, "learning_rate": 3.7384070900988026e-07, "loss": 0.191, "step": 8764 }, { "epoch": 1.66, "grad_norm": 10.5, "learning_rate": 3.7343789685582666e-07, "loss": 0.1562, "step": 8765 }, { "epoch": 1.66, "grad_norm": 11.3125, "learning_rate": 3.7303528431586023e-07, "loss": 0.1878, "step": 8766 }, { "epoch": 1.66, "grad_norm": 9.375, "learning_rate": 3.7263287142777134e-07, "loss": 0.227, "step": 8767 }, { "epoch": 1.66, "grad_norm": 9.25, "learning_rate": 3.72230658229335e-07, "loss": 0.1545, "step": 8768 }, { "epoch": 1.66, "grad_norm": 9.5, "learning_rate": 3.718286447583055e-07, "loss": 0.1953, "step": 8769 }, { "epoch": 1.66, "grad_norm": 11.0625, "learning_rate": 3.714268310524188e-07, "loss": 0.2659, "step": 8770 }, { "epoch": 1.66, "grad_norm": 10.1875, "learning_rate": 3.7102521714939027e-07, "loss": 0.1615, "step": 8771 }, { "epoch": 1.66, "grad_norm": 13.0625, "learning_rate": 3.7062380308692024e-07, "loss": 0.1574, "step": 8772 }, { "epoch": 1.66, "grad_norm": 9.375, "learning_rate": 3.702225889026875e-07, "loss": 0.1849, "step": 8773 }, { "epoch": 1.66, "grad_norm": 9.5, "learning_rate": 3.6982157463435335e-07, "loss": 0.1502, "step": 8774 }, { "epoch": 1.66, "grad_norm": 12.6875, "learning_rate": 3.694207603195582e-07, "loss": 0.2043, "step": 8775 }, { "epoch": 1.66, "grad_norm": 10.625, "learning_rate": 3.6902014599592694e-07, "loss": 0.2091, "step": 8776 }, { "epoch": 1.66, "grad_norm": 10.25, "learning_rate": 3.686197317010634e-07, "loss": 0.1659, "step": 8777 }, { "epoch": 1.66, "grad_norm": 12.75, "learning_rate": 3.6821951747255364e-07, "loss": 0.2335, "step": 8778 }, { "epoch": 1.66, "grad_norm": 10.4375, "learning_rate": 3.678195033479642e-07, "loss": 0.2306, "step": 8779 }, { "epoch": 1.66, "grad_norm": 9.375, "learning_rate": 3.6741968936484353e-07, "loss": 0.215, "step": 8780 }, { "epoch": 1.66, "grad_norm": 9.75, "learning_rate": 3.670200755607209e-07, "loss": 0.1896, "step": 8781 }, { "epoch": 1.66, "grad_norm": 11.375, "learning_rate": 3.6662066197310665e-07, "loss": 0.1711, "step": 8782 }, { "epoch": 1.66, "grad_norm": 8.0625, "learning_rate": 3.6622144863949274e-07, "loss": 0.1542, "step": 8783 }, { "epoch": 1.66, "grad_norm": 11.6875, "learning_rate": 3.658224355973519e-07, "loss": 0.1903, "step": 8784 }, { "epoch": 1.66, "grad_norm": 13.5625, "learning_rate": 3.654236228841393e-07, "loss": 0.2432, "step": 8785 }, { "epoch": 1.66, "grad_norm": 10.9375, "learning_rate": 3.6502501053728823e-07, "loss": 0.1778, "step": 8786 }, { "epoch": 1.66, "grad_norm": 12.125, "learning_rate": 3.646265985942174e-07, "loss": 0.1556, "step": 8787 }, { "epoch": 1.66, "grad_norm": 13.0625, "learning_rate": 3.6422838709232357e-07, "loss": 0.1792, "step": 8788 }, { "epoch": 1.66, "grad_norm": 10.0, "learning_rate": 3.638303760689857e-07, "loss": 0.2071, "step": 8789 }, { "epoch": 1.66, "grad_norm": 12.0, "learning_rate": 3.634325655615639e-07, "loss": 0.2219, "step": 8790 }, { "epoch": 1.66, "grad_norm": 12.0, "learning_rate": 3.6303495560739977e-07, "loss": 0.2407, "step": 8791 }, { "epoch": 1.66, "grad_norm": 13.375, "learning_rate": 3.626375462438156e-07, "loss": 0.1987, "step": 8792 }, { "epoch": 1.66, "grad_norm": 12.8125, "learning_rate": 3.622403375081149e-07, "loss": 0.2589, "step": 8793 }, { "epoch": 1.66, "grad_norm": 14.625, "learning_rate": 3.618433294375828e-07, "loss": 0.2057, "step": 8794 }, { "epoch": 1.66, "grad_norm": 9.875, "learning_rate": 3.614465220694849e-07, "loss": 0.2168, "step": 8795 }, { "epoch": 1.66, "grad_norm": 9.5, "learning_rate": 3.610499154410685e-07, "loss": 0.182, "step": 8796 }, { "epoch": 1.66, "grad_norm": 9.25, "learning_rate": 3.606535095895619e-07, "loss": 0.1919, "step": 8797 }, { "epoch": 1.66, "grad_norm": 14.0625, "learning_rate": 3.6025730455217457e-07, "loss": 0.1991, "step": 8798 }, { "epoch": 1.66, "grad_norm": 12.375, "learning_rate": 3.5986130036609696e-07, "loss": 0.1821, "step": 8799 }, { "epoch": 1.66, "grad_norm": 8.8125, "learning_rate": 3.5946549706850077e-07, "loss": 0.1602, "step": 8800 }, { "epoch": 1.66, "grad_norm": 12.5, "learning_rate": 3.590698946965393e-07, "loss": 0.1811, "step": 8801 }, { "epoch": 1.67, "grad_norm": 17.125, "learning_rate": 3.586744932873462e-07, "loss": 0.2004, "step": 8802 }, { "epoch": 1.67, "grad_norm": 11.375, "learning_rate": 3.582792928780368e-07, "loss": 0.194, "step": 8803 }, { "epoch": 1.67, "grad_norm": 11.5, "learning_rate": 3.5788429350570724e-07, "loss": 0.2104, "step": 8804 }, { "epoch": 1.67, "grad_norm": 10.1875, "learning_rate": 3.574894952074348e-07, "loss": 0.2325, "step": 8805 }, { "epoch": 1.67, "grad_norm": 11.6875, "learning_rate": 3.5709489802027877e-07, "loss": 0.1868, "step": 8806 }, { "epoch": 1.67, "grad_norm": 10.75, "learning_rate": 3.5670050198127775e-07, "loss": 0.205, "step": 8807 }, { "epoch": 1.67, "grad_norm": 14.625, "learning_rate": 3.5630630712745335e-07, "loss": 0.2526, "step": 8808 }, { "epoch": 1.67, "grad_norm": 14.9375, "learning_rate": 3.559123134958073e-07, "loss": 0.2345, "step": 8809 }, { "epoch": 1.67, "grad_norm": 10.25, "learning_rate": 3.5551852112332226e-07, "loss": 0.2095, "step": 8810 }, { "epoch": 1.67, "grad_norm": 7.46875, "learning_rate": 3.551249300469628e-07, "loss": 0.1639, "step": 8811 }, { "epoch": 1.67, "grad_norm": 11.375, "learning_rate": 3.5473154030367383e-07, "loss": 0.2276, "step": 8812 }, { "epoch": 1.67, "grad_norm": 11.6875, "learning_rate": 3.54338351930382e-07, "loss": 0.1769, "step": 8813 }, { "epoch": 1.67, "grad_norm": 16.0, "learning_rate": 3.539453649639943e-07, "loss": 0.2031, "step": 8814 }, { "epoch": 1.67, "grad_norm": 12.125, "learning_rate": 3.5355257944139964e-07, "loss": 0.1966, "step": 8815 }, { "epoch": 1.67, "grad_norm": 11.125, "learning_rate": 3.531599953994674e-07, "loss": 0.2368, "step": 8816 }, { "epoch": 1.67, "grad_norm": 10.125, "learning_rate": 3.527676128750485e-07, "loss": 0.1988, "step": 8817 }, { "epoch": 1.67, "grad_norm": 14.1875, "learning_rate": 3.523754319049744e-07, "loss": 0.2419, "step": 8818 }, { "epoch": 1.67, "grad_norm": 9.625, "learning_rate": 3.519834525260582e-07, "loss": 0.1827, "step": 8819 }, { "epoch": 1.67, "grad_norm": 12.25, "learning_rate": 3.515916747750939e-07, "loss": 0.217, "step": 8820 }, { "epoch": 1.67, "grad_norm": 14.5625, "learning_rate": 3.512000986888564e-07, "loss": 0.2182, "step": 8821 }, { "epoch": 1.67, "grad_norm": 13.875, "learning_rate": 3.508087243041014e-07, "loss": 0.1765, "step": 8822 }, { "epoch": 1.67, "grad_norm": 8.8125, "learning_rate": 3.5041755165756724e-07, "loss": 0.1763, "step": 8823 }, { "epoch": 1.67, "grad_norm": 9.9375, "learning_rate": 3.5002658078597124e-07, "loss": 0.2119, "step": 8824 }, { "epoch": 1.67, "grad_norm": 11.75, "learning_rate": 3.4963581172601237e-07, "loss": 0.1922, "step": 8825 }, { "epoch": 1.67, "grad_norm": 12.9375, "learning_rate": 3.4924524451437096e-07, "loss": 0.2697, "step": 8826 }, { "epoch": 1.67, "grad_norm": 10.125, "learning_rate": 3.4885487918771025e-07, "loss": 0.2306, "step": 8827 }, { "epoch": 1.67, "grad_norm": 14.4375, "learning_rate": 3.484647157826704e-07, "loss": 0.1928, "step": 8828 }, { "epoch": 1.67, "grad_norm": 18.25, "learning_rate": 3.480747543358759e-07, "loss": 0.194, "step": 8829 }, { "epoch": 1.67, "grad_norm": 17.5, "learning_rate": 3.476849948839303e-07, "loss": 0.1718, "step": 8830 }, { "epoch": 1.67, "grad_norm": 8.9375, "learning_rate": 3.472954374634213e-07, "loss": 0.2055, "step": 8831 }, { "epoch": 1.67, "grad_norm": 11.5, "learning_rate": 3.4690608211091383e-07, "loss": 0.2218, "step": 8832 }, { "epoch": 1.67, "grad_norm": 9.125, "learning_rate": 3.465169288629555e-07, "loss": 0.1852, "step": 8833 }, { "epoch": 1.67, "grad_norm": 13.5, "learning_rate": 3.461279777560753e-07, "loss": 0.2133, "step": 8834 }, { "epoch": 1.67, "grad_norm": 15.6875, "learning_rate": 3.4573922882678354e-07, "loss": 0.1933, "step": 8835 }, { "epoch": 1.67, "grad_norm": 10.75, "learning_rate": 3.453506821115701e-07, "loss": 0.2013, "step": 8836 }, { "epoch": 1.67, "grad_norm": 11.625, "learning_rate": 3.4496233764690664e-07, "loss": 0.2138, "step": 8837 }, { "epoch": 1.67, "grad_norm": 9.1875, "learning_rate": 3.4457419546924634e-07, "loss": 0.1788, "step": 8838 }, { "epoch": 1.67, "grad_norm": 14.125, "learning_rate": 3.4418625561502383e-07, "loss": 0.243, "step": 8839 }, { "epoch": 1.67, "grad_norm": 15.0625, "learning_rate": 3.4379851812065213e-07, "loss": 0.2157, "step": 8840 }, { "epoch": 1.67, "grad_norm": 19.25, "learning_rate": 3.434109830225271e-07, "loss": 0.1873, "step": 8841 }, { "epoch": 1.67, "grad_norm": 15.5, "learning_rate": 3.4302365035702753e-07, "loss": 0.2005, "step": 8842 }, { "epoch": 1.67, "grad_norm": 9.9375, "learning_rate": 3.426365201605095e-07, "loss": 0.1807, "step": 8843 }, { "epoch": 1.67, "grad_norm": 14.25, "learning_rate": 3.422495924693117e-07, "loss": 0.259, "step": 8844 }, { "epoch": 1.67, "grad_norm": 11.625, "learning_rate": 3.418628673197541e-07, "loss": 0.2166, "step": 8845 }, { "epoch": 1.67, "grad_norm": 13.8125, "learning_rate": 3.4147634474813866e-07, "loss": 0.1712, "step": 8846 }, { "epoch": 1.67, "grad_norm": 11.5625, "learning_rate": 3.410900247907453e-07, "loss": 0.2528, "step": 8847 }, { "epoch": 1.67, "grad_norm": 7.9375, "learning_rate": 3.4070390748383725e-07, "loss": 0.2059, "step": 8848 }, { "epoch": 1.67, "grad_norm": 10.5, "learning_rate": 3.403179928636591e-07, "loss": 0.2163, "step": 8849 }, { "epoch": 1.67, "grad_norm": 18.75, "learning_rate": 3.3993228096643544e-07, "loss": 0.1958, "step": 8850 }, { "epoch": 1.67, "grad_norm": 9.625, "learning_rate": 3.395467718283707e-07, "loss": 0.1306, "step": 8851 }, { "epoch": 1.67, "grad_norm": 12.5, "learning_rate": 3.391614654856518e-07, "loss": 0.2368, "step": 8852 }, { "epoch": 1.67, "grad_norm": 10.875, "learning_rate": 3.3877636197444693e-07, "loss": 0.188, "step": 8853 }, { "epoch": 1.67, "grad_norm": 13.125, "learning_rate": 3.38391461330905e-07, "loss": 0.1906, "step": 8854 }, { "epoch": 1.68, "grad_norm": 12.1875, "learning_rate": 3.3800676359115436e-07, "loss": 0.2192, "step": 8855 }, { "epoch": 1.68, "grad_norm": 13.6875, "learning_rate": 3.3762226879130553e-07, "loss": 0.2128, "step": 8856 }, { "epoch": 1.68, "grad_norm": 10.5625, "learning_rate": 3.372379769674505e-07, "loss": 0.2296, "step": 8857 }, { "epoch": 1.68, "grad_norm": 9.3125, "learning_rate": 3.3685388815566213e-07, "loss": 0.182, "step": 8858 }, { "epoch": 1.68, "grad_norm": 14.625, "learning_rate": 3.364700023919917e-07, "loss": 0.1933, "step": 8859 }, { "epoch": 1.68, "grad_norm": 9.375, "learning_rate": 3.360863197124756e-07, "loss": 0.1457, "step": 8860 }, { "epoch": 1.68, "grad_norm": 10.375, "learning_rate": 3.357028401531279e-07, "loss": 0.1531, "step": 8861 }, { "epoch": 1.68, "grad_norm": 9.5625, "learning_rate": 3.353195637499454e-07, "loss": 0.1481, "step": 8862 }, { "epoch": 1.68, "grad_norm": 14.75, "learning_rate": 3.3493649053890325e-07, "loss": 0.2199, "step": 8863 }, { "epoch": 1.68, "grad_norm": 9.0625, "learning_rate": 3.3455362055596155e-07, "loss": 0.1811, "step": 8864 }, { "epoch": 1.68, "grad_norm": 14.0, "learning_rate": 3.341709538370583e-07, "loss": 0.2263, "step": 8865 }, { "epoch": 1.68, "grad_norm": 13.25, "learning_rate": 3.3378849041811405e-07, "loss": 0.2092, "step": 8866 }, { "epoch": 1.68, "grad_norm": 11.3125, "learning_rate": 3.3340623033502774e-07, "loss": 0.1869, "step": 8867 }, { "epoch": 1.68, "grad_norm": 10.3125, "learning_rate": 3.330241736236822e-07, "loss": 0.1642, "step": 8868 }, { "epoch": 1.68, "grad_norm": 12.4375, "learning_rate": 3.326423203199408e-07, "loss": 0.2423, "step": 8869 }, { "epoch": 1.68, "grad_norm": 10.0, "learning_rate": 3.3226067045964484e-07, "loss": 0.1983, "step": 8870 }, { "epoch": 1.68, "grad_norm": 12.0, "learning_rate": 3.3187922407862035e-07, "loss": 0.2375, "step": 8871 }, { "epoch": 1.68, "grad_norm": 28.75, "learning_rate": 3.314979812126717e-07, "loss": 0.2126, "step": 8872 }, { "epoch": 1.68, "grad_norm": 8.3125, "learning_rate": 3.311169418975865e-07, "loss": 0.1952, "step": 8873 }, { "epoch": 1.68, "grad_norm": 9.25, "learning_rate": 3.3073610616912915e-07, "loss": 0.1615, "step": 8874 }, { "epoch": 1.68, "grad_norm": 10.5625, "learning_rate": 3.303554740630499e-07, "loss": 0.1641, "step": 8875 }, { "epoch": 1.68, "grad_norm": 9.3125, "learning_rate": 3.299750456150766e-07, "loss": 0.191, "step": 8876 }, { "epoch": 1.68, "grad_norm": 13.125, "learning_rate": 3.2959482086091957e-07, "loss": 0.23, "step": 8877 }, { "epoch": 1.68, "grad_norm": 10.9375, "learning_rate": 3.2921479983626796e-07, "loss": 0.1575, "step": 8878 }, { "epoch": 1.68, "grad_norm": 13.5625, "learning_rate": 3.288349825767945e-07, "loss": 0.2311, "step": 8879 }, { "epoch": 1.68, "grad_norm": 11.0, "learning_rate": 3.2845536911815136e-07, "loss": 0.2073, "step": 8880 }, { "epoch": 1.68, "grad_norm": 9.875, "learning_rate": 3.280759594959715e-07, "loss": 0.2088, "step": 8881 }, { "epoch": 1.68, "grad_norm": 12.375, "learning_rate": 3.276967537458689e-07, "loss": 0.2223, "step": 8882 }, { "epoch": 1.68, "grad_norm": 11.3125, "learning_rate": 3.273177519034387e-07, "loss": 0.2407, "step": 8883 }, { "epoch": 1.68, "grad_norm": 10.25, "learning_rate": 3.269389540042564e-07, "loss": 0.1619, "step": 8884 }, { "epoch": 1.68, "grad_norm": 10.6875, "learning_rate": 3.2656036008387904e-07, "loss": 0.178, "step": 8885 }, { "epoch": 1.68, "grad_norm": 12.5625, "learning_rate": 3.2618197017784375e-07, "loss": 0.1848, "step": 8886 }, { "epoch": 1.68, "grad_norm": 8.5625, "learning_rate": 3.258037843216691e-07, "loss": 0.1747, "step": 8887 }, { "epoch": 1.68, "grad_norm": 8.75, "learning_rate": 3.2542580255085415e-07, "loss": 0.2157, "step": 8888 }, { "epoch": 1.68, "grad_norm": 14.4375, "learning_rate": 3.250480249008789e-07, "loss": 0.2174, "step": 8889 }, { "epoch": 1.68, "grad_norm": 9.5, "learning_rate": 3.2467045140720437e-07, "loss": 0.1588, "step": 8890 }, { "epoch": 1.68, "grad_norm": 14.0, "learning_rate": 3.24293082105272e-07, "loss": 0.2083, "step": 8891 }, { "epoch": 1.68, "grad_norm": 13.6875, "learning_rate": 3.2391591703050453e-07, "loss": 0.234, "step": 8892 }, { "epoch": 1.68, "grad_norm": 15.5, "learning_rate": 3.2353895621830556e-07, "loss": 0.1332, "step": 8893 }, { "epoch": 1.68, "grad_norm": 10.0, "learning_rate": 3.231621997040588e-07, "loss": 0.1557, "step": 8894 }, { "epoch": 1.68, "grad_norm": 8.875, "learning_rate": 3.227856475231295e-07, "loss": 0.1471, "step": 8895 }, { "epoch": 1.68, "grad_norm": 10.625, "learning_rate": 3.2240929971086354e-07, "loss": 0.178, "step": 8896 }, { "epoch": 1.68, "grad_norm": 10.25, "learning_rate": 3.220331563025875e-07, "loss": 0.1661, "step": 8897 }, { "epoch": 1.68, "grad_norm": 9.0625, "learning_rate": 3.2165721733360885e-07, "loss": 0.1676, "step": 8898 }, { "epoch": 1.68, "grad_norm": 15.625, "learning_rate": 3.2128148283921606e-07, "loss": 0.2679, "step": 8899 }, { "epoch": 1.68, "grad_norm": 20.375, "learning_rate": 3.209059528546779e-07, "loss": 0.2489, "step": 8900 }, { "epoch": 1.68, "grad_norm": 14.6875, "learning_rate": 3.205306274152442e-07, "loss": 0.1988, "step": 8901 }, { "epoch": 1.68, "grad_norm": 11.9375, "learning_rate": 3.2015550655614625e-07, "loss": 0.2154, "step": 8902 }, { "epoch": 1.68, "grad_norm": 18.875, "learning_rate": 3.1978059031259475e-07, "loss": 0.1688, "step": 8903 }, { "epoch": 1.68, "grad_norm": 12.625, "learning_rate": 3.194058787197826e-07, "loss": 0.1975, "step": 8904 }, { "epoch": 1.68, "grad_norm": 12.9375, "learning_rate": 3.190313718128826e-07, "loss": 0.2601, "step": 8905 }, { "epoch": 1.68, "grad_norm": 18.625, "learning_rate": 3.1865706962704857e-07, "loss": 0.2256, "step": 8906 }, { "epoch": 1.68, "grad_norm": 13.1875, "learning_rate": 3.1828297219741537e-07, "loss": 0.1929, "step": 8907 }, { "epoch": 1.69, "grad_norm": 16.25, "learning_rate": 3.1790907955909803e-07, "loss": 0.2218, "step": 8908 }, { "epoch": 1.69, "grad_norm": 11.75, "learning_rate": 3.175353917471932e-07, "loss": 0.1911, "step": 8909 }, { "epoch": 1.69, "grad_norm": 17.25, "learning_rate": 3.1716190879677753e-07, "loss": 0.1915, "step": 8910 }, { "epoch": 1.69, "grad_norm": 15.6875, "learning_rate": 3.1678863074290885e-07, "loss": 0.1782, "step": 8911 }, { "epoch": 1.69, "grad_norm": 11.8125, "learning_rate": 3.1641555762062543e-07, "loss": 0.1931, "step": 8912 }, { "epoch": 1.69, "grad_norm": 9.625, "learning_rate": 3.160426894649471e-07, "loss": 0.1758, "step": 8913 }, { "epoch": 1.69, "grad_norm": 12.375, "learning_rate": 3.156700263108728e-07, "loss": 0.2054, "step": 8914 }, { "epoch": 1.69, "grad_norm": 10.875, "learning_rate": 3.1529756819338534e-07, "loss": 0.1818, "step": 8915 }, { "epoch": 1.69, "grad_norm": 10.375, "learning_rate": 3.149253151474443e-07, "loss": 0.1882, "step": 8916 }, { "epoch": 1.69, "grad_norm": 14.4375, "learning_rate": 3.1455326720799287e-07, "loss": 0.1931, "step": 8917 }, { "epoch": 1.69, "grad_norm": 13.5625, "learning_rate": 3.1418142440995336e-07, "loss": 0.2043, "step": 8918 }, { "epoch": 1.69, "grad_norm": 9.0625, "learning_rate": 3.1380978678823085e-07, "loss": 0.1883, "step": 8919 }, { "epoch": 1.69, "grad_norm": 12.5625, "learning_rate": 3.134383543777089e-07, "loss": 0.1865, "step": 8920 }, { "epoch": 1.69, "grad_norm": 32.75, "learning_rate": 3.1306712721325266e-07, "loss": 0.2272, "step": 8921 }, { "epoch": 1.69, "grad_norm": 13.1875, "learning_rate": 3.126961053297081e-07, "loss": 0.2698, "step": 8922 }, { "epoch": 1.69, "grad_norm": 14.9375, "learning_rate": 3.123252887619033e-07, "loss": 0.2118, "step": 8923 }, { "epoch": 1.69, "grad_norm": 10.5, "learning_rate": 3.1195467754464415e-07, "loss": 0.1849, "step": 8924 }, { "epoch": 1.69, "grad_norm": 19.0, "learning_rate": 3.1158427171271874e-07, "loss": 0.2059, "step": 8925 }, { "epoch": 1.69, "grad_norm": 10.0625, "learning_rate": 3.112140713008979e-07, "loss": 0.1784, "step": 8926 }, { "epoch": 1.69, "grad_norm": 9.6875, "learning_rate": 3.108440763439291e-07, "loss": 0.1986, "step": 8927 }, { "epoch": 1.69, "grad_norm": 16.5, "learning_rate": 3.1047428687654366e-07, "loss": 0.1637, "step": 8928 }, { "epoch": 1.69, "grad_norm": 14.75, "learning_rate": 3.10104702933452e-07, "loss": 0.2502, "step": 8929 }, { "epoch": 1.69, "grad_norm": 10.5, "learning_rate": 3.097353245493473e-07, "loss": 0.2024, "step": 8930 }, { "epoch": 1.69, "grad_norm": 12.0, "learning_rate": 3.093661517589006e-07, "loss": 0.2011, "step": 8931 }, { "epoch": 1.69, "grad_norm": 14.875, "learning_rate": 3.0899718459676545e-07, "loss": 0.1936, "step": 8932 }, { "epoch": 1.69, "grad_norm": 16.625, "learning_rate": 3.086284230975753e-07, "loss": 0.1825, "step": 8933 }, { "epoch": 1.69, "grad_norm": 13.375, "learning_rate": 3.0825986729594635e-07, "loss": 0.2518, "step": 8934 }, { "epoch": 1.69, "grad_norm": 11.0625, "learning_rate": 3.078915172264718e-07, "loss": 0.1872, "step": 8935 }, { "epoch": 1.69, "grad_norm": 9.375, "learning_rate": 3.075233729237281e-07, "loss": 0.1715, "step": 8936 }, { "epoch": 1.69, "grad_norm": 12.0625, "learning_rate": 3.07155434422273e-07, "loss": 0.2543, "step": 8937 }, { "epoch": 1.69, "grad_norm": 10.8125, "learning_rate": 3.0678770175664313e-07, "loss": 0.1439, "step": 8938 }, { "epoch": 1.69, "grad_norm": 18.875, "learning_rate": 3.06420174961356e-07, "loss": 0.2813, "step": 8939 }, { "epoch": 1.69, "grad_norm": 9.25, "learning_rate": 3.060528540709101e-07, "loss": 0.2096, "step": 8940 }, { "epoch": 1.69, "grad_norm": 11.375, "learning_rate": 3.056857391197857e-07, "loss": 0.2274, "step": 8941 }, { "epoch": 1.69, "grad_norm": 11.875, "learning_rate": 3.0531883014244286e-07, "loss": 0.1771, "step": 8942 }, { "epoch": 1.69, "grad_norm": 12.9375, "learning_rate": 3.0495212717332134e-07, "loss": 0.2247, "step": 8943 }, { "epoch": 1.69, "grad_norm": 11.375, "learning_rate": 3.045856302468422e-07, "loss": 0.1739, "step": 8944 }, { "epoch": 1.69, "grad_norm": 10.8125, "learning_rate": 3.0421933939740853e-07, "loss": 0.1917, "step": 8945 }, { "epoch": 1.69, "grad_norm": 10.125, "learning_rate": 3.038532546594031e-07, "loss": 0.2173, "step": 8946 }, { "epoch": 1.69, "grad_norm": 11.5, "learning_rate": 3.034873760671875e-07, "loss": 0.2131, "step": 8947 }, { "epoch": 1.69, "grad_norm": 11.5, "learning_rate": 3.0312170365510737e-07, "loss": 0.1721, "step": 8948 }, { "epoch": 1.69, "grad_norm": 11.3125, "learning_rate": 3.027562374574869e-07, "loss": 0.1845, "step": 8949 }, { "epoch": 1.69, "grad_norm": 10.9375, "learning_rate": 3.023909775086314e-07, "loss": 0.1805, "step": 8950 }, { "epoch": 1.69, "grad_norm": 37.0, "learning_rate": 3.020259238428255e-07, "loss": 0.2009, "step": 8951 }, { "epoch": 1.69, "grad_norm": 6.9375, "learning_rate": 3.0166107649433727e-07, "loss": 0.1347, "step": 8952 }, { "epoch": 1.69, "grad_norm": 10.4375, "learning_rate": 3.012964354974135e-07, "loss": 0.1961, "step": 8953 }, { "epoch": 1.69, "grad_norm": 10.3125, "learning_rate": 3.009320008862815e-07, "loss": 0.166, "step": 8954 }, { "epoch": 1.69, "grad_norm": 13.25, "learning_rate": 3.005677726951492e-07, "loss": 0.1981, "step": 8955 }, { "epoch": 1.69, "grad_norm": 10.25, "learning_rate": 3.0020375095820676e-07, "loss": 0.2631, "step": 8956 }, { "epoch": 1.69, "grad_norm": 9.375, "learning_rate": 2.998399357096241e-07, "loss": 0.1842, "step": 8957 }, { "epoch": 1.69, "grad_norm": 10.75, "learning_rate": 2.994763269835496e-07, "loss": 0.2161, "step": 8958 }, { "epoch": 1.69, "grad_norm": 15.0625, "learning_rate": 2.9911292481411555e-07, "loss": 0.2099, "step": 8959 }, { "epoch": 1.69, "grad_norm": 15.3125, "learning_rate": 2.9874972923543357e-07, "loss": 0.2051, "step": 8960 }, { "epoch": 1.7, "grad_norm": 11.6875, "learning_rate": 2.9838674028159567e-07, "loss": 0.2132, "step": 8961 }, { "epoch": 1.7, "grad_norm": 13.25, "learning_rate": 2.9802395798667313e-07, "loss": 0.1987, "step": 8962 }, { "epoch": 1.7, "grad_norm": 12.4375, "learning_rate": 2.97661382384721e-07, "loss": 0.1815, "step": 8963 }, { "epoch": 1.7, "grad_norm": 11.25, "learning_rate": 2.9729901350977255e-07, "loss": 0.2226, "step": 8964 }, { "epoch": 1.7, "grad_norm": 14.0, "learning_rate": 2.969368513958426e-07, "loss": 0.2329, "step": 8965 }, { "epoch": 1.7, "grad_norm": 19.25, "learning_rate": 2.965748960769249e-07, "loss": 0.2458, "step": 8966 }, { "epoch": 1.7, "grad_norm": 10.75, "learning_rate": 2.9621314758699644e-07, "loss": 0.2079, "step": 8967 }, { "epoch": 1.7, "grad_norm": 11.25, "learning_rate": 2.958516059600133e-07, "loss": 0.2097, "step": 8968 }, { "epoch": 1.7, "grad_norm": 12.0625, "learning_rate": 2.9549027122991236e-07, "loss": 0.2011, "step": 8969 }, { "epoch": 1.7, "grad_norm": 11.1875, "learning_rate": 2.9512914343061057e-07, "loss": 0.2494, "step": 8970 }, { "epoch": 1.7, "grad_norm": 12.0625, "learning_rate": 2.9476822259600636e-07, "loss": 0.2249, "step": 8971 }, { "epoch": 1.7, "grad_norm": 12.5, "learning_rate": 2.944075087599782e-07, "loss": 0.24, "step": 8972 }, { "epoch": 1.7, "grad_norm": 16.875, "learning_rate": 2.9404700195638516e-07, "loss": 0.2183, "step": 8973 }, { "epoch": 1.7, "grad_norm": 7.375, "learning_rate": 2.9368670221906673e-07, "loss": 0.1615, "step": 8974 }, { "epoch": 1.7, "grad_norm": 10.875, "learning_rate": 2.933266095818435e-07, "loss": 0.1458, "step": 8975 }, { "epoch": 1.7, "grad_norm": 13.75, "learning_rate": 2.929667240785164e-07, "loss": 0.1767, "step": 8976 }, { "epoch": 1.7, "grad_norm": 13.875, "learning_rate": 2.926070457428665e-07, "loss": 0.201, "step": 8977 }, { "epoch": 1.7, "grad_norm": 8.5, "learning_rate": 2.922475746086556e-07, "loss": 0.2006, "step": 8978 }, { "epoch": 1.7, "grad_norm": 11.3125, "learning_rate": 2.918883107096268e-07, "loss": 0.2253, "step": 8979 }, { "epoch": 1.7, "grad_norm": 12.3125, "learning_rate": 2.9152925407950284e-07, "loss": 0.2001, "step": 8980 }, { "epoch": 1.7, "grad_norm": 13.6875, "learning_rate": 2.9117040475198697e-07, "loss": 0.2429, "step": 8981 }, { "epoch": 1.7, "grad_norm": 8.8125, "learning_rate": 2.9081176276076373e-07, "loss": 0.2106, "step": 8982 }, { "epoch": 1.7, "grad_norm": 12.875, "learning_rate": 2.904533281394975e-07, "loss": 0.2278, "step": 8983 }, { "epoch": 1.7, "grad_norm": 9.6875, "learning_rate": 2.900951009218339e-07, "loss": 0.1874, "step": 8984 }, { "epoch": 1.7, "grad_norm": 9.25, "learning_rate": 2.897370811413983e-07, "loss": 0.2224, "step": 8985 }, { "epoch": 1.7, "grad_norm": 11.5625, "learning_rate": 2.89379268831797e-07, "loss": 0.2476, "step": 8986 }, { "epoch": 1.7, "grad_norm": 17.625, "learning_rate": 2.8902166402661693e-07, "loss": 0.2447, "step": 8987 }, { "epoch": 1.7, "grad_norm": 11.375, "learning_rate": 2.8866426675942497e-07, "loss": 0.2262, "step": 8988 }, { "epoch": 1.7, "grad_norm": 9.375, "learning_rate": 2.883070770637694e-07, "loss": 0.1824, "step": 8989 }, { "epoch": 1.7, "grad_norm": 12.25, "learning_rate": 2.8795009497317866e-07, "loss": 0.2173, "step": 8990 }, { "epoch": 1.7, "grad_norm": 16.125, "learning_rate": 2.875933205211612e-07, "loss": 0.2438, "step": 8991 }, { "epoch": 1.7, "grad_norm": 12.25, "learning_rate": 2.8723675374120655e-07, "loss": 0.2045, "step": 8992 }, { "epoch": 1.7, "grad_norm": 11.25, "learning_rate": 2.868803946667845e-07, "loss": 0.1727, "step": 8993 }, { "epoch": 1.7, "grad_norm": 10.125, "learning_rate": 2.865242433313456e-07, "loss": 0.1723, "step": 8994 }, { "epoch": 1.7, "grad_norm": 8.6875, "learning_rate": 2.8616829976832045e-07, "loss": 0.1537, "step": 8995 }, { "epoch": 1.7, "grad_norm": 9.0625, "learning_rate": 2.858125640111206e-07, "loss": 0.1504, "step": 8996 }, { "epoch": 1.7, "grad_norm": 11.625, "learning_rate": 2.85457036093138e-07, "loss": 0.2019, "step": 8997 }, { "epoch": 1.7, "grad_norm": 11.8125, "learning_rate": 2.851017160477448e-07, "loss": 0.2151, "step": 8998 }, { "epoch": 1.7, "grad_norm": 14.25, "learning_rate": 2.847466039082941e-07, "loss": 0.1866, "step": 8999 }, { "epoch": 1.7, "grad_norm": 12.6875, "learning_rate": 2.843916997081189e-07, "loss": 0.236, "step": 9000 }, { "epoch": 1.7, "grad_norm": 13.3125, "learning_rate": 2.840370034805334e-07, "loss": 0.198, "step": 9001 }, { "epoch": 1.7, "grad_norm": 11.375, "learning_rate": 2.83682515258831e-07, "loss": 0.2167, "step": 9002 }, { "epoch": 1.7, "grad_norm": 11.5625, "learning_rate": 2.833282350762881e-07, "loss": 0.1883, "step": 9003 }, { "epoch": 1.7, "grad_norm": 11.125, "learning_rate": 2.8297416296615835e-07, "loss": 0.1991, "step": 9004 }, { "epoch": 1.7, "grad_norm": 9.875, "learning_rate": 2.82620298961678e-07, "loss": 0.165, "step": 9005 }, { "epoch": 1.7, "grad_norm": 11.75, "learning_rate": 2.82266643096063e-07, "loss": 0.1972, "step": 9006 }, { "epoch": 1.7, "grad_norm": 12.5625, "learning_rate": 2.819131954025112e-07, "loss": 0.2348, "step": 9007 }, { "epoch": 1.7, "grad_norm": 12.0, "learning_rate": 2.8155995591419793e-07, "loss": 0.2264, "step": 9008 }, { "epoch": 1.7, "grad_norm": 12.5625, "learning_rate": 2.812069246642818e-07, "loss": 0.1965, "step": 9009 }, { "epoch": 1.7, "grad_norm": 12.375, "learning_rate": 2.8085410168590036e-07, "loss": 0.1876, "step": 9010 }, { "epoch": 1.7, "grad_norm": 17.125, "learning_rate": 2.805014870121725e-07, "loss": 0.1971, "step": 9011 }, { "epoch": 1.7, "grad_norm": 11.625, "learning_rate": 2.8014908067619665e-07, "loss": 0.2685, "step": 9012 }, { "epoch": 1.71, "grad_norm": 9.5, "learning_rate": 2.79796882711052e-07, "loss": 0.1729, "step": 9013 }, { "epoch": 1.71, "grad_norm": 13.875, "learning_rate": 2.794448931497995e-07, "loss": 0.2627, "step": 9014 }, { "epoch": 1.71, "grad_norm": 11.5625, "learning_rate": 2.7909311202547807e-07, "loss": 0.1948, "step": 9015 }, { "epoch": 1.71, "grad_norm": 12.5625, "learning_rate": 2.7874153937110905e-07, "loss": 0.225, "step": 9016 }, { "epoch": 1.71, "grad_norm": 13.5, "learning_rate": 2.7839017521969244e-07, "loss": 0.1836, "step": 9017 }, { "epoch": 1.71, "grad_norm": 13.125, "learning_rate": 2.780390196042118e-07, "loss": 0.1524, "step": 9018 }, { "epoch": 1.71, "grad_norm": 14.375, "learning_rate": 2.776880725576275e-07, "loss": 0.2561, "step": 9019 }, { "epoch": 1.71, "grad_norm": 9.4375, "learning_rate": 2.7733733411288237e-07, "loss": 0.1557, "step": 9020 }, { "epoch": 1.71, "grad_norm": 8.9375, "learning_rate": 2.769868043028984e-07, "loss": 0.1939, "step": 9021 }, { "epoch": 1.71, "grad_norm": 12.25, "learning_rate": 2.766364831605806e-07, "loss": 0.1779, "step": 9022 }, { "epoch": 1.71, "grad_norm": 12.875, "learning_rate": 2.762863707188107e-07, "loss": 0.2124, "step": 9023 }, { "epoch": 1.71, "grad_norm": 11.375, "learning_rate": 2.7593646701045336e-07, "loss": 0.1968, "step": 9024 }, { "epoch": 1.71, "grad_norm": 11.3125, "learning_rate": 2.755867720683536e-07, "loss": 0.2227, "step": 9025 }, { "epoch": 1.71, "grad_norm": 8.5625, "learning_rate": 2.7523728592533603e-07, "loss": 0.1672, "step": 9026 }, { "epoch": 1.71, "grad_norm": 12.0, "learning_rate": 2.7488800861420526e-07, "loss": 0.1783, "step": 9027 }, { "epoch": 1.71, "grad_norm": 12.125, "learning_rate": 2.7453894016774703e-07, "loss": 0.188, "step": 9028 }, { "epoch": 1.71, "grad_norm": 12.1875, "learning_rate": 2.7419008061872784e-07, "loss": 0.1345, "step": 9029 }, { "epoch": 1.71, "grad_norm": 9.4375, "learning_rate": 2.7384142999989457e-07, "loss": 0.1484, "step": 9030 }, { "epoch": 1.71, "grad_norm": 10.4375, "learning_rate": 2.734929883439727e-07, "loss": 0.1807, "step": 9031 }, { "epoch": 1.71, "grad_norm": 15.8125, "learning_rate": 2.731447556836694e-07, "loss": 0.2183, "step": 9032 }, { "epoch": 1.71, "grad_norm": 13.5625, "learning_rate": 2.7279673205167363e-07, "loss": 0.2046, "step": 9033 }, { "epoch": 1.71, "grad_norm": 11.9375, "learning_rate": 2.72448917480653e-07, "loss": 0.2168, "step": 9034 }, { "epoch": 1.71, "grad_norm": 15.8125, "learning_rate": 2.7210131200325504e-07, "loss": 0.237, "step": 9035 }, { "epoch": 1.71, "grad_norm": 10.9375, "learning_rate": 2.717539156521085e-07, "loss": 0.1948, "step": 9036 }, { "epoch": 1.71, "grad_norm": 11.4375, "learning_rate": 2.714067284598235e-07, "loss": 0.2727, "step": 9037 }, { "epoch": 1.71, "grad_norm": 13.9375, "learning_rate": 2.710597504589882e-07, "loss": 0.2163, "step": 9038 }, { "epoch": 1.71, "grad_norm": 12.875, "learning_rate": 2.707129816821727e-07, "loss": 0.2332, "step": 9039 }, { "epoch": 1.71, "grad_norm": 8.8125, "learning_rate": 2.7036642216192787e-07, "loss": 0.1928, "step": 9040 }, { "epoch": 1.71, "grad_norm": 13.5625, "learning_rate": 2.700200719307841e-07, "loss": 0.2237, "step": 9041 }, { "epoch": 1.71, "grad_norm": 10.125, "learning_rate": 2.696739310212515e-07, "loss": 0.1798, "step": 9042 }, { "epoch": 1.71, "grad_norm": 11.9375, "learning_rate": 2.6932799946582125e-07, "loss": 0.2692, "step": 9043 }, { "epoch": 1.71, "grad_norm": 10.25, "learning_rate": 2.6898227729696606e-07, "loss": 0.2396, "step": 9044 }, { "epoch": 1.71, "grad_norm": 10.25, "learning_rate": 2.6863676454713745e-07, "loss": 0.1915, "step": 9045 }, { "epoch": 1.71, "grad_norm": 9.125, "learning_rate": 2.6829146124876694e-07, "loss": 0.1766, "step": 9046 }, { "epoch": 1.71, "grad_norm": 9.4375, "learning_rate": 2.6794636743426724e-07, "loss": 0.2193, "step": 9047 }, { "epoch": 1.71, "grad_norm": 13.0625, "learning_rate": 2.6760148313603214e-07, "loss": 0.2325, "step": 9048 }, { "epoch": 1.71, "grad_norm": 11.625, "learning_rate": 2.6725680838643463e-07, "loss": 0.2137, "step": 9049 }, { "epoch": 1.71, "grad_norm": 10.6875, "learning_rate": 2.669123432178272e-07, "loss": 0.2704, "step": 9050 }, { "epoch": 1.71, "grad_norm": 11.5625, "learning_rate": 2.665680876625454e-07, "loss": 0.233, "step": 9051 }, { "epoch": 1.71, "grad_norm": 9.625, "learning_rate": 2.6622404175290245e-07, "loss": 0.2299, "step": 9052 }, { "epoch": 1.71, "grad_norm": 9.8125, "learning_rate": 2.65880205521194e-07, "loss": 0.1656, "step": 9053 }, { "epoch": 1.71, "grad_norm": 9.25, "learning_rate": 2.65536578999693e-07, "loss": 0.2042, "step": 9054 }, { "epoch": 1.71, "grad_norm": 12.25, "learning_rate": 2.651931622206563e-07, "loss": 0.2452, "step": 9055 }, { "epoch": 1.71, "grad_norm": 15.6875, "learning_rate": 2.648499552163189e-07, "loss": 0.1887, "step": 9056 }, { "epoch": 1.71, "grad_norm": 11.75, "learning_rate": 2.645069580188972e-07, "loss": 0.2546, "step": 9057 }, { "epoch": 1.71, "grad_norm": 10.625, "learning_rate": 2.6416417066058574e-07, "loss": 0.22, "step": 9058 }, { "epoch": 1.71, "grad_norm": 12.75, "learning_rate": 2.6382159317356245e-07, "loss": 0.2597, "step": 9059 }, { "epoch": 1.71, "grad_norm": 12.5, "learning_rate": 2.634792255899837e-07, "loss": 0.2563, "step": 9060 }, { "epoch": 1.71, "grad_norm": 12.1875, "learning_rate": 2.6313706794198647e-07, "loss": 0.212, "step": 9061 }, { "epoch": 1.71, "grad_norm": 11.6875, "learning_rate": 2.6279512026168795e-07, "loss": 0.1767, "step": 9062 }, { "epoch": 1.71, "grad_norm": 11.1875, "learning_rate": 2.6245338258118563e-07, "loss": 0.1753, "step": 9063 }, { "epoch": 1.71, "grad_norm": 10.75, "learning_rate": 2.62111854932558e-07, "loss": 0.2082, "step": 9064 }, { "epoch": 1.71, "grad_norm": 10.6875, "learning_rate": 2.6177053734786265e-07, "loss": 0.2181, "step": 9065 }, { "epoch": 1.72, "grad_norm": 10.5, "learning_rate": 2.6142942985913856e-07, "loss": 0.216, "step": 9066 }, { "epoch": 1.72, "grad_norm": 11.8125, "learning_rate": 2.6108853249840386e-07, "loss": 0.1352, "step": 9067 }, { "epoch": 1.72, "grad_norm": 11.75, "learning_rate": 2.607478452976586e-07, "loss": 0.205, "step": 9068 }, { "epoch": 1.72, "grad_norm": 13.5625, "learning_rate": 2.6040736828888024e-07, "loss": 0.1778, "step": 9069 }, { "epoch": 1.72, "grad_norm": 10.5, "learning_rate": 2.600671015040301e-07, "loss": 0.1865, "step": 9070 }, { "epoch": 1.72, "grad_norm": 10.875, "learning_rate": 2.5972704497504745e-07, "loss": 0.1958, "step": 9071 }, { "epoch": 1.72, "grad_norm": 12.25, "learning_rate": 2.5938719873385256e-07, "loss": 0.2367, "step": 9072 }, { "epoch": 1.72, "grad_norm": 15.0, "learning_rate": 2.590475628123454e-07, "loss": 0.1769, "step": 9073 }, { "epoch": 1.72, "grad_norm": 10.6875, "learning_rate": 2.587081372424066e-07, "loss": 0.1616, "step": 9074 }, { "epoch": 1.72, "grad_norm": 13.25, "learning_rate": 2.583689220558974e-07, "loss": 0.1904, "step": 9075 }, { "epoch": 1.72, "grad_norm": 8.8125, "learning_rate": 2.580299172846587e-07, "loss": 0.1733, "step": 9076 }, { "epoch": 1.72, "grad_norm": 11.0, "learning_rate": 2.57691122960512e-07, "loss": 0.1355, "step": 9077 }, { "epoch": 1.72, "grad_norm": 9.375, "learning_rate": 2.573525391152587e-07, "loss": 0.1917, "step": 9078 }, { "epoch": 1.72, "grad_norm": 17.75, "learning_rate": 2.57014165780681e-07, "loss": 0.2153, "step": 9079 }, { "epoch": 1.72, "grad_norm": 9.5625, "learning_rate": 2.566760029885404e-07, "loss": 0.1748, "step": 9080 }, { "epoch": 1.72, "grad_norm": 14.8125, "learning_rate": 2.563380507705801e-07, "loss": 0.1736, "step": 9081 }, { "epoch": 1.72, "grad_norm": 12.3125, "learning_rate": 2.56000309158522e-07, "loss": 0.194, "step": 9082 }, { "epoch": 1.72, "grad_norm": 13.1875, "learning_rate": 2.5566277818406896e-07, "loss": 0.2841, "step": 9083 }, { "epoch": 1.72, "grad_norm": 12.875, "learning_rate": 2.553254578789044e-07, "loss": 0.2239, "step": 9084 }, { "epoch": 1.72, "grad_norm": 9.1875, "learning_rate": 2.549883482746912e-07, "loss": 0.1836, "step": 9085 }, { "epoch": 1.72, "grad_norm": 14.125, "learning_rate": 2.54651449403073e-07, "loss": 0.2204, "step": 9086 }, { "epoch": 1.72, "grad_norm": 10.5, "learning_rate": 2.543147612956734e-07, "loss": 0.1905, "step": 9087 }, { "epoch": 1.72, "grad_norm": 14.4375, "learning_rate": 2.539782839840965e-07, "loss": 0.2015, "step": 9088 }, { "epoch": 1.72, "grad_norm": 8.625, "learning_rate": 2.536420174999263e-07, "loss": 0.194, "step": 9089 }, { "epoch": 1.72, "grad_norm": 10.6875, "learning_rate": 2.5330596187472716e-07, "loss": 0.2091, "step": 9090 }, { "epoch": 1.72, "grad_norm": 10.625, "learning_rate": 2.5297011714004363e-07, "loss": 0.2479, "step": 9091 }, { "epoch": 1.72, "grad_norm": 12.5, "learning_rate": 2.5263448332740054e-07, "loss": 0.2162, "step": 9092 }, { "epoch": 1.72, "grad_norm": 14.125, "learning_rate": 2.522990604683029e-07, "loss": 0.2763, "step": 9093 }, { "epoch": 1.72, "grad_norm": 12.8125, "learning_rate": 2.519638485942355e-07, "loss": 0.2076, "step": 9094 }, { "epoch": 1.72, "grad_norm": 12.375, "learning_rate": 2.5162884773666406e-07, "loss": 0.2188, "step": 9095 }, { "epoch": 1.72, "grad_norm": 10.375, "learning_rate": 2.512940579270337e-07, "loss": 0.2155, "step": 9096 }, { "epoch": 1.72, "grad_norm": 11.9375, "learning_rate": 2.5095947919677066e-07, "loss": 0.1844, "step": 9097 }, { "epoch": 1.72, "grad_norm": 9.1875, "learning_rate": 2.5062511157728066e-07, "loss": 0.1447, "step": 9098 }, { "epoch": 1.72, "grad_norm": 15.6875, "learning_rate": 2.5029095509994965e-07, "loss": 0.2253, "step": 9099 }, { "epoch": 1.72, "grad_norm": 9.1875, "learning_rate": 2.49957009796144e-07, "loss": 0.1543, "step": 9100 }, { "epoch": 1.72, "grad_norm": 14.0, "learning_rate": 2.4962327569721003e-07, "loss": 0.1944, "step": 9101 }, { "epoch": 1.72, "grad_norm": 9.875, "learning_rate": 2.492897528344748e-07, "loss": 0.1466, "step": 9102 }, { "epoch": 1.72, "grad_norm": 9.875, "learning_rate": 2.489564412392448e-07, "loss": 0.1843, "step": 9103 }, { "epoch": 1.72, "grad_norm": 10.8125, "learning_rate": 2.486233409428068e-07, "loss": 0.2128, "step": 9104 }, { "epoch": 1.72, "grad_norm": 13.75, "learning_rate": 2.482904519764279e-07, "loss": 0.2804, "step": 9105 }, { "epoch": 1.72, "grad_norm": 13.875, "learning_rate": 2.4795777437135653e-07, "loss": 0.1942, "step": 9106 }, { "epoch": 1.72, "grad_norm": 13.8125, "learning_rate": 2.476253081588187e-07, "loss": 0.2113, "step": 9107 }, { "epoch": 1.72, "grad_norm": 11.375, "learning_rate": 2.4729305337002295e-07, "loss": 0.1515, "step": 9108 }, { "epoch": 1.72, "grad_norm": 12.375, "learning_rate": 2.4696101003615607e-07, "loss": 0.2186, "step": 9109 }, { "epoch": 1.72, "grad_norm": 15.8125, "learning_rate": 2.4662917818838755e-07, "loss": 0.272, "step": 9110 }, { "epoch": 1.72, "grad_norm": 10.4375, "learning_rate": 2.462975578578641e-07, "loss": 0.2197, "step": 9111 }, { "epoch": 1.72, "grad_norm": 11.9375, "learning_rate": 2.4596614907571424e-07, "loss": 0.2265, "step": 9112 }, { "epoch": 1.72, "grad_norm": 13.9375, "learning_rate": 2.4563495187304636e-07, "loss": 0.2361, "step": 9113 }, { "epoch": 1.72, "grad_norm": 11.375, "learning_rate": 2.4530396628094976e-07, "loss": 0.1793, "step": 9114 }, { "epoch": 1.72, "grad_norm": 16.375, "learning_rate": 2.4497319233049213e-07, "loss": 0.2629, "step": 9115 }, { "epoch": 1.72, "grad_norm": 9.375, "learning_rate": 2.446426300527219e-07, "loss": 0.1937, "step": 9116 }, { "epoch": 1.72, "grad_norm": 8.9375, "learning_rate": 2.4431227947866896e-07, "loss": 0.186, "step": 9117 }, { "epoch": 1.72, "grad_norm": 12.6875, "learning_rate": 2.4398214063934277e-07, "loss": 0.1983, "step": 9118 }, { "epoch": 1.73, "grad_norm": 12.875, "learning_rate": 2.43652213565731e-07, "loss": 0.2203, "step": 9119 }, { "epoch": 1.73, "grad_norm": 14.875, "learning_rate": 2.43322498288803e-07, "loss": 0.214, "step": 9120 }, { "epoch": 1.73, "grad_norm": 16.0, "learning_rate": 2.429929948395093e-07, "loss": 0.1777, "step": 9121 }, { "epoch": 1.73, "grad_norm": 12.5625, "learning_rate": 2.426637032487794e-07, "loss": 0.1966, "step": 9122 }, { "epoch": 1.73, "grad_norm": 9.4375, "learning_rate": 2.423346235475221e-07, "loss": 0.1588, "step": 9123 }, { "epoch": 1.73, "grad_norm": 14.75, "learning_rate": 2.4200575576662696e-07, "loss": 0.2523, "step": 9124 }, { "epoch": 1.73, "grad_norm": 13.4375, "learning_rate": 2.4167709993696497e-07, "loss": 0.2364, "step": 9125 }, { "epoch": 1.73, "grad_norm": 12.8125, "learning_rate": 2.413486560893852e-07, "loss": 0.2084, "step": 9126 }, { "epoch": 1.73, "grad_norm": 11.625, "learning_rate": 2.4102042425471734e-07, "loss": 0.2134, "step": 9127 }, { "epoch": 1.73, "grad_norm": 11.0625, "learning_rate": 2.406924044637726e-07, "loss": 0.222, "step": 9128 }, { "epoch": 1.73, "grad_norm": 12.9375, "learning_rate": 2.403645967473414e-07, "loss": 0.2125, "step": 9129 }, { "epoch": 1.73, "grad_norm": 17.625, "learning_rate": 2.4003700113619273e-07, "loss": 0.3109, "step": 9130 }, { "epoch": 1.73, "grad_norm": 11.3125, "learning_rate": 2.3970961766107746e-07, "loss": 0.1992, "step": 9131 }, { "epoch": 1.73, "grad_norm": 22.375, "learning_rate": 2.393824463527267e-07, "loss": 0.2355, "step": 9132 }, { "epoch": 1.73, "grad_norm": 8.4375, "learning_rate": 2.390554872418513e-07, "loss": 0.1867, "step": 9133 }, { "epoch": 1.73, "grad_norm": 11.25, "learning_rate": 2.387287403591407e-07, "loss": 0.1918, "step": 9134 }, { "epoch": 1.73, "grad_norm": 10.25, "learning_rate": 2.3840220573526612e-07, "loss": 0.2144, "step": 9135 }, { "epoch": 1.73, "grad_norm": 11.875, "learning_rate": 2.3807588340087866e-07, "loss": 0.1959, "step": 9136 }, { "epoch": 1.73, "grad_norm": 13.625, "learning_rate": 2.3774977338660992e-07, "loss": 0.2673, "step": 9137 }, { "epoch": 1.73, "grad_norm": 11.9375, "learning_rate": 2.374238757230693e-07, "loss": 0.1778, "step": 9138 }, { "epoch": 1.73, "grad_norm": 13.5, "learning_rate": 2.3709819044084896e-07, "loss": 0.2384, "step": 9139 }, { "epoch": 1.73, "grad_norm": 34.25, "learning_rate": 2.367727175705198e-07, "loss": 0.2224, "step": 9140 }, { "epoch": 1.73, "grad_norm": 8.4375, "learning_rate": 2.364474571426334e-07, "loss": 0.2154, "step": 9141 }, { "epoch": 1.73, "grad_norm": 12.3125, "learning_rate": 2.3612240918771955e-07, "loss": 0.2239, "step": 9142 }, { "epoch": 1.73, "grad_norm": 11.25, "learning_rate": 2.3579757373629108e-07, "loss": 0.1719, "step": 9143 }, { "epoch": 1.73, "grad_norm": 9.5, "learning_rate": 2.3547295081883886e-07, "loss": 0.2027, "step": 9144 }, { "epoch": 1.73, "grad_norm": 16.375, "learning_rate": 2.3514854046583435e-07, "loss": 0.1739, "step": 9145 }, { "epoch": 1.73, "grad_norm": 15.25, "learning_rate": 2.348243427077282e-07, "loss": 0.2823, "step": 9146 }, { "epoch": 1.73, "grad_norm": 18.375, "learning_rate": 2.3450035757495272e-07, "loss": 0.2965, "step": 9147 }, { "epoch": 1.73, "grad_norm": 11.5625, "learning_rate": 2.3417658509791941e-07, "loss": 0.193, "step": 9148 }, { "epoch": 1.73, "grad_norm": 12.875, "learning_rate": 2.3385302530701954e-07, "loss": 0.2025, "step": 9149 }, { "epoch": 1.73, "grad_norm": 11.875, "learning_rate": 2.335296782326249e-07, "loss": 0.1754, "step": 9150 }, { "epoch": 1.73, "grad_norm": 16.5, "learning_rate": 2.3320654390508707e-07, "loss": 0.2824, "step": 9151 }, { "epoch": 1.73, "grad_norm": 10.3125, "learning_rate": 2.3288362235473815e-07, "loss": 0.1643, "step": 9152 }, { "epoch": 1.73, "grad_norm": 19.5, "learning_rate": 2.3256091361188838e-07, "loss": 0.1943, "step": 9153 }, { "epoch": 1.73, "grad_norm": 9.4375, "learning_rate": 2.3223841770683096e-07, "loss": 0.2399, "step": 9154 }, { "epoch": 1.73, "grad_norm": 11.125, "learning_rate": 2.3191613466983702e-07, "loss": 0.2309, "step": 9155 }, { "epoch": 1.73, "grad_norm": 7.78125, "learning_rate": 2.31594064531159e-07, "loss": 0.1387, "step": 9156 }, { "epoch": 1.73, "grad_norm": 11.8125, "learning_rate": 2.3127220732102746e-07, "loss": 0.182, "step": 9157 }, { "epoch": 1.73, "grad_norm": 10.0625, "learning_rate": 2.3095056306965485e-07, "loss": 0.1942, "step": 9158 }, { "epoch": 1.73, "grad_norm": 12.75, "learning_rate": 2.3062913180723316e-07, "loss": 0.2112, "step": 9159 }, { "epoch": 1.73, "grad_norm": 8.0625, "learning_rate": 2.3030791356393384e-07, "loss": 0.1492, "step": 9160 }, { "epoch": 1.73, "grad_norm": 8.5, "learning_rate": 2.2998690836990883e-07, "loss": 0.2116, "step": 9161 }, { "epoch": 1.73, "grad_norm": 10.5, "learning_rate": 2.296661162552899e-07, "loss": 0.1746, "step": 9162 }, { "epoch": 1.73, "grad_norm": 38.0, "learning_rate": 2.2934553725018903e-07, "loss": 0.2391, "step": 9163 }, { "epoch": 1.73, "grad_norm": 14.6875, "learning_rate": 2.2902517138469776e-07, "loss": 0.2483, "step": 9164 }, { "epoch": 1.73, "grad_norm": 17.125, "learning_rate": 2.287050186888881e-07, "loss": 0.1963, "step": 9165 }, { "epoch": 1.73, "grad_norm": 12.1875, "learning_rate": 2.283850791928116e-07, "loss": 0.2335, "step": 9166 }, { "epoch": 1.73, "grad_norm": 14.1875, "learning_rate": 2.2806535292650007e-07, "loss": 0.286, "step": 9167 }, { "epoch": 1.73, "grad_norm": 21.875, "learning_rate": 2.2774583991996556e-07, "loss": 0.2558, "step": 9168 }, { "epoch": 1.73, "grad_norm": 11.125, "learning_rate": 2.2742654020319943e-07, "loss": 0.1867, "step": 9169 }, { "epoch": 1.73, "grad_norm": 8.8125, "learning_rate": 2.2710745380617377e-07, "loss": 0.219, "step": 9170 }, { "epoch": 1.73, "grad_norm": 10.3125, "learning_rate": 2.2678858075884019e-07, "loss": 0.186, "step": 9171 }, { "epoch": 1.74, "grad_norm": 11.75, "learning_rate": 2.2646992109113002e-07, "loss": 0.1956, "step": 9172 }, { "epoch": 1.74, "grad_norm": 50.75, "learning_rate": 2.2615147483295518e-07, "loss": 0.2237, "step": 9173 }, { "epoch": 1.74, "grad_norm": 11.5, "learning_rate": 2.2583324201420732e-07, "loss": 0.1529, "step": 9174 }, { "epoch": 1.74, "grad_norm": 13.0625, "learning_rate": 2.2551522266475779e-07, "loss": 0.2418, "step": 9175 }, { "epoch": 1.74, "grad_norm": 11.6875, "learning_rate": 2.2519741681445829e-07, "loss": 0.2045, "step": 9176 }, { "epoch": 1.74, "grad_norm": 11.125, "learning_rate": 2.248798244931405e-07, "loss": 0.2095, "step": 9177 }, { "epoch": 1.74, "grad_norm": 12.8125, "learning_rate": 2.245624457306153e-07, "loss": 0.2879, "step": 9178 }, { "epoch": 1.74, "grad_norm": 11.75, "learning_rate": 2.2424528055667473e-07, "loss": 0.1908, "step": 9179 }, { "epoch": 1.74, "grad_norm": 10.625, "learning_rate": 2.2392832900108962e-07, "loss": 0.1837, "step": 9180 }, { "epoch": 1.74, "grad_norm": 11.25, "learning_rate": 2.236115910936118e-07, "loss": 0.2127, "step": 9181 }, { "epoch": 1.74, "grad_norm": 12.3125, "learning_rate": 2.232950668639719e-07, "loss": 0.1442, "step": 9182 }, { "epoch": 1.74, "grad_norm": 11.1875, "learning_rate": 2.229787563418817e-07, "loss": 0.2088, "step": 9183 }, { "epoch": 1.74, "grad_norm": 12.6875, "learning_rate": 2.226626595570322e-07, "loss": 0.175, "step": 9184 }, { "epoch": 1.74, "grad_norm": 11.125, "learning_rate": 2.223467765390941e-07, "loss": 0.203, "step": 9185 }, { "epoch": 1.74, "grad_norm": 11.6875, "learning_rate": 2.2203110731771865e-07, "loss": 0.2121, "step": 9186 }, { "epoch": 1.74, "grad_norm": 12.25, "learning_rate": 2.217156519225369e-07, "loss": 0.1923, "step": 9187 }, { "epoch": 1.74, "grad_norm": 16.75, "learning_rate": 2.2140041038315963e-07, "loss": 0.1786, "step": 9188 }, { "epoch": 1.74, "grad_norm": 19.625, "learning_rate": 2.210853827291773e-07, "loss": 0.2305, "step": 9189 }, { "epoch": 1.74, "grad_norm": 9.375, "learning_rate": 2.2077056899016125e-07, "loss": 0.1496, "step": 9190 }, { "epoch": 1.74, "grad_norm": 10.5625, "learning_rate": 2.2045596919566174e-07, "loss": 0.1715, "step": 9191 }, { "epoch": 1.74, "grad_norm": 15.3125, "learning_rate": 2.2014158337520935e-07, "loss": 0.2474, "step": 9192 }, { "epoch": 1.74, "grad_norm": 10.5625, "learning_rate": 2.1982741155831405e-07, "loss": 0.2116, "step": 9193 }, { "epoch": 1.74, "grad_norm": 12.375, "learning_rate": 2.1951345377446755e-07, "loss": 0.2191, "step": 9194 }, { "epoch": 1.74, "grad_norm": 8.8125, "learning_rate": 2.19199710053139e-07, "loss": 0.1879, "step": 9195 }, { "epoch": 1.74, "grad_norm": 11.25, "learning_rate": 2.1888618042377906e-07, "loss": 0.1809, "step": 9196 }, { "epoch": 1.74, "grad_norm": 8.25, "learning_rate": 2.1857286491581696e-07, "loss": 0.1675, "step": 9197 }, { "epoch": 1.74, "grad_norm": 14.0, "learning_rate": 2.1825976355866446e-07, "loss": 0.2032, "step": 9198 }, { "epoch": 1.74, "grad_norm": 10.125, "learning_rate": 2.1794687638170996e-07, "loss": 0.1952, "step": 9199 }, { "epoch": 1.74, "grad_norm": 11.625, "learning_rate": 2.1763420341432362e-07, "loss": 0.2169, "step": 9200 }, { "epoch": 1.74, "grad_norm": 7.84375, "learning_rate": 2.1732174468585498e-07, "loss": 0.1812, "step": 9201 }, { "epoch": 1.74, "grad_norm": 21.625, "learning_rate": 2.1700950022563448e-07, "loss": 0.2461, "step": 9202 }, { "epoch": 1.74, "grad_norm": 11.0625, "learning_rate": 2.1669747006297066e-07, "loss": 0.1654, "step": 9203 }, { "epoch": 1.74, "grad_norm": 11.1875, "learning_rate": 2.163856542271528e-07, "loss": 0.1891, "step": 9204 }, { "epoch": 1.74, "grad_norm": 11.8125, "learning_rate": 2.1607405274745058e-07, "loss": 0.1759, "step": 9205 }, { "epoch": 1.74, "grad_norm": 9.625, "learning_rate": 2.1576266565311394e-07, "loss": 0.1798, "step": 9206 }, { "epoch": 1.74, "grad_norm": 12.125, "learning_rate": 2.1545149297337003e-07, "loss": 0.1735, "step": 9207 }, { "epoch": 1.74, "grad_norm": 10.9375, "learning_rate": 2.1514053473742856e-07, "loss": 0.2695, "step": 9208 }, { "epoch": 1.74, "grad_norm": 11.5625, "learning_rate": 2.1482979097447898e-07, "loss": 0.2031, "step": 9209 }, { "epoch": 1.74, "grad_norm": 10.5, "learning_rate": 2.1451926171368874e-07, "loss": 0.1656, "step": 9210 }, { "epoch": 1.74, "grad_norm": 13.25, "learning_rate": 2.1420894698420707e-07, "loss": 0.1951, "step": 9211 }, { "epoch": 1.74, "grad_norm": 10.75, "learning_rate": 2.1389884681516121e-07, "loss": 0.1953, "step": 9212 }, { "epoch": 1.74, "grad_norm": 13.4375, "learning_rate": 2.1358896123566147e-07, "loss": 0.2407, "step": 9213 }, { "epoch": 1.74, "grad_norm": 9.4375, "learning_rate": 2.1327929027479377e-07, "loss": 0.1945, "step": 9214 }, { "epoch": 1.74, "grad_norm": 10.4375, "learning_rate": 2.129698339616265e-07, "loss": 0.2213, "step": 9215 }, { "epoch": 1.74, "grad_norm": 9.8125, "learning_rate": 2.1266059232520814e-07, "loss": 0.191, "step": 9216 }, { "epoch": 1.74, "grad_norm": 13.25, "learning_rate": 2.1235156539456624e-07, "loss": 0.2008, "step": 9217 }, { "epoch": 1.74, "grad_norm": 12.4375, "learning_rate": 2.1204275319870738e-07, "loss": 0.2556, "step": 9218 }, { "epoch": 1.74, "grad_norm": 14.25, "learning_rate": 2.1173415576661915e-07, "loss": 0.2601, "step": 9219 }, { "epoch": 1.74, "grad_norm": 11.5, "learning_rate": 2.1142577312726898e-07, "loss": 0.258, "step": 9220 }, { "epoch": 1.74, "grad_norm": 11.6875, "learning_rate": 2.1111760530960422e-07, "loss": 0.2155, "step": 9221 }, { "epoch": 1.74, "grad_norm": 10.875, "learning_rate": 2.1080965234255095e-07, "loss": 0.1884, "step": 9222 }, { "epoch": 1.74, "grad_norm": 11.5, "learning_rate": 2.105019142550152e-07, "loss": 0.1702, "step": 9223 }, { "epoch": 1.74, "grad_norm": 9.1875, "learning_rate": 2.1019439107588497e-07, "loss": 0.1737, "step": 9224 }, { "epoch": 1.75, "grad_norm": 9.9375, "learning_rate": 2.0988708283402632e-07, "loss": 0.1524, "step": 9225 }, { "epoch": 1.75, "grad_norm": 6.46875, "learning_rate": 2.095799895582837e-07, "loss": 0.1448, "step": 9226 }, { "epoch": 1.75, "grad_norm": 10.25, "learning_rate": 2.0927311127748485e-07, "loss": 0.1745, "step": 9227 }, { "epoch": 1.75, "grad_norm": 10.6875, "learning_rate": 2.089664480204348e-07, "loss": 0.2231, "step": 9228 }, { "epoch": 1.75, "grad_norm": 10.8125, "learning_rate": 2.0865999981591972e-07, "loss": 0.2329, "step": 9229 }, { "epoch": 1.75, "grad_norm": 14.125, "learning_rate": 2.083537666927038e-07, "loss": 0.2077, "step": 9230 }, { "epoch": 1.75, "grad_norm": 15.6875, "learning_rate": 2.080477486795332e-07, "loss": 0.217, "step": 9231 }, { "epoch": 1.75, "grad_norm": 13.25, "learning_rate": 2.0774194580513247e-07, "loss": 0.2292, "step": 9232 }, { "epoch": 1.75, "grad_norm": 10.4375, "learning_rate": 2.0743635809820723e-07, "loss": 0.226, "step": 9233 }, { "epoch": 1.75, "grad_norm": 9.0625, "learning_rate": 2.0713098558744067e-07, "loss": 0.1753, "step": 9234 }, { "epoch": 1.75, "grad_norm": 14.9375, "learning_rate": 2.0682582830149816e-07, "loss": 0.2291, "step": 9235 }, { "epoch": 1.75, "grad_norm": 8.5625, "learning_rate": 2.0652088626902434e-07, "loss": 0.1497, "step": 9236 }, { "epoch": 1.75, "grad_norm": 13.1875, "learning_rate": 2.0621615951864181e-07, "loss": 0.2634, "step": 9237 }, { "epoch": 1.75, "grad_norm": 9.125, "learning_rate": 2.0591164807895553e-07, "loss": 0.1935, "step": 9238 }, { "epoch": 1.75, "grad_norm": 15.6875, "learning_rate": 2.056073519785487e-07, "loss": 0.2312, "step": 9239 }, { "epoch": 1.75, "grad_norm": 10.75, "learning_rate": 2.0530327124598514e-07, "loss": 0.2293, "step": 9240 }, { "epoch": 1.75, "grad_norm": 10.5, "learning_rate": 2.0499940590980704e-07, "loss": 0.211, "step": 9241 }, { "epoch": 1.75, "grad_norm": 11.6875, "learning_rate": 2.0469575599853798e-07, "loss": 0.2371, "step": 9242 }, { "epoch": 1.75, "grad_norm": 11.9375, "learning_rate": 2.0439232154068066e-07, "loss": 0.2648, "step": 9243 }, { "epoch": 1.75, "grad_norm": 11.5, "learning_rate": 2.040891025647182e-07, "loss": 0.1842, "step": 9244 }, { "epoch": 1.75, "grad_norm": 10.875, "learning_rate": 2.0378609909911112e-07, "loss": 0.1689, "step": 9245 }, { "epoch": 1.75, "grad_norm": 10.375, "learning_rate": 2.034833111723028e-07, "loss": 0.2143, "step": 9246 }, { "epoch": 1.75, "grad_norm": 13.625, "learning_rate": 2.0318073881271493e-07, "loss": 0.189, "step": 9247 }, { "epoch": 1.75, "grad_norm": 8.5, "learning_rate": 2.0287838204874927e-07, "loss": 0.1774, "step": 9248 }, { "epoch": 1.75, "grad_norm": 10.625, "learning_rate": 2.0257624090878615e-07, "loss": 0.2266, "step": 9249 }, { "epoch": 1.75, "grad_norm": 13.125, "learning_rate": 2.022743154211876e-07, "loss": 0.1997, "step": 9250 }, { "epoch": 1.75, "grad_norm": 10.4375, "learning_rate": 2.01972605614294e-07, "loss": 0.2047, "step": 9251 }, { "epoch": 1.75, "grad_norm": 11.25, "learning_rate": 2.016711115164263e-07, "loss": 0.2007, "step": 9252 }, { "epoch": 1.75, "grad_norm": 10.75, "learning_rate": 2.013698331558847e-07, "loss": 0.1693, "step": 9253 }, { "epoch": 1.75, "grad_norm": 11.6875, "learning_rate": 2.0106877056094931e-07, "loss": 0.2296, "step": 9254 }, { "epoch": 1.75, "grad_norm": 12.25, "learning_rate": 2.0076792375988003e-07, "loss": 0.172, "step": 9255 }, { "epoch": 1.75, "grad_norm": 9.9375, "learning_rate": 2.004672927809162e-07, "loss": 0.2218, "step": 9256 }, { "epoch": 1.75, "grad_norm": 8.6875, "learning_rate": 2.001668776522775e-07, "loss": 0.1583, "step": 9257 }, { "epoch": 1.75, "grad_norm": 8.4375, "learning_rate": 1.99866678402163e-07, "loss": 0.1791, "step": 9258 }, { "epoch": 1.75, "grad_norm": 9.875, "learning_rate": 1.9956669505875133e-07, "loss": 0.2038, "step": 9259 }, { "epoch": 1.75, "grad_norm": 15.75, "learning_rate": 1.99266927650201e-07, "loss": 0.1779, "step": 9260 }, { "epoch": 1.75, "grad_norm": 11.3125, "learning_rate": 1.9896737620465038e-07, "loss": 0.2253, "step": 9261 }, { "epoch": 1.75, "grad_norm": 12.125, "learning_rate": 1.9866804075021756e-07, "loss": 0.2046, "step": 9262 }, { "epoch": 1.75, "grad_norm": 10.5, "learning_rate": 1.98368921315e-07, "loss": 0.1999, "step": 9263 }, { "epoch": 1.75, "grad_norm": 11.0, "learning_rate": 1.9807001792707558e-07, "loss": 0.2135, "step": 9264 }, { "epoch": 1.75, "grad_norm": 13.125, "learning_rate": 1.9777133061450126e-07, "loss": 0.2409, "step": 9265 }, { "epoch": 1.75, "grad_norm": 13.9375, "learning_rate": 1.974728594053138e-07, "loss": 0.1491, "step": 9266 }, { "epoch": 1.75, "grad_norm": 8.875, "learning_rate": 1.9717460432752994e-07, "loss": 0.1991, "step": 9267 }, { "epoch": 1.75, "grad_norm": 10.125, "learning_rate": 1.968765654091459e-07, "loss": 0.1982, "step": 9268 }, { "epoch": 1.75, "grad_norm": 10.0625, "learning_rate": 1.9657874267813793e-07, "loss": 0.16, "step": 9269 }, { "epoch": 1.75, "grad_norm": 9.625, "learning_rate": 1.9628113616246169e-07, "loss": 0.1943, "step": 9270 }, { "epoch": 1.75, "grad_norm": 13.1875, "learning_rate": 1.9598374589005237e-07, "loss": 0.2198, "step": 9271 }, { "epoch": 1.75, "grad_norm": 9.625, "learning_rate": 1.956865718888254e-07, "loss": 0.1962, "step": 9272 }, { "epoch": 1.75, "grad_norm": 12.375, "learning_rate": 1.953896141866757e-07, "loss": 0.2166, "step": 9273 }, { "epoch": 1.75, "grad_norm": 12.625, "learning_rate": 1.950928728114776e-07, "loss": 0.2031, "step": 9274 }, { "epoch": 1.75, "grad_norm": 10.9375, "learning_rate": 1.9479634779108524e-07, "loss": 0.2122, "step": 9275 }, { "epoch": 1.75, "grad_norm": 12.3125, "learning_rate": 1.9450003915333277e-07, "loss": 0.1539, "step": 9276 }, { "epoch": 1.75, "grad_norm": 14.75, "learning_rate": 1.9420394692603372e-07, "loss": 0.2123, "step": 9277 }, { "epoch": 1.76, "grad_norm": 14.625, "learning_rate": 1.9390807113698117e-07, "loss": 0.2375, "step": 9278 }, { "epoch": 1.76, "grad_norm": 8.0625, "learning_rate": 1.9361241181394875e-07, "loss": 0.1774, "step": 9279 }, { "epoch": 1.76, "grad_norm": 9.5625, "learning_rate": 1.9331696898468837e-07, "loss": 0.1818, "step": 9280 }, { "epoch": 1.76, "grad_norm": 16.625, "learning_rate": 1.930217426769329e-07, "loss": 0.1996, "step": 9281 }, { "epoch": 1.76, "grad_norm": 9.4375, "learning_rate": 1.9272673291839405e-07, "loss": 0.19, "step": 9282 }, { "epoch": 1.76, "grad_norm": 14.625, "learning_rate": 1.9243193973676355e-07, "loss": 0.24, "step": 9283 }, { "epoch": 1.76, "grad_norm": 10.4375, "learning_rate": 1.9213736315971316e-07, "loss": 0.1764, "step": 9284 }, { "epoch": 1.76, "grad_norm": 10.0, "learning_rate": 1.918430032148927e-07, "loss": 0.1975, "step": 9285 }, { "epoch": 1.76, "grad_norm": 11.5, "learning_rate": 1.9154885992993506e-07, "loss": 0.2284, "step": 9286 }, { "epoch": 1.76, "grad_norm": 14.9375, "learning_rate": 1.9125493333244844e-07, "loss": 0.2319, "step": 9287 }, { "epoch": 1.76, "grad_norm": 11.4375, "learning_rate": 1.909612234500241e-07, "loss": 0.2193, "step": 9288 }, { "epoch": 1.76, "grad_norm": 14.125, "learning_rate": 1.9066773031023055e-07, "loss": 0.2129, "step": 9289 }, { "epoch": 1.76, "grad_norm": 9.3125, "learning_rate": 1.9037445394061909e-07, "loss": 0.1758, "step": 9290 }, { "epoch": 1.76, "grad_norm": 10.625, "learning_rate": 1.9008139436871688e-07, "loss": 0.1878, "step": 9291 }, { "epoch": 1.76, "grad_norm": 12.0625, "learning_rate": 1.8978855162203325e-07, "loss": 0.2376, "step": 9292 }, { "epoch": 1.76, "grad_norm": 12.1875, "learning_rate": 1.8949592572805654e-07, "loss": 0.2096, "step": 9293 }, { "epoch": 1.76, "grad_norm": 13.625, "learning_rate": 1.8920351671425447e-07, "loss": 0.2172, "step": 9294 }, { "epoch": 1.76, "grad_norm": 18.375, "learning_rate": 1.8891132460807455e-07, "loss": 0.2301, "step": 9295 }, { "epoch": 1.76, "grad_norm": 10.125, "learning_rate": 1.8861934943694398e-07, "loss": 0.1714, "step": 9296 }, { "epoch": 1.76, "grad_norm": 13.25, "learning_rate": 1.883275912282703e-07, "loss": 0.2208, "step": 9297 }, { "epoch": 1.76, "grad_norm": 9.6875, "learning_rate": 1.8803605000943915e-07, "loss": 0.1601, "step": 9298 }, { "epoch": 1.76, "grad_norm": 10.125, "learning_rate": 1.8774472580781717e-07, "loss": 0.1901, "step": 9299 }, { "epoch": 1.76, "grad_norm": 12.375, "learning_rate": 1.8745361865074918e-07, "loss": 0.2497, "step": 9300 }, { "epoch": 1.76, "grad_norm": 12.25, "learning_rate": 1.8716272856556222e-07, "loss": 0.1919, "step": 9301 }, { "epoch": 1.76, "grad_norm": 10.875, "learning_rate": 1.8687205557955974e-07, "loss": 0.2166, "step": 9302 }, { "epoch": 1.76, "grad_norm": 14.75, "learning_rate": 1.865815997200268e-07, "loss": 0.2092, "step": 9303 }, { "epoch": 1.76, "grad_norm": 12.0, "learning_rate": 1.862913610142275e-07, "loss": 0.1891, "step": 9304 }, { "epoch": 1.76, "grad_norm": 12.75, "learning_rate": 1.8600133948940697e-07, "loss": 0.1901, "step": 9305 }, { "epoch": 1.76, "grad_norm": 13.0625, "learning_rate": 1.8571153517278672e-07, "loss": 0.2286, "step": 9306 }, { "epoch": 1.76, "grad_norm": 12.125, "learning_rate": 1.8542194809157062e-07, "loss": 0.1909, "step": 9307 }, { "epoch": 1.76, "grad_norm": 10.3125, "learning_rate": 1.8513257827294183e-07, "loss": 0.2062, "step": 9308 }, { "epoch": 1.76, "grad_norm": 10.0625, "learning_rate": 1.848434257440629e-07, "loss": 0.2166, "step": 9309 }, { "epoch": 1.76, "grad_norm": 9.625, "learning_rate": 1.8455449053207452e-07, "loss": 0.1809, "step": 9310 }, { "epoch": 1.76, "grad_norm": 15.375, "learning_rate": 1.8426577266409811e-07, "loss": 0.2021, "step": 9311 }, { "epoch": 1.76, "grad_norm": 13.4375, "learning_rate": 1.8397727216723616e-07, "loss": 0.2724, "step": 9312 }, { "epoch": 1.76, "grad_norm": 8.8125, "learning_rate": 1.8368898906856892e-07, "loss": 0.2133, "step": 9313 }, { "epoch": 1.76, "grad_norm": 12.8125, "learning_rate": 1.8340092339515585e-07, "loss": 0.2405, "step": 9314 }, { "epoch": 1.76, "grad_norm": 12.8125, "learning_rate": 1.8311307517403704e-07, "loss": 0.2135, "step": 9315 }, { "epoch": 1.76, "grad_norm": 16.625, "learning_rate": 1.828254444322325e-07, "loss": 0.2712, "step": 9316 }, { "epoch": 1.76, "grad_norm": 14.3125, "learning_rate": 1.8253803119674145e-07, "loss": 0.2083, "step": 9317 }, { "epoch": 1.76, "grad_norm": 11.6875, "learning_rate": 1.8225083549454126e-07, "loss": 0.261, "step": 9318 }, { "epoch": 1.76, "grad_norm": 11.8125, "learning_rate": 1.8196385735259143e-07, "loss": 0.2068, "step": 9319 }, { "epoch": 1.76, "grad_norm": 14.0, "learning_rate": 1.8167709679782957e-07, "loss": 0.2195, "step": 9320 }, { "epoch": 1.76, "grad_norm": 11.5, "learning_rate": 1.8139055385717252e-07, "loss": 0.246, "step": 9321 }, { "epoch": 1.76, "grad_norm": 14.625, "learning_rate": 1.8110422855751707e-07, "loss": 0.2026, "step": 9322 }, { "epoch": 1.76, "grad_norm": 14.0625, "learning_rate": 1.8081812092574037e-07, "loss": 0.1854, "step": 9323 }, { "epoch": 1.76, "grad_norm": 31.625, "learning_rate": 1.8053223098869866e-07, "loss": 0.2245, "step": 9324 }, { "epoch": 1.76, "grad_norm": 14.3125, "learning_rate": 1.802465587732269e-07, "loss": 0.2608, "step": 9325 }, { "epoch": 1.76, "grad_norm": 10.4375, "learning_rate": 1.7996110430614e-07, "loss": 0.2016, "step": 9326 }, { "epoch": 1.76, "grad_norm": 13.1875, "learning_rate": 1.796758676142335e-07, "loss": 0.2154, "step": 9327 }, { "epoch": 1.76, "grad_norm": 9.3125, "learning_rate": 1.793908487242821e-07, "loss": 0.1912, "step": 9328 }, { "epoch": 1.76, "grad_norm": 9.0, "learning_rate": 1.7910604766303852e-07, "loss": 0.1485, "step": 9329 }, { "epoch": 1.76, "grad_norm": 14.625, "learning_rate": 1.788214644572367e-07, "loss": 0.2446, "step": 9330 }, { "epoch": 1.77, "grad_norm": 12.625, "learning_rate": 1.7853709913358995e-07, "loss": 0.1564, "step": 9331 }, { "epoch": 1.77, "grad_norm": 11.5, "learning_rate": 1.7825295171879087e-07, "loss": 0.1781, "step": 9332 }, { "epoch": 1.77, "grad_norm": 10.75, "learning_rate": 1.779690222395103e-07, "loss": 0.2502, "step": 9333 }, { "epoch": 1.77, "grad_norm": 14.0, "learning_rate": 1.7768531072240137e-07, "loss": 0.1945, "step": 9334 }, { "epoch": 1.77, "grad_norm": 13.75, "learning_rate": 1.7740181719409444e-07, "loss": 0.3201, "step": 9335 }, { "epoch": 1.77, "grad_norm": 11.8125, "learning_rate": 1.77118541681201e-07, "loss": 0.1272, "step": 9336 }, { "epoch": 1.77, "grad_norm": 10.125, "learning_rate": 1.7683548421030954e-07, "loss": 0.2121, "step": 9337 }, { "epoch": 1.77, "grad_norm": 12.8125, "learning_rate": 1.7655264480799157e-07, "loss": 0.2429, "step": 9338 }, { "epoch": 1.77, "grad_norm": 10.25, "learning_rate": 1.762700235007958e-07, "loss": 0.1752, "step": 9339 }, { "epoch": 1.77, "grad_norm": 13.6875, "learning_rate": 1.7598762031525108e-07, "loss": 0.2021, "step": 9340 }, { "epoch": 1.77, "grad_norm": 11.0625, "learning_rate": 1.7570543527786533e-07, "loss": 0.2122, "step": 9341 }, { "epoch": 1.77, "grad_norm": 14.0625, "learning_rate": 1.754234684151271e-07, "loss": 0.2321, "step": 9342 }, { "epoch": 1.77, "grad_norm": 11.8125, "learning_rate": 1.7514171975350352e-07, "loss": 0.2222, "step": 9343 }, { "epoch": 1.77, "grad_norm": 17.0, "learning_rate": 1.7486018931944155e-07, "loss": 0.2406, "step": 9344 }, { "epoch": 1.77, "grad_norm": 17.75, "learning_rate": 1.7457887713936722e-07, "loss": 0.2012, "step": 9345 }, { "epoch": 1.77, "grad_norm": 8.25, "learning_rate": 1.742977832396872e-07, "loss": 0.179, "step": 9346 }, { "epoch": 1.77, "grad_norm": 17.875, "learning_rate": 1.7401690764678625e-07, "loss": 0.189, "step": 9347 }, { "epoch": 1.77, "grad_norm": 11.8125, "learning_rate": 1.7373625038702962e-07, "loss": 0.2539, "step": 9348 }, { "epoch": 1.77, "grad_norm": 10.625, "learning_rate": 1.7345581148676182e-07, "loss": 0.2043, "step": 9349 }, { "epoch": 1.77, "grad_norm": 7.84375, "learning_rate": 1.731755909723068e-07, "loss": 0.1239, "step": 9350 }, { "epoch": 1.77, "grad_norm": 8.5625, "learning_rate": 1.7289558886996793e-07, "loss": 0.1814, "step": 9351 }, { "epoch": 1.77, "grad_norm": 9.5, "learning_rate": 1.726158052060284e-07, "loss": 0.1866, "step": 9352 }, { "epoch": 1.77, "grad_norm": 12.8125, "learning_rate": 1.723362400067502e-07, "loss": 0.2054, "step": 9353 }, { "epoch": 1.77, "grad_norm": 9.5625, "learning_rate": 1.7205689329837598e-07, "loss": 0.1768, "step": 9354 }, { "epoch": 1.77, "grad_norm": 13.125, "learning_rate": 1.7177776510712647e-07, "loss": 0.1412, "step": 9355 }, { "epoch": 1.77, "grad_norm": 10.25, "learning_rate": 1.7149885545920313e-07, "loss": 0.2222, "step": 9356 }, { "epoch": 1.77, "grad_norm": 12.9375, "learning_rate": 1.712201643807862e-07, "loss": 0.2039, "step": 9357 }, { "epoch": 1.77, "grad_norm": 11.875, "learning_rate": 1.7094169189803583e-07, "loss": 0.2299, "step": 9358 }, { "epoch": 1.77, "grad_norm": 24.5, "learning_rate": 1.7066343803709086e-07, "loss": 0.2257, "step": 9359 }, { "epoch": 1.77, "grad_norm": 11.25, "learning_rate": 1.7038540282407063e-07, "loss": 0.1996, "step": 9360 }, { "epoch": 1.77, "grad_norm": 18.125, "learning_rate": 1.7010758628507318e-07, "loss": 0.2151, "step": 9361 }, { "epoch": 1.77, "grad_norm": 20.375, "learning_rate": 1.6982998844617655e-07, "loss": 0.2329, "step": 9362 }, { "epoch": 1.77, "grad_norm": 19.625, "learning_rate": 1.6955260933343764e-07, "loss": 0.2671, "step": 9363 }, { "epoch": 1.77, "grad_norm": 18.125, "learning_rate": 1.69275448972894e-07, "loss": 0.2054, "step": 9364 }, { "epoch": 1.77, "grad_norm": 12.25, "learning_rate": 1.6899850739056089e-07, "loss": 0.1828, "step": 9365 }, { "epoch": 1.77, "grad_norm": 10.5625, "learning_rate": 1.6872178461243476e-07, "loss": 0.2243, "step": 9366 }, { "epoch": 1.77, "grad_norm": 13.6875, "learning_rate": 1.684452806644904e-07, "loss": 0.2451, "step": 9367 }, { "epoch": 1.77, "grad_norm": 10.5, "learning_rate": 1.6816899557268256e-07, "loss": 0.1965, "step": 9368 }, { "epoch": 1.77, "grad_norm": 13.6875, "learning_rate": 1.67892929362945e-07, "loss": 0.1963, "step": 9369 }, { "epoch": 1.77, "grad_norm": 11.3125, "learning_rate": 1.6761708206119164e-07, "loss": 0.2344, "step": 9370 }, { "epoch": 1.77, "grad_norm": 12.875, "learning_rate": 1.6734145369331517e-07, "loss": 0.2025, "step": 9371 }, { "epoch": 1.77, "grad_norm": 9.75, "learning_rate": 1.6706604428518818e-07, "loss": 0.2105, "step": 9372 }, { "epoch": 1.77, "grad_norm": 12.75, "learning_rate": 1.6679085386266197e-07, "loss": 0.1736, "step": 9373 }, { "epoch": 1.77, "grad_norm": 10.625, "learning_rate": 1.665158824515692e-07, "loss": 0.2391, "step": 9374 }, { "epoch": 1.77, "grad_norm": 9.1875, "learning_rate": 1.6624113007771946e-07, "loss": 0.1868, "step": 9375 }, { "epoch": 1.77, "grad_norm": 10.875, "learning_rate": 1.6596659676690303e-07, "loss": 0.1823, "step": 9376 }, { "epoch": 1.77, "grad_norm": 10.875, "learning_rate": 1.6569228254489006e-07, "loss": 0.1729, "step": 9377 }, { "epoch": 1.77, "grad_norm": 13.3125, "learning_rate": 1.6541818743742914e-07, "loss": 0.1984, "step": 9378 }, { "epoch": 1.77, "grad_norm": 12.0625, "learning_rate": 1.651443114702489e-07, "loss": 0.2161, "step": 9379 }, { "epoch": 1.77, "grad_norm": 10.625, "learning_rate": 1.6487065466905733e-07, "loss": 0.1895, "step": 9380 }, { "epoch": 1.77, "grad_norm": 12.875, "learning_rate": 1.6459721705954167e-07, "loss": 0.2434, "step": 9381 }, { "epoch": 1.77, "grad_norm": 12.1875, "learning_rate": 1.6432399866736892e-07, "loss": 0.1942, "step": 9382 }, { "epoch": 1.78, "grad_norm": 9.9375, "learning_rate": 1.6405099951818494e-07, "loss": 0.1559, "step": 9383 }, { "epoch": 1.78, "grad_norm": 17.125, "learning_rate": 1.637782196376153e-07, "loss": 0.1874, "step": 9384 }, { "epoch": 1.78, "grad_norm": 9.8125, "learning_rate": 1.6350565905126597e-07, "loss": 0.1824, "step": 9385 }, { "epoch": 1.78, "grad_norm": 12.1875, "learning_rate": 1.6323331778472063e-07, "loss": 0.2141, "step": 9386 }, { "epoch": 1.78, "grad_norm": 13.125, "learning_rate": 1.62961195863543e-07, "loss": 0.1935, "step": 9387 }, { "epoch": 1.78, "grad_norm": 13.0625, "learning_rate": 1.626892933132762e-07, "loss": 0.1956, "step": 9388 }, { "epoch": 1.78, "grad_norm": 12.0, "learning_rate": 1.6241761015944435e-07, "loss": 0.1948, "step": 9389 }, { "epoch": 1.78, "grad_norm": 11.5, "learning_rate": 1.6214614642754782e-07, "loss": 0.1549, "step": 9390 }, { "epoch": 1.78, "grad_norm": 10.4375, "learning_rate": 1.6187490214306905e-07, "loss": 0.2304, "step": 9391 }, { "epoch": 1.78, "grad_norm": 11.0, "learning_rate": 1.616038773314682e-07, "loss": 0.2168, "step": 9392 }, { "epoch": 1.78, "grad_norm": 7.90625, "learning_rate": 1.613330720181866e-07, "loss": 0.1778, "step": 9393 }, { "epoch": 1.78, "grad_norm": 15.6875, "learning_rate": 1.6106248622864334e-07, "loss": 0.1473, "step": 9394 }, { "epoch": 1.78, "grad_norm": 10.9375, "learning_rate": 1.6079211998823703e-07, "loss": 0.2121, "step": 9395 }, { "epoch": 1.78, "grad_norm": 11.125, "learning_rate": 1.6052197332234704e-07, "loss": 0.2093, "step": 9396 }, { "epoch": 1.78, "grad_norm": 10.625, "learning_rate": 1.6025204625633145e-07, "loss": 0.2026, "step": 9397 }, { "epoch": 1.78, "grad_norm": 10.3125, "learning_rate": 1.5998233881552638e-07, "loss": 0.248, "step": 9398 }, { "epoch": 1.78, "grad_norm": 10.5, "learning_rate": 1.597128510252488e-07, "loss": 0.1885, "step": 9399 }, { "epoch": 1.78, "grad_norm": 17.25, "learning_rate": 1.5944358291079515e-07, "loss": 0.2078, "step": 9400 }, { "epoch": 1.78, "grad_norm": 16.125, "learning_rate": 1.591745344974413e-07, "loss": 0.2103, "step": 9401 }, { "epoch": 1.78, "grad_norm": 16.25, "learning_rate": 1.5890570581044096e-07, "loss": 0.2599, "step": 9402 }, { "epoch": 1.78, "grad_norm": 12.25, "learning_rate": 1.586370968750281e-07, "loss": 0.2364, "step": 9403 }, { "epoch": 1.78, "grad_norm": 14.9375, "learning_rate": 1.5836870771641756e-07, "loss": 0.2497, "step": 9404 }, { "epoch": 1.78, "grad_norm": 9.8125, "learning_rate": 1.5810053835980137e-07, "loss": 0.2234, "step": 9405 }, { "epoch": 1.78, "grad_norm": 14.625, "learning_rate": 1.578325888303514e-07, "loss": 0.2106, "step": 9406 }, { "epoch": 1.78, "grad_norm": 18.5, "learning_rate": 1.5756485915322028e-07, "loss": 0.2578, "step": 9407 }, { "epoch": 1.78, "grad_norm": 12.5, "learning_rate": 1.57297349353539e-07, "loss": 0.2008, "step": 9408 }, { "epoch": 1.78, "grad_norm": 13.625, "learning_rate": 1.5703005945641698e-07, "loss": 0.2029, "step": 9409 }, { "epoch": 1.78, "grad_norm": 13.5, "learning_rate": 1.5676298948694407e-07, "loss": 0.2376, "step": 9410 }, { "epoch": 1.78, "grad_norm": 13.75, "learning_rate": 1.5649613947019e-07, "loss": 0.2246, "step": 9411 }, { "epoch": 1.78, "grad_norm": 10.75, "learning_rate": 1.562295094312033e-07, "loss": 0.2261, "step": 9412 }, { "epoch": 1.78, "grad_norm": 19.625, "learning_rate": 1.5596309939501092e-07, "loss": 0.2321, "step": 9413 }, { "epoch": 1.78, "grad_norm": 11.0, "learning_rate": 1.556969093866198e-07, "loss": 0.1787, "step": 9414 }, { "epoch": 1.78, "grad_norm": 10.6875, "learning_rate": 1.5543093943101745e-07, "loss": 0.2148, "step": 9415 }, { "epoch": 1.78, "grad_norm": 9.4375, "learning_rate": 1.551651895531697e-07, "loss": 0.1881, "step": 9416 }, { "epoch": 1.78, "grad_norm": 9.0625, "learning_rate": 1.548996597780203e-07, "loss": 0.1913, "step": 9417 }, { "epoch": 1.78, "grad_norm": 13.8125, "learning_rate": 1.5463435013049533e-07, "loss": 0.2248, "step": 9418 }, { "epoch": 1.78, "grad_norm": 11.4375, "learning_rate": 1.5436926063549768e-07, "loss": 0.1543, "step": 9419 }, { "epoch": 1.78, "grad_norm": 7.78125, "learning_rate": 1.5410439131791106e-07, "loss": 0.1528, "step": 9420 }, { "epoch": 1.78, "grad_norm": 10.125, "learning_rate": 1.5383974220259723e-07, "loss": 0.1872, "step": 9421 }, { "epoch": 1.78, "grad_norm": 11.25, "learning_rate": 1.5357531331439856e-07, "loss": 0.2163, "step": 9422 }, { "epoch": 1.78, "grad_norm": 8.1875, "learning_rate": 1.5331110467813632e-07, "loss": 0.1682, "step": 9423 }, { "epoch": 1.78, "grad_norm": 19.375, "learning_rate": 1.5304711631861118e-07, "loss": 0.1759, "step": 9424 }, { "epoch": 1.78, "grad_norm": 13.5625, "learning_rate": 1.527833482606017e-07, "loss": 0.2653, "step": 9425 }, { "epoch": 1.78, "grad_norm": 8.0625, "learning_rate": 1.5251980052886832e-07, "loss": 0.1887, "step": 9426 }, { "epoch": 1.78, "grad_norm": 11.3125, "learning_rate": 1.5225647314814906e-07, "loss": 0.2269, "step": 9427 }, { "epoch": 1.78, "grad_norm": 14.0, "learning_rate": 1.5199336614316163e-07, "loss": 0.1842, "step": 9428 }, { "epoch": 1.78, "grad_norm": 10.0625, "learning_rate": 1.517304795386032e-07, "loss": 0.1817, "step": 9429 }, { "epoch": 1.78, "grad_norm": 9.6875, "learning_rate": 1.5146781335914988e-07, "loss": 0.1688, "step": 9430 }, { "epoch": 1.78, "grad_norm": 14.8125, "learning_rate": 1.512053676294578e-07, "loss": 0.2535, "step": 9431 }, { "epoch": 1.78, "grad_norm": 10.3125, "learning_rate": 1.5094314237416141e-07, "loss": 0.2004, "step": 9432 }, { "epoch": 1.78, "grad_norm": 9.8125, "learning_rate": 1.5068113761787572e-07, "loss": 0.1952, "step": 9433 }, { "epoch": 1.78, "grad_norm": 10.25, "learning_rate": 1.5041935338519387e-07, "loss": 0.1807, "step": 9434 }, { "epoch": 1.78, "grad_norm": 13.5, "learning_rate": 1.5015778970068923e-07, "loss": 0.2829, "step": 9435 }, { "epoch": 1.79, "grad_norm": 13.1875, "learning_rate": 1.4989644658891272e-07, "loss": 0.2784, "step": 9436 }, { "epoch": 1.79, "grad_norm": 14.75, "learning_rate": 1.4963532407439725e-07, "loss": 0.25, "step": 9437 }, { "epoch": 1.79, "grad_norm": 11.0, "learning_rate": 1.4937442218165314e-07, "loss": 0.2168, "step": 9438 }, { "epoch": 1.79, "grad_norm": 10.3125, "learning_rate": 1.4911374093517062e-07, "loss": 0.2029, "step": 9439 }, { "epoch": 1.79, "grad_norm": 13.6875, "learning_rate": 1.4885328035941866e-07, "loss": 0.1939, "step": 9440 }, { "epoch": 1.79, "grad_norm": 10.6875, "learning_rate": 1.485930404788463e-07, "loss": 0.2353, "step": 9441 }, { "epoch": 1.79, "grad_norm": 12.1875, "learning_rate": 1.483330213178813e-07, "loss": 0.2633, "step": 9442 }, { "epoch": 1.79, "grad_norm": 10.0, "learning_rate": 1.4807322290093102e-07, "loss": 0.1701, "step": 9443 }, { "epoch": 1.79, "grad_norm": 15.0625, "learning_rate": 1.4781364525238205e-07, "loss": 0.2059, "step": 9444 }, { "epoch": 1.79, "grad_norm": 11.8125, "learning_rate": 1.4755428839659997e-07, "loss": 0.2304, "step": 9445 }, { "epoch": 1.79, "grad_norm": 13.75, "learning_rate": 1.4729515235792967e-07, "loss": 0.1476, "step": 9446 }, { "epoch": 1.79, "grad_norm": 11.9375, "learning_rate": 1.4703623716069592e-07, "loss": 0.2193, "step": 9447 }, { "epoch": 1.79, "grad_norm": 10.75, "learning_rate": 1.4677754282920203e-07, "loss": 0.2124, "step": 9448 }, { "epoch": 1.79, "grad_norm": 10.9375, "learning_rate": 1.4651906938773108e-07, "loss": 0.1511, "step": 9449 }, { "epoch": 1.79, "grad_norm": 10.1875, "learning_rate": 1.4626081686054506e-07, "loss": 0.2031, "step": 9450 }, { "epoch": 1.79, "grad_norm": 14.0625, "learning_rate": 1.4600278527188537e-07, "loss": 0.2654, "step": 9451 }, { "epoch": 1.79, "grad_norm": 10.75, "learning_rate": 1.457449746459727e-07, "loss": 0.1595, "step": 9452 }, { "epoch": 1.79, "grad_norm": 8.875, "learning_rate": 1.4548738500700682e-07, "loss": 0.1773, "step": 9453 }, { "epoch": 1.79, "grad_norm": 17.25, "learning_rate": 1.45230016379167e-07, "loss": 0.1967, "step": 9454 }, { "epoch": 1.79, "grad_norm": 10.875, "learning_rate": 1.4497286878661198e-07, "loss": 0.2014, "step": 9455 }, { "epoch": 1.79, "grad_norm": 13.1875, "learning_rate": 1.4471594225347912e-07, "loss": 0.2167, "step": 9456 }, { "epoch": 1.79, "grad_norm": 9.9375, "learning_rate": 1.444592368038855e-07, "loss": 0.186, "step": 9457 }, { "epoch": 1.79, "grad_norm": 14.5, "learning_rate": 1.442027524619269e-07, "loss": 0.2951, "step": 9458 }, { "epoch": 1.79, "grad_norm": 15.625, "learning_rate": 1.4394648925167925e-07, "loss": 0.2716, "step": 9459 }, { "epoch": 1.79, "grad_norm": 15.9375, "learning_rate": 1.4369044719719728e-07, "loss": 0.2739, "step": 9460 }, { "epoch": 1.79, "grad_norm": 10.0, "learning_rate": 1.434346263225142e-07, "loss": 0.221, "step": 9461 }, { "epoch": 1.79, "grad_norm": 7.53125, "learning_rate": 1.4317902665164392e-07, "loss": 0.1142, "step": 9462 }, { "epoch": 1.79, "grad_norm": 11.0, "learning_rate": 1.4292364820857856e-07, "loss": 0.1892, "step": 9463 }, { "epoch": 1.79, "grad_norm": 10.125, "learning_rate": 1.426684910172893e-07, "loss": 0.1945, "step": 9464 }, { "epoch": 1.79, "grad_norm": 11.3125, "learning_rate": 1.4241355510172778e-07, "loss": 0.2302, "step": 9465 }, { "epoch": 1.79, "grad_norm": 11.625, "learning_rate": 1.4215884048582345e-07, "loss": 0.2455, "step": 9466 }, { "epoch": 1.79, "grad_norm": 12.875, "learning_rate": 1.419043471934861e-07, "loss": 0.1898, "step": 9467 }, { "epoch": 1.79, "grad_norm": 17.375, "learning_rate": 1.4165007524860386e-07, "loss": 0.2442, "step": 9468 }, { "epoch": 1.79, "grad_norm": 9.1875, "learning_rate": 1.4139602467504453e-07, "loss": 0.1209, "step": 9469 }, { "epoch": 1.79, "grad_norm": 9.8125, "learning_rate": 1.411421954966552e-07, "loss": 0.1755, "step": 9470 }, { "epoch": 1.79, "grad_norm": 14.0625, "learning_rate": 1.4088858773726232e-07, "loss": 0.2454, "step": 9471 }, { "epoch": 1.79, "grad_norm": 11.0, "learning_rate": 1.4063520142067076e-07, "loss": 0.2363, "step": 9472 }, { "epoch": 1.79, "grad_norm": 10.125, "learning_rate": 1.403820365706654e-07, "loss": 0.1691, "step": 9473 }, { "epoch": 1.79, "grad_norm": 10.25, "learning_rate": 1.4012909321101027e-07, "loss": 0.2153, "step": 9474 }, { "epoch": 1.79, "grad_norm": 14.75, "learning_rate": 1.3987637136544802e-07, "loss": 0.219, "step": 9475 }, { "epoch": 1.79, "grad_norm": 16.5, "learning_rate": 1.396238710577011e-07, "loss": 0.2396, "step": 9476 }, { "epoch": 1.79, "grad_norm": 40.0, "learning_rate": 1.3937159231147134e-07, "loss": 0.1988, "step": 9477 }, { "epoch": 1.79, "grad_norm": 11.8125, "learning_rate": 1.3911953515043874e-07, "loss": 0.1941, "step": 9478 }, { "epoch": 1.79, "grad_norm": 14.25, "learning_rate": 1.3886769959826352e-07, "loss": 0.184, "step": 9479 }, { "epoch": 1.79, "grad_norm": 10.1875, "learning_rate": 1.3861608567858453e-07, "loss": 0.2239, "step": 9480 }, { "epoch": 1.79, "grad_norm": 10.1875, "learning_rate": 1.383646934150207e-07, "loss": 0.1942, "step": 9481 }, { "epoch": 1.79, "grad_norm": 9.875, "learning_rate": 1.3811352283116863e-07, "loss": 0.1785, "step": 9482 }, { "epoch": 1.79, "grad_norm": 9.6875, "learning_rate": 1.378625739506051e-07, "loss": 0.1731, "step": 9483 }, { "epoch": 1.79, "grad_norm": 14.4375, "learning_rate": 1.3761184679688595e-07, "loss": 0.2155, "step": 9484 }, { "epoch": 1.79, "grad_norm": 8.5, "learning_rate": 1.3736134139354707e-07, "loss": 0.1863, "step": 9485 }, { "epoch": 1.79, "grad_norm": 12.8125, "learning_rate": 1.3711105776410134e-07, "loss": 0.1798, "step": 9486 }, { "epoch": 1.79, "grad_norm": 10.375, "learning_rate": 1.3686099593204271e-07, "loss": 0.2317, "step": 9487 }, { "epoch": 1.79, "grad_norm": 10.8125, "learning_rate": 1.366111559208444e-07, "loss": 0.2178, "step": 9488 }, { "epoch": 1.8, "grad_norm": 13.875, "learning_rate": 1.3636153775395705e-07, "loss": 0.2553, "step": 9489 }, { "epoch": 1.8, "grad_norm": 16.25, "learning_rate": 1.3611214145481223e-07, "loss": 0.232, "step": 9490 }, { "epoch": 1.8, "grad_norm": 10.5, "learning_rate": 1.358629670468195e-07, "loss": 0.1994, "step": 9491 }, { "epoch": 1.8, "grad_norm": 17.625, "learning_rate": 1.356140145533691e-07, "loss": 0.2377, "step": 9492 }, { "epoch": 1.8, "grad_norm": 12.75, "learning_rate": 1.3536528399782839e-07, "loss": 0.2056, "step": 9493 }, { "epoch": 1.8, "grad_norm": 14.0625, "learning_rate": 1.3511677540354567e-07, "loss": 0.2375, "step": 9494 }, { "epoch": 1.8, "grad_norm": 12.1875, "learning_rate": 1.3486848879384702e-07, "loss": 0.1883, "step": 9495 }, { "epoch": 1.8, "grad_norm": 10.1875, "learning_rate": 1.3462042419203957e-07, "loss": 0.1741, "step": 9496 }, { "epoch": 1.8, "grad_norm": 12.5, "learning_rate": 1.3437258162140727e-07, "loss": 0.1675, "step": 9497 }, { "epoch": 1.8, "grad_norm": 17.375, "learning_rate": 1.341249611052145e-07, "loss": 0.2408, "step": 9498 }, { "epoch": 1.8, "grad_norm": 14.1875, "learning_rate": 1.3387756266670522e-07, "loss": 0.2143, "step": 9499 }, { "epoch": 1.8, "grad_norm": 10.8125, "learning_rate": 1.336303863291019e-07, "loss": 0.1976, "step": 9500 }, { "epoch": 1.8, "grad_norm": 14.0625, "learning_rate": 1.3338343211560605e-07, "loss": 0.248, "step": 9501 }, { "epoch": 1.8, "grad_norm": 9.8125, "learning_rate": 1.3313670004939772e-07, "loss": 0.191, "step": 9502 }, { "epoch": 1.8, "grad_norm": 15.0625, "learning_rate": 1.3289019015363835e-07, "loss": 0.2073, "step": 9503 }, { "epoch": 1.8, "grad_norm": 13.6875, "learning_rate": 1.3264390245146697e-07, "loss": 0.2486, "step": 9504 }, { "epoch": 1.8, "grad_norm": 10.875, "learning_rate": 1.323978369660006e-07, "loss": 0.1976, "step": 9505 }, { "epoch": 1.8, "grad_norm": 11.5, "learning_rate": 1.321519937203375e-07, "loss": 0.2074, "step": 9506 }, { "epoch": 1.8, "grad_norm": 16.625, "learning_rate": 1.3190637273755414e-07, "loss": 0.2362, "step": 9507 }, { "epoch": 1.8, "grad_norm": 11.8125, "learning_rate": 1.3166097404070683e-07, "loss": 0.156, "step": 9508 }, { "epoch": 1.8, "grad_norm": 14.3125, "learning_rate": 1.3141579765282936e-07, "loss": 0.2135, "step": 9509 }, { "epoch": 1.8, "grad_norm": 11.6875, "learning_rate": 1.3117084359693637e-07, "loss": 0.1905, "step": 9510 }, { "epoch": 1.8, "grad_norm": 12.0625, "learning_rate": 1.3092611189602088e-07, "loss": 0.1726, "step": 9511 }, { "epoch": 1.8, "grad_norm": 15.5625, "learning_rate": 1.3068160257305534e-07, "loss": 0.2055, "step": 9512 }, { "epoch": 1.8, "grad_norm": 7.9375, "learning_rate": 1.3043731565099027e-07, "loss": 0.1993, "step": 9513 }, { "epoch": 1.8, "grad_norm": 9.375, "learning_rate": 1.3019325115275733e-07, "loss": 0.2358, "step": 9514 }, { "epoch": 1.8, "grad_norm": 15.6875, "learning_rate": 1.2994940910126514e-07, "loss": 0.2159, "step": 9515 }, { "epoch": 1.8, "grad_norm": 11.875, "learning_rate": 1.2970578951940345e-07, "loss": 0.211, "step": 9516 }, { "epoch": 1.8, "grad_norm": 14.0, "learning_rate": 1.2946239243003867e-07, "loss": 0.2871, "step": 9517 }, { "epoch": 1.8, "grad_norm": 13.4375, "learning_rate": 1.2921921785601894e-07, "loss": 0.2077, "step": 9518 }, { "epoch": 1.8, "grad_norm": 19.375, "learning_rate": 1.2897626582017042e-07, "loss": 0.2618, "step": 9519 }, { "epoch": 1.8, "grad_norm": 10.0625, "learning_rate": 1.2873353634529711e-07, "loss": 0.1697, "step": 9520 }, { "epoch": 1.8, "grad_norm": 14.5625, "learning_rate": 1.2849102945418469e-07, "loss": 0.2153, "step": 9521 }, { "epoch": 1.8, "grad_norm": 10.4375, "learning_rate": 1.2824874516959574e-07, "loss": 0.2608, "step": 9522 }, { "epoch": 1.8, "grad_norm": 9.3125, "learning_rate": 1.280066835142732e-07, "loss": 0.2169, "step": 9523 }, { "epoch": 1.8, "grad_norm": 9.875, "learning_rate": 1.277648445109378e-07, "loss": 0.1735, "step": 9524 }, { "epoch": 1.8, "grad_norm": 14.75, "learning_rate": 1.2752322818229134e-07, "loss": 0.1811, "step": 9525 }, { "epoch": 1.8, "grad_norm": 13.1875, "learning_rate": 1.2728183455101295e-07, "loss": 0.199, "step": 9526 }, { "epoch": 1.8, "grad_norm": 11.8125, "learning_rate": 1.2704066363976253e-07, "loss": 0.2472, "step": 9527 }, { "epoch": 1.8, "grad_norm": 11.3125, "learning_rate": 1.2679971547117642e-07, "loss": 0.1993, "step": 9528 }, { "epoch": 1.8, "grad_norm": 11.625, "learning_rate": 1.2655899006787291e-07, "loss": 0.2057, "step": 9529 }, { "epoch": 1.8, "grad_norm": 11.875, "learning_rate": 1.2631848745244784e-07, "loss": 0.1668, "step": 9530 }, { "epoch": 1.8, "grad_norm": 10.6875, "learning_rate": 1.2607820764747676e-07, "loss": 0.1976, "step": 9531 }, { "epoch": 1.8, "grad_norm": 12.6875, "learning_rate": 1.2583815067551352e-07, "loss": 0.1985, "step": 9532 }, { "epoch": 1.8, "grad_norm": 8.5, "learning_rate": 1.2559831655909182e-07, "loss": 0.1537, "step": 9533 }, { "epoch": 1.8, "grad_norm": 14.6875, "learning_rate": 1.2535870532072447e-07, "loss": 0.231, "step": 9534 }, { "epoch": 1.8, "grad_norm": 11.8125, "learning_rate": 1.2511931698290265e-07, "loss": 0.1759, "step": 9535 }, { "epoch": 1.8, "grad_norm": 11.5625, "learning_rate": 1.2488015156809696e-07, "loss": 0.1738, "step": 9536 }, { "epoch": 1.8, "grad_norm": 16.375, "learning_rate": 1.2464120909875753e-07, "loss": 0.2895, "step": 9537 }, { "epoch": 1.8, "grad_norm": 9.875, "learning_rate": 1.244024895973131e-07, "loss": 0.2333, "step": 9538 }, { "epoch": 1.8, "grad_norm": 11.25, "learning_rate": 1.2416399308617155e-07, "loss": 0.216, "step": 9539 }, { "epoch": 1.8, "grad_norm": 12.0625, "learning_rate": 1.2392571958771943e-07, "loss": 0.1862, "step": 9540 }, { "epoch": 1.8, "grad_norm": 11.875, "learning_rate": 1.236876691243233e-07, "loss": 0.2799, "step": 9541 }, { "epoch": 1.81, "grad_norm": 13.5625, "learning_rate": 1.2344984171832803e-07, "loss": 0.2799, "step": 9542 }, { "epoch": 1.81, "grad_norm": 10.875, "learning_rate": 1.2321223739205774e-07, "loss": 0.1836, "step": 9543 }, { "epoch": 1.81, "grad_norm": 12.125, "learning_rate": 1.2297485616781568e-07, "loss": 0.1586, "step": 9544 }, { "epoch": 1.81, "grad_norm": 14.75, "learning_rate": 1.2273769806788433e-07, "loss": 0.2671, "step": 9545 }, { "epoch": 1.81, "grad_norm": 12.4375, "learning_rate": 1.2250076311452447e-07, "loss": 0.2204, "step": 9546 }, { "epoch": 1.81, "grad_norm": 14.125, "learning_rate": 1.222640513299772e-07, "loss": 0.2504, "step": 9547 }, { "epoch": 1.81, "grad_norm": 11.3125, "learning_rate": 1.2202756273646144e-07, "loss": 0.1853, "step": 9548 }, { "epoch": 1.81, "grad_norm": 13.375, "learning_rate": 1.217912973561758e-07, "loss": 0.2228, "step": 9549 }, { "epoch": 1.81, "grad_norm": 9.0625, "learning_rate": 1.215552552112978e-07, "loss": 0.2037, "step": 9550 }, { "epoch": 1.81, "grad_norm": 8.0625, "learning_rate": 1.213194363239842e-07, "loss": 0.1627, "step": 9551 }, { "epoch": 1.81, "grad_norm": 15.3125, "learning_rate": 1.2108384071637057e-07, "loss": 0.1822, "step": 9552 }, { "epoch": 1.81, "grad_norm": 13.6875, "learning_rate": 1.2084846841057124e-07, "loss": 0.1973, "step": 9553 }, { "epoch": 1.81, "grad_norm": 8.5, "learning_rate": 1.2061331942868016e-07, "loss": 0.1732, "step": 9554 }, { "epoch": 1.81, "grad_norm": 13.5625, "learning_rate": 1.2037839379277023e-07, "loss": 0.2155, "step": 9555 }, { "epoch": 1.81, "grad_norm": 12.5, "learning_rate": 1.2014369152489296e-07, "loss": 0.2047, "step": 9556 }, { "epoch": 1.81, "grad_norm": 11.9375, "learning_rate": 1.1990921264707938e-07, "loss": 0.1711, "step": 9557 }, { "epoch": 1.81, "grad_norm": 13.375, "learning_rate": 1.1967495718133908e-07, "loss": 0.264, "step": 9558 }, { "epoch": 1.81, "grad_norm": 13.6875, "learning_rate": 1.1944092514966088e-07, "loss": 0.2052, "step": 9559 }, { "epoch": 1.81, "grad_norm": 11.375, "learning_rate": 1.1920711657401301e-07, "loss": 0.2561, "step": 9560 }, { "epoch": 1.81, "grad_norm": 11.625, "learning_rate": 1.1897353147634238e-07, "loss": 0.2039, "step": 9561 }, { "epoch": 1.81, "grad_norm": 8.9375, "learning_rate": 1.187401698785745e-07, "loss": 0.1707, "step": 9562 }, { "epoch": 1.81, "grad_norm": 12.8125, "learning_rate": 1.1850703180261464e-07, "loss": 0.1764, "step": 9563 }, { "epoch": 1.81, "grad_norm": 14.5625, "learning_rate": 1.1827411727034665e-07, "loss": 0.2634, "step": 9564 }, { "epoch": 1.81, "grad_norm": 12.125, "learning_rate": 1.1804142630363418e-07, "loss": 0.1518, "step": 9565 }, { "epoch": 1.81, "grad_norm": 10.9375, "learning_rate": 1.178089589243181e-07, "loss": 0.2224, "step": 9566 }, { "epoch": 1.81, "grad_norm": 10.3125, "learning_rate": 1.1757671515422037e-07, "loss": 0.2071, "step": 9567 }, { "epoch": 1.81, "grad_norm": 13.75, "learning_rate": 1.1734469501513995e-07, "loss": 0.1825, "step": 9568 }, { "epoch": 1.81, "grad_norm": 11.9375, "learning_rate": 1.1711289852885749e-07, "loss": 0.231, "step": 9569 }, { "epoch": 1.81, "grad_norm": 10.5625, "learning_rate": 1.1688132571713002e-07, "loss": 0.1558, "step": 9570 }, { "epoch": 1.81, "grad_norm": 13.3125, "learning_rate": 1.1664997660169431e-07, "loss": 0.2199, "step": 9571 }, { "epoch": 1.81, "grad_norm": 12.1875, "learning_rate": 1.1641885120426716e-07, "loss": 0.2018, "step": 9572 }, { "epoch": 1.81, "grad_norm": 11.125, "learning_rate": 1.1618794954654316e-07, "loss": 0.2076, "step": 9573 }, { "epoch": 1.81, "grad_norm": 13.3125, "learning_rate": 1.1595727165019665e-07, "loss": 0.2908, "step": 9574 }, { "epoch": 1.81, "grad_norm": 8.8125, "learning_rate": 1.1572681753688003e-07, "loss": 0.1946, "step": 9575 }, { "epoch": 1.81, "grad_norm": 11.625, "learning_rate": 1.1549658722822655e-07, "loss": 0.1558, "step": 9576 }, { "epoch": 1.81, "grad_norm": 10.125, "learning_rate": 1.1526658074584612e-07, "loss": 0.1882, "step": 9577 }, { "epoch": 1.81, "grad_norm": 11.0625, "learning_rate": 1.1503679811132956e-07, "loss": 0.2234, "step": 9578 }, { "epoch": 1.81, "grad_norm": 11.0, "learning_rate": 1.1480723934624488e-07, "loss": 0.2064, "step": 9579 }, { "epoch": 1.81, "grad_norm": 11.0, "learning_rate": 1.1457790447214152e-07, "loss": 0.214, "step": 9580 }, { "epoch": 1.81, "grad_norm": 19.875, "learning_rate": 1.1434879351054501e-07, "loss": 0.2103, "step": 9581 }, { "epoch": 1.81, "grad_norm": 9.6875, "learning_rate": 1.1411990648296234e-07, "loss": 0.2015, "step": 9582 }, { "epoch": 1.81, "grad_norm": 15.1875, "learning_rate": 1.138912434108777e-07, "loss": 0.2243, "step": 9583 }, { "epoch": 1.81, "grad_norm": 10.4375, "learning_rate": 1.1366280431575588e-07, "loss": 0.2287, "step": 9584 }, { "epoch": 1.81, "grad_norm": 20.375, "learning_rate": 1.1343458921903916e-07, "loss": 0.2432, "step": 9585 }, { "epoch": 1.81, "grad_norm": 10.25, "learning_rate": 1.132065981421493e-07, "loss": 0.2282, "step": 9586 }, { "epoch": 1.81, "grad_norm": 15.4375, "learning_rate": 1.1297883110648777e-07, "loss": 0.2085, "step": 9587 }, { "epoch": 1.81, "grad_norm": 13.6875, "learning_rate": 1.1275128813343417e-07, "loss": 0.2414, "step": 9588 }, { "epoch": 1.81, "grad_norm": 13.0, "learning_rate": 1.1252396924434695e-07, "loss": 0.1608, "step": 9589 }, { "epoch": 1.81, "grad_norm": 9.5625, "learning_rate": 1.1229687446056348e-07, "loss": 0.2115, "step": 9590 }, { "epoch": 1.81, "grad_norm": 8.4375, "learning_rate": 1.1207000380340172e-07, "loss": 0.177, "step": 9591 }, { "epoch": 1.81, "grad_norm": 28.375, "learning_rate": 1.1184335729415685e-07, "loss": 0.2934, "step": 9592 }, { "epoch": 1.81, "grad_norm": 10.1875, "learning_rate": 1.1161693495410298e-07, "loss": 0.1769, "step": 9593 }, { "epoch": 1.81, "grad_norm": 9.125, "learning_rate": 1.1139073680449363e-07, "loss": 0.1768, "step": 9594 }, { "epoch": 1.82, "grad_norm": 10.375, "learning_rate": 1.1116476286656214e-07, "loss": 0.1648, "step": 9595 }, { "epoch": 1.82, "grad_norm": 10.3125, "learning_rate": 1.1093901316152011e-07, "loss": 0.2183, "step": 9596 }, { "epoch": 1.82, "grad_norm": 12.6875, "learning_rate": 1.1071348771055701e-07, "loss": 0.2507, "step": 9597 }, { "epoch": 1.82, "grad_norm": 10.4375, "learning_rate": 1.1048818653484284e-07, "loss": 0.1821, "step": 9598 }, { "epoch": 1.82, "grad_norm": 12.0, "learning_rate": 1.1026310965552594e-07, "loss": 0.2347, "step": 9599 }, { "epoch": 1.82, "grad_norm": 12.1875, "learning_rate": 1.1003825709373389e-07, "loss": 0.2258, "step": 9600 }, { "epoch": 1.82, "grad_norm": 9.4375, "learning_rate": 1.0981362887057201e-07, "loss": 0.1744, "step": 9601 }, { "epoch": 1.82, "grad_norm": 12.8125, "learning_rate": 1.0958922500712648e-07, "loss": 0.2115, "step": 9602 }, { "epoch": 1.82, "grad_norm": 14.1875, "learning_rate": 1.0936504552446103e-07, "loss": 0.2066, "step": 9603 }, { "epoch": 1.82, "grad_norm": 15.3125, "learning_rate": 1.0914109044361881e-07, "loss": 0.1608, "step": 9604 }, { "epoch": 1.82, "grad_norm": 18.875, "learning_rate": 1.0891735978562107e-07, "loss": 0.2288, "step": 9605 }, { "epoch": 1.82, "grad_norm": 12.8125, "learning_rate": 1.0869385357146989e-07, "loss": 0.2549, "step": 9606 }, { "epoch": 1.82, "grad_norm": 7.53125, "learning_rate": 1.0847057182214488e-07, "loss": 0.1463, "step": 9607 }, { "epoch": 1.82, "grad_norm": 9.875, "learning_rate": 1.0824751455860427e-07, "loss": 0.2032, "step": 9608 }, { "epoch": 1.82, "grad_norm": 12.3125, "learning_rate": 1.0802468180178605e-07, "loss": 0.263, "step": 9609 }, { "epoch": 1.82, "grad_norm": 14.125, "learning_rate": 1.0780207357260709e-07, "loss": 0.2092, "step": 9610 }, { "epoch": 1.82, "grad_norm": 9.625, "learning_rate": 1.0757968989196316e-07, "loss": 0.2282, "step": 9611 }, { "epoch": 1.82, "grad_norm": 11.75, "learning_rate": 1.0735753078072786e-07, "loss": 0.2404, "step": 9612 }, { "epoch": 1.82, "grad_norm": 9.375, "learning_rate": 1.0713559625975534e-07, "loss": 0.1211, "step": 9613 }, { "epoch": 1.82, "grad_norm": 13.3125, "learning_rate": 1.0691388634987782e-07, "loss": 0.2631, "step": 9614 }, { "epoch": 1.82, "grad_norm": 13.0, "learning_rate": 1.066924010719067e-07, "loss": 0.2054, "step": 9615 }, { "epoch": 1.82, "grad_norm": 12.3125, "learning_rate": 1.0647114044663148e-07, "loss": 0.2142, "step": 9616 }, { "epoch": 1.82, "grad_norm": 10.125, "learning_rate": 1.0625010449482193e-07, "loss": 0.1848, "step": 9617 }, { "epoch": 1.82, "grad_norm": 12.5625, "learning_rate": 1.0602929323722588e-07, "loss": 0.1654, "step": 9618 }, { "epoch": 1.82, "grad_norm": 20.125, "learning_rate": 1.0580870669457011e-07, "loss": 0.1548, "step": 9619 }, { "epoch": 1.82, "grad_norm": 7.96875, "learning_rate": 1.0558834488756025e-07, "loss": 0.165, "step": 9620 }, { "epoch": 1.82, "grad_norm": 9.9375, "learning_rate": 1.0536820783688145e-07, "loss": 0.2318, "step": 9621 }, { "epoch": 1.82, "grad_norm": 13.0, "learning_rate": 1.0514829556319716e-07, "loss": 0.2445, "step": 9622 }, { "epoch": 1.82, "grad_norm": 11.375, "learning_rate": 1.0492860808714977e-07, "loss": 0.1912, "step": 9623 }, { "epoch": 1.82, "grad_norm": 9.875, "learning_rate": 1.0470914542936084e-07, "loss": 0.2076, "step": 9624 }, { "epoch": 1.82, "grad_norm": 14.1875, "learning_rate": 1.0448990761043054e-07, "loss": 0.1997, "step": 9625 }, { "epoch": 1.82, "grad_norm": 10.5625, "learning_rate": 1.0427089465093798e-07, "loss": 0.2229, "step": 9626 }, { "epoch": 1.82, "grad_norm": 12.1875, "learning_rate": 1.0405210657144171e-07, "loss": 0.1748, "step": 9627 }, { "epoch": 1.82, "grad_norm": 15.375, "learning_rate": 1.0383354339247837e-07, "loss": 0.2058, "step": 9628 }, { "epoch": 1.82, "grad_norm": 17.375, "learning_rate": 1.0361520513456375e-07, "loss": 0.1707, "step": 9629 }, { "epoch": 1.82, "grad_norm": 20.125, "learning_rate": 1.0339709181819285e-07, "loss": 0.2636, "step": 9630 }, { "epoch": 1.82, "grad_norm": 11.0, "learning_rate": 1.0317920346383902e-07, "loss": 0.2188, "step": 9631 }, { "epoch": 1.82, "grad_norm": 9.75, "learning_rate": 1.0296154009195508e-07, "loss": 0.2195, "step": 9632 }, { "epoch": 1.82, "grad_norm": 8.9375, "learning_rate": 1.0274410172297272e-07, "loss": 0.1493, "step": 9633 }, { "epoch": 1.82, "grad_norm": 20.625, "learning_rate": 1.0252688837730146e-07, "loss": 0.2163, "step": 9634 }, { "epoch": 1.82, "grad_norm": 16.25, "learning_rate": 1.0230990007533109e-07, "loss": 0.2204, "step": 9635 }, { "epoch": 1.82, "grad_norm": 8.4375, "learning_rate": 1.020931368374295e-07, "loss": 0.1501, "step": 9636 }, { "epoch": 1.82, "grad_norm": 10.9375, "learning_rate": 1.0187659868394345e-07, "loss": 0.1774, "step": 9637 }, { "epoch": 1.82, "grad_norm": 13.0625, "learning_rate": 1.0166028563519891e-07, "loss": 0.2178, "step": 9638 }, { "epoch": 1.82, "grad_norm": 12.8125, "learning_rate": 1.014441977115005e-07, "loss": 0.2292, "step": 9639 }, { "epoch": 1.82, "grad_norm": 14.5625, "learning_rate": 1.0122833493313195e-07, "loss": 0.2726, "step": 9640 }, { "epoch": 1.82, "grad_norm": 47.5, "learning_rate": 1.010126973203554e-07, "loss": 0.2429, "step": 9641 }, { "epoch": 1.82, "grad_norm": 16.625, "learning_rate": 1.0079728489341189e-07, "loss": 0.2041, "step": 9642 }, { "epoch": 1.82, "grad_norm": 10.0625, "learning_rate": 1.0058209767252219e-07, "loss": 0.1987, "step": 9643 }, { "epoch": 1.82, "grad_norm": 11.875, "learning_rate": 1.0036713567788459e-07, "loss": 0.2493, "step": 9644 }, { "epoch": 1.82, "grad_norm": 15.75, "learning_rate": 1.0015239892967766e-07, "loss": 0.2299, "step": 9645 }, { "epoch": 1.82, "grad_norm": 10.5, "learning_rate": 9.993788744805748e-08, "loss": 0.223, "step": 9646 }, { "epoch": 1.82, "grad_norm": 13.1875, "learning_rate": 9.972360125315995e-08, "loss": 0.2068, "step": 9647 }, { "epoch": 1.83, "grad_norm": 10.25, "learning_rate": 9.950954036509919e-08, "loss": 0.2301, "step": 9648 }, { "epoch": 1.83, "grad_norm": 11.0625, "learning_rate": 9.929570480396888e-08, "loss": 0.1887, "step": 9649 }, { "epoch": 1.83, "grad_norm": 13.0625, "learning_rate": 9.908209458984075e-08, "loss": 0.1845, "step": 9650 }, { "epoch": 1.83, "grad_norm": 12.1875, "learning_rate": 9.886870974276597e-08, "loss": 0.2571, "step": 9651 }, { "epoch": 1.83, "grad_norm": 11.1875, "learning_rate": 9.865555028277435e-08, "loss": 0.2359, "step": 9652 }, { "epoch": 1.83, "grad_norm": 11.6875, "learning_rate": 9.844261622987433e-08, "loss": 0.2528, "step": 9653 }, { "epoch": 1.83, "grad_norm": 13.0, "learning_rate": 9.822990760405355e-08, "loss": 0.2343, "step": 9654 }, { "epoch": 1.83, "grad_norm": 10.3125, "learning_rate": 9.80174244252785e-08, "loss": 0.1915, "step": 9655 }, { "epoch": 1.83, "grad_norm": 9.1875, "learning_rate": 9.78051667134941e-08, "loss": 0.1712, "step": 9656 }, { "epoch": 1.83, "grad_norm": 12.0, "learning_rate": 9.759313448862439e-08, "loss": 0.2049, "step": 9657 }, { "epoch": 1.83, "grad_norm": 9.125, "learning_rate": 9.738132777057208e-08, "loss": 0.1914, "step": 9658 }, { "epoch": 1.83, "grad_norm": 11.3125, "learning_rate": 9.71697465792193e-08, "loss": 0.2317, "step": 9659 }, { "epoch": 1.83, "grad_norm": 12.9375, "learning_rate": 9.6958390934426e-08, "loss": 0.2127, "step": 9660 }, { "epoch": 1.83, "grad_norm": 10.75, "learning_rate": 9.674726085603187e-08, "loss": 0.2492, "step": 9661 }, { "epoch": 1.83, "grad_norm": 9.75, "learning_rate": 9.653635636385522e-08, "loss": 0.1679, "step": 9662 }, { "epoch": 1.83, "grad_norm": 9.75, "learning_rate": 9.632567747769244e-08, "loss": 0.1758, "step": 9663 }, { "epoch": 1.83, "grad_norm": 13.6875, "learning_rate": 9.611522421731967e-08, "loss": 0.2039, "step": 9664 }, { "epoch": 1.83, "grad_norm": 10.6875, "learning_rate": 9.590499660249165e-08, "loss": 0.1747, "step": 9665 }, { "epoch": 1.83, "grad_norm": 10.625, "learning_rate": 9.569499465294174e-08, "loss": 0.1886, "step": 9666 }, { "epoch": 1.83, "grad_norm": 16.125, "learning_rate": 9.54852183883817e-08, "loss": 0.2568, "step": 9667 }, { "epoch": 1.83, "grad_norm": 9.875, "learning_rate": 9.527566782850383e-08, "loss": 0.2, "step": 9668 }, { "epoch": 1.83, "grad_norm": 12.125, "learning_rate": 9.506634299297685e-08, "loss": 0.1896, "step": 9669 }, { "epoch": 1.83, "grad_norm": 10.5625, "learning_rate": 9.485724390144974e-08, "loss": 0.2204, "step": 9670 }, { "epoch": 1.83, "grad_norm": 8.625, "learning_rate": 9.464837057355014e-08, "loss": 0.1348, "step": 9671 }, { "epoch": 1.83, "grad_norm": 15.625, "learning_rate": 9.44397230288846e-08, "loss": 0.1943, "step": 9672 }, { "epoch": 1.83, "grad_norm": 10.375, "learning_rate": 9.4231301287038e-08, "loss": 0.1767, "step": 9673 }, { "epoch": 1.83, "grad_norm": 11.875, "learning_rate": 9.402310536757441e-08, "loss": 0.2403, "step": 9674 }, { "epoch": 1.83, "grad_norm": 12.5625, "learning_rate": 9.381513529003572e-08, "loss": 0.2085, "step": 9675 }, { "epoch": 1.83, "grad_norm": 13.75, "learning_rate": 9.360739107394517e-08, "loss": 0.2278, "step": 9676 }, { "epoch": 1.83, "grad_norm": 11.6875, "learning_rate": 9.339987273880163e-08, "loss": 0.2313, "step": 9677 }, { "epoch": 1.83, "grad_norm": 11.5, "learning_rate": 9.319258030408451e-08, "loss": 0.2323, "step": 9678 }, { "epoch": 1.83, "grad_norm": 12.125, "learning_rate": 9.298551378925213e-08, "loss": 0.2166, "step": 9679 }, { "epoch": 1.83, "grad_norm": 10.9375, "learning_rate": 9.277867321374145e-08, "loss": 0.1984, "step": 9680 }, { "epoch": 1.83, "grad_norm": 28.25, "learning_rate": 9.257205859696722e-08, "loss": 0.2619, "step": 9681 }, { "epoch": 1.83, "grad_norm": 7.875, "learning_rate": 9.236566995832392e-08, "loss": 0.1981, "step": 9682 }, { "epoch": 1.83, "grad_norm": 10.375, "learning_rate": 9.215950731718493e-08, "loss": 0.1849, "step": 9683 }, { "epoch": 1.83, "grad_norm": 13.0625, "learning_rate": 9.195357069290229e-08, "loss": 0.2546, "step": 9684 }, { "epoch": 1.83, "grad_norm": 14.75, "learning_rate": 9.174786010480608e-08, "loss": 0.2639, "step": 9685 }, { "epoch": 1.83, "grad_norm": 9.625, "learning_rate": 9.154237557220557e-08, "loss": 0.2367, "step": 9686 }, { "epoch": 1.83, "grad_norm": 8.875, "learning_rate": 9.133711711439036e-08, "loss": 0.149, "step": 9687 }, { "epoch": 1.83, "grad_norm": 8.625, "learning_rate": 9.113208475062613e-08, "loss": 0.2018, "step": 9688 }, { "epoch": 1.83, "grad_norm": 10.9375, "learning_rate": 9.092727850015859e-08, "loss": 0.1696, "step": 9689 }, { "epoch": 1.83, "grad_norm": 9.875, "learning_rate": 9.07226983822132e-08, "loss": 0.1819, "step": 9690 }, { "epoch": 1.83, "grad_norm": 13.0, "learning_rate": 9.051834441599321e-08, "loss": 0.2403, "step": 9691 }, { "epoch": 1.83, "grad_norm": 14.5, "learning_rate": 9.031421662067996e-08, "loss": 0.2566, "step": 9692 }, { "epoch": 1.83, "grad_norm": 9.8125, "learning_rate": 9.011031501543476e-08, "loss": 0.1837, "step": 9693 }, { "epoch": 1.83, "grad_norm": 15.6875, "learning_rate": 8.990663961939705e-08, "loss": 0.2655, "step": 9694 }, { "epoch": 1.83, "grad_norm": 12.1875, "learning_rate": 8.970319045168596e-08, "loss": 0.2087, "step": 9695 }, { "epoch": 1.83, "grad_norm": 9.5, "learning_rate": 8.949996753139789e-08, "loss": 0.1903, "step": 9696 }, { "epoch": 1.83, "grad_norm": 11.0, "learning_rate": 8.929697087760869e-08, "loss": 0.2289, "step": 9697 }, { "epoch": 1.83, "grad_norm": 14.875, "learning_rate": 8.909420050937368e-08, "loss": 0.2128, "step": 9698 }, { "epoch": 1.83, "grad_norm": 20.125, "learning_rate": 8.889165644572623e-08, "loss": 0.2429, "step": 9699 }, { "epoch": 1.83, "grad_norm": 13.8125, "learning_rate": 8.868933870567752e-08, "loss": 0.1347, "step": 9700 }, { "epoch": 1.84, "grad_norm": 13.75, "learning_rate": 8.84872473082199e-08, "loss": 0.2505, "step": 9701 }, { "epoch": 1.84, "grad_norm": 30.625, "learning_rate": 8.82853822723223e-08, "loss": 0.2036, "step": 9702 }, { "epoch": 1.84, "grad_norm": 13.875, "learning_rate": 8.808374361693378e-08, "loss": 0.2115, "step": 9703 }, { "epoch": 1.84, "grad_norm": 11.5, "learning_rate": 8.788233136098056e-08, "loss": 0.1984, "step": 9704 }, { "epoch": 1.84, "grad_norm": 12.8125, "learning_rate": 8.768114552336976e-08, "loss": 0.2377, "step": 9705 }, { "epoch": 1.84, "grad_norm": 7.96875, "learning_rate": 8.748018612298542e-08, "loss": 0.151, "step": 9706 }, { "epoch": 1.84, "grad_norm": 14.4375, "learning_rate": 8.727945317869135e-08, "loss": 0.2259, "step": 9707 }, { "epoch": 1.84, "grad_norm": 13.3125, "learning_rate": 8.707894670932914e-08, "loss": 0.2367, "step": 9708 }, { "epoch": 1.84, "grad_norm": 12.1875, "learning_rate": 8.687866673372041e-08, "loss": 0.1334, "step": 9709 }, { "epoch": 1.84, "grad_norm": 10.6875, "learning_rate": 8.667861327066456e-08, "loss": 0.1907, "step": 9710 }, { "epoch": 1.84, "grad_norm": 7.21875, "learning_rate": 8.647878633893991e-08, "loss": 0.1549, "step": 9711 }, { "epoch": 1.84, "grad_norm": 14.25, "learning_rate": 8.627918595730394e-08, "loss": 0.2583, "step": 9712 }, { "epoch": 1.84, "grad_norm": 13.5625, "learning_rate": 8.607981214449251e-08, "loss": 0.3008, "step": 9713 }, { "epoch": 1.84, "grad_norm": 11.5625, "learning_rate": 8.58806649192201e-08, "loss": 0.2002, "step": 9714 }, { "epoch": 1.84, "grad_norm": 9.5, "learning_rate": 8.568174430017978e-08, "loss": 0.1562, "step": 9715 }, { "epoch": 1.84, "grad_norm": 10.125, "learning_rate": 8.548305030604414e-08, "loss": 0.1961, "step": 9716 }, { "epoch": 1.84, "grad_norm": 18.875, "learning_rate": 8.528458295546382e-08, "loss": 0.1978, "step": 9717 }, { "epoch": 1.84, "grad_norm": 15.8125, "learning_rate": 8.508634226706836e-08, "loss": 0.2495, "step": 9718 }, { "epoch": 1.84, "grad_norm": 18.125, "learning_rate": 8.488832825946591e-08, "loss": 0.2047, "step": 9719 }, { "epoch": 1.84, "grad_norm": 11.0625, "learning_rate": 8.469054095124357e-08, "loss": 0.1727, "step": 9720 }, { "epoch": 1.84, "grad_norm": 10.4375, "learning_rate": 8.449298036096704e-08, "loss": 0.1917, "step": 9721 }, { "epoch": 1.84, "grad_norm": 11.75, "learning_rate": 8.429564650718063e-08, "loss": 0.2204, "step": 9722 }, { "epoch": 1.84, "grad_norm": 9.875, "learning_rate": 8.40985394084079e-08, "loss": 0.1908, "step": 9723 }, { "epoch": 1.84, "grad_norm": 15.5625, "learning_rate": 8.390165908315013e-08, "loss": 0.1831, "step": 9724 }, { "epoch": 1.84, "grad_norm": 8.75, "learning_rate": 8.370500554988836e-08, "loss": 0.1645, "step": 9725 }, { "epoch": 1.84, "grad_norm": 17.125, "learning_rate": 8.350857882708175e-08, "loss": 0.2555, "step": 9726 }, { "epoch": 1.84, "grad_norm": 11.0625, "learning_rate": 8.331237893316801e-08, "loss": 0.202, "step": 9727 }, { "epoch": 1.84, "grad_norm": 15.3125, "learning_rate": 8.311640588656411e-08, "loss": 0.1682, "step": 9728 }, { "epoch": 1.84, "grad_norm": 12.75, "learning_rate": 8.29206597056656e-08, "loss": 0.2112, "step": 9729 }, { "epoch": 1.84, "grad_norm": 10.6875, "learning_rate": 8.27251404088461e-08, "loss": 0.2208, "step": 9730 }, { "epoch": 1.84, "grad_norm": 17.75, "learning_rate": 8.252984801445901e-08, "loss": 0.1941, "step": 9731 }, { "epoch": 1.84, "grad_norm": 7.9375, "learning_rate": 8.233478254083549e-08, "loss": 0.1392, "step": 9732 }, { "epoch": 1.84, "grad_norm": 14.75, "learning_rate": 8.213994400628589e-08, "loss": 0.2412, "step": 9733 }, { "epoch": 1.84, "grad_norm": 9.8125, "learning_rate": 8.194533242909891e-08, "loss": 0.2046, "step": 9734 }, { "epoch": 1.84, "grad_norm": 11.875, "learning_rate": 8.175094782754245e-08, "loss": 0.1927, "step": 9735 }, { "epoch": 1.84, "grad_norm": 11.625, "learning_rate": 8.155679021986274e-08, "loss": 0.2702, "step": 9736 }, { "epoch": 1.84, "grad_norm": 11.75, "learning_rate": 8.136285962428491e-08, "loss": 0.2322, "step": 9737 }, { "epoch": 1.84, "grad_norm": 15.125, "learning_rate": 8.116915605901222e-08, "loss": 0.1903, "step": 9738 }, { "epoch": 1.84, "grad_norm": 12.4375, "learning_rate": 8.097567954222729e-08, "loss": 0.1962, "step": 9739 }, { "epoch": 1.84, "grad_norm": 9.5, "learning_rate": 8.07824300920912e-08, "loss": 0.1542, "step": 9740 }, { "epoch": 1.84, "grad_norm": 8.625, "learning_rate": 8.058940772674384e-08, "loss": 0.1465, "step": 9741 }, { "epoch": 1.84, "grad_norm": 11.0, "learning_rate": 8.039661246430353e-08, "loss": 0.1327, "step": 9742 }, { "epoch": 1.84, "grad_norm": 11.1875, "learning_rate": 8.020404432286716e-08, "loss": 0.2268, "step": 9743 }, { "epoch": 1.84, "grad_norm": 14.5, "learning_rate": 8.001170332051111e-08, "loss": 0.2272, "step": 9744 }, { "epoch": 1.84, "grad_norm": 11.125, "learning_rate": 7.981958947528929e-08, "loss": 0.1407, "step": 9745 }, { "epoch": 1.84, "grad_norm": 11.375, "learning_rate": 7.962770280523502e-08, "loss": 0.1946, "step": 9746 }, { "epoch": 1.84, "grad_norm": 13.375, "learning_rate": 7.94360433283603e-08, "loss": 0.1917, "step": 9747 }, { "epoch": 1.84, "grad_norm": 8.375, "learning_rate": 7.924461106265546e-08, "loss": 0.176, "step": 9748 }, { "epoch": 1.84, "grad_norm": 12.4375, "learning_rate": 7.905340602608974e-08, "loss": 0.2438, "step": 9749 }, { "epoch": 1.84, "grad_norm": 13.6875, "learning_rate": 7.8862428236611e-08, "loss": 0.186, "step": 9750 }, { "epoch": 1.84, "grad_norm": 10.4375, "learning_rate": 7.867167771214573e-08, "loss": 0.168, "step": 9751 }, { "epoch": 1.84, "grad_norm": 15.0, "learning_rate": 7.848115447059907e-08, "loss": 0.2385, "step": 9752 }, { "epoch": 1.84, "grad_norm": 10.3125, "learning_rate": 7.829085852985501e-08, "loss": 0.21, "step": 9753 }, { "epoch": 1.85, "grad_norm": 10.375, "learning_rate": 7.810078990777598e-08, "loss": 0.1927, "step": 9754 }, { "epoch": 1.85, "grad_norm": 14.75, "learning_rate": 7.791094862220294e-08, "loss": 0.1716, "step": 9755 }, { "epoch": 1.85, "grad_norm": 12.25, "learning_rate": 7.772133469095639e-08, "loss": 0.2315, "step": 9756 }, { "epoch": 1.85, "grad_norm": 11.6875, "learning_rate": 7.753194813183429e-08, "loss": 0.2219, "step": 9757 }, { "epoch": 1.85, "grad_norm": 11.4375, "learning_rate": 7.734278896261383e-08, "loss": 0.162, "step": 9758 }, { "epoch": 1.85, "grad_norm": 10.8125, "learning_rate": 7.715385720105051e-08, "loss": 0.2043, "step": 9759 }, { "epoch": 1.85, "grad_norm": 11.875, "learning_rate": 7.696515286487988e-08, "loss": 0.1932, "step": 9760 }, { "epoch": 1.85, "grad_norm": 12.875, "learning_rate": 7.677667597181415e-08, "loss": 0.1904, "step": 9761 }, { "epoch": 1.85, "grad_norm": 12.9375, "learning_rate": 7.658842653954501e-08, "loss": 0.2312, "step": 9762 }, { "epoch": 1.85, "grad_norm": 11.375, "learning_rate": 7.640040458574305e-08, "loss": 0.2214, "step": 9763 }, { "epoch": 1.85, "grad_norm": 11.0, "learning_rate": 7.621261012805802e-08, "loss": 0.2184, "step": 9764 }, { "epoch": 1.85, "grad_norm": 11.125, "learning_rate": 7.602504318411696e-08, "loss": 0.2261, "step": 9765 }, { "epoch": 1.85, "grad_norm": 10.875, "learning_rate": 7.583770377152577e-08, "loss": 0.2131, "step": 9766 }, { "epoch": 1.85, "grad_norm": 12.375, "learning_rate": 7.565059190787038e-08, "loss": 0.1727, "step": 9767 }, { "epoch": 1.85, "grad_norm": 10.6875, "learning_rate": 7.546370761071425e-08, "loss": 0.2548, "step": 9768 }, { "epoch": 1.85, "grad_norm": 10.375, "learning_rate": 7.527705089759945e-08, "loss": 0.1971, "step": 9769 }, { "epoch": 1.85, "grad_norm": 11.125, "learning_rate": 7.509062178604615e-08, "loss": 0.19, "step": 9770 }, { "epoch": 1.85, "grad_norm": 10.0, "learning_rate": 7.490442029355561e-08, "loss": 0.1964, "step": 9771 }, { "epoch": 1.85, "grad_norm": 12.1875, "learning_rate": 7.471844643760468e-08, "loss": 0.2247, "step": 9772 }, { "epoch": 1.85, "grad_norm": 9.9375, "learning_rate": 7.453270023565024e-08, "loss": 0.2134, "step": 9773 }, { "epoch": 1.85, "grad_norm": 11.9375, "learning_rate": 7.434718170512805e-08, "loss": 0.2114, "step": 9774 }, { "epoch": 1.85, "grad_norm": 13.0, "learning_rate": 7.416189086345255e-08, "loss": 0.197, "step": 9775 }, { "epoch": 1.85, "grad_norm": 16.625, "learning_rate": 7.39768277280159e-08, "loss": 0.2078, "step": 9776 }, { "epoch": 1.85, "grad_norm": 6.9375, "learning_rate": 7.379199231618922e-08, "loss": 0.1276, "step": 9777 }, { "epoch": 1.85, "grad_norm": 8.625, "learning_rate": 7.360738464532335e-08, "loss": 0.205, "step": 9778 }, { "epoch": 1.85, "grad_norm": 9.1875, "learning_rate": 7.342300473274639e-08, "loss": 0.211, "step": 9779 }, { "epoch": 1.85, "grad_norm": 12.25, "learning_rate": 7.32388525957653e-08, "loss": 0.2184, "step": 9780 }, { "epoch": 1.85, "grad_norm": 13.9375, "learning_rate": 7.305492825166571e-08, "loss": 0.2421, "step": 9781 }, { "epoch": 1.85, "grad_norm": 10.0, "learning_rate": 7.287123171771298e-08, "loss": 0.2143, "step": 9782 }, { "epoch": 1.85, "grad_norm": 12.0, "learning_rate": 7.268776301114944e-08, "loss": 0.2023, "step": 9783 }, { "epoch": 1.85, "grad_norm": 12.3125, "learning_rate": 7.250452214919712e-08, "loss": 0.2315, "step": 9784 }, { "epoch": 1.85, "grad_norm": 11.1875, "learning_rate": 7.232150914905561e-08, "loss": 0.2481, "step": 9785 }, { "epoch": 1.85, "grad_norm": 9.3125, "learning_rate": 7.213872402790451e-08, "loss": 0.201, "step": 9786 }, { "epoch": 1.85, "grad_norm": 11.3125, "learning_rate": 7.195616680290146e-08, "loss": 0.2601, "step": 9787 }, { "epoch": 1.85, "grad_norm": 9.75, "learning_rate": 7.177383749118167e-08, "loss": 0.192, "step": 9788 }, { "epoch": 1.85, "grad_norm": 10.8125, "learning_rate": 7.159173610986059e-08, "loss": 0.2265, "step": 9789 }, { "epoch": 1.85, "grad_norm": 8.0, "learning_rate": 7.140986267603123e-08, "loss": 0.1615, "step": 9790 }, { "epoch": 1.85, "grad_norm": 14.3125, "learning_rate": 7.122821720676604e-08, "loss": 0.2574, "step": 9791 }, { "epoch": 1.85, "grad_norm": 12.125, "learning_rate": 7.10467997191147e-08, "loss": 0.232, "step": 9792 }, { "epoch": 1.85, "grad_norm": 11.5625, "learning_rate": 7.086561023010696e-08, "loss": 0.1754, "step": 9793 }, { "epoch": 1.85, "grad_norm": 13.0625, "learning_rate": 7.06846487567503e-08, "loss": 0.1822, "step": 9794 }, { "epoch": 1.85, "grad_norm": 11.5, "learning_rate": 7.050391531603112e-08, "loss": 0.1973, "step": 9795 }, { "epoch": 1.85, "grad_norm": 11.5625, "learning_rate": 7.032340992491394e-08, "loss": 0.1724, "step": 9796 }, { "epoch": 1.85, "grad_norm": 12.0625, "learning_rate": 7.014313260034295e-08, "loss": 0.1624, "step": 9797 }, { "epoch": 1.85, "grad_norm": 14.6875, "learning_rate": 6.996308335923991e-08, "loss": 0.2497, "step": 9798 }, { "epoch": 1.85, "grad_norm": 10.3125, "learning_rate": 6.978326221850518e-08, "loss": 0.1728, "step": 9799 }, { "epoch": 1.85, "grad_norm": 9.0, "learning_rate": 6.96036691950186e-08, "loss": 0.19, "step": 9800 }, { "epoch": 1.85, "grad_norm": 10.6875, "learning_rate": 6.942430430563779e-08, "loss": 0.2147, "step": 9801 }, { "epoch": 1.85, "grad_norm": 9.5625, "learning_rate": 6.924516756719957e-08, "loss": 0.1778, "step": 9802 }, { "epoch": 1.85, "grad_norm": 12.0, "learning_rate": 6.906625899651797e-08, "loss": 0.2269, "step": 9803 }, { "epoch": 1.85, "grad_norm": 8.875, "learning_rate": 6.888757861038764e-08, "loss": 0.1887, "step": 9804 }, { "epoch": 1.85, "grad_norm": 12.6875, "learning_rate": 6.870912642558014e-08, "loss": 0.1963, "step": 9805 }, { "epoch": 1.86, "grad_norm": 10.5625, "learning_rate": 6.853090245884709e-08, "loss": 0.2146, "step": 9806 }, { "epoch": 1.86, "grad_norm": 12.9375, "learning_rate": 6.835290672691647e-08, "loss": 0.2235, "step": 9807 }, { "epoch": 1.86, "grad_norm": 9.25, "learning_rate": 6.817513924649743e-08, "loss": 0.2184, "step": 9808 }, { "epoch": 1.86, "grad_norm": 9.0, "learning_rate": 6.79976000342758e-08, "loss": 0.2213, "step": 9809 }, { "epoch": 1.86, "grad_norm": 10.4375, "learning_rate": 6.782028910691713e-08, "loss": 0.2236, "step": 9810 }, { "epoch": 1.86, "grad_norm": 8.625, "learning_rate": 6.764320648106476e-08, "loss": 0.192, "step": 9811 }, { "epoch": 1.86, "grad_norm": 13.0, "learning_rate": 6.746635217334124e-08, "loss": 0.1705, "step": 9812 }, { "epoch": 1.86, "grad_norm": 9.5, "learning_rate": 6.72897262003469e-08, "loss": 0.2046, "step": 9813 }, { "epoch": 1.86, "grad_norm": 13.25, "learning_rate": 6.711332857866126e-08, "loss": 0.1724, "step": 9814 }, { "epoch": 1.86, "grad_norm": 9.625, "learning_rate": 6.693715932484246e-08, "loss": 0.2074, "step": 9815 }, { "epoch": 1.86, "grad_norm": 16.125, "learning_rate": 6.676121845542671e-08, "loss": 0.1694, "step": 9816 }, { "epoch": 1.86, "grad_norm": 13.0, "learning_rate": 6.658550598692915e-08, "loss": 0.2181, "step": 9817 }, { "epoch": 1.86, "grad_norm": 12.0625, "learning_rate": 6.64100219358435e-08, "loss": 0.2355, "step": 9818 }, { "epoch": 1.86, "grad_norm": 9.25, "learning_rate": 6.623476631864161e-08, "loss": 0.1766, "step": 9819 }, { "epoch": 1.86, "grad_norm": 15.5, "learning_rate": 6.605973915177477e-08, "loss": 0.2222, "step": 9820 }, { "epoch": 1.86, "grad_norm": 12.3125, "learning_rate": 6.588494045167176e-08, "loss": 0.2165, "step": 9821 }, { "epoch": 1.86, "grad_norm": 11.125, "learning_rate": 6.571037023474059e-08, "loss": 0.2042, "step": 9822 }, { "epoch": 1.86, "grad_norm": 15.6875, "learning_rate": 6.553602851736757e-08, "loss": 0.1796, "step": 9823 }, { "epoch": 1.86, "grad_norm": 9.4375, "learning_rate": 6.536191531591768e-08, "loss": 0.1779, "step": 9824 }, { "epoch": 1.86, "grad_norm": 10.75, "learning_rate": 6.518803064673451e-08, "loss": 0.2304, "step": 9825 }, { "epoch": 1.86, "grad_norm": 15.375, "learning_rate": 6.50143745261403e-08, "loss": 0.2353, "step": 9826 }, { "epoch": 1.86, "grad_norm": 11.75, "learning_rate": 6.484094697043503e-08, "loss": 0.2512, "step": 9827 }, { "epoch": 1.86, "grad_norm": 10.9375, "learning_rate": 6.46677479958982e-08, "loss": 0.1598, "step": 9828 }, { "epoch": 1.86, "grad_norm": 10.8125, "learning_rate": 6.449477761878765e-08, "loss": 0.2053, "step": 9829 }, { "epoch": 1.86, "grad_norm": 16.75, "learning_rate": 6.432203585533925e-08, "loss": 0.2597, "step": 9830 }, { "epoch": 1.86, "grad_norm": 13.0, "learning_rate": 6.41495227217681e-08, "loss": 0.2411, "step": 9831 }, { "epoch": 1.86, "grad_norm": 10.125, "learning_rate": 6.39772382342671e-08, "loss": 0.1928, "step": 9832 }, { "epoch": 1.86, "grad_norm": 10.125, "learning_rate": 6.380518240900829e-08, "loss": 0.2187, "step": 9833 }, { "epoch": 1.86, "grad_norm": 12.5625, "learning_rate": 6.363335526214209e-08, "loss": 0.2367, "step": 9834 }, { "epoch": 1.86, "grad_norm": 14.0, "learning_rate": 6.346175680979726e-08, "loss": 0.2187, "step": 9835 }, { "epoch": 1.86, "grad_norm": 10.125, "learning_rate": 6.32903870680815e-08, "loss": 0.2085, "step": 9836 }, { "epoch": 1.86, "grad_norm": 8.75, "learning_rate": 6.31192460530805e-08, "loss": 0.1487, "step": 9837 }, { "epoch": 1.86, "grad_norm": 10.375, "learning_rate": 6.294833378085897e-08, "loss": 0.2132, "step": 9838 }, { "epoch": 1.86, "grad_norm": 8.875, "learning_rate": 6.277765026745986e-08, "loss": 0.1457, "step": 9839 }, { "epoch": 1.86, "grad_norm": 10.6875, "learning_rate": 6.260719552890454e-08, "loss": 0.2038, "step": 9840 }, { "epoch": 1.86, "grad_norm": 9.625, "learning_rate": 6.243696958119355e-08, "loss": 0.1517, "step": 9841 }, { "epoch": 1.86, "grad_norm": 9.3125, "learning_rate": 6.226697244030495e-08, "loss": 0.1741, "step": 9842 }, { "epoch": 1.86, "grad_norm": 17.625, "learning_rate": 6.209720412219594e-08, "loss": 0.2231, "step": 9843 }, { "epoch": 1.86, "grad_norm": 11.0625, "learning_rate": 6.192766464280298e-08, "loss": 0.2259, "step": 9844 }, { "epoch": 1.86, "grad_norm": 12.8125, "learning_rate": 6.175835401803914e-08, "loss": 0.1917, "step": 9845 }, { "epoch": 1.86, "grad_norm": 9.6875, "learning_rate": 6.158927226379785e-08, "loss": 0.1881, "step": 9846 }, { "epoch": 1.86, "grad_norm": 11.6875, "learning_rate": 6.142041939594944e-08, "loss": 0.1891, "step": 9847 }, { "epoch": 1.86, "grad_norm": 9.5625, "learning_rate": 6.125179543034516e-08, "loss": 0.1779, "step": 9848 }, { "epoch": 1.86, "grad_norm": 8.5, "learning_rate": 6.108340038281174e-08, "loss": 0.2192, "step": 9849 }, { "epoch": 1.86, "grad_norm": 17.125, "learning_rate": 6.091523426915658e-08, "loss": 0.1711, "step": 9850 }, { "epoch": 1.86, "grad_norm": 8.5, "learning_rate": 6.074729710516453e-08, "loss": 0.1596, "step": 9851 }, { "epoch": 1.86, "grad_norm": 18.625, "learning_rate": 6.057958890660048e-08, "loss": 0.206, "step": 9852 }, { "epoch": 1.86, "grad_norm": 11.625, "learning_rate": 6.041210968920547e-08, "loss": 0.1675, "step": 9853 }, { "epoch": 1.86, "grad_norm": 16.5, "learning_rate": 6.024485946870023e-08, "loss": 0.2243, "step": 9854 }, { "epoch": 1.86, "grad_norm": 8.6875, "learning_rate": 6.007783826078528e-08, "loss": 0.2111, "step": 9855 }, { "epoch": 1.86, "grad_norm": 10.5, "learning_rate": 5.991104608113724e-08, "loss": 0.2268, "step": 9856 }, { "epoch": 1.86, "grad_norm": 10.25, "learning_rate": 5.974448294541302e-08, "loss": 0.1681, "step": 9857 }, { "epoch": 1.86, "grad_norm": 12.125, "learning_rate": 5.957814886924679e-08, "loss": 0.1956, "step": 9858 }, { "epoch": 1.87, "grad_norm": 11.0625, "learning_rate": 5.9412043868252724e-08, "loss": 0.1845, "step": 9859 }, { "epoch": 1.87, "grad_norm": 11.375, "learning_rate": 5.924616795802196e-08, "loss": 0.1822, "step": 9860 }, { "epoch": 1.87, "grad_norm": 10.25, "learning_rate": 5.9080521154125104e-08, "loss": 0.1547, "step": 9861 }, { "epoch": 1.87, "grad_norm": 12.0625, "learning_rate": 5.891510347211027e-08, "loss": 0.2262, "step": 9862 }, { "epoch": 1.87, "grad_norm": 9.625, "learning_rate": 5.874991492750559e-08, "loss": 0.1401, "step": 9863 }, { "epoch": 1.87, "grad_norm": 11.4375, "learning_rate": 5.8584955535816446e-08, "loss": 0.1987, "step": 9864 }, { "epoch": 1.87, "grad_norm": 17.375, "learning_rate": 5.842022531252711e-08, "loss": 0.1748, "step": 9865 }, { "epoch": 1.87, "grad_norm": 20.625, "learning_rate": 5.8255724273099946e-08, "loss": 0.2116, "step": 9866 }, { "epoch": 1.87, "grad_norm": 16.125, "learning_rate": 5.8091452432977314e-08, "loss": 0.2336, "step": 9867 }, { "epoch": 1.87, "grad_norm": 10.375, "learning_rate": 5.792740980757772e-08, "loss": 0.2069, "step": 9868 }, { "epoch": 1.87, "grad_norm": 11.125, "learning_rate": 5.776359641229939e-08, "loss": 0.176, "step": 9869 }, { "epoch": 1.87, "grad_norm": 11.3125, "learning_rate": 5.760001226251977e-08, "loss": 0.2582, "step": 9870 }, { "epoch": 1.87, "grad_norm": 10.125, "learning_rate": 5.743665737359405e-08, "loss": 0.1984, "step": 9871 }, { "epoch": 1.87, "grad_norm": 10.0625, "learning_rate": 5.727353176085498e-08, "loss": 0.1909, "step": 9872 }, { "epoch": 1.87, "grad_norm": 13.25, "learning_rate": 5.7110635439615037e-08, "loss": 0.2493, "step": 9873 }, { "epoch": 1.87, "grad_norm": 12.25, "learning_rate": 5.6947968425165314e-08, "loss": 0.1837, "step": 9874 }, { "epoch": 1.87, "grad_norm": 10.5625, "learning_rate": 5.678553073277443e-08, "loss": 0.194, "step": 9875 }, { "epoch": 1.87, "grad_norm": 13.4375, "learning_rate": 5.662332237769019e-08, "loss": 0.238, "step": 9876 }, { "epoch": 1.87, "grad_norm": 11.8125, "learning_rate": 5.646134337513792e-08, "loss": 0.2187, "step": 9877 }, { "epoch": 1.87, "grad_norm": 13.5, "learning_rate": 5.6299593740322956e-08, "loss": 0.216, "step": 9878 }, { "epoch": 1.87, "grad_norm": 12.625, "learning_rate": 5.6138073488428145e-08, "loss": 0.2308, "step": 9879 }, { "epoch": 1.87, "grad_norm": 9.25, "learning_rate": 5.5976782634614424e-08, "loss": 0.1522, "step": 9880 }, { "epoch": 1.87, "grad_norm": 31.5, "learning_rate": 5.58157211940219e-08, "loss": 0.1946, "step": 9881 }, { "epoch": 1.87, "grad_norm": 11.4375, "learning_rate": 5.565488918176931e-08, "loss": 0.1842, "step": 9882 }, { "epoch": 1.87, "grad_norm": 11.6875, "learning_rate": 5.5494286612953194e-08, "loss": 0.2409, "step": 9883 }, { "epoch": 1.87, "grad_norm": 9.25, "learning_rate": 5.533391350264844e-08, "loss": 0.1539, "step": 9884 }, { "epoch": 1.87, "grad_norm": 13.0625, "learning_rate": 5.5173769865909385e-08, "loss": 0.2018, "step": 9885 }, { "epoch": 1.87, "grad_norm": 10.125, "learning_rate": 5.5013855717768184e-08, "loss": 0.2074, "step": 9886 }, { "epoch": 1.87, "grad_norm": 12.9375, "learning_rate": 5.485417107323504e-08, "loss": 0.2047, "step": 9887 }, { "epoch": 1.87, "grad_norm": 11.875, "learning_rate": 5.469471594729936e-08, "loss": 0.1402, "step": 9888 }, { "epoch": 1.87, "grad_norm": 9.5625, "learning_rate": 5.453549035492889e-08, "loss": 0.2255, "step": 9889 }, { "epoch": 1.87, "grad_norm": 8.3125, "learning_rate": 5.437649431106973e-08, "loss": 0.1382, "step": 9890 }, { "epoch": 1.87, "grad_norm": 11.125, "learning_rate": 5.421772783064577e-08, "loss": 0.2107, "step": 9891 }, { "epoch": 1.87, "grad_norm": 9.6875, "learning_rate": 5.405919092856038e-08, "loss": 0.1815, "step": 9892 }, { "epoch": 1.87, "grad_norm": 11.3125, "learning_rate": 5.390088361969525e-08, "loss": 0.1625, "step": 9893 }, { "epoch": 1.87, "grad_norm": 13.625, "learning_rate": 5.37428059189099e-08, "loss": 0.3032, "step": 9894 }, { "epoch": 1.87, "grad_norm": 12.625, "learning_rate": 5.35849578410419e-08, "loss": 0.2397, "step": 9895 }, { "epoch": 1.87, "grad_norm": 11.125, "learning_rate": 5.342733940090938e-08, "loss": 0.25, "step": 9896 }, { "epoch": 1.87, "grad_norm": 12.8125, "learning_rate": 5.3269950613306366e-08, "loss": 0.2432, "step": 9897 }, { "epoch": 1.87, "grad_norm": 9.125, "learning_rate": 5.311279149300741e-08, "loss": 0.1717, "step": 9898 }, { "epoch": 1.87, "grad_norm": 12.0625, "learning_rate": 5.2955862054763496e-08, "loss": 0.2136, "step": 9899 }, { "epoch": 1.87, "grad_norm": 13.8125, "learning_rate": 5.279916231330617e-08, "loss": 0.2416, "step": 9900 }, { "epoch": 1.87, "grad_norm": 13.9375, "learning_rate": 5.2642692283343676e-08, "loss": 0.212, "step": 9901 }, { "epoch": 1.87, "grad_norm": 8.6875, "learning_rate": 5.2486451979563705e-08, "loss": 0.2326, "step": 9902 }, { "epoch": 1.87, "grad_norm": 10.6875, "learning_rate": 5.233044141663202e-08, "loss": 0.1802, "step": 9903 }, { "epoch": 1.87, "grad_norm": 10.1875, "learning_rate": 5.217466060919302e-08, "loss": 0.1947, "step": 9904 }, { "epoch": 1.87, "grad_norm": 13.875, "learning_rate": 5.20191095718689e-08, "loss": 0.2039, "step": 9905 }, { "epoch": 1.87, "grad_norm": 13.4375, "learning_rate": 5.186378831926131e-08, "loss": 0.2224, "step": 9906 }, { "epoch": 1.87, "grad_norm": 9.125, "learning_rate": 5.170869686594943e-08, "loss": 0.1734, "step": 9907 }, { "epoch": 1.87, "grad_norm": 13.75, "learning_rate": 5.15538352264916e-08, "loss": 0.1978, "step": 9908 }, { "epoch": 1.87, "grad_norm": 15.8125, "learning_rate": 5.139920341542371e-08, "loss": 0.2013, "step": 9909 }, { "epoch": 1.87, "grad_norm": 12.9375, "learning_rate": 5.124480144726107e-08, "loss": 0.2945, "step": 9910 }, { "epoch": 1.87, "grad_norm": 17.125, "learning_rate": 5.109062933649683e-08, "loss": 0.2323, "step": 9911 }, { "epoch": 1.88, "grad_norm": 10.8125, "learning_rate": 5.0936687097602466e-08, "loss": 0.2079, "step": 9912 }, { "epoch": 1.88, "grad_norm": 13.0625, "learning_rate": 5.078297474502808e-08, "loss": 0.2202, "step": 9913 }, { "epoch": 1.88, "grad_norm": 11.25, "learning_rate": 5.0629492293202686e-08, "loss": 0.2418, "step": 9914 }, { "epoch": 1.88, "grad_norm": 16.875, "learning_rate": 5.047623975653254e-08, "loss": 0.2688, "step": 9915 }, { "epoch": 1.88, "grad_norm": 9.0, "learning_rate": 5.032321714940336e-08, "loss": 0.1617, "step": 9916 }, { "epoch": 1.88, "grad_norm": 10.8125, "learning_rate": 5.0170424486179205e-08, "loss": 0.2659, "step": 9917 }, { "epoch": 1.88, "grad_norm": 9.625, "learning_rate": 5.001786178120166e-08, "loss": 0.2497, "step": 9918 }, { "epoch": 1.88, "grad_norm": 17.375, "learning_rate": 4.9865529048791775e-08, "loss": 0.2638, "step": 9919 }, { "epoch": 1.88, "grad_norm": 20.125, "learning_rate": 4.9713426303248105e-08, "loss": 0.2274, "step": 9920 }, { "epoch": 1.88, "grad_norm": 10.25, "learning_rate": 4.956155355884867e-08, "loss": 0.2116, "step": 9921 }, { "epoch": 1.88, "grad_norm": 12.8125, "learning_rate": 4.9409910829849026e-08, "loss": 0.2332, "step": 9922 }, { "epoch": 1.88, "grad_norm": 11.125, "learning_rate": 4.925849813048361e-08, "loss": 0.1992, "step": 9923 }, { "epoch": 1.88, "grad_norm": 11.875, "learning_rate": 4.910731547496467e-08, "loss": 0.2065, "step": 9924 }, { "epoch": 1.88, "grad_norm": 10.125, "learning_rate": 4.895636287748362e-08, "loss": 0.1941, "step": 9925 }, { "epoch": 1.88, "grad_norm": 12.5, "learning_rate": 4.880564035220997e-08, "loss": 0.1984, "step": 9926 }, { "epoch": 1.88, "grad_norm": 17.125, "learning_rate": 4.865514791329129e-08, "loss": 0.1812, "step": 9927 }, { "epoch": 1.88, "grad_norm": 12.1875, "learning_rate": 4.850488557485433e-08, "loss": 0.1704, "step": 9928 }, { "epoch": 1.88, "grad_norm": 14.25, "learning_rate": 4.8354853351003096e-08, "loss": 0.2221, "step": 9929 }, { "epoch": 1.88, "grad_norm": 13.0, "learning_rate": 4.8205051255821314e-08, "loss": 0.1984, "step": 9930 }, { "epoch": 1.88, "grad_norm": 9.6875, "learning_rate": 4.8055479303370236e-08, "loss": 0.1748, "step": 9931 }, { "epoch": 1.88, "grad_norm": 11.25, "learning_rate": 4.7906137507689466e-08, "loss": 0.2285, "step": 9932 }, { "epoch": 1.88, "grad_norm": 10.6875, "learning_rate": 4.775702588279779e-08, "loss": 0.1646, "step": 9933 }, { "epoch": 1.88, "grad_norm": 17.625, "learning_rate": 4.760814444269124e-08, "loss": 0.2169, "step": 9934 }, { "epoch": 1.88, "grad_norm": 11.3125, "learning_rate": 4.74594932013453e-08, "loss": 0.1752, "step": 9935 }, { "epoch": 1.88, "grad_norm": 9.5625, "learning_rate": 4.73110721727138e-08, "loss": 0.2024, "step": 9936 }, { "epoch": 1.88, "grad_norm": 12.875, "learning_rate": 4.7162881370727834e-08, "loss": 0.178, "step": 9937 }, { "epoch": 1.88, "grad_norm": 11.9375, "learning_rate": 4.701492080929792e-08, "loss": 0.2133, "step": 9938 }, { "epoch": 1.88, "grad_norm": 10.75, "learning_rate": 4.68671905023127e-08, "loss": 0.2124, "step": 9939 }, { "epoch": 1.88, "grad_norm": 17.25, "learning_rate": 4.671969046363939e-08, "loss": 0.2146, "step": 9940 }, { "epoch": 1.88, "grad_norm": 9.0625, "learning_rate": 4.6572420707122755e-08, "loss": 0.2127, "step": 9941 }, { "epoch": 1.88, "grad_norm": 9.9375, "learning_rate": 4.64253812465873e-08, "loss": 0.1668, "step": 9942 }, { "epoch": 1.88, "grad_norm": 10.4375, "learning_rate": 4.627857209583475e-08, "loss": 0.1471, "step": 9943 }, { "epoch": 1.88, "grad_norm": 10.625, "learning_rate": 4.6131993268646035e-08, "loss": 0.1817, "step": 9944 }, { "epoch": 1.88, "grad_norm": 10.875, "learning_rate": 4.598564477877959e-08, "loss": 0.2028, "step": 9945 }, { "epoch": 1.88, "grad_norm": 14.6875, "learning_rate": 4.583952663997249e-08, "loss": 0.1743, "step": 9946 }, { "epoch": 1.88, "grad_norm": 13.9375, "learning_rate": 4.5693638865941524e-08, "loss": 0.2053, "step": 9947 }, { "epoch": 1.88, "grad_norm": 13.5625, "learning_rate": 4.554798147037992e-08, "loss": 0.2033, "step": 9948 }, { "epoch": 1.88, "grad_norm": 12.1875, "learning_rate": 4.540255446696007e-08, "loss": 0.278, "step": 9949 }, { "epoch": 1.88, "grad_norm": 12.625, "learning_rate": 4.525735786933299e-08, "loss": 0.1935, "step": 9950 }, { "epoch": 1.88, "grad_norm": 13.125, "learning_rate": 4.511239169112808e-08, "loss": 0.2166, "step": 9951 }, { "epoch": 1.88, "grad_norm": 13.4375, "learning_rate": 4.496765594595248e-08, "loss": 0.1795, "step": 9952 }, { "epoch": 1.88, "grad_norm": 14.8125, "learning_rate": 4.482315064739229e-08, "loss": 0.1965, "step": 9953 }, { "epoch": 1.88, "grad_norm": 15.625, "learning_rate": 4.467887580901137e-08, "loss": 0.223, "step": 9954 }, { "epoch": 1.88, "grad_norm": 8.1875, "learning_rate": 4.4534831444353336e-08, "loss": 0.1515, "step": 9955 }, { "epoch": 1.88, "grad_norm": 9.8125, "learning_rate": 4.4391017566938474e-08, "loss": 0.1921, "step": 9956 }, { "epoch": 1.88, "grad_norm": 13.5625, "learning_rate": 4.4247434190266005e-08, "loss": 0.1867, "step": 9957 }, { "epoch": 1.88, "grad_norm": 24.25, "learning_rate": 4.4104081327814296e-08, "loss": 0.3076, "step": 9958 }, { "epoch": 1.88, "grad_norm": 18.625, "learning_rate": 4.3960958993039536e-08, "loss": 0.2005, "step": 9959 }, { "epoch": 1.88, "grad_norm": 13.625, "learning_rate": 4.381806719937542e-08, "loss": 0.181, "step": 9960 }, { "epoch": 1.88, "grad_norm": 7.65625, "learning_rate": 4.367540596023484e-08, "loss": 0.1348, "step": 9961 }, { "epoch": 1.88, "grad_norm": 10.625, "learning_rate": 4.353297528900957e-08, "loss": 0.2013, "step": 9962 }, { "epoch": 1.88, "grad_norm": 15.3125, "learning_rate": 4.33907751990692e-08, "loss": 0.2502, "step": 9963 }, { "epoch": 1.88, "grad_norm": 12.375, "learning_rate": 4.324880570376139e-08, "loss": 0.229, "step": 9964 }, { "epoch": 1.89, "grad_norm": 11.8125, "learning_rate": 4.310706681641186e-08, "loss": 0.1653, "step": 9965 }, { "epoch": 1.89, "grad_norm": 9.6875, "learning_rate": 4.29655585503258e-08, "loss": 0.1981, "step": 9966 }, { "epoch": 1.89, "grad_norm": 10.125, "learning_rate": 4.282428091878649e-08, "loss": 0.17, "step": 9967 }, { "epoch": 1.89, "grad_norm": 26.25, "learning_rate": 4.268323393505441e-08, "loss": 0.2303, "step": 9968 }, { "epoch": 1.89, "grad_norm": 11.25, "learning_rate": 4.2542417612369545e-08, "loss": 0.2125, "step": 9969 }, { "epoch": 1.89, "grad_norm": 14.0625, "learning_rate": 4.240183196395048e-08, "loss": 0.2528, "step": 9970 }, { "epoch": 1.89, "grad_norm": 15.3125, "learning_rate": 4.2261477002993044e-08, "loss": 0.2527, "step": 9971 }, { "epoch": 1.89, "grad_norm": 9.5625, "learning_rate": 4.212135274267143e-08, "loss": 0.205, "step": 9972 }, { "epoch": 1.89, "grad_norm": 9.75, "learning_rate": 4.198145919613955e-08, "loss": 0.1768, "step": 9973 }, { "epoch": 1.89, "grad_norm": 9.25, "learning_rate": 4.1841796376528856e-08, "loss": 0.199, "step": 9974 }, { "epoch": 1.89, "grad_norm": 9.875, "learning_rate": 4.170236429694857e-08, "loss": 0.2564, "step": 9975 }, { "epoch": 1.89, "grad_norm": 13.0, "learning_rate": 4.156316297048657e-08, "loss": 0.1992, "step": 9976 }, { "epoch": 1.89, "grad_norm": 13.0625, "learning_rate": 4.1424192410209615e-08, "loss": 0.263, "step": 9977 }, { "epoch": 1.89, "grad_norm": 9.375, "learning_rate": 4.128545262916312e-08, "loss": 0.2455, "step": 9978 }, { "epoch": 1.89, "grad_norm": 9.375, "learning_rate": 4.114694364036886e-08, "loss": 0.1788, "step": 9979 }, { "epoch": 1.89, "grad_norm": 12.125, "learning_rate": 4.100866545682924e-08, "loss": 0.2033, "step": 9980 }, { "epoch": 1.89, "grad_norm": 12.875, "learning_rate": 4.08706180915236e-08, "loss": 0.2082, "step": 9981 }, { "epoch": 1.89, "grad_norm": 8.1875, "learning_rate": 4.073280155741044e-08, "loss": 0.1618, "step": 9982 }, { "epoch": 1.89, "grad_norm": 15.8125, "learning_rate": 4.0595215867425265e-08, "loss": 0.2522, "step": 9983 }, { "epoch": 1.89, "grad_norm": 12.4375, "learning_rate": 4.0457861034483846e-08, "loss": 0.2277, "step": 9984 }, { "epoch": 1.89, "grad_norm": 18.0, "learning_rate": 4.032073707147893e-08, "loss": 0.2116, "step": 9985 }, { "epoch": 1.89, "grad_norm": 8.0625, "learning_rate": 4.0183843991282166e-08, "loss": 0.1574, "step": 9986 }, { "epoch": 1.89, "grad_norm": 15.8125, "learning_rate": 4.0047181806742716e-08, "loss": 0.1809, "step": 9987 }, { "epoch": 1.89, "grad_norm": 14.0625, "learning_rate": 3.991075053068894e-08, "loss": 0.242, "step": 9988 }, { "epoch": 1.89, "grad_norm": 9.25, "learning_rate": 3.977455017592724e-08, "loss": 0.1751, "step": 9989 }, { "epoch": 1.89, "grad_norm": 13.9375, "learning_rate": 3.963858075524269e-08, "loss": 0.2681, "step": 9990 }, { "epoch": 1.89, "grad_norm": 12.5, "learning_rate": 3.9502842281397834e-08, "loss": 0.2568, "step": 9991 }, { "epoch": 1.89, "grad_norm": 14.625, "learning_rate": 3.936733476713417e-08, "loss": 0.2212, "step": 9992 }, { "epoch": 1.89, "grad_norm": 14.75, "learning_rate": 3.923205822517151e-08, "loss": 0.2172, "step": 9993 }, { "epoch": 1.89, "grad_norm": 12.5625, "learning_rate": 3.909701266820803e-08, "loss": 0.237, "step": 9994 }, { "epoch": 1.89, "grad_norm": 15.125, "learning_rate": 3.8962198108919714e-08, "loss": 0.2321, "step": 9995 }, { "epoch": 1.89, "grad_norm": 13.625, "learning_rate": 3.8827614559961145e-08, "loss": 0.2518, "step": 9996 }, { "epoch": 1.89, "grad_norm": 12.6875, "learning_rate": 3.869326203396584e-08, "loss": 0.2055, "step": 9997 }, { "epoch": 1.89, "grad_norm": 13.6875, "learning_rate": 3.855914054354398e-08, "loss": 0.2543, "step": 9998 }, { "epoch": 1.89, "grad_norm": 8.9375, "learning_rate": 3.8425250101286336e-08, "loss": 0.1663, "step": 9999 }, { "epoch": 1.89, "grad_norm": 9.75, "learning_rate": 3.829159071976035e-08, "loss": 0.1724, "step": 10000 }, { "epoch": 1.89, "grad_norm": 12.9375, "learning_rate": 3.81581624115121e-08, "loss": 0.1969, "step": 10001 }, { "epoch": 1.89, "grad_norm": 12.875, "learning_rate": 3.802496518906629e-08, "loss": 0.2207, "step": 10002 }, { "epoch": 1.89, "grad_norm": 11.625, "learning_rate": 3.7891999064925687e-08, "loss": 0.2228, "step": 10003 }, { "epoch": 1.89, "grad_norm": 8.75, "learning_rate": 3.7759264051571435e-08, "loss": 0.2029, "step": 10004 }, { "epoch": 1.89, "grad_norm": 12.8125, "learning_rate": 3.7626760161462995e-08, "loss": 0.2432, "step": 10005 }, { "epoch": 1.89, "grad_norm": 15.625, "learning_rate": 3.749448740703793e-08, "loss": 0.2715, "step": 10006 }, { "epoch": 1.89, "grad_norm": 12.3125, "learning_rate": 3.7362445800712686e-08, "loss": 0.1873, "step": 10007 }, { "epoch": 1.89, "grad_norm": 15.875, "learning_rate": 3.723063535488125e-08, "loss": 0.1821, "step": 10008 }, { "epoch": 1.89, "grad_norm": 15.125, "learning_rate": 3.709905608191622e-08, "loss": 0.2118, "step": 10009 }, { "epoch": 1.89, "grad_norm": 11.75, "learning_rate": 3.69677079941691e-08, "loss": 0.2075, "step": 10010 }, { "epoch": 1.89, "grad_norm": 13.125, "learning_rate": 3.6836591103968364e-08, "loss": 0.2472, "step": 10011 }, { "epoch": 1.89, "grad_norm": 14.9375, "learning_rate": 3.670570542362223e-08, "loss": 0.1714, "step": 10012 }, { "epoch": 1.89, "grad_norm": 10.25, "learning_rate": 3.6575050965416414e-08, "loss": 0.2183, "step": 10013 }, { "epoch": 1.89, "grad_norm": 15.3125, "learning_rate": 3.6444627741614724e-08, "loss": 0.1972, "step": 10014 }, { "epoch": 1.89, "grad_norm": 9.625, "learning_rate": 3.631443576445987e-08, "loss": 0.2134, "step": 10015 }, { "epoch": 1.89, "grad_norm": 11.375, "learning_rate": 3.618447504617262e-08, "loss": 0.203, "step": 10016 }, { "epoch": 1.89, "grad_norm": 14.5625, "learning_rate": 3.6054745598952114e-08, "loss": 0.2287, "step": 10017 }, { "epoch": 1.9, "grad_norm": 12.625, "learning_rate": 3.5925247434975276e-08, "loss": 0.2013, "step": 10018 }, { "epoch": 1.9, "grad_norm": 9.375, "learning_rate": 3.5795980566397934e-08, "loss": 0.2018, "step": 10019 }, { "epoch": 1.9, "grad_norm": 15.625, "learning_rate": 3.5666945005354006e-08, "loss": 0.2006, "step": 10020 }, { "epoch": 1.9, "grad_norm": 15.5625, "learning_rate": 3.5538140763955455e-08, "loss": 0.2212, "step": 10021 }, { "epoch": 1.9, "grad_norm": 11.3125, "learning_rate": 3.540956785429317e-08, "loss": 0.1956, "step": 10022 }, { "epoch": 1.9, "grad_norm": 8.4375, "learning_rate": 3.528122628843583e-08, "loss": 0.211, "step": 10023 }, { "epoch": 1.9, "grad_norm": 12.1875, "learning_rate": 3.5153116078430196e-08, "loss": 0.209, "step": 10024 }, { "epoch": 1.9, "grad_norm": 12.25, "learning_rate": 3.5025237236301636e-08, "loss": 0.2206, "step": 10025 }, { "epoch": 1.9, "grad_norm": 12.5, "learning_rate": 3.489758977405388e-08, "loss": 0.2274, "step": 10026 }, { "epoch": 1.9, "grad_norm": 11.0625, "learning_rate": 3.4770173703668755e-08, "loss": 0.1876, "step": 10027 }, { "epoch": 1.9, "grad_norm": 13.3125, "learning_rate": 3.464298903710667e-08, "loss": 0.1957, "step": 10028 }, { "epoch": 1.9, "grad_norm": 9.3125, "learning_rate": 3.451603578630558e-08, "loss": 0.1965, "step": 10029 }, { "epoch": 1.9, "grad_norm": 10.625, "learning_rate": 3.4389313963182904e-08, "loss": 0.1591, "step": 10030 }, { "epoch": 1.9, "grad_norm": 14.5625, "learning_rate": 3.426282357963301e-08, "loss": 0.2326, "step": 10031 }, { "epoch": 1.9, "grad_norm": 23.25, "learning_rate": 3.413656464752918e-08, "loss": 0.2578, "step": 10032 }, { "epoch": 1.9, "grad_norm": 10.25, "learning_rate": 3.40105371787236e-08, "loss": 0.1828, "step": 10033 }, { "epoch": 1.9, "grad_norm": 13.75, "learning_rate": 3.388474118504514e-08, "loss": 0.2388, "step": 10034 }, { "epoch": 1.9, "grad_norm": 13.125, "learning_rate": 3.375917667830325e-08, "loss": 0.1932, "step": 10035 }, { "epoch": 1.9, "grad_norm": 13.25, "learning_rate": 3.363384367028294e-08, "loss": 0.2196, "step": 10036 }, { "epoch": 1.9, "grad_norm": 11.5, "learning_rate": 3.350874217274952e-08, "loss": 0.2296, "step": 10037 }, { "epoch": 1.9, "grad_norm": 10.75, "learning_rate": 3.3383872197445546e-08, "loss": 0.2099, "step": 10038 }, { "epoch": 1.9, "grad_norm": 14.5625, "learning_rate": 3.325923375609302e-08, "loss": 0.2001, "step": 10039 }, { "epoch": 1.9, "grad_norm": 11.375, "learning_rate": 3.313482686039038e-08, "loss": 0.2188, "step": 10040 }, { "epoch": 1.9, "grad_norm": 10.25, "learning_rate": 3.3010651522015766e-08, "loss": 0.2089, "step": 10041 }, { "epoch": 1.9, "grad_norm": 11.125, "learning_rate": 3.288670775262515e-08, "loss": 0.2566, "step": 10042 }, { "epoch": 1.9, "grad_norm": 10.625, "learning_rate": 3.2762995563852815e-08, "loss": 0.2141, "step": 10043 }, { "epoch": 1.9, "grad_norm": 15.25, "learning_rate": 3.263951496731116e-08, "loss": 0.2532, "step": 10044 }, { "epoch": 1.9, "grad_norm": 13.875, "learning_rate": 3.25162659745909e-08, "loss": 0.236, "step": 10045 }, { "epoch": 1.9, "grad_norm": 6.5625, "learning_rate": 3.239324859726112e-08, "loss": 0.1363, "step": 10046 }, { "epoch": 1.9, "grad_norm": 12.0625, "learning_rate": 3.2270462846869244e-08, "loss": 0.2722, "step": 10047 }, { "epoch": 1.9, "grad_norm": 10.4375, "learning_rate": 3.21479087349405e-08, "loss": 0.2514, "step": 10048 }, { "epoch": 1.9, "grad_norm": 12.9375, "learning_rate": 3.2025586272978746e-08, "loss": 0.2048, "step": 10049 }, { "epoch": 1.9, "grad_norm": 16.375, "learning_rate": 3.190349547246646e-08, "loss": 0.2568, "step": 10050 }, { "epoch": 1.9, "grad_norm": 11.25, "learning_rate": 3.1781636344863644e-08, "loss": 0.182, "step": 10051 }, { "epoch": 1.9, "grad_norm": 11.5, "learning_rate": 3.166000890160836e-08, "loss": 0.1865, "step": 10052 }, { "epoch": 1.9, "grad_norm": 9.6875, "learning_rate": 3.1538613154118155e-08, "loss": 0.2227, "step": 10053 }, { "epoch": 1.9, "grad_norm": 11.3125, "learning_rate": 3.14174491137878e-08, "loss": 0.2177, "step": 10054 }, { "epoch": 1.9, "grad_norm": 7.875, "learning_rate": 3.129651679199069e-08, "loss": 0.1593, "step": 10055 }, { "epoch": 1.9, "grad_norm": 9.625, "learning_rate": 3.117581620007804e-08, "loss": 0.1412, "step": 10056 }, { "epoch": 1.9, "grad_norm": 11.25, "learning_rate": 3.1055347349379925e-08, "loss": 0.2277, "step": 10057 }, { "epoch": 1.9, "grad_norm": 16.125, "learning_rate": 3.0935110251204827e-08, "loss": 0.2502, "step": 10058 }, { "epoch": 1.9, "grad_norm": 16.125, "learning_rate": 3.081510491683842e-08, "loss": 0.213, "step": 10059 }, { "epoch": 1.9, "grad_norm": 10.75, "learning_rate": 3.069533135754532e-08, "loss": 0.2449, "step": 10060 }, { "epoch": 1.9, "grad_norm": 12.25, "learning_rate": 3.057578958456847e-08, "loss": 0.2543, "step": 10061 }, { "epoch": 1.9, "grad_norm": 12.1875, "learning_rate": 3.0456479609128896e-08, "loss": 0.1967, "step": 10062 }, { "epoch": 1.9, "grad_norm": 10.5625, "learning_rate": 3.0337401442425976e-08, "loss": 0.2029, "step": 10063 }, { "epoch": 1.9, "grad_norm": 13.5, "learning_rate": 3.021855509563687e-08, "loss": 0.2144, "step": 10064 }, { "epoch": 1.9, "grad_norm": 12.3125, "learning_rate": 3.0099940579917665e-08, "loss": 0.2466, "step": 10065 }, { "epoch": 1.9, "grad_norm": 12.1875, "learning_rate": 2.9981557906402503e-08, "loss": 0.2066, "step": 10066 }, { "epoch": 1.9, "grad_norm": 11.0, "learning_rate": 2.9863407086203054e-08, "loss": 0.2257, "step": 10067 }, { "epoch": 1.9, "grad_norm": 15.3125, "learning_rate": 2.9745488130409893e-08, "loss": 0.3013, "step": 10068 }, { "epoch": 1.9, "grad_norm": 18.375, "learning_rate": 2.962780105009222e-08, "loss": 0.1792, "step": 10069 }, { "epoch": 1.9, "grad_norm": 9.25, "learning_rate": 2.9510345856296753e-08, "loss": 0.1989, "step": 10070 }, { "epoch": 1.91, "grad_norm": 10.75, "learning_rate": 2.939312256004856e-08, "loss": 0.2626, "step": 10071 }, { "epoch": 1.91, "grad_norm": 9.625, "learning_rate": 2.9276131172350786e-08, "loss": 0.1642, "step": 10072 }, { "epoch": 1.91, "grad_norm": 12.125, "learning_rate": 2.9159371704185757e-08, "loss": 0.1927, "step": 10073 }, { "epoch": 1.91, "grad_norm": 8.9375, "learning_rate": 2.904284416651276e-08, "loss": 0.1654, "step": 10074 }, { "epoch": 1.91, "grad_norm": 10.75, "learning_rate": 2.8926548570269997e-08, "loss": 0.2345, "step": 10075 }, { "epoch": 1.91, "grad_norm": 13.25, "learning_rate": 2.8810484926373726e-08, "loss": 0.2292, "step": 10076 }, { "epoch": 1.91, "grad_norm": 13.125, "learning_rate": 2.869465324571885e-08, "loss": 0.1843, "step": 10077 }, { "epoch": 1.91, "grad_norm": 11.0, "learning_rate": 2.857905353917806e-08, "loss": 0.2427, "step": 10078 }, { "epoch": 1.91, "grad_norm": 9.1875, "learning_rate": 2.8463685817602117e-08, "loss": 0.1931, "step": 10079 }, { "epoch": 1.91, "grad_norm": 12.9375, "learning_rate": 2.834855009182014e-08, "loss": 0.2337, "step": 10080 }, { "epoch": 1.91, "grad_norm": 11.8125, "learning_rate": 2.8233646372640144e-08, "loss": 0.2072, "step": 10081 }, { "epoch": 1.91, "grad_norm": 13.5, "learning_rate": 2.811897467084712e-08, "loss": 0.2487, "step": 10082 }, { "epoch": 1.91, "grad_norm": 11.25, "learning_rate": 2.8004534997205234e-08, "loss": 0.178, "step": 10083 }, { "epoch": 1.91, "grad_norm": 15.6875, "learning_rate": 2.7890327362457004e-08, "loss": 0.1656, "step": 10084 }, { "epoch": 1.91, "grad_norm": 8.0, "learning_rate": 2.7776351777322463e-08, "loss": 0.155, "step": 10085 }, { "epoch": 1.91, "grad_norm": 12.9375, "learning_rate": 2.7662608252499723e-08, "loss": 0.2109, "step": 10086 }, { "epoch": 1.91, "grad_norm": 12.0, "learning_rate": 2.7549096798666074e-08, "loss": 0.1896, "step": 10087 }, { "epoch": 1.91, "grad_norm": 11.1875, "learning_rate": 2.743581742647633e-08, "loss": 0.1927, "step": 10088 }, { "epoch": 1.91, "grad_norm": 10.9375, "learning_rate": 2.7322770146563926e-08, "loss": 0.2047, "step": 10089 }, { "epoch": 1.91, "grad_norm": 11.4375, "learning_rate": 2.720995496953982e-08, "loss": 0.2287, "step": 10090 }, { "epoch": 1.91, "grad_norm": 9.0625, "learning_rate": 2.7097371905994153e-08, "loss": 0.1752, "step": 10091 }, { "epoch": 1.91, "grad_norm": 10.0, "learning_rate": 2.69850209664943e-08, "loss": 0.1956, "step": 10092 }, { "epoch": 1.91, "grad_norm": 8.875, "learning_rate": 2.6872902161586835e-08, "loss": 0.1484, "step": 10093 }, { "epoch": 1.91, "grad_norm": 10.75, "learning_rate": 2.6761015501795274e-08, "loss": 0.1685, "step": 10094 }, { "epoch": 1.91, "grad_norm": 16.75, "learning_rate": 2.6649360997622887e-08, "loss": 0.2304, "step": 10095 }, { "epoch": 1.91, "grad_norm": 11.5625, "learning_rate": 2.6537938659549622e-08, "loss": 0.1882, "step": 10096 }, { "epoch": 1.91, "grad_norm": 11.9375, "learning_rate": 2.6426748498034893e-08, "loss": 0.2533, "step": 10097 }, { "epoch": 1.91, "grad_norm": 13.1875, "learning_rate": 2.6315790523515626e-08, "loss": 0.2486, "step": 10098 }, { "epoch": 1.91, "grad_norm": 30.125, "learning_rate": 2.6205064746407107e-08, "loss": 0.2085, "step": 10099 }, { "epoch": 1.91, "grad_norm": 11.125, "learning_rate": 2.6094571177102958e-08, "loss": 0.1752, "step": 10100 }, { "epoch": 1.91, "grad_norm": 15.1875, "learning_rate": 2.598430982597461e-08, "loss": 0.2333, "step": 10101 }, { "epoch": 1.91, "grad_norm": 13.375, "learning_rate": 2.5874280703372113e-08, "loss": 0.1688, "step": 10102 }, { "epoch": 1.91, "grad_norm": 14.625, "learning_rate": 2.5764483819623596e-08, "loss": 0.2802, "step": 10103 }, { "epoch": 1.91, "grad_norm": 9.75, "learning_rate": 2.5654919185035543e-08, "loss": 0.2211, "step": 10104 }, { "epoch": 1.91, "grad_norm": 13.8125, "learning_rate": 2.554558680989222e-08, "loss": 0.1976, "step": 10105 }, { "epoch": 1.91, "grad_norm": 10.1875, "learning_rate": 2.5436486704456255e-08, "loss": 0.1725, "step": 10106 }, { "epoch": 1.91, "grad_norm": 12.1875, "learning_rate": 2.5327618878968618e-08, "loss": 0.2482, "step": 10107 }, { "epoch": 1.91, "grad_norm": 22.875, "learning_rate": 2.521898334364864e-08, "loss": 0.1599, "step": 10108 }, { "epoch": 1.91, "grad_norm": 9.375, "learning_rate": 2.511058010869344e-08, "loss": 0.1726, "step": 10109 }, { "epoch": 1.91, "grad_norm": 10.5625, "learning_rate": 2.5002409184278485e-08, "loss": 0.2074, "step": 10110 }, { "epoch": 1.91, "grad_norm": 11.125, "learning_rate": 2.4894470580557327e-08, "loss": 0.2323, "step": 10111 }, { "epoch": 1.91, "grad_norm": 14.625, "learning_rate": 2.4786764307662414e-08, "loss": 0.2129, "step": 10112 }, { "epoch": 1.91, "grad_norm": 10.6875, "learning_rate": 2.4679290375703156e-08, "loss": 0.1685, "step": 10113 }, { "epoch": 1.91, "grad_norm": 11.0, "learning_rate": 2.4572048794768154e-08, "loss": 0.163, "step": 10114 }, { "epoch": 1.91, "grad_norm": 10.0, "learning_rate": 2.446503957492352e-08, "loss": 0.2114, "step": 10115 }, { "epoch": 1.91, "grad_norm": 8.375, "learning_rate": 2.4358262726214278e-08, "loss": 0.1167, "step": 10116 }, { "epoch": 1.91, "grad_norm": 16.25, "learning_rate": 2.4251718258663236e-08, "loss": 0.232, "step": 10117 }, { "epoch": 1.91, "grad_norm": 11.9375, "learning_rate": 2.4145406182271292e-08, "loss": 0.2211, "step": 10118 }, { "epoch": 1.91, "grad_norm": 10.375, "learning_rate": 2.403932650701768e-08, "loss": 0.2005, "step": 10119 }, { "epoch": 1.91, "grad_norm": 10.1875, "learning_rate": 2.3933479242859715e-08, "loss": 0.1843, "step": 10120 }, { "epoch": 1.91, "grad_norm": 10.75, "learning_rate": 2.382786439973278e-08, "loss": 0.2357, "step": 10121 }, { "epoch": 1.91, "grad_norm": 10.375, "learning_rate": 2.372248198755117e-08, "loss": 0.1785, "step": 10122 }, { "epoch": 1.91, "grad_norm": 12.0, "learning_rate": 2.361733201620642e-08, "loss": 0.2466, "step": 10123 }, { "epoch": 1.92, "grad_norm": 9.5625, "learning_rate": 2.351241449556868e-08, "loss": 0.1664, "step": 10124 }, { "epoch": 1.92, "grad_norm": 10.8125, "learning_rate": 2.3407729435486194e-08, "loss": 0.217, "step": 10125 }, { "epoch": 1.92, "grad_norm": 12.0625, "learning_rate": 2.3303276845785538e-08, "loss": 0.2295, "step": 10126 }, { "epoch": 1.92, "grad_norm": 11.0625, "learning_rate": 2.3199056736271653e-08, "loss": 0.1844, "step": 10127 }, { "epoch": 1.92, "grad_norm": 9.75, "learning_rate": 2.3095069116726988e-08, "loss": 0.1587, "step": 10128 }, { "epoch": 1.92, "grad_norm": 12.9375, "learning_rate": 2.2991313996912345e-08, "loss": 0.2109, "step": 10129 }, { "epoch": 1.92, "grad_norm": 13.125, "learning_rate": 2.2887791386567436e-08, "loss": 0.2613, "step": 10130 }, { "epoch": 1.92, "grad_norm": 11.8125, "learning_rate": 2.278450129540949e-08, "loss": 0.1908, "step": 10131 }, { "epoch": 1.92, "grad_norm": 11.1875, "learning_rate": 2.2681443733133802e-08, "loss": 0.2182, "step": 10132 }, { "epoch": 1.92, "grad_norm": 10.75, "learning_rate": 2.25786187094143e-08, "loss": 0.1724, "step": 10133 }, { "epoch": 1.92, "grad_norm": 14.0625, "learning_rate": 2.2476026233902716e-08, "loss": 0.2632, "step": 10134 }, { "epoch": 1.92, "grad_norm": 14.25, "learning_rate": 2.2373666316229393e-08, "loss": 0.2278, "step": 10135 }, { "epoch": 1.92, "grad_norm": 14.25, "learning_rate": 2.2271538966002203e-08, "loss": 0.1613, "step": 10136 }, { "epoch": 1.92, "grad_norm": 11.875, "learning_rate": 2.216964419280765e-08, "loss": 0.2436, "step": 10137 }, { "epoch": 1.92, "grad_norm": 12.625, "learning_rate": 2.2067982006210575e-08, "loss": 0.1714, "step": 10138 }, { "epoch": 1.92, "grad_norm": 20.5, "learning_rate": 2.196655241575335e-08, "loss": 0.1861, "step": 10139 }, { "epoch": 1.92, "grad_norm": 11.0, "learning_rate": 2.186535543095669e-08, "loss": 0.249, "step": 10140 }, { "epoch": 1.92, "grad_norm": 9.75, "learning_rate": 2.1764391061319945e-08, "loss": 0.1536, "step": 10141 }, { "epoch": 1.92, "grad_norm": 9.125, "learning_rate": 2.1663659316320806e-08, "loss": 0.1574, "step": 10142 }, { "epoch": 1.92, "grad_norm": 16.25, "learning_rate": 2.1563160205413934e-08, "loss": 0.2266, "step": 10143 }, { "epoch": 1.92, "grad_norm": 16.125, "learning_rate": 2.1462893738033175e-08, "loss": 0.2677, "step": 10144 }, { "epoch": 1.92, "grad_norm": 9.25, "learning_rate": 2.1362859923589885e-08, "loss": 0.212, "step": 10145 }, { "epoch": 1.92, "grad_norm": 10.375, "learning_rate": 2.1263058771474887e-08, "loss": 0.2192, "step": 10146 }, { "epoch": 1.92, "grad_norm": 10.1875, "learning_rate": 2.1163490291055132e-08, "loss": 0.1569, "step": 10147 }, { "epoch": 1.92, "grad_norm": 9.3125, "learning_rate": 2.1064154491677313e-08, "loss": 0.1736, "step": 10148 }, { "epoch": 1.92, "grad_norm": 11.375, "learning_rate": 2.0965051382665913e-08, "loss": 0.211, "step": 10149 }, { "epoch": 1.92, "grad_norm": 11.75, "learning_rate": 2.0866180973323493e-08, "loss": 0.2181, "step": 10150 }, { "epoch": 1.92, "grad_norm": 10.6875, "learning_rate": 2.0767543272930123e-08, "loss": 0.2293, "step": 10151 }, { "epoch": 1.92, "grad_norm": 13.875, "learning_rate": 2.066913829074535e-08, "loss": 0.2023, "step": 10152 }, { "epoch": 1.92, "grad_norm": 10.25, "learning_rate": 2.057096603600567e-08, "loss": 0.1784, "step": 10153 }, { "epoch": 1.92, "grad_norm": 14.0, "learning_rate": 2.0473026517926487e-08, "loss": 0.2379, "step": 10154 }, { "epoch": 1.92, "grad_norm": 10.125, "learning_rate": 2.0375319745701004e-08, "loss": 0.1873, "step": 10155 }, { "epoch": 1.92, "grad_norm": 14.5, "learning_rate": 2.02778457285005e-08, "loss": 0.1624, "step": 10156 }, { "epoch": 1.92, "grad_norm": 12.875, "learning_rate": 2.0180604475474874e-08, "loss": 0.2379, "step": 10157 }, { "epoch": 1.92, "grad_norm": 11.9375, "learning_rate": 2.0083595995751825e-08, "loss": 0.1754, "step": 10158 }, { "epoch": 1.92, "grad_norm": 13.0625, "learning_rate": 1.9986820298437128e-08, "loss": 0.2011, "step": 10159 }, { "epoch": 1.92, "grad_norm": 11.3125, "learning_rate": 1.98902773926149e-08, "loss": 0.2568, "step": 10160 }, { "epoch": 1.92, "grad_norm": 9.375, "learning_rate": 1.979396728734706e-08, "loss": 0.1765, "step": 10161 }, { "epoch": 1.92, "grad_norm": 13.875, "learning_rate": 1.9697889991674712e-08, "loss": 0.2299, "step": 10162 }, { "epoch": 1.92, "grad_norm": 12.75, "learning_rate": 1.9602045514615364e-08, "loss": 0.1657, "step": 10163 }, { "epoch": 1.92, "grad_norm": 10.5625, "learning_rate": 1.9506433865166263e-08, "loss": 0.1691, "step": 10164 }, { "epoch": 1.92, "grad_norm": 13.8125, "learning_rate": 1.9411055052302175e-08, "loss": 0.2783, "step": 10165 }, { "epoch": 1.92, "grad_norm": 10.75, "learning_rate": 1.9315909084975937e-08, "loss": 0.1746, "step": 10166 }, { "epoch": 1.92, "grad_norm": 10.3125, "learning_rate": 1.9220995972118183e-08, "loss": 0.1756, "step": 10167 }, { "epoch": 1.92, "grad_norm": 12.0625, "learning_rate": 1.9126315722639012e-08, "loss": 0.2357, "step": 10168 }, { "epoch": 1.92, "grad_norm": 11.4375, "learning_rate": 1.9031868345425196e-08, "loss": 0.2954, "step": 10169 }, { "epoch": 1.92, "grad_norm": 11.0, "learning_rate": 1.8937653849342153e-08, "loss": 0.2173, "step": 10170 }, { "epoch": 1.92, "grad_norm": 13.125, "learning_rate": 1.8843672243233635e-08, "loss": 0.1608, "step": 10171 }, { "epoch": 1.92, "grad_norm": 12.8125, "learning_rate": 1.8749923535921754e-08, "loss": 0.2016, "step": 10172 }, { "epoch": 1.92, "grad_norm": 77.0, "learning_rate": 1.865640773620614e-08, "loss": 0.2096, "step": 10173 }, { "epoch": 1.92, "grad_norm": 11.8125, "learning_rate": 1.856312485286449e-08, "loss": 0.1786, "step": 10174 }, { "epoch": 1.92, "grad_norm": 8.125, "learning_rate": 1.847007489465369e-08, "loss": 0.1552, "step": 10175 }, { "epoch": 1.93, "grad_norm": 10.6875, "learning_rate": 1.8377257870307585e-08, "loss": 0.1681, "step": 10176 }, { "epoch": 1.93, "grad_norm": 8.3125, "learning_rate": 1.8284673788538933e-08, "loss": 0.1847, "step": 10177 }, { "epoch": 1.93, "grad_norm": 10.75, "learning_rate": 1.8192322658037998e-08, "loss": 0.1261, "step": 10178 }, { "epoch": 1.93, "grad_norm": 10.625, "learning_rate": 1.8100204487473684e-08, "loss": 0.1458, "step": 10179 }, { "epoch": 1.93, "grad_norm": 9.375, "learning_rate": 1.800831928549296e-08, "loss": 0.1589, "step": 10180 }, { "epoch": 1.93, "grad_norm": 10.5, "learning_rate": 1.791666706072087e-08, "loss": 0.1811, "step": 10181 }, { "epoch": 1.93, "grad_norm": 11.75, "learning_rate": 1.7825247821759984e-08, "loss": 0.2443, "step": 10182 }, { "epoch": 1.93, "grad_norm": 19.375, "learning_rate": 1.773406157719232e-08, "loss": 0.2097, "step": 10183 }, { "epoch": 1.93, "grad_norm": 11.4375, "learning_rate": 1.7643108335576586e-08, "loss": 0.1895, "step": 10184 }, { "epoch": 1.93, "grad_norm": 17.25, "learning_rate": 1.755238810545096e-08, "loss": 0.2484, "step": 10185 }, { "epoch": 1.93, "grad_norm": 10.8125, "learning_rate": 1.7461900895330285e-08, "loss": 0.199, "step": 10186 }, { "epoch": 1.93, "grad_norm": 12.875, "learning_rate": 1.737164671370917e-08, "loss": 0.2352, "step": 10187 }, { "epoch": 1.93, "grad_norm": 10.125, "learning_rate": 1.728162556905888e-08, "loss": 0.195, "step": 10188 }, { "epoch": 1.93, "grad_norm": 11.875, "learning_rate": 1.719183746982961e-08, "loss": 0.2312, "step": 10189 }, { "epoch": 1.93, "grad_norm": 15.25, "learning_rate": 1.7102282424449613e-08, "loss": 0.2044, "step": 10190 }, { "epoch": 1.93, "grad_norm": 11.0625, "learning_rate": 1.7012960441324943e-08, "loss": 0.1346, "step": 10191 }, { "epoch": 1.93, "grad_norm": 10.4375, "learning_rate": 1.6923871528840284e-08, "loss": 0.1889, "step": 10192 }, { "epoch": 1.93, "grad_norm": 12.5625, "learning_rate": 1.6835015695357837e-08, "loss": 0.2317, "step": 10193 }, { "epoch": 1.93, "grad_norm": 11.1875, "learning_rate": 1.6746392949218703e-08, "loss": 0.1743, "step": 10194 }, { "epoch": 1.93, "grad_norm": 36.0, "learning_rate": 1.6658003298740954e-08, "loss": 0.1975, "step": 10195 }, { "epoch": 1.93, "grad_norm": 12.5625, "learning_rate": 1.6569846752222117e-08, "loss": 0.1689, "step": 10196 }, { "epoch": 1.93, "grad_norm": 9.875, "learning_rate": 1.648192331793669e-08, "loss": 0.2212, "step": 10197 }, { "epoch": 1.93, "grad_norm": 8.6875, "learning_rate": 1.6394233004138338e-08, "loss": 0.1883, "step": 10198 }, { "epoch": 1.93, "grad_norm": 10.5625, "learning_rate": 1.6306775819057708e-08, "loss": 0.208, "step": 10199 }, { "epoch": 1.93, "grad_norm": 14.5625, "learning_rate": 1.621955177090434e-08, "loss": 0.2294, "step": 10200 }, { "epoch": 1.93, "grad_norm": 12.3125, "learning_rate": 1.6132560867865853e-08, "loss": 0.2554, "step": 10201 }, { "epoch": 1.93, "grad_norm": 11.8125, "learning_rate": 1.604580311810766e-08, "loss": 0.2776, "step": 10202 }, { "epoch": 1.93, "grad_norm": 20.75, "learning_rate": 1.5959278529773802e-08, "loss": 0.2812, "step": 10203 }, { "epoch": 1.93, "grad_norm": 12.5, "learning_rate": 1.587298711098556e-08, "loss": 0.2477, "step": 10204 }, { "epoch": 1.93, "grad_norm": 12.5, "learning_rate": 1.5786928869843122e-08, "loss": 0.2402, "step": 10205 }, { "epoch": 1.93, "grad_norm": 13.375, "learning_rate": 1.5701103814424468e-08, "loss": 0.2276, "step": 10206 }, { "epoch": 1.93, "grad_norm": 13.375, "learning_rate": 1.5615511952785655e-08, "loss": 0.2125, "step": 10207 }, { "epoch": 1.93, "grad_norm": 12.0625, "learning_rate": 1.553015329296137e-08, "loss": 0.1779, "step": 10208 }, { "epoch": 1.93, "grad_norm": 11.4375, "learning_rate": 1.5445027842963535e-08, "loss": 0.1877, "step": 10209 }, { "epoch": 1.93, "grad_norm": 13.625, "learning_rate": 1.53601356107827e-08, "loss": 0.22, "step": 10210 }, { "epoch": 1.93, "grad_norm": 12.875, "learning_rate": 1.5275476604387494e-08, "loss": 0.2705, "step": 10211 }, { "epoch": 1.93, "grad_norm": 9.5625, "learning_rate": 1.519105083172462e-08, "loss": 0.1568, "step": 10212 }, { "epoch": 1.93, "grad_norm": 11.9375, "learning_rate": 1.510685830071884e-08, "loss": 0.2175, "step": 10213 }, { "epoch": 1.93, "grad_norm": 11.875, "learning_rate": 1.5022899019273006e-08, "loss": 0.2407, "step": 10214 }, { "epoch": 1.93, "grad_norm": 13.75, "learning_rate": 1.4939172995268313e-08, "loss": 0.2432, "step": 10215 }, { "epoch": 1.93, "grad_norm": 13.5, "learning_rate": 1.4855680236563752e-08, "loss": 0.2489, "step": 10216 }, { "epoch": 1.93, "grad_norm": 9.625, "learning_rate": 1.4772420750996664e-08, "loss": 0.2145, "step": 10217 }, { "epoch": 1.93, "grad_norm": 11.25, "learning_rate": 1.4689394546381908e-08, "loss": 0.2337, "step": 10218 }, { "epoch": 1.93, "grad_norm": 9.1875, "learning_rate": 1.4606601630513528e-08, "loss": 0.1897, "step": 10219 }, { "epoch": 1.93, "grad_norm": 13.9375, "learning_rate": 1.4524042011162809e-08, "loss": 0.2414, "step": 10220 }, { "epoch": 1.93, "grad_norm": 8.875, "learning_rate": 1.4441715696079384e-08, "loss": 0.1637, "step": 10221 }, { "epoch": 1.93, "grad_norm": 10.0625, "learning_rate": 1.4359622692990683e-08, "loss": 0.2008, "step": 10222 }, { "epoch": 1.93, "grad_norm": 10.75, "learning_rate": 1.427776300960304e-08, "loss": 0.1976, "step": 10223 }, { "epoch": 1.93, "grad_norm": 8.5, "learning_rate": 1.4196136653600034e-08, "loss": 0.1824, "step": 10224 }, { "epoch": 1.93, "grad_norm": 9.8125, "learning_rate": 1.4114743632643868e-08, "loss": 0.1714, "step": 10225 }, { "epoch": 1.93, "grad_norm": 9.875, "learning_rate": 1.403358395437454e-08, "loss": 0.1812, "step": 10226 }, { "epoch": 1.93, "grad_norm": 11.25, "learning_rate": 1.3952657626410404e-08, "loss": 0.1594, "step": 10227 }, { "epoch": 1.93, "grad_norm": 8.5625, "learning_rate": 1.3871964656347603e-08, "loss": 0.1905, "step": 10228 }, { "epoch": 1.94, "grad_norm": 10.9375, "learning_rate": 1.3791505051760357e-08, "loss": 0.2022, "step": 10229 }, { "epoch": 1.94, "grad_norm": 13.125, "learning_rate": 1.3711278820202067e-08, "loss": 0.1798, "step": 10230 }, { "epoch": 1.94, "grad_norm": 8.5625, "learning_rate": 1.3631285969202267e-08, "loss": 0.1534, "step": 10231 }, { "epoch": 1.94, "grad_norm": 9.1875, "learning_rate": 1.355152650627023e-08, "loss": 0.2045, "step": 10232 }, { "epoch": 1.94, "grad_norm": 11.0625, "learning_rate": 1.3472000438892462e-08, "loss": 0.2272, "step": 10233 }, { "epoch": 1.94, "grad_norm": 11.5625, "learning_rate": 1.3392707774534109e-08, "loss": 0.2286, "step": 10234 }, { "epoch": 1.94, "grad_norm": 26.5, "learning_rate": 1.33136485206381e-08, "loss": 0.2629, "step": 10235 }, { "epoch": 1.94, "grad_norm": 12.4375, "learning_rate": 1.323482268462517e-08, "loss": 0.15, "step": 10236 }, { "epoch": 1.94, "grad_norm": 10.1875, "learning_rate": 1.3156230273894954e-08, "loss": 0.161, "step": 10237 }, { "epoch": 1.94, "grad_norm": 12.375, "learning_rate": 1.307787129582433e-08, "loss": 0.2558, "step": 10238 }, { "epoch": 1.94, "grad_norm": 11.625, "learning_rate": 1.29997457577688e-08, "loss": 0.248, "step": 10239 }, { "epoch": 1.94, "grad_norm": 9.1875, "learning_rate": 1.292185366706139e-08, "loss": 0.1521, "step": 10240 }, { "epoch": 1.94, "grad_norm": 7.625, "learning_rate": 1.2844195031014307e-08, "loss": 0.1528, "step": 10241 }, { "epoch": 1.94, "grad_norm": 10.0625, "learning_rate": 1.2766769856916439e-08, "loss": 0.1562, "step": 10242 }, { "epoch": 1.94, "grad_norm": 13.875, "learning_rate": 1.2689578152036142e-08, "loss": 0.2118, "step": 10243 }, { "epoch": 1.94, "grad_norm": 11.1875, "learning_rate": 1.2612619923618451e-08, "loss": 0.2082, "step": 10244 }, { "epoch": 1.94, "grad_norm": 11.0, "learning_rate": 1.2535895178887592e-08, "loss": 0.1661, "step": 10245 }, { "epoch": 1.94, "grad_norm": 9.8125, "learning_rate": 1.2459403925045577e-08, "loss": 0.2321, "step": 10246 }, { "epoch": 1.94, "grad_norm": 13.5, "learning_rate": 1.238314616927222e-08, "loss": 0.2075, "step": 10247 }, { "epoch": 1.94, "grad_norm": 15.0625, "learning_rate": 1.2307121918725685e-08, "loss": 0.2454, "step": 10248 }, { "epoch": 1.94, "grad_norm": 9.6875, "learning_rate": 1.2231331180542206e-08, "loss": 0.2115, "step": 10249 }, { "epoch": 1.94, "grad_norm": 11.5, "learning_rate": 1.2155773961835814e-08, "loss": 0.2344, "step": 10250 }, { "epoch": 1.94, "grad_norm": 12.75, "learning_rate": 1.2080450269698896e-08, "loss": 0.2606, "step": 10251 }, { "epoch": 1.94, "grad_norm": 12.5, "learning_rate": 1.200536011120218e-08, "loss": 0.2197, "step": 10252 }, { "epoch": 1.94, "grad_norm": 10.0, "learning_rate": 1.1930503493394197e-08, "loss": 0.1898, "step": 10253 }, { "epoch": 1.94, "grad_norm": 12.0625, "learning_rate": 1.1855880423300714e-08, "loss": 0.202, "step": 10254 }, { "epoch": 1.94, "grad_norm": 13.8125, "learning_rate": 1.178149090792724e-08, "loss": 0.2259, "step": 10255 }, { "epoch": 1.94, "grad_norm": 12.25, "learning_rate": 1.1707334954255967e-08, "loss": 0.2211, "step": 10256 }, { "epoch": 1.94, "grad_norm": 9.3125, "learning_rate": 1.163341256924827e-08, "loss": 0.2087, "step": 10257 }, { "epoch": 1.94, "grad_norm": 14.5, "learning_rate": 1.1559723759842211e-08, "loss": 0.2173, "step": 10258 }, { "epoch": 1.94, "grad_norm": 10.125, "learning_rate": 1.1486268532955313e-08, "loss": 0.2051, "step": 10259 }, { "epoch": 1.94, "grad_norm": 17.375, "learning_rate": 1.1413046895482616e-08, "loss": 0.2505, "step": 10260 }, { "epoch": 1.94, "grad_norm": 13.1875, "learning_rate": 1.134005885429723e-08, "loss": 0.2287, "step": 10261 }, { "epoch": 1.94, "grad_norm": 11.8125, "learning_rate": 1.1267304416249791e-08, "loss": 0.1885, "step": 10262 }, { "epoch": 1.94, "grad_norm": 29.0, "learning_rate": 1.119478358817011e-08, "loss": 0.2353, "step": 10263 }, { "epoch": 1.94, "grad_norm": 13.125, "learning_rate": 1.1122496376865244e-08, "loss": 0.1974, "step": 10264 }, { "epoch": 1.94, "grad_norm": 11.375, "learning_rate": 1.1050442789120874e-08, "loss": 0.2225, "step": 10265 }, { "epoch": 1.94, "grad_norm": 13.5, "learning_rate": 1.0978622831699926e-08, "loss": 0.1726, "step": 10266 }, { "epoch": 1.94, "grad_norm": 14.0, "learning_rate": 1.0907036511344504e-08, "loss": 0.2048, "step": 10267 }, { "epoch": 1.94, "grad_norm": 12.8125, "learning_rate": 1.083568383477368e-08, "loss": 0.2368, "step": 10268 }, { "epoch": 1.94, "grad_norm": 13.875, "learning_rate": 1.0764564808685707e-08, "loss": 0.2349, "step": 10269 }, { "epoch": 1.94, "grad_norm": 11.5, "learning_rate": 1.0693679439755523e-08, "loss": 0.2533, "step": 10270 }, { "epoch": 1.94, "grad_norm": 10.125, "learning_rate": 1.0623027734637803e-08, "loss": 0.1898, "step": 10271 }, { "epoch": 1.94, "grad_norm": 17.75, "learning_rate": 1.0552609699963912e-08, "loss": 0.1822, "step": 10272 }, { "epoch": 1.94, "grad_norm": 23.5, "learning_rate": 1.0482425342343838e-08, "loss": 0.2094, "step": 10273 }, { "epoch": 1.94, "grad_norm": 13.875, "learning_rate": 1.0412474668365369e-08, "loss": 0.2124, "step": 10274 }, { "epoch": 1.94, "grad_norm": 16.75, "learning_rate": 1.0342757684595195e-08, "loss": 0.229, "step": 10275 }, { "epoch": 1.94, "grad_norm": 10.4375, "learning_rate": 1.0273274397576971e-08, "loss": 0.2187, "step": 10276 }, { "epoch": 1.94, "grad_norm": 10.6875, "learning_rate": 1.0204024813832703e-08, "loss": 0.1924, "step": 10277 }, { "epoch": 1.94, "grad_norm": 12.0, "learning_rate": 1.01350089398633e-08, "loss": 0.214, "step": 10278 }, { "epoch": 1.94, "grad_norm": 13.0625, "learning_rate": 1.006622678214636e-08, "loss": 0.1841, "step": 10279 }, { "epoch": 1.94, "grad_norm": 11.5625, "learning_rate": 9.99767834713894e-09, "loss": 0.2163, "step": 10280 }, { "epoch": 1.94, "grad_norm": 9.0, "learning_rate": 9.929363641274781e-09, "loss": 0.1848, "step": 10281 }, { "epoch": 1.95, "grad_norm": 11.6875, "learning_rate": 9.86128267096681e-09, "loss": 0.2568, "step": 10282 }, { "epoch": 1.95, "grad_norm": 19.625, "learning_rate": 9.793435442605748e-09, "loss": 0.1877, "step": 10283 }, { "epoch": 1.95, "grad_norm": 12.375, "learning_rate": 9.725821962559834e-09, "loss": 0.2182, "step": 10284 }, { "epoch": 1.95, "grad_norm": 10.75, "learning_rate": 9.658442237175935e-09, "loss": 0.1968, "step": 10285 }, { "epoch": 1.95, "grad_norm": 12.25, "learning_rate": 9.591296272778439e-09, "loss": 0.2172, "step": 10286 }, { "epoch": 1.95, "grad_norm": 14.625, "learning_rate": 9.524384075670912e-09, "loss": 0.2147, "step": 10287 }, { "epoch": 1.95, "grad_norm": 10.6875, "learning_rate": 9.457705652133331e-09, "loss": 0.2092, "step": 10288 }, { "epoch": 1.95, "grad_norm": 8.4375, "learning_rate": 9.391261008425134e-09, "loss": 0.2191, "step": 10289 }, { "epoch": 1.95, "grad_norm": 14.8125, "learning_rate": 9.325050150783277e-09, "loss": 0.2701, "step": 10290 }, { "epoch": 1.95, "grad_norm": 11.3125, "learning_rate": 9.259073085422509e-09, "loss": 0.2325, "step": 10291 }, { "epoch": 1.95, "grad_norm": 12.125, "learning_rate": 9.193329818535934e-09, "loss": 0.2335, "step": 10292 }, { "epoch": 1.95, "grad_norm": 11.75, "learning_rate": 9.127820356295002e-09, "loss": 0.1922, "step": 10293 }, { "epoch": 1.95, "grad_norm": 14.125, "learning_rate": 9.062544704848687e-09, "loss": 0.2419, "step": 10294 }, { "epoch": 1.95, "grad_norm": 11.3125, "learning_rate": 8.997502870324026e-09, "loss": 0.2723, "step": 10295 }, { "epoch": 1.95, "grad_norm": 11.625, "learning_rate": 8.932694858826696e-09, "loss": 0.1858, "step": 10296 }, { "epoch": 1.95, "grad_norm": 13.5, "learning_rate": 8.868120676439607e-09, "loss": 0.2335, "step": 10297 }, { "epoch": 1.95, "grad_norm": 9.9375, "learning_rate": 8.803780329224576e-09, "loss": 0.1948, "step": 10298 }, { "epoch": 1.95, "grad_norm": 10.5, "learning_rate": 8.739673823220662e-09, "loss": 0.2021, "step": 10299 }, { "epoch": 1.95, "grad_norm": 11.375, "learning_rate": 8.675801164445553e-09, "loss": 0.2209, "step": 10300 }, { "epoch": 1.95, "grad_norm": 8.0625, "learning_rate": 8.612162358895005e-09, "loss": 0.1422, "step": 10301 }, { "epoch": 1.95, "grad_norm": 10.0625, "learning_rate": 8.54875741254202e-09, "loss": 0.206, "step": 10302 }, { "epoch": 1.95, "grad_norm": 11.375, "learning_rate": 8.48558633133878e-09, "loss": 0.1789, "step": 10303 }, { "epoch": 1.95, "grad_norm": 11.9375, "learning_rate": 8.422649121214433e-09, "loss": 0.2027, "step": 10304 }, { "epoch": 1.95, "grad_norm": 9.8125, "learning_rate": 8.359945788077306e-09, "loss": 0.1957, "step": 10305 }, { "epoch": 1.95, "grad_norm": 9.75, "learning_rate": 8.297476337812971e-09, "loss": 0.1569, "step": 10306 }, { "epoch": 1.95, "grad_norm": 11.1875, "learning_rate": 8.235240776285069e-09, "loss": 0.1865, "step": 10307 }, { "epoch": 1.95, "grad_norm": 13.0625, "learning_rate": 8.173239109335595e-09, "loss": 0.2235, "step": 10308 }, { "epoch": 1.95, "grad_norm": 10.125, "learning_rate": 8.111471342784338e-09, "loss": 0.176, "step": 10309 }, { "epoch": 1.95, "grad_norm": 12.375, "learning_rate": 8.049937482429437e-09, "loss": 0.2205, "step": 10310 }, { "epoch": 1.95, "grad_norm": 12.0, "learning_rate": 7.988637534046828e-09, "loss": 0.2155, "step": 10311 }, { "epoch": 1.95, "grad_norm": 13.5625, "learning_rate": 7.927571503390518e-09, "loss": 0.2512, "step": 10312 }, { "epoch": 1.95, "grad_norm": 15.25, "learning_rate": 7.86673939619259e-09, "loss": 0.2329, "step": 10313 }, { "epoch": 1.95, "grad_norm": 14.8125, "learning_rate": 7.8061412181632e-09, "loss": 0.2187, "step": 10314 }, { "epoch": 1.95, "grad_norm": 7.78125, "learning_rate": 7.74577697499057e-09, "loss": 0.1714, "step": 10315 }, { "epoch": 1.95, "grad_norm": 14.25, "learning_rate": 7.68564667234073e-09, "loss": 0.2876, "step": 10316 }, { "epoch": 1.95, "grad_norm": 10.9375, "learning_rate": 7.62575031585805e-09, "loss": 0.1825, "step": 10317 }, { "epoch": 1.95, "grad_norm": 13.125, "learning_rate": 7.566087911165254e-09, "loss": 0.2175, "step": 10318 }, { "epoch": 1.95, "grad_norm": 10.75, "learning_rate": 7.506659463861754e-09, "loss": 0.172, "step": 10319 }, { "epoch": 1.95, "grad_norm": 12.0, "learning_rate": 7.447464979526697e-09, "loss": 0.221, "step": 10320 }, { "epoch": 1.95, "grad_norm": 11.5, "learning_rate": 7.388504463716195e-09, "loss": 0.204, "step": 10321 }, { "epoch": 1.95, "grad_norm": 11.8125, "learning_rate": 7.329777921964709e-09, "loss": 0.2186, "step": 10322 }, { "epoch": 1.95, "grad_norm": 14.375, "learning_rate": 7.271285359784774e-09, "loss": 0.2444, "step": 10323 }, { "epoch": 1.95, "grad_norm": 11.5625, "learning_rate": 7.213026782666999e-09, "loss": 0.2196, "step": 10324 }, { "epoch": 1.95, "grad_norm": 10.9375, "learning_rate": 7.155002196079508e-09, "loss": 0.1878, "step": 10325 }, { "epoch": 1.95, "grad_norm": 11.875, "learning_rate": 7.097211605469611e-09, "loss": 0.1752, "step": 10326 }, { "epoch": 1.95, "grad_norm": 9.625, "learning_rate": 7.039655016261304e-09, "loss": 0.204, "step": 10327 }, { "epoch": 1.95, "grad_norm": 12.875, "learning_rate": 6.982332433857486e-09, "loss": 0.2078, "step": 10328 }, { "epoch": 1.95, "grad_norm": 14.375, "learning_rate": 6.9252438636391304e-09, "loss": 0.2315, "step": 10329 }, { "epoch": 1.95, "grad_norm": 13.6875, "learning_rate": 6.868389310964452e-09, "loss": 0.1891, "step": 10330 }, { "epoch": 1.95, "grad_norm": 8.75, "learning_rate": 6.811768781170569e-09, "loss": 0.2127, "step": 10331 }, { "epoch": 1.95, "grad_norm": 12.8125, "learning_rate": 6.75538227957212e-09, "loss": 0.2364, "step": 10332 }, { "epoch": 1.95, "grad_norm": 8.5625, "learning_rate": 6.699229811462371e-09, "loss": 0.1563, "step": 10333 }, { "epoch": 1.95, "grad_norm": 19.25, "learning_rate": 6.64331138211155e-09, "loss": 0.2405, "step": 10334 }, { "epoch": 1.96, "grad_norm": 9.75, "learning_rate": 6.587626996768792e-09, "loss": 0.2197, "step": 10335 }, { "epoch": 1.96, "grad_norm": 11.9375, "learning_rate": 6.5321766606610274e-09, "loss": 0.164, "step": 10336 }, { "epoch": 1.96, "grad_norm": 12.8125, "learning_rate": 6.4769603789932575e-09, "loss": 0.1801, "step": 10337 }, { "epoch": 1.96, "grad_norm": 12.0, "learning_rate": 6.42197815694856e-09, "loss": 0.2137, "step": 10338 }, { "epoch": 1.96, "grad_norm": 11.8125, "learning_rate": 6.367229999687808e-09, "loss": 0.1917, "step": 10339 }, { "epoch": 1.96, "grad_norm": 11.0, "learning_rate": 6.3127159123502206e-09, "loss": 0.2326, "step": 10340 }, { "epoch": 1.96, "grad_norm": 10.5, "learning_rate": 6.25843590005254e-09, "loss": 0.1848, "step": 10341 }, { "epoch": 1.96, "grad_norm": 9.9375, "learning_rate": 6.204389967890412e-09, "loss": 0.186, "step": 10342 }, { "epoch": 1.96, "grad_norm": 13.25, "learning_rate": 6.150578120936445e-09, "loss": 0.1642, "step": 10343 }, { "epoch": 1.96, "grad_norm": 10.5, "learning_rate": 6.097000364241879e-09, "loss": 0.1605, "step": 10344 }, { "epoch": 1.96, "grad_norm": 16.5, "learning_rate": 6.043656702836298e-09, "loss": 0.2423, "step": 10345 }, { "epoch": 1.96, "grad_norm": 16.0, "learning_rate": 5.990547141726533e-09, "loss": 0.1492, "step": 10346 }, { "epoch": 1.96, "grad_norm": 10.9375, "learning_rate": 5.937671685897761e-09, "loss": 0.1691, "step": 10347 }, { "epoch": 1.96, "grad_norm": 9.625, "learning_rate": 5.885030340313513e-09, "loss": 0.2012, "step": 10348 }, { "epoch": 1.96, "grad_norm": 12.9375, "learning_rate": 5.8326231099148364e-09, "loss": 0.2303, "step": 10349 }, { "epoch": 1.96, "grad_norm": 11.9375, "learning_rate": 5.780449999621407e-09, "loss": 0.233, "step": 10350 }, { "epoch": 1.96, "grad_norm": 9.625, "learning_rate": 5.728511014330141e-09, "loss": 0.1761, "step": 10351 }, { "epoch": 1.96, "grad_norm": 13.875, "learning_rate": 5.676806158916581e-09, "loss": 0.1565, "step": 10352 }, { "epoch": 1.96, "grad_norm": 11.1875, "learning_rate": 5.625335438234347e-09, "loss": 0.2127, "step": 10353 }, { "epoch": 1.96, "grad_norm": 13.6875, "learning_rate": 5.574098857114296e-09, "loss": 0.2025, "step": 10354 }, { "epoch": 1.96, "grad_norm": 11.4375, "learning_rate": 5.52309642036647e-09, "loss": 0.2243, "step": 10355 }, { "epoch": 1.96, "grad_norm": 10.1875, "learning_rate": 5.472328132777593e-09, "loss": 0.1766, "step": 10356 }, { "epoch": 1.96, "grad_norm": 27.625, "learning_rate": 5.421793999114134e-09, "loss": 0.1668, "step": 10357 }, { "epoch": 1.96, "grad_norm": 12.125, "learning_rate": 5.371494024118406e-09, "loss": 0.1622, "step": 10358 }, { "epoch": 1.96, "grad_norm": 13.1875, "learning_rate": 5.321428212512747e-09, "loss": 0.1708, "step": 10359 }, { "epoch": 1.96, "grad_norm": 11.125, "learning_rate": 5.2715965689967285e-09, "loss": 0.1547, "step": 10360 }, { "epoch": 1.96, "grad_norm": 11.25, "learning_rate": 5.2219990982471655e-09, "loss": 0.1861, "step": 10361 }, { "epoch": 1.96, "grad_norm": 10.5, "learning_rate": 5.172635804920334e-09, "loss": 0.1781, "step": 10362 }, { "epoch": 1.96, "grad_norm": 9.625, "learning_rate": 5.123506693649471e-09, "loss": 0.1882, "step": 10363 }, { "epoch": 1.96, "grad_norm": 10.1875, "learning_rate": 5.074611769046444e-09, "loss": 0.2084, "step": 10364 }, { "epoch": 1.96, "grad_norm": 14.25, "learning_rate": 5.025951035700638e-09, "loss": 0.3308, "step": 10365 }, { "epoch": 1.96, "grad_norm": 10.5625, "learning_rate": 4.977524498179786e-09, "loss": 0.219, "step": 10366 }, { "epoch": 1.96, "grad_norm": 9.9375, "learning_rate": 4.929332161029421e-09, "loss": 0.1732, "step": 10367 }, { "epoch": 1.96, "grad_norm": 12.75, "learning_rate": 4.881374028773422e-09, "loss": 0.258, "step": 10368 }, { "epoch": 1.96, "grad_norm": 13.125, "learning_rate": 4.8336501059131905e-09, "loss": 0.2581, "step": 10369 }, { "epoch": 1.96, "grad_norm": 10.6875, "learning_rate": 4.786160396928752e-09, "loss": 0.2602, "step": 10370 }, { "epoch": 1.96, "grad_norm": 13.125, "learning_rate": 4.738904906277653e-09, "loss": 0.1727, "step": 10371 }, { "epoch": 1.96, "grad_norm": 24.125, "learning_rate": 4.6918836383957885e-09, "loss": 0.1827, "step": 10372 }, { "epoch": 1.96, "grad_norm": 13.375, "learning_rate": 4.645096597696575e-09, "loss": 0.2043, "step": 10373 }, { "epoch": 1.96, "grad_norm": 12.5, "learning_rate": 4.598543788572052e-09, "loss": 0.25, "step": 10374 }, { "epoch": 1.96, "grad_norm": 12.0, "learning_rate": 4.552225215391781e-09, "loss": 0.2023, "step": 10375 }, { "epoch": 1.96, "grad_norm": 12.875, "learning_rate": 4.506140882503951e-09, "loss": 0.1556, "step": 10376 }, { "epoch": 1.96, "grad_norm": 10.3125, "learning_rate": 4.46029079423399e-09, "loss": 0.1924, "step": 10377 }, { "epoch": 1.96, "grad_norm": 14.375, "learning_rate": 4.4146749548856785e-09, "loss": 0.2558, "step": 10378 }, { "epoch": 1.96, "grad_norm": 14.1875, "learning_rate": 4.369293368741145e-09, "loss": 0.183, "step": 10379 }, { "epoch": 1.96, "grad_norm": 12.0, "learning_rate": 4.324146040059762e-09, "loss": 0.2122, "step": 10380 }, { "epoch": 1.96, "grad_norm": 16.5, "learning_rate": 4.279232973080083e-09, "loss": 0.2933, "step": 10381 }, { "epoch": 1.96, "grad_norm": 12.0625, "learning_rate": 4.2345541720173465e-09, "loss": 0.2462, "step": 10382 }, { "epoch": 1.96, "grad_norm": 14.6875, "learning_rate": 4.19010964106542e-09, "loss": 0.1906, "step": 10383 }, { "epoch": 1.96, "grad_norm": 10.3125, "learning_rate": 4.145899384396524e-09, "loss": 0.227, "step": 10384 }, { "epoch": 1.96, "grad_norm": 11.8125, "learning_rate": 4.101923406160668e-09, "loss": 0.2175, "step": 10385 }, { "epoch": 1.96, "grad_norm": 9.625, "learning_rate": 4.05818171048511e-09, "loss": 0.1849, "step": 10386 }, { "epoch": 1.96, "grad_norm": 17.0, "learning_rate": 4.0146743014762844e-09, "loss": 0.2029, "step": 10387 }, { "epoch": 1.97, "grad_norm": 10.1875, "learning_rate": 3.971401183217871e-09, "loss": 0.166, "step": 10388 }, { "epoch": 1.97, "grad_norm": 9.8125, "learning_rate": 3.9283623597721755e-09, "loss": 0.1843, "step": 10389 }, { "epoch": 1.97, "grad_norm": 12.0625, "learning_rate": 3.885557835178466e-09, "loss": 0.175, "step": 10390 }, { "epoch": 1.97, "grad_norm": 11.0, "learning_rate": 3.8429876134551955e-09, "loss": 0.1673, "step": 10391 }, { "epoch": 1.97, "grad_norm": 15.375, "learning_rate": 3.8006516985980565e-09, "loss": 0.2457, "step": 10392 }, { "epoch": 1.97, "grad_norm": 13.5, "learning_rate": 3.758550094581092e-09, "loss": 0.2438, "step": 10393 }, { "epoch": 1.97, "grad_norm": 13.0, "learning_rate": 3.7166828053561422e-09, "loss": 0.2362, "step": 10394 }, { "epoch": 1.97, "grad_norm": 9.625, "learning_rate": 3.6750498348531174e-09, "loss": 0.2179, "step": 10395 }, { "epoch": 1.97, "grad_norm": 20.125, "learning_rate": 3.633651186980558e-09, "loss": 0.2326, "step": 10396 }, { "epoch": 1.97, "grad_norm": 11.8125, "learning_rate": 3.59248686562369e-09, "loss": 0.2234, "step": 10397 }, { "epoch": 1.97, "grad_norm": 16.625, "learning_rate": 3.551556874646922e-09, "loss": 0.2623, "step": 10398 }, { "epoch": 1.97, "grad_norm": 13.375, "learning_rate": 3.5108612178919034e-09, "loss": 0.2636, "step": 10399 }, { "epoch": 1.97, "grad_norm": 12.75, "learning_rate": 3.470399899178911e-09, "loss": 0.1832, "step": 10400 }, { "epoch": 1.97, "grad_norm": 10.5625, "learning_rate": 3.430172922306019e-09, "loss": 0.2498, "step": 10401 }, { "epoch": 1.97, "grad_norm": 39.75, "learning_rate": 3.390180291048817e-09, "loss": 0.1729, "step": 10402 }, { "epoch": 1.97, "grad_norm": 11.3125, "learning_rate": 3.350422009161525e-09, "loss": 0.2285, "step": 10403 }, { "epoch": 1.97, "grad_norm": 10.4375, "learning_rate": 3.310898080376157e-09, "loss": 0.2214, "step": 10404 }, { "epoch": 1.97, "grad_norm": 10.625, "learning_rate": 3.271608508402524e-09, "loss": 0.1312, "step": 10405 }, { "epoch": 1.97, "grad_norm": 12.8125, "learning_rate": 3.2325532969287865e-09, "loss": 0.2302, "step": 10406 }, { "epoch": 1.97, "grad_norm": 12.25, "learning_rate": 3.1937324496211784e-09, "loss": 0.1887, "step": 10407 }, { "epoch": 1.97, "grad_norm": 11.375, "learning_rate": 3.1551459701231745e-09, "loss": 0.1946, "step": 10408 }, { "epoch": 1.97, "grad_norm": 13.0625, "learning_rate": 3.1167938620571545e-09, "loss": 0.2302, "step": 10409 }, { "epoch": 1.97, "grad_norm": 9.0, "learning_rate": 3.0786761290230173e-09, "loss": 0.1981, "step": 10410 }, { "epoch": 1.97, "grad_norm": 12.875, "learning_rate": 3.0407927745984554e-09, "loss": 0.2091, "step": 10411 }, { "epoch": 1.97, "grad_norm": 14.8125, "learning_rate": 3.0031438023400696e-09, "loss": 0.2162, "step": 10412 }, { "epoch": 1.97, "grad_norm": 12.125, "learning_rate": 2.965729215781421e-09, "loss": 0.2104, "step": 10413 }, { "epoch": 1.97, "grad_norm": 8.375, "learning_rate": 2.928549018434701e-09, "loss": 0.205, "step": 10414 }, { "epoch": 1.97, "grad_norm": 12.5, "learning_rate": 2.8916032137898953e-09, "loss": 0.2285, "step": 10415 }, { "epoch": 1.97, "grad_norm": 11.75, "learning_rate": 2.8548918053147856e-09, "loss": 0.19, "step": 10416 }, { "epoch": 1.97, "grad_norm": 10.6875, "learning_rate": 2.818414796455504e-09, "loss": 0.1958, "step": 10417 }, { "epoch": 1.97, "grad_norm": 20.375, "learning_rate": 2.782172190635979e-09, "loss": 0.2095, "step": 10418 }, { "epoch": 1.97, "grad_norm": 9.875, "learning_rate": 2.746163991258488e-09, "loss": 0.2279, "step": 10419 }, { "epoch": 1.97, "grad_norm": 10.6875, "learning_rate": 2.7103902017025505e-09, "loss": 0.1507, "step": 10420 }, { "epoch": 1.97, "grad_norm": 14.5625, "learning_rate": 2.6748508253265916e-09, "loss": 0.191, "step": 10421 }, { "epoch": 1.97, "grad_norm": 9.625, "learning_rate": 2.6395458654662752e-09, "loss": 0.2141, "step": 10422 }, { "epoch": 1.97, "grad_norm": 10.625, "learning_rate": 2.604475325435618e-09, "loss": 0.2148, "step": 10423 }, { "epoch": 1.97, "grad_norm": 15.9375, "learning_rate": 2.5696392085267084e-09, "loss": 0.2223, "step": 10424 }, { "epoch": 1.97, "grad_norm": 9.5, "learning_rate": 2.5350375180094313e-09, "loss": 0.1549, "step": 10425 }, { "epoch": 1.97, "grad_norm": 11.125, "learning_rate": 2.500670257131743e-09, "loss": 0.1718, "step": 10426 }, { "epoch": 1.97, "grad_norm": 11.9375, "learning_rate": 2.4665374291196755e-09, "loss": 0.2091, "step": 10427 }, { "epoch": 1.97, "grad_norm": 15.8125, "learning_rate": 2.4326390371770537e-09, "loss": 0.2708, "step": 10428 }, { "epoch": 1.97, "grad_norm": 14.125, "learning_rate": 2.3989750844857772e-09, "loss": 0.2478, "step": 10429 }, { "epoch": 1.97, "grad_norm": 20.75, "learning_rate": 2.3655455742060962e-09, "loss": 0.2939, "step": 10430 }, { "epoch": 1.97, "grad_norm": 13.8125, "learning_rate": 2.3323505094755006e-09, "loss": 0.2475, "step": 10431 }, { "epoch": 1.97, "grad_norm": 14.9375, "learning_rate": 2.2993898934103862e-09, "loss": 0.2465, "step": 10432 }, { "epoch": 1.97, "grad_norm": 10.6875, "learning_rate": 2.2666637291043904e-09, "loss": 0.1932, "step": 10433 }, { "epoch": 1.97, "grad_norm": 13.5625, "learning_rate": 2.2341720196292215e-09, "loss": 0.185, "step": 10434 }, { "epoch": 1.97, "grad_norm": 10.125, "learning_rate": 2.201914768035218e-09, "loss": 0.2018, "step": 10435 }, { "epoch": 1.97, "grad_norm": 9.875, "learning_rate": 2.169891977349958e-09, "loss": 0.2013, "step": 10436 }, { "epoch": 1.97, "grad_norm": 11.5625, "learning_rate": 2.1381036505796483e-09, "loss": 0.1671, "step": 10437 }, { "epoch": 1.97, "grad_norm": 10.6875, "learning_rate": 2.106549790707735e-09, "loss": 0.207, "step": 10438 }, { "epoch": 1.97, "grad_norm": 11.5625, "learning_rate": 2.075230400696293e-09, "loss": 0.2221, "step": 10439 }, { "epoch": 1.97, "grad_norm": 22.375, "learning_rate": 2.0441454834854715e-09, "loss": 0.2377, "step": 10440 }, { "epoch": 1.98, "grad_norm": 12.125, "learning_rate": 2.0132950419926577e-09, "loss": 0.184, "step": 10441 }, { "epoch": 1.98, "grad_norm": 23.875, "learning_rate": 1.9826790791138696e-09, "loss": 0.2271, "step": 10442 }, { "epoch": 1.98, "grad_norm": 11.25, "learning_rate": 1.9522975977229186e-09, "loss": 0.1915, "step": 10443 }, { "epoch": 1.98, "grad_norm": 14.8125, "learning_rate": 1.9221506006714132e-09, "loss": 0.2521, "step": 10444 }, { "epoch": 1.98, "grad_norm": 12.4375, "learning_rate": 1.892238090789866e-09, "loss": 0.2871, "step": 10445 }, { "epoch": 1.98, "grad_norm": 11.375, "learning_rate": 1.8625600708851998e-09, "loss": 0.2096, "step": 10446 }, { "epoch": 1.98, "grad_norm": 9.875, "learning_rate": 1.833116543743796e-09, "loss": 0.1493, "step": 10447 }, { "epoch": 1.98, "grad_norm": 11.375, "learning_rate": 1.8039075121292772e-09, "loss": 0.1992, "step": 10448 }, { "epoch": 1.98, "grad_norm": 16.125, "learning_rate": 1.7749329787833392e-09, "loss": 0.1965, "step": 10449 }, { "epoch": 1.98, "grad_norm": 12.8125, "learning_rate": 1.746192946425751e-09, "loss": 0.1918, "step": 10450 }, { "epoch": 1.98, "grad_norm": 7.4375, "learning_rate": 1.717687417754077e-09, "loss": 0.1501, "step": 10451 }, { "epoch": 1.98, "grad_norm": 9.4375, "learning_rate": 1.6894163954445098e-09, "loss": 0.1883, "step": 10452 }, { "epoch": 1.98, "grad_norm": 11.125, "learning_rate": 1.6613798821504822e-09, "loss": 0.1603, "step": 10453 }, { "epoch": 1.98, "grad_norm": 10.5625, "learning_rate": 1.6335778805035008e-09, "loss": 0.2148, "step": 10454 }, { "epoch": 1.98, "grad_norm": 11.1875, "learning_rate": 1.6060103931136995e-09, "loss": 0.1704, "step": 10455 }, { "epoch": 1.98, "grad_norm": 12.125, "learning_rate": 1.5786774225684531e-09, "loss": 0.2115, "step": 10456 }, { "epoch": 1.98, "grad_norm": 13.8125, "learning_rate": 1.5515789714334871e-09, "loss": 0.1996, "step": 10457 }, { "epoch": 1.98, "grad_norm": 13.8125, "learning_rate": 1.5247150422526003e-09, "loss": 0.2363, "step": 10458 }, { "epoch": 1.98, "grad_norm": 14.6875, "learning_rate": 1.4980856375471088e-09, "loss": 0.2443, "step": 10459 }, { "epoch": 1.98, "grad_norm": 11.875, "learning_rate": 1.4716907598169571e-09, "loss": 0.1751, "step": 10460 }, { "epoch": 1.98, "grad_norm": 13.375, "learning_rate": 1.4455304115393308e-09, "loss": 0.2625, "step": 10461 }, { "epoch": 1.98, "grad_norm": 13.5625, "learning_rate": 1.4196045951705984e-09, "loss": 0.2398, "step": 10462 }, { "epoch": 1.98, "grad_norm": 15.8125, "learning_rate": 1.3939133131435356e-09, "loss": 0.2149, "step": 10463 }, { "epoch": 1.98, "grad_norm": 10.125, "learning_rate": 1.3684565678701022e-09, "loss": 0.1988, "step": 10464 }, { "epoch": 1.98, "grad_norm": 12.3125, "learning_rate": 1.3432343617397758e-09, "loss": 0.1568, "step": 10465 }, { "epoch": 1.98, "grad_norm": 9.875, "learning_rate": 1.3182466971201069e-09, "loss": 0.1907, "step": 10466 }, { "epoch": 1.98, "grad_norm": 12.5, "learning_rate": 1.2934935763564415e-09, "loss": 0.1943, "step": 10467 }, { "epoch": 1.98, "grad_norm": 7.28125, "learning_rate": 1.2689750017727543e-09, "loss": 0.1455, "step": 10468 }, { "epoch": 1.98, "grad_norm": 9.4375, "learning_rate": 1.244690975669982e-09, "loss": 0.1707, "step": 10469 }, { "epoch": 1.98, "grad_norm": 11.75, "learning_rate": 1.2206415003276906e-09, "loss": 0.2522, "step": 10470 }, { "epoch": 1.98, "grad_norm": 13.0, "learning_rate": 1.1968265780037958e-09, "loss": 0.2259, "step": 10471 }, { "epoch": 1.98, "grad_norm": 13.9375, "learning_rate": 1.1732462109331765e-09, "loss": 0.2711, "step": 10472 }, { "epoch": 1.98, "grad_norm": 10.125, "learning_rate": 1.1499004013296178e-09, "loss": 0.2151, "step": 10473 }, { "epoch": 1.98, "grad_norm": 10.375, "learning_rate": 1.126789151384422e-09, "loss": 0.1395, "step": 10474 }, { "epoch": 1.98, "grad_norm": 8.6875, "learning_rate": 1.1039124632666876e-09, "loss": 0.1627, "step": 10475 }, { "epoch": 1.98, "grad_norm": 8.625, "learning_rate": 1.0812703391241408e-09, "loss": 0.1408, "step": 10476 }, { "epoch": 1.98, "grad_norm": 9.4375, "learning_rate": 1.0588627810820263e-09, "loss": 0.1881, "step": 10477 }, { "epoch": 1.98, "grad_norm": 9.5, "learning_rate": 1.0366897912439388e-09, "loss": 0.1544, "step": 10478 }, { "epoch": 1.98, "grad_norm": 14.6875, "learning_rate": 1.0147513716904368e-09, "loss": 0.2267, "step": 10479 }, { "epoch": 1.98, "grad_norm": 11.0, "learning_rate": 9.93047524481816e-10, "loss": 0.2841, "step": 10480 }, { "epoch": 1.98, "grad_norm": 8.1875, "learning_rate": 9.715782516545036e-10, "loss": 0.1168, "step": 10481 }, { "epoch": 1.98, "grad_norm": 9.8125, "learning_rate": 9.503435552241091e-10, "loss": 0.1995, "step": 10482 }, { "epoch": 1.98, "grad_norm": 12.75, "learning_rate": 9.293434371837607e-10, "loss": 0.2045, "step": 10483 }, { "epoch": 1.98, "grad_norm": 11.625, "learning_rate": 9.085778995049366e-10, "loss": 0.1301, "step": 10484 }, { "epoch": 1.98, "grad_norm": 10.8125, "learning_rate": 8.880469441366335e-10, "loss": 0.1534, "step": 10485 }, { "epoch": 1.98, "grad_norm": 12.125, "learning_rate": 8.677505730061986e-10, "loss": 0.2551, "step": 10486 }, { "epoch": 1.98, "grad_norm": 11.75, "learning_rate": 8.47688788018497e-10, "loss": 0.2642, "step": 10487 }, { "epoch": 1.98, "grad_norm": 8.5625, "learning_rate": 8.278615910567445e-10, "loss": 0.164, "step": 10488 }, { "epoch": 1.98, "grad_norm": 11.8125, "learning_rate": 8.0826898398223e-10, "loss": 0.2039, "step": 10489 }, { "epoch": 1.98, "grad_norm": 12.0, "learning_rate": 7.88910968634038e-10, "loss": 0.2352, "step": 10490 }, { "epoch": 1.98, "grad_norm": 10.9375, "learning_rate": 7.697875468290483e-10, "loss": 0.2108, "step": 10491 }, { "epoch": 1.98, "grad_norm": 11.5625, "learning_rate": 7.508987203627693e-10, "loss": 0.1578, "step": 10492 }, { "epoch": 1.98, "grad_norm": 14.0, "learning_rate": 7.32244491007672e-10, "loss": 0.1906, "step": 10493 }, { "epoch": 1.99, "grad_norm": 14.125, "learning_rate": 7.138248605148557e-10, "loss": 0.2382, "step": 10494 }, { "epoch": 1.99, "grad_norm": 8.375, "learning_rate": 6.956398306137701e-10, "loss": 0.1541, "step": 10495 }, { "epoch": 1.99, "grad_norm": 10.125, "learning_rate": 6.776894030111058e-10, "loss": 0.1764, "step": 10496 }, { "epoch": 1.99, "grad_norm": 12.6875, "learning_rate": 6.599735793916262e-10, "loss": 0.2861, "step": 10497 }, { "epoch": 1.99, "grad_norm": 10.8125, "learning_rate": 6.424923614184453e-10, "loss": 0.2131, "step": 10498 }, { "epoch": 1.99, "grad_norm": 11.75, "learning_rate": 6.252457507327503e-10, "loss": 0.1745, "step": 10499 }, { "epoch": 1.99, "grad_norm": 15.0625, "learning_rate": 6.082337489529689e-10, "loss": 0.2383, "step": 10500 }, { "epoch": 1.99, "grad_norm": 10.375, "learning_rate": 5.914563576761567e-10, "loss": 0.2352, "step": 10501 }, { "epoch": 1.99, "grad_norm": 9.875, "learning_rate": 5.749135784771653e-10, "loss": 0.1931, "step": 10502 }, { "epoch": 1.99, "grad_norm": 15.3125, "learning_rate": 5.586054129089191e-10, "loss": 0.2793, "step": 10503 }, { "epoch": 1.99, "grad_norm": 10.0625, "learning_rate": 5.425318625018605e-10, "loss": 0.1974, "step": 10504 }, { "epoch": 1.99, "grad_norm": 17.75, "learning_rate": 5.266929287653377e-10, "loss": 0.1958, "step": 10505 }, { "epoch": 1.99, "grad_norm": 12.25, "learning_rate": 5.110886131853843e-10, "loss": 0.2629, "step": 10506 }, { "epoch": 1.99, "grad_norm": 11.25, "learning_rate": 4.957189172274946e-10, "loss": 0.1971, "step": 10507 }, { "epoch": 1.99, "grad_norm": 15.125, "learning_rate": 4.805838423338483e-10, "loss": 0.2337, "step": 10508 }, { "epoch": 1.99, "grad_norm": 19.375, "learning_rate": 4.656833899252533e-10, "loss": 0.1918, "step": 10509 }, { "epoch": 1.99, "grad_norm": 15.9375, "learning_rate": 4.5101756140031317e-10, "loss": 0.2052, "step": 10510 }, { "epoch": 1.99, "grad_norm": 14.0625, "learning_rate": 4.3658635813598195e-10, "loss": 0.1856, "step": 10511 }, { "epoch": 1.99, "grad_norm": 11.125, "learning_rate": 4.223897814867317e-10, "loss": 0.266, "step": 10512 }, { "epoch": 1.99, "grad_norm": 8.9375, "learning_rate": 4.084278327848301e-10, "loss": 0.2247, "step": 10513 }, { "epoch": 1.99, "grad_norm": 12.0, "learning_rate": 3.94700513341173e-10, "loss": 0.25, "step": 10514 }, { "epoch": 1.99, "grad_norm": 11.4375, "learning_rate": 3.812078244441741e-10, "loss": 0.2369, "step": 10515 }, { "epoch": 1.99, "grad_norm": 13.3125, "learning_rate": 3.6794976736059805e-10, "loss": 0.1559, "step": 10516 }, { "epoch": 1.99, "grad_norm": 10.5625, "learning_rate": 3.5492634333444964e-10, "loss": 0.2259, "step": 10517 }, { "epoch": 1.99, "grad_norm": 9.3125, "learning_rate": 3.4213755358863956e-10, "loss": 0.1874, "step": 10518 }, { "epoch": 1.99, "grad_norm": 14.75, "learning_rate": 3.2958339932331886e-10, "loss": 0.2648, "step": 10519 }, { "epoch": 1.99, "grad_norm": 12.25, "learning_rate": 3.172638817172669e-10, "loss": 0.188, "step": 10520 }, { "epoch": 1.99, "grad_norm": 18.875, "learning_rate": 3.051790019265033e-10, "loss": 0.2054, "step": 10521 }, { "epoch": 1.99, "grad_norm": 10.4375, "learning_rate": 2.9332876108567607e-10, "loss": 0.1737, "step": 10522 }, { "epoch": 1.99, "grad_norm": 9.5, "learning_rate": 2.817131603069512e-10, "loss": 0.1421, "step": 10523 }, { "epoch": 1.99, "grad_norm": 9.4375, "learning_rate": 2.703322006808451e-10, "loss": 0.1739, "step": 10524 }, { "epoch": 1.99, "grad_norm": 12.6875, "learning_rate": 2.591858832753924e-10, "loss": 0.2613, "step": 10525 }, { "epoch": 1.99, "grad_norm": 12.75, "learning_rate": 2.482742091369783e-10, "loss": 0.2037, "step": 10526 }, { "epoch": 1.99, "grad_norm": 10.9375, "learning_rate": 2.375971792900611e-10, "loss": 0.1853, "step": 10527 }, { "epoch": 1.99, "grad_norm": 9.375, "learning_rate": 2.271547947363395e-10, "loss": 0.1932, "step": 10528 }, { "epoch": 1.99, "grad_norm": 10.5, "learning_rate": 2.1694705645669556e-10, "loss": 0.2232, "step": 10529 }, { "epoch": 1.99, "grad_norm": 10.4375, "learning_rate": 2.0697396540869662e-10, "loss": 0.1696, "step": 10530 }, { "epoch": 1.99, "grad_norm": 14.3125, "learning_rate": 1.9723552252881584e-10, "loss": 0.2272, "step": 10531 }, { "epoch": 1.99, "grad_norm": 12.0625, "learning_rate": 1.8773172873104428e-10, "loss": 0.1976, "step": 10532 }, { "epoch": 1.99, "grad_norm": 14.125, "learning_rate": 1.7846258490772372e-10, "loss": 0.196, "step": 10533 }, { "epoch": 1.99, "grad_norm": 12.1875, "learning_rate": 1.6942809192843635e-10, "loss": 0.1897, "step": 10534 }, { "epoch": 1.99, "grad_norm": 17.5, "learning_rate": 1.6062825064167009e-10, "loss": 0.2247, "step": 10535 }, { "epoch": 1.99, "grad_norm": 15.9375, "learning_rate": 1.5206306187315334e-10, "loss": 0.1858, "step": 10536 }, { "epoch": 1.99, "grad_norm": 12.3125, "learning_rate": 1.4373252642724267e-10, "loss": 0.2121, "step": 10537 }, { "epoch": 1.99, "grad_norm": 8.9375, "learning_rate": 1.3563664508525754e-10, "loss": 0.2227, "step": 10538 }, { "epoch": 1.99, "grad_norm": 16.875, "learning_rate": 1.2777541860797827e-10, "loss": 0.2403, "step": 10539 }, { "epoch": 1.99, "grad_norm": 20.0, "learning_rate": 1.2014884773259296e-10, "loss": 0.2808, "step": 10540 }, { "epoch": 1.99, "grad_norm": 11.6875, "learning_rate": 1.1275693317519543e-10, "loss": 0.2126, "step": 10541 }, { "epoch": 1.99, "grad_norm": 12.375, "learning_rate": 1.0559967562995266e-10, "loss": 0.1987, "step": 10542 }, { "epoch": 1.99, "grad_norm": 9.0, "learning_rate": 9.867707576827201e-11, "loss": 0.2224, "step": 10543 }, { "epoch": 1.99, "grad_norm": 12.375, "learning_rate": 9.198913424018907e-11, "loss": 0.2322, "step": 10544 }, { "epoch": 1.99, "grad_norm": 10.0625, "learning_rate": 8.553585167353495e-11, "loss": 0.2287, "step": 10545 }, { "epoch": 2.0, "grad_norm": 9.625, "learning_rate": 7.931722867393632e-11, "loss": 0.2033, "step": 10546 }, { "epoch": 2.0, "grad_norm": 12.4375, "learning_rate": 7.333326582509292e-11, "loss": 0.1708, "step": 10547 }, { "epoch": 2.0, "grad_norm": 13.0, "learning_rate": 6.758396368850007e-11, "loss": 0.1371, "step": 10548 }, { "epoch": 2.0, "grad_norm": 13.6875, "learning_rate": 6.206932280428124e-11, "loss": 0.2166, "step": 10549 }, { "epoch": 2.0, "grad_norm": 10.8125, "learning_rate": 5.6789343690077934e-11, "loss": 0.1953, "step": 10550 }, { "epoch": 2.0, "grad_norm": 10.1875, "learning_rate": 5.174402684104962e-11, "loss": 0.1504, "step": 10551 }, { "epoch": 2.0, "grad_norm": 14.3125, "learning_rate": 4.6933372730983974e-11, "loss": 0.2645, "step": 10552 }, { "epoch": 2.0, "grad_norm": 11.6875, "learning_rate": 4.235738181174176e-11, "loss": 0.2137, "step": 10553 }, { "epoch": 2.0, "grad_norm": 11.875, "learning_rate": 3.8016054512424184e-11, "loss": 0.1539, "step": 10554 }, { "epoch": 2.0, "grad_norm": 10.875, "learning_rate": 3.390939124076065e-11, "loss": 0.1687, "step": 10555 }, { "epoch": 2.0, "grad_norm": 16.125, "learning_rate": 3.003739238199854e-11, "loss": 0.2521, "step": 10556 }, { "epoch": 2.0, "grad_norm": 10.875, "learning_rate": 2.6400058300013466e-11, "loss": 0.2154, "step": 10557 }, { "epoch": 2.0, "grad_norm": 12.375, "learning_rate": 2.299738933592144e-11, "loss": 0.2336, "step": 10558 }, { "epoch": 2.0, "grad_norm": 12.25, "learning_rate": 1.9829385809466696e-11, "loss": 0.239, "step": 10559 }, { "epoch": 2.0, "grad_norm": 10.8125, "learning_rate": 1.6896048017356338e-11, "loss": 0.2755, "step": 10560 }, { "epoch": 2.0, "grad_norm": 13.625, "learning_rate": 1.4197376235480787e-11, "loss": 0.2158, "step": 10561 }, { "epoch": 2.0, "grad_norm": 11.5, "learning_rate": 1.173337071724845e-11, "loss": 0.2049, "step": 10562 }, { "epoch": 2.0, "grad_norm": 17.5, "learning_rate": 9.504031693308158e-12, "loss": 0.2667, "step": 10563 }, { "epoch": 2.0, "grad_norm": 13.5625, "learning_rate": 7.509359373492064e-12, "loss": 0.2298, "step": 10564 }, { "epoch": 2.0, "grad_norm": 10.625, "learning_rate": 5.7493539451503045e-12, "loss": 0.1871, "step": 10565 }, { "epoch": 2.0, "grad_norm": 10.0, "learning_rate": 4.224015572873441e-12, "loss": 0.1569, "step": 10566 }, { "epoch": 2.0, "grad_norm": 14.375, "learning_rate": 2.9333444001578003e-12, "loss": 0.2829, "step": 10567 }, { "epoch": 2.0, "grad_norm": 7.4375, "learning_rate": 1.877340548295248e-12, "loss": 0.1284, "step": 10568 }, { "epoch": 2.0, "grad_norm": 17.75, "learning_rate": 1.0560041163731883e-12, "loss": 0.1974, "step": 10569 }, { "epoch": 2.0, "grad_norm": 20.25, "learning_rate": 4.693351809970104e-13, "loss": 0.2055, "step": 10570 }, { "epoch": 2.0, "grad_norm": 12.375, "learning_rate": 1.173337979554212e-13, "loss": 0.2311, "step": 10571 }, { "epoch": 2.0, "grad_norm": 8.9375, "learning_rate": 0.0, "loss": 0.1962, "step": 10572 }, { "epoch": 2.0, "step": 10572, "total_flos": 1.0441111681567293e+19, "train_loss": 0.26366888968653857, "train_runtime": 128184.5541, "train_samples_per_second": 10.557, "train_steps_per_second": 0.082 } ], "logging_steps": 1.0, "max_steps": 10572, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 2000, "total_flos": 1.0441111681567293e+19, "train_batch_size": 2, "trial_name": null, "trial_params": null }